Network management method and related device
By grouping APs and STAs using channel parameter information and clustering techniques, the method addresses signal interference in WiFi networks, improving network performance by reducing interference and enhancing transmission capacity.
Patent Information
- Application Number
- CN202011063602.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-09-30
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2040-09-30
AI Technical Summary
Existing WiFi networks face significant wireless signal interference due to the increasing density of access points (APs) and users, leading to high packet error rates and limited concurrent transmission capacity without proper management of AP-STA connections.
A network management method that groups APs and STAs based on channel parameter information using centralized or distributed systems, employing clustering algorithms and spectral clustering techniques to reduce interference and improve network performance.
The method effectively reduces interference between user devices by optimizing AP-STA connections, enhancing network transmission performance through improved channel allocation and management.
Smart Images

Figure CN114339778B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of communication technology, and in particular to a network management method and related devices. Background Art
[0002] With the popularity of wireless fidelity (WiFi) technology and the growing demand for its applications, future WiFi networks will need to deploy a large number of wireless access points (APs) to serve a large number of access users. The rapidly increasing density of APs and users has made the problem of wireless signal interference between devices more and more prominent. Without proper management measures, signal interference will lead to a high packet error rate and prevent the interfered device from accessing the channel, thus limiting the network's concurrent transmission capacity.
[0003] Currently, APs use the least congested channel scan (LCCS) technology to periodically scan orthogonal channels, determine the connection status of each channel, and configure the AP to the channel with the least number of AP connections.
[0004] From the above solution, we can see that in LCCS technology, AP only allocates channels based on AP measurement information, or in other words, when a fixed AP is connected to a station (STA), the AP adjusts the channel accessed by the AP to reduce interference. However, there is no corresponding planning and management for the connection between AP and STA, resulting in large signal interference between users served by each AP, resulting in low network transmission performance. Summary of the invention
[0005] The embodiments of the present application provide a network management method and related devices for implementing management and planning of STAs connected to an AP, which can reduce interference between user devices and thus improve network transmission performance.
[0006] A first aspect of an embodiment of the present application provides a network management method, the network management method comprising:
[0007] The first communication device obtains first channel parameter information, which includes channel parameter information between a first AP in a first network and a STA within a signal coverage range of the first AP; then, the first communication device groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups, wherein the first STA includes part or all of the STAs within the signal coverage range of the first AP; the first communication device sends a grouping result of the multiple groups, which is used to indicate the AP and STA included in each of the multiple groups.
[0008] In this embodiment, the first communication device groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups, and sends the grouping results of the multiple groups, so that the AP or STA receives the grouping result and can adjust the connection in combination with the group to which it belongs, thereby realizing the management and planning of the STA connected to the AP. Since the multiple groups are divided by the first communication device according to the first channel parameter information, the management and planning of the STA connected to the AP through the solution of the embodiment of the present application can reduce interference between user devices, thereby improving the performance of network transmission.
[0009] In one possible implementation, the first communication device is a server, the first AP includes all APs of the first network, and the first STA includes all STAs of the first network; the first communication device obtains first channel parameter information, including: the server receives the first channel parameter information sent by a control device, the first channel parameter information includes channel parameter information between each AP of all APs of the first network and the STA within the signal coverage range of each AP, and the first network is a network controlled by the control device; the first communication device sends the grouping results of the multiple groups, including: the server sends the grouping results of the multiple groups to the control device.
[0010] In this possible implementation, a centralized network management system is used, and the centralized network management system includes a control device and a server. The server in the centralized network management system implements grouping of all APs and all STAs in the first network, and sends the grouping results of the multiple groups to the control device, and then the control device manages and plans the STAs connected to the APs in the first network.
[0011] In another possible implementation, the first channel parameter information includes a channel matrix W, where the channel matrix W is a matrix of K rows by L columns, K is the number of STAs included in the first network, L is the number of APs included in the first network, and the channel matrix w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, and j is an integer greater than or equal to 1 and less than or equal to L; the first communication device groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups, including: the server calculates a first intermediate matrix D according to the channel matrix W, the first intermediate matrix D 1 is a K-row by K-column matrix, D 1 The elements a on the main diagonal of iis the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All the off-diagonal elements of D are zero; 2 is a matrix of L rows by L columns, D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix of K rows by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K The elements in are all zero; the server determines a second intermediate matrix according to the channel matrix W and the first intermediate matrix D, each row vector in the second intermediate matrix corresponds to a network node, and the network node is a STA or AP in the first network; the server clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result, and the first clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group in the multiple groups; the server determines the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters.
[0012] In this possible implementation, the specific content of the first channel parameter information and the grouping process of the server grouping all APs and all STAs of the first network are provided. Specifically, the network nodes of the first network are divided and grouped by clustering to achieve grouping of all APs and all STAs of the first network, thereby improving the feasibility of the solution.
[0013] In another possible implementation, the channel matrix W is an edge weight matrix W of a weighted bipartite graph, the weighted bipartite graph includes first type vertices and second type vertices, the first type vertices represent STAs in the first network, the second type vertices represent APs in the first network, the elements of the edge weight matrix W are edge weights of the weighted bipartite graph, the edge weights of the weighted bipartite graph are channel gains between STAs and APs in the first network, and the first intermediate matrix is the degree matrix of the weighted bipartite graph.
[0014] In this possible implementation, the first channel parameter information is represented by an edge weight matrix of a weighted bipartite graph, so that the server can group all APs and all STAs of the first network by combining the graph method and the spectral clustering method. In this way, the signal interference between sub-networks formed by each group in the multiple groups obtained is smaller, and the effective strength within the sub-network is larger.
[0015] In another possible implementation, the server determines the second intermediate matrix according to the channel matrix W and the first intermediate matrix D, including: the server determines the third intermediate matrix R according to the channel matrix W and the first intermediate matrix D, W T is the transpose of W; the server performs singular value decomposition on the third intermediate matrix R to obtain the left singular matrix of the third intermediate matrix R and the right singular matrix of the third intermediate matrix R; wherein the third intermediate matrix R=UΣV T , the Σ is a diagonal matrix, the V T is the transpose of the right singular matrix V; the main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the left singular matrix U, and the main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the right singular matrix V, r is an integer greater than or equal to 1 and less than or equal to min(K, L), and min(K, L) refers to the minimum value of K and L; the server selects the corresponding M column vectors from the left singular matrix U in descending order of the singular values in the diagonal matrix Σ to obtain the fourth intermediate matrix U M , and select corresponding M column vectors from the right singular matrix V in descending order of the singular values in the diagonal matrix Σ to obtain the fifth intermediate matrix V M , M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), log 2 (K+L) refers to finding the logarithm of K+L with base 2; the server calculates the second intermediate matrix Z according to the fourth intermediate matrix, the fifth intermediate matrix and the first intermediate matrix,
[0016] In this possible implementation, the specific process of the server determining the second intermediate matrix Z is shown. The M column vectors selected by the server from the left singular matrix U and the M column vectors selected from the right singular matrix V according to the size of the singular values contain more characteristic information characterizing the weighted bipartite graph. The server clusters the second intermediate matrix Z by spectral clustering, thereby achieving a better grouping effect of the server on the STAs and APs in the first network. For example, the sum of the effective signal strengths in the subnetwork formed by the APs and STAs in each group is large, and the signal interference strength between the network nodes (APs or STAs) in the subnetwork and those outside the subnetwork is low.
[0017] In another possible implementation, the first communication device clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result, including: the server clusters the row vectors of the second intermediate matrix according to the clustering algorithm The K+L row vectors included are clustered to obtain row vectors of multiple clusters, where one cluster in the multiple clusters corresponds to one group in the multiple groups; the server determines the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters, including: the server divides the network nodes corresponding to the row vectors of the same cluster in the row vectors of the multiple clusters into the same group.
[0018] In this possible implementation, the server clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain the row vectors included in each cluster, and then determines the network nodes included in each group in combination with the network nodes corresponding to each row vector, thereby realizing the grouping of all APs and all STAs of the first network.
[0019] In another possible implementation, the first communication device is the first AP, and the first STA includes the STA that has established a connection with the first AP within the signal coverage of the first AP; the method also includes: the first AP adjusts the STA connected to the first AP according to the group to which the first AP belongs.
[0020] In this possible implementation, a distributed network management system is used. The first AP in the distributed network management system groups the first AP and the STA connected to the first AP. The first AP then adjusts the STA connected to the first AP to implement connection management and planning for the STA connected to the first AP, reduce interference between user devices, and thus improve network transmission performance.
[0021] In another possible implementation, the first AP is the pth AP in the first network, and the first channel parameter information includes the pth column vector of the channel matrix W. w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, j is an integer greater than or equal to 1 and less than or equal to L, p is an integer less than or equal to L, L is the number of APs included in the first network, and K is the number of STAs included in the first network; the first communication device groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups, including: the first AP calculates the p+Kth main diagonal element b of the first intermediate matrix D according to the pth column vector of the channel matrix W p ;
[0022] Among them, the first intermediate matrix The D 1 is a K-row by K-column matrix. D 1 The elements a on the main diagonal of iis the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All the off-diagonal elements of D are zero; 2 is a matrix of L rows by L columns, D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix multiplied by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero;
[0023] The first AP obtains the first K main diagonal elements of the first intermediate matrix D;
[0024] The first AP is based on the pth column vector of the channel matrix W, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, each row vector of the sixth intermediate matrix Q corresponds to a network node, which is the AP or STA in the first network, the i-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the i-th STA, and the K+j-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the j-th AP; the first AP clusters the target row vector according to the clustering algorithm to obtain a second clustering result; wherein the target row vector includes the K+pth row vector corresponding to the first AP and the row vector corresponding to the first STA in the first K row vectors of the sixth intermediate matrix Q, the second clustering result includes row vectors of multiple clusters, one cluster in the multiple clusters corresponds to one group of the multiple groups; the first AP determines the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters.
[0025] In this implementation, a grouping process in which the first AP groups the first AP and STAs connected to the first AP in a distributed network management system is shown, which provides a specific implementation method for the implementation of the solution and improves the feasibility and diversity of the solution.
[0026] In another possible implementation, the first AP calculates the channel matrix W based on the p-th row vector and the p+K-th main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, including:
[0027] The first AP is based on the pth column vector of the channel matrix W, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K main diagonal elements in the first intermediate matrix D determine the pth row vector of the third intermediate matrix R The second intermediate matrix The W T is the transpose of W;
[0028] The first AP is based on the stochastic gradient descent algorithm and the pth row vector of the third intermediate matrix R. Determine the first matrix X M p row vectors and the second matrix Y M , the first matrix X M The similarity with the third matrix is greater than or equal to the first preset threshold, and the second matrix Y M The degree of similarity with the fourth matrix is greater than or equal to a second preset threshold;
[0029] The third matrix is a matrix obtained by selecting M column vectors from a left singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the left singular matrix has a corresponding singular value, and the M column vectors of the third matrix are M column vectors selected from the left singular matrix in descending order of singular value size;
[0030] The fourth matrix is a matrix obtained by selecting M column vectors from a right singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the right singular matrix has a corresponding singular value, and the M column vectors of the fourth matrix are M column vectors selected from the right singular matrix in descending order of singular value size;
[0031] The M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2 (K+L) refers to finding the logarithm of K+L with base 2, and min(K, L) refers to taking the minimum value of K and L;
[0032] The first AP is based on the p+Kth main diagonal element b of the first intermediate matrix D. p , the first K main diagonal elements in the first intermediate matrix D, the first matrix X M The p-th row vector and the second matrix Y M Calculate the sixth intermediate matrix The first K row vectors and the K+pth row vector of .
[0033] In this possible implementation method, a specific implementation process for the first AP to determine the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q is provided. Since the first K row vectors of the sixth intermediate matrix Q respectively correspond to the STAs of the first network and the K+pth row vector corresponds to the first AP, this facilitates the subsequent clustering of the row vectors corresponding to the first AP and the STAs connected to the first AP by the first AP, so as to realize the grouping of the first AP and the STAs connected to the first AP.
[0034] In another possible implementation, the first AP clusters the target row vector according to a clustering algorithm to obtain a second clustering result, including: the first AP generates a first clustering center according to a first random seed, the first clustering center includes clustering centers corresponding to the multiple clusters, and each cluster corresponds to a cluster center; the first AP determines the cluster to which each row vector in the target row vector belongs according to the first clustering center, and obtains the row vector included in each cluster in the multiple clusters; the first AP determines the network nodes included in the multiple groups according to the row vectors of the multiple clusters, including: when a first preset condition is met, the first AP divides the network nodes corresponding to the row vector of the same cluster in the multiple clusters into the same group.
[0035] In this possible implementation, a clustering process of the target row vector by the first AP according to a clustering algorithm is provided, thereby improving the feasibility of the solution.
[0036] In another possible implementation, when the first preset condition is not met, the method further includes:
[0037] The first AP calculates the first average vector corresponding to the row vector of each cluster in the multiple clusters according to the gossip protocol; the first AP uses the first average vector of the row vector of each cluster in the multiple clusters as the cluster centers corresponding to the multiple clusters respectively to obtain second cluster centers; the first AP determines the cluster to which each row vector in the target row vector belongs according to the second cluster centers; when the second preset condition is met, the first AP divides the network nodes corresponding to the row vector of the same cluster in the multiple clusters into the same group.
[0038] In this possible implementation, when the first preset condition is not met, the first AP clusters the target row vector again according to the clustering algorithm to group the first AP and the STAs connected to the first AP.
[0039] In another possible implementation, the first preset condition includes:
[0040] The update times of the cluster centers of the clusters corresponding to the multiple groups are greater than or equal to a third preset threshold.
[0041] In this possible implementation, the first preset condition may be set to the number of iterations of the above clustering, that is, the number of updates of the cluster centers corresponding to the plurality of clusters.
[0042] In another possible implementation, the first preset condition includes:
[0043] The first convergence accuracy is less than or equal to the preset convergence accuracy;
[0044] The first convergence accuracy is a ratio of an absolute value of the first difference to the first global error, and the first difference is a difference between the first global error and a preset initialization error;
[0045] The first global error is the sum of first local errors respectively calculated by all APs in the first network;
[0046] The first local error of the first AP includes a sum of errors corresponding to each of the multiple clusters determined by the first AP;
[0047] The error corresponding to each cluster is the sum of errors between row vectors included in each cluster and a first average vector corresponding to each cluster. The first average vector corresponding to each cluster is calculated by the first AP according to a gossip protocol.
[0048] In this possible implementation, another first preset condition is provided. Since the first AP does not have global interference information of the first network, the first AP can obtain the first local error of other APs through the gossip protocol, and determine the first global error through the first local errors of all APs, and determine the first convergence accuracy according to the first global error and the preset initialization error, that is, the first convergence accuracy is equal to the absolute value of the difference between the first global error and the preset initialization error. The first AP then compares the first convergence accuracy with the preset convergence accuracy. If it is less than or equal to the preset convergence accuracy, the first AP can determine that the clustering iteration of the target row vector has reached a convergence degree, so that the first AP divides the multiple groups according to the row vectors of the multiple clusters, so that the signal interference between the sub-networks corresponding to each group in the multiple groups is small, and the effective strength within the sub-network is large.
[0049] In another possible implementation, the first AP adjusts the STA connected to the first AP according to the group to which the first AP belongs, including: the first AP sends information about the group to which the first AP belongs, and the information about the group to which the first AP belongs is used by the STA in the first network to determine whether to access the first AP.
[0050] In this possible implementation, a specific adjustment method is provided for the first AP to adjust the STA connected to the first AP according to the group to which the first AP belongs. That is, the STA determines whether to connect to the first AP according to the received information of the group to which the first AP belongs, thereby realizing connection management and planning of the STA connected to the first AP.
[0051] In another possible implementation, the information of the group to which the first AP belongs includes the first group number of the group to which the first AP belongs; the first AP sends the information of the group to which the first AP belongs, including: the first AP sends a first beacon frame, and the first beacon frame carries the first group number.
[0052] In this possible implementation, a method is provided in which the first AP carries the group number of the group to which the first AP belongs through the beacon frame of the first AP, so that the STA of the first network determines whether to connect to the first AP through the beacon frame.
[0053] In another possible implementation, the first AP adjusts the STA connected to the first AP according to the group to which the first AP belongs, including:
[0054] The first AP determines whether the second STA that has established a connection with the first AP belongs to the same group;
[0055] If yes, the first AP keeps providing services for the second STA;
[0056] If not, the first AP determines the group to which the second STA belongs, and sends first indication information to the third AP;
[0057] The first indication information is used to indicate that the third AP provides a service for the second STA, and the third AP and the second STA belong to the same group.
[0058] In this possible implementation, another implementation is provided in which the first AP adjusts the STA connected to the first AP according to the group to which the first AP belongs. The first AP determines whether the second STA that has established a connection with the first AP belongs to the same group. If not, the first AP can send indication information to a third AP that belongs to the same group as the second STA, and the third AP provides services for the second STA, thereby realizing connection management and planning of the STA connected to the first AP. Moreover, this implementation is transparent to the STA, and the STA is not aware of it, so that improvements to the STA can be avoided.
[0059] In another possible implementation, the method also includes: the first AP receives first group information sent by the fourth AP, the first group information includes the group information to which the fourth AP belongs and the identifier of the fourth AP, and the fourth AP includes APs in the first network other than the first AP; the first AP determines, based on the first group information, that the fifth AP belongs to the same group as the first AP, and the fifth AP is a partial AP in the fourth AP; the first AP determines whether the priority of the first AP is higher than the priority of the fifth AP based on the identifier of the fifth AP, the identifier of the first AP and a preset AP priority rule; if the priority of the first AP is higher than the priority of the fifth AP, the first AP selects the first channel by scanning channels, and adjusts the channel accessed by the first AP to the first channel; the first AP sends first channel allocation information to the fifth AP, and the first channel allocation information is used to instruct the fifth AP to access the first channel.
[0060] In this possible implementation, a process is provided in which a first AP allocates information in the same group in a distributed network management system, and notifies multiple APs in the same group of channel allocation information, thereby achieving multi-AP joint transmission to improve network throughput and network transmission performance.
[0061] In another possible implementation, if the priority of the first AP is lower than the priority of the fifth AP, the method also includes: the first AP receives second channel allocation information sent by the sixth AP, the second channel allocation information is information of the second channel selected by the sixth AP, and the sixth AP is the AP with the highest priority among the fifth APs; the first AP adjusts the channel accessed by the first AP to the second channel according to the second channel allocation information.
[0062] In this possible implementation, a process is provided in which other APs in the same group select channels and notify the first AP of the channel allocation information in a distributed network management system, so that the first AP can adjust the access channel, thereby realizing multi-AP joint transmission to improve network throughput and network transmission performance.
[0063] A second aspect of an embodiment of the present application provides a network management method, the network management method comprising:
[0064] The control device sends first channel parameter information to the server, and the first channel parameter information includes channel parameter information between each AP of all APs in the first network and the STA within the signal coverage range of each AP, and the first network is a network controlled by the control device; the control device receives grouping results of multiple groups from the server, and the multiple groups include all STAs and all APs in the first network; the control device determines the APs and STAs included in each of the multiple groups according to the grouping results; the control device sends second grouping information, and the second grouping information includes information about the groups to which the APs of the first network respectively belong and information about the groups to which the STAs of the first network respectively belong.
[0065] In this embodiment, in a centralized network management system, the control device collects global interference information in the first network and reports it to the server, then receives the grouping results of multiple groups sent by the server, and then the control device sends the grouping information of multiple groups, so that the AP of the first network adjusts the connected STA according to the grouping information, thereby realizing the management and planning of the STA connected to the AP. Since the multiple groups are divided by the server according to the first channel parameter information, the management and planning of the STA connected to the AP through the solution of the embodiment of the present application can reduce interference between user devices, thereby improving the performance of network transmission.
[0066] In another possible implementation, the method also includes: the control device allocates the same channel to APs of the same group among the multiple groups; the control device sends second group information, including: the control device sends the second group information, and the second group information includes channel allocation information corresponding to the multiple groups respectively.
[0067] In this possible implementation, the control device may further allocate the same channel to APs in the same group so that the APs in the same group can access the same channel, thereby achieving multi-AP joint transmission and improving network throughput and network transmission performance.
[0068] In another possible implementation, the control device allocates the same channel to APs in the same group among the multiple groups, including: the control device allocates a corresponding channel to the APs in each group among the multiple groups according to the first channel parameter information, wherein the control device allocates the same channel to APs in the same group.
[0069] In this possible implementation, the control device may allocate channels to APs in the same group in combination with the first channel parameter information, so that the allocated channels conform to the current channel status, thereby improving network transmission performance.
[0070] A third aspect of an embodiment of the present application provides a network management method, the network management method comprising:
[0071] The second AP sends second channel parameter information to the control device, and the second channel parameter information includes channel parameter information between the second AP and the STA within the signal coverage range of the second AP; the second AP receives second grouping information sent by the control device, and the second grouping information includes information about the groups to which the APs of the first network respectively belong and information about the groups to which the STAs of the first network respectively belong; the second AP determines the group to which the second AP belongs based on the second grouping information; the second AP adjusts the STA connected to the second AP based on the group to which the second AP belongs.
[0072] In this embodiment, the second AP obtains the second channel parameter information and sends it to the control device; then, the control device sends the second grouping information to the second AP, and the second AP adjusts the STA connected to the second AP according to the second grouping information, thereby realizing the connection management and planning of the STA connected to the second AP by the second AP. In addition, since the multiple groups are divided by the server according to the first channel parameter information, the management and planning of the STA connected to the AP through the solution of the embodiment of the present application can reduce the interference between user devices, thereby improving the performance of network transmission.
[0073] In a possible implementation, the second grouping information further includes third channel allocation information, and the third channel allocation information includes information about a third channel allocated by the control device to the second AP; and the method further includes:
[0074] The second AP adjusts the channel accessed by the second AP to the third channel according to the third channel allocation information.
[0075] In this possible implementation, the second AP can adjust the channel accessed by the second AP according to the information of the third channel allocated by the control device, so that multiple APs in the same group can access the same channel to achieve multi-AP joint transmission and improve network throughput.
[0076] In another possible implementation, the second AP adjusts the STA connected to the second AP according to the group to which the second AP belongs, including: the second AP sends information about the group to which the second AP belongs, and the information about the group to which the second AP belongs is used to instruct the STA in the first network to determine whether to access the second AP, and the first network is a network controlled by the control device.
[0077] In this possible implementation manner, a second AP is provided to adjust management and planning of STAs connected to the second AP.
[0078] In another possible implementation manner, the information of the group to which the second AP belongs includes a second group number of the group to which the second AP belongs; and the second AP sends the information of the group to which the second AP belongs, including:
[0079] The second AP sends a second beacon frame, where the second beacon frame carries the second group number.
[0080] In this possible implementation, the second AP may carry the second group number via a beacon frame to notify STAs and other APs in the first network of the group to which the second AP belongs, so that STAs in the same group are connected to the second AP.
[0081] In another possible implementation, the second AP adjusts the STA connected to the second AP according to the group to which the second AP belongs, including: the second AP determines the STA included in the group to which the second AP belongs; the second AP determines whether the second AP and a third STA belong to the same group, and the third STA is a STA that has established a connection with the second AP; if so, the second AP continues to provide services for the third STA; if not, the second AP determines a seventh AP according to the group to which the third STA belongs, and sends third indication information to the seventh AP; wherein the third indication information is used to indicate that the seventh AP provides services for the third STA, and the seventh AP and the third STA belong to the same group.
[0082] In this possible implementation, another implementation is provided in which the second AP adjusts the STA connected to the second AP according to the group to which the second AP belongs. The second AP determines whether the third STA that has established a connection with the second AP belongs to the same group. If not, the second AP can send indication information to the seventh AP that belongs to the same group as the third STA, and the seventh AP provides services for the third STA, thereby realizing connection management and planning of the STA connected to the second AP. Moreover, this implementation is transparent to the STA, and the STA is not aware of it, so that improvements to the STA can be avoided.
[0083] A fourth aspect of the embodiments of the present application provides a network management method, the network management method comprising:
[0084] The fourth STA receives the information of the group to which the second AP belongs, which is sent by the second AP; then, the fourth STA determines that the second AP and the fourth STA belong to the same group according to the information of the group to which the second AP belongs; and the fourth STA adjusts the AP to which it is connected to to the second AP.
[0085] In this possible implementation, the fourth STA determines that the second AP and the fourth STA belong to the same group based on the information of the group to which the second AP belongs and the group to which the fourth STA belongs, and adjusts the AP connected to the fourth STA to the second AP, thereby implementing connection management and planning for STAs connected to the AP. In addition, since the multiple groups are divided by the server based on the first channel parameter information, the management and planning of STAs connected to the AP through the solution of the embodiment of the present application can reduce interference between user devices, thereby improving the performance of network transmission.
[0086] In a possible implementation manner, the information of the group to which the second AP belongs includes a second group number of the group to which the second AP belongs; and the fourth STA receives the information of the group to which the second AP belongs sent by the second AP, including:
[0087] The fourth STA receives a second beacon frame sent by the second AP, where the first beacon frame carries the second group number;
[0088] The fourth STA determines, according to the second group number, that the second AP and the fourth STA belong to the same group;
[0089] If the second group number is consistent with the group number of the group to which the fourth STA belongs, the fourth STA determines that the second AP and the fourth STA belong to the same group.
[0090] In this possible implementation, a method is provided for the fourth STA to determine the group to which the second AP belongs by receiving the second beacon frame of the second AP, and to determine whether to connect to the second AP according to the group to which the second AP belongs and the group to which the fourth STA belongs.
[0091] In another possible implementation, the method further includes:
[0092] The fourth STA receives information of the group to which the fourth STA belongs, which is sent by the eighth AP. The eighth AP is the AP to which the fourth STA was connected before the connection adjustment.
[0093] In this possible implementation, the fourth STA first obtains the information of the group to which the fourth STA belongs. Specifically, the eighth AP that belongs to the same group as the fourth STA may obtain the information of the group to which the fourth STA belongs.
[0094] A fifth aspect of an embodiment of the present application provides a network management system, the network management system comprising a server, a control device, a second AP and a fourth STA;
[0095] The second AP is used to send second channel parameter information to the control device;
[0096] The control device is used to receive second channel parameter information sent by the second AP; send first channel parameter information to the server, where the first channel parameter information includes channel parameter information between each AP of all APs in the first network and a STA within the signal coverage range of each AP;
[0097] The server is configured to receive first channel parameter information sent by a control device; group all APs of a first network and all STAs of the first network according to the first channel parameter information to obtain a plurality of groups; and send a grouping result of the plurality of groups to the control device, the grouping result being used to indicate the APs and STAs included in each of the plurality of groups;
[0098] The control device is used to receive the grouping results of the multiple groups; determine the AP and STA included in each group of the multiple groups according to the grouping results; send second grouping information, the second grouping information including information of the groups to which the APs of the first network respectively belong and information of the groups to which the STAs of the first network respectively belong;
[0099] The second AP is further used to determine the group to which the second AP belongs according to the second grouping information, and adjust the STA connected to the second AP according to the group to which the second AP belongs.
[0100] In a possible implementation, the first channel parameter information includes a channel matrix W, where the channel matrix W is a matrix of K rows by L columns, K is the number of STAs included in the first network, L is the number of APs included in the first network, and the channel matrix w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, j is an integer greater than or equal to 1 and less than or equal to L, and the server is specifically used for:
[0101] The first intermediate matrix D is calculated based on the channel matrix W. D 1 is a K-row by K-column matrix, D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All the off-diagonal elements of D are zero; 2 is a matrix of L rows by L columns, D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix of K rows by L columns, 0K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K The elements in are all zero; a second intermediate matrix is determined according to the channel matrix W and the first intermediate matrix D, each row vector in the second intermediate matrix corresponds to a network node, and the network node is a STA or AP in the first network; the row vectors of the second intermediate matrix are clustered according to a clustering algorithm to obtain a first clustering result, and the first clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group in the multiple groups; the network nodes respectively included in the multiple groups are determined according to the row vectors of the multiple clusters.
[0102] In another possible implementation, the channel matrix W is an edge weight matrix W of a weighted bipartite graph, the weighted bipartite graph includes first type vertices and second type vertices, the first type vertices represent STAs in the first network, the second type vertices represent APs in the first network, the elements of the edge weight matrix W are edge weights of the weighted bipartite graph, the edge weights of the weighted bipartite graph are channel gains between STAs and APs in the first network, and the first intermediate matrix is the degree matrix of the weighted bipartite graph.
[0103] In another possible implementation, the server is specifically used for:
[0104] The server determines a third intermediate matrix R according to the channel matrix W and the first intermediate matrix D, W T is the transpose of W; performing singular value decomposition on the third intermediate matrix R to obtain the left singular matrix of the third intermediate matrix R and the right singular matrix of the third intermediate matrix R; wherein the third intermediate matrix R = UΣV T , the Σ is a diagonal matrix, the V T is the transpose of the right singular matrix V; the main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the left singular matrix U, and the main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the right singular matrix V, r is an integer greater than or equal to 1 and less than or equal to min(K, L), min(K, L) refers to the minimum value of K and L; select the corresponding M column vectors from the left singular matrix U in descending order of the singular values in the diagonal matrix Σ to obtain the fourth intermediate matrix U M , and select corresponding M column vectors from the right singular matrix V in descending order of the singular values in the diagonal matrix Σ to obtain the fifth intermediate matrix V M , M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), log 2(K+L) means taking the logarithm of K+L with base 2 as the base; the second intermediate matrix Z is calculated based on the fourth intermediate matrix, the fifth intermediate matrix and the first intermediate matrix,
[0105] In another possible implementation, the server is specifically used for:
[0106] According to the clustering algorithm, the second intermediate matrix Clustering the K+L row vectors included to obtain row vectors of multiple clusters, where one cluster in the multiple clusters corresponds to one group in the multiple groups;
[0107] The network nodes corresponding to the row vectors of the same cluster in the row vectors of multiple clusters are divided into the same group.
[0108] In another possible implementation, the control device is further used for:
[0109] Allocating a same channel to APs in the same group among the multiple groups;
[0110] The control device is specifically used for:
[0111] Sending the second grouping information to the second AP, where the second grouping information includes third channel allocation information, where the third channel allocation information includes information about a third channel allocated by the control device to the second AP;
[0112] The second AP is also used for:
[0113] The channel accessed by the second AP is adjusted to the third channel according to the third channel allocation information included in the second grouping information.
[0114] In another possible implementation, the control device is specifically used for:
[0115] A corresponding channel is allocated to the APs of each group in the multiple groups according to the first channel parameter information, wherein the control device allocates the same channel to the APs of the same group.
[0116] In another possible implementation, the control device is specifically used for:
[0117] A corresponding channel is allocated to the APs of each group in the multiple groups according to the first channel parameter information, wherein the control device allocates the same channel to the APs of the same group.
[0118] In another possible implementation manner, the second AP is specifically used for:
[0119] The information of the group to which the second AP belongs is sent, where the information of the group to which the second AP belongs is used to instruct the STA in the first network to determine whether to access the second AP, and the first network is a network controlled by the control device.
[0120] The fourth STA is used to receive information about the group to which the second AP belongs, sent by the second AP; determine that the second AP and the fourth STA belong to the same group according to the information about the group to which the second AP belongs; and adjust the AP to which it is connected to to the second AP.
[0121] In another possible implementation manner, the second AP is specifically used for:
[0122] The second AP sends a second beacon frame, where the first beacon frame carries a second group number of the group to which the second AP belongs;
[0123] The fourth STA is specifically used for:
[0124] Determine, according to the second group number, that the second AP and the fourth STA belong to the same group;
[0125] If the second group number is consistent with the group number of the group to which the fourth STA belongs, it is determined that the second AP and the fourth STA belong to the same group.
[0126] In another possible implementation, the fourth STA is further used for:
[0127] The information of the group to which the fourth STA belongs is received from an eighth AP, where the eighth AP is the AP to which the fourth STA was connected before the connection adjustment.
[0128] In another possible implementation manner, the second AP is specifically used for:
[0129] Determine the STAs included in the group to which the second AP belongs;
[0130] Determine whether the second AP and the third STA belong to the same group, the third STA being a STA that has established a connection with the second AP;
[0131] If yes, keep providing service to the third STA;
[0132] If not, determine the seventh AP according to the group to which the third STA belongs, and send third indication information to the seventh AP; wherein the third indication information is used to indicate that the seventh AP provides services for the third STA, and the seventh AP and the third STA belong to the same group.
[0133] A sixth aspect of an embodiment of the present application provides a network management system, the network management system comprising a first AP and a first STA;
[0134] The first AP is used to obtain first channel parameter information, where the first channel parameter information includes channel parameter information between a first AP in a first network and an STA within a signal coverage range of the first AP; grouping the first AP and the first STA according to the first channel parameter information to obtain multiple groups, where the first STA includes STAs within the signal coverage range of the first AP and that have established a connection with the first AP; sending grouping results of the multiple groups, where the grouping results are used to indicate the AP and STA included in each of the multiple groups; and adjusting the STA connected to the first AP according to the group to which the first AP belongs.
[0135] In a possible implementation, the first AP is the pth AP in the first network, and the first channel parameter information includes the pth column vector of the channel matrix W. w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, j is an integer greater than or equal to 1 and less than or equal to L, p is an integer less than or equal to L, L is the number of APs included in the first network, and K is the number of STAs included in the first network; the first AP is specifically used for:
[0136] The p+Kth main diagonal element b of the first intermediate matrix D is calculated based on the pth column vector of the channel matrix W. p ;
[0137] Among them, the first intermediate matrix The D 1 is a K-row by K-column matrix. D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All the off-diagonal elements of D are zero; 2 is a matrix of L rows by L columns, D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix multiplied by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero;
[0138] Obtain the first K main diagonal elements of the first intermediate matrix D;
[0139] According to the p-th column vector of the channel matrix W, the p+K-th main diagonal element b of the first intermediate matrix D p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, each row vector of the sixth intermediate matrix Q corresponds to a network node, and the network node is an AP or STA in the first network. The i-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the i-th STA, and the K+j-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the j-th AP;
[0140] Clustering the target row vector according to the clustering algorithm to obtain a second clustering result; wherein the target row vector includes the K+pth row vector corresponding to the first AP and the row vector corresponding to the first STA in the first K row vectors of the sixth intermediate matrix Q, and the second clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group of the multiple groups;
[0141] The network nodes respectively included in the plurality of groups are determined according to the row vectors of the plurality of clusters.
[0142] In another possible implementation manner, the first AP is specifically used for:
[0143] According to the p-th column vector of the channel matrix W, the p+K-th main diagonal element b of the first intermediate matrix D p The first K main diagonal elements in the first intermediate matrix D and the first K main diagonal elements in the first intermediate matrix D determine the pth row vector of the third intermediate matrix R The second intermediate matrix The W T is the transpose of W;
[0144] According to the stochastic gradient descent algorithm and the pth row vector of the third intermediate matrix R Determine the first matrix X M p row vectors and the second matrix Y M , the first matrix X M The similarity with the third matrix is greater than or equal to the first preset threshold, and the second matrix Y M The degree of similarity with the fourth matrix is greater than or equal to a second preset threshold;
[0145] The third matrix is a matrix obtained by selecting M column vectors from a left singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the left singular matrix has a corresponding singular value, and the M column vectors of the third matrix are M column vectors selected from the left singular matrix in descending order of singular value size;
[0146] The fourth matrix is a matrix obtained by selecting M column vectors from a right singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the right singular matrix has a corresponding singular value, and the M column vectors of the fourth matrix are M column vectors selected from the right singular matrix in descending order of singular value size;
[0147] The M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2 (K+L) refers to finding the logarithm of K+L with base 2, and min(K, L) refers to taking the minimum value of K and L;
[0148] According to the p+Kth main diagonal element b of the first intermediate matrix D p , the first K main diagonal elements in the first intermediate matrix D, the first matrix X M The p-th row vector and the second matrix Y M Calculate the sixth intermediate matrix The first K row vectors and the K+pth row vector of .
[0149] In another possible implementation manner, the first AP is specifically used for:
[0150] Generate a first cluster center according to the first random seed, the first cluster center including cluster centers corresponding to the multiple clusters respectively, and each cluster corresponds to a cluster center;
[0151] Determine, according to the first cluster center, the cluster to which each row vector in the target row vector belongs, and obtain the row vector included in each cluster in the multiple clusters;
[0152] Determining the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters includes: when a first preset condition is met, the first AP divides the network nodes corresponding to the row vector of the same cluster in the multiple clusters into the same group.
[0153] In another possible implementation manner, when the first preset condition is not met, the first AP is further configured to:
[0154] Calculate a first average vector corresponding to the row vector of each cluster in the plurality of clusters according to the gossip protocol;
[0155] Using the first average vector of the row vector of each cluster in the multiple clusters as the cluster centers corresponding to the multiple clusters, respectively, to obtain second cluster centers;
[0156] The cluster to which each row vector in the target row vector belongs is determined according to the second cluster center; when the second preset condition is met, the first AP divides the network nodes corresponding to the row vectors of the same cluster in the multiple clusters into the same group.
[0157] In another possible implementation, the first preset condition includes:
[0158] The update times of the cluster centers of the clusters corresponding to the multiple groups are greater than or equal to a third preset threshold.
[0159] In another possible implementation, the first preset condition includes:
[0160] The first convergence accuracy is less than or equal to the preset convergence accuracy;
[0161] The first convergence accuracy is a ratio of an absolute value of the first difference to the first global error, and the first difference is a difference between the first global error and a preset initialization error;
[0162] The first global error is the sum of first local errors respectively calculated by all APs in the first network;
[0163] The first local error of the first AP includes a sum of errors corresponding to each of the multiple clusters determined by the first AP;
[0164] The error corresponding to each cluster is the sum of errors between row vectors included in each cluster and the first average vector corresponding to each cluster. The first average vector corresponding to each cluster is calculated by the first AP according to the gossip protocol.
[0165] In another possible implementation manner, the first AP is specifically used for:
[0166] Sending information about the group to which the first AP belongs, where the information about the group to which the first AP belongs is used by STAs in the first network to determine whether to access the first AP;
[0167] The network management system also includes a fifth STA, which has no connection with the first AP;
[0168] Determining, according to information of the group to which the first AP belongs and the group to which the fifth STA belongs, that the first AP and the fifth STA belong to the same group;
[0169] The fifth STA adjusts the AP to which the fifth STA is connected to the first AP.
[0170] In another possible implementation manner, the information of the group to which the first AP belongs includes a first group number of the group to which the first AP belongs; and the first AP is specifically configured to:
[0171] Sending a first beacon frame, where the first beacon frame carries the first group number;
[0172] The fifth STA is specifically used for:
[0173] Determine a first group number of a group to which the first AP belongs according to the first beacon frame;
[0174] The fifth STA determines, according to the first group number and the group number to which the fifth STA belongs, that the first AP and the fifth STA belong to the same group.
[0175] In another possible implementation, the network management system further includes a second STA; and the first AP is specifically configured to:
[0176] Determine whether the second STA that has established a connection with the first AP belongs to the same group;
[0177] If yes, continue to provide service for the second STA;
[0178] If not, determining the group to which the second STA belongs, and sending first indication information to the third AP;
[0179] The first indication information is used to indicate that the third AP provides a service for the second STA, and the third AP and the second STA belong to the same group.
[0180] In another possible implementation, the network management system further includes a fifth AP; and the first AP is further used for:
[0181] receiving first grouping information sent by a fourth AP, the first grouping information including grouping information to which the fourth AP belongs and an identifier of the fourth AP, the fourth AP including APs other than the first AP in the first network;
[0182] Determining, according to the first grouping information, that the fifth AP and the first AP belong to the same group, and that the fifth AP is a partial AP in the fourth AP; the first AP determines, according to the identifier of the fifth AP, the identifier of the first AP, and a preset AP priority rule, whether the priority of the first AP is higher than the priority of the fifth AP;
[0183] If the priority of the first AP is higher than the priority of the fifth AP, selecting the first channel by scanning channels, and adjusting the channel accessed by the first AP to the first channel;
[0184] Sending first channel allocation information to the fifth AP, where the first channel allocation information is used to instruct the fifth AP to access the first channel;
[0185] The fifth AP is used to determine the first channel according to the first channel allocation information, and adjust the channel accessed by the fifth AP to the first channel.
[0186] In another possible implementation manner, the first AP is further used for:
[0187] If the priority of the first AP is lower than the priority of the fifth AP, receiving the second channel allocation information sent by the sixth AP, the second channel allocation information is information of the second channel selected by the sixth AP, and the sixth AP is the AP with the highest priority among the fifth APs;
[0188] The channel accessed by the first AP is adjusted to the second channel according to the second channel allocation information.
[0189] A seventh aspect of an embodiment of the present application provides a first communication device, the first communication device comprising:
[0190] A transceiver module, configured to obtain first channel parameter information, where the first channel parameter information includes channel parameter information between a first AP in a first network and a STA within a signal coverage range of the first AP;
[0191] A grouping module, configured to group the first AP and the first STA according to the first channel parameter information to obtain a plurality of groups, wherein the first STA includes some or all of the STAs within the signal coverage of the first AP;
[0192] The transceiver module is used to send the grouping result of the multiple groups, and the grouping result is used to indicate the AP and STA included in each group of the multiple groups.
[0193] In a possible implementation, the communication device is a server, the first AP includes all APs of the first network, and the first STA includes all STAs of the first network; and the transceiver module is specifically used for:
[0194] Receive the first channel parameter information sent by the control device, where the first channel parameter information includes channel parameter information between each AP of all APs in the first network and a STA within a signal coverage range of each AP, where the first network is a network controlled by the control device;
[0195] The transceiver module is specifically used for:
[0196] The grouping result of the plurality of groups is sent to the control device.
[0197] In another possible implementation, the first channel parameter information includes a channel matrix W, where the channel matrix W is a matrix of K rows by L columns, K is the number of STAs included in the first network, L is the number of APs included in the first network, and the channel matrix w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, and j is an integer greater than or equal to 1 and less than or equal to L;
[0198] The grouping module is specifically used for:
[0199] The first intermediate matrix D is calculated based on the channel matrix W. D 1 is a K-row by K-column matrix, D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All the off-diagonal elements of D are zero; 2 is a matrix of L rows by L columns, D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix of K rows by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero;
[0200] Determine a second intermediate matrix according to the channel matrix W and the first intermediate matrix D, wherein each row vector in the second intermediate matrix corresponds to a network node, and the network node is a STA or an AP in the first network;
[0201] Clustering the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result, wherein the first clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one of the multiple groups;
[0202] The network nodes respectively included in the plurality of groups are determined according to the row vectors of the plurality of clusters.
[0203] In another possible implementation, the channel matrix W is an edge weight matrix W of a weighted bipartite graph, the weighted bipartite graph includes first type vertices and second type vertices, the first type vertices represent STAs in the first network, the second type vertices represent APs in the first network, the elements of the edge weight matrix W are edge weights of the weighted bipartite graph, the edge weights of the weighted bipartite graph are channel gains between STAs and APs in the first network, and the first intermediate matrix is the degree matrix of the weighted bipartite graph.
[0204] In another possible implementation, the grouping module is specifically used for:
[0205] Determine a third intermediate matrix R according to the channel matrix W and the first intermediate matrix D, W T is the transpose of W;
[0206] The third intermediate matrix R is subjected to singular value decomposition to obtain a left singular matrix of the third intermediate matrix R and a right singular matrix of the third intermediate matrix R; wherein the third intermediate matrix R = UΣV T , the Σ is a diagonal matrix, the V T is the transpose of the right singular matrix V;
[0207] The main diagonal element corresponding to the r-th column vector of the diagonal matrix Σ is the singular value corresponding to the r-th column vector of the left singular matrix U, and the main diagonal element corresponding to the r-th column vector of the diagonal matrix Σ is the singular value corresponding to the r-th column vector of the right singular matrix V, where r is an integer greater than or equal to 1 and less than or equal to min(K, L), and min(K, L) refers to the minimum value of K and L;
[0208] According to the order of the singular values in the diagonal matrix Σ, the corresponding M column vectors are selected from the left singular matrix U from large to small to obtain the fourth intermediate matrix U M , and select corresponding M column vectors from the right singular matrix V in descending order of the singular values in the diagonal matrix Σ to obtain the fifth intermediate matrix V M , where M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2 (K+L) means taking the logarithm of K+L with base 2;
[0209] The second intermediate matrix Z is calculated based on the fourth intermediate matrix, the fifth intermediate matrix and the first intermediate matrix.
[0210] In another possible implementation, the grouping module is specifically used for:
[0211] According to the clustering algorithm, the second intermediate matrix Clustering the K+L row vectors included to obtain row vectors of multiple clusters, where one cluster in the multiple clusters corresponds to one group in the multiple groups;
[0212] The network nodes corresponding to the row vectors of the same cluster in the row vectors of multiple clusters are divided into the same group.
[0213] In another possible implementation, the communication device is the first AP, the first STA includes a STA that has established a connection with the first AP within a signal coverage range of the first AP; the first AP further includes a group management module, and the group management module is used to:
[0214] The STAs connected to the first AP are adjusted according to the group to which the first AP belongs.
[0215] In another possible implementation, the communication device is the pth AP in the first network, and the first channel parameter information includes the pth column vector of the channel matrix W. w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, j is an integer greater than or equal to 1 and less than or equal to L, p is an integer less than or equal to L, L is the number of APs included in the first network, and K is the number of STAs included in the first network;
[0216] The grouping module is specifically used for:
[0217] The p+Kth main diagonal element b of the first intermediate matrix D is calculated based on the pth column vector of the channel matrix W. p ;
[0218] Among them, the first intermediate matrix The D 1 is a K-row by K-column matrix. The D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, and the D 1 The off-diagonal elements of are all zero; 2 is a matrix of L rows by L columns, The D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix multiplied by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero;
[0219] Obtain the first K main diagonal elements of the first intermediate matrix D;
[0220] According to the p-th column vector of the channel matrix W, the p+K-th main diagonal element b of the first intermediate matrix D p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, each row vector of the sixth intermediate matrix Q corresponds to a network node, and the network node is an AP or STA in the first network. The i-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the i-th STA, and the K+j-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the j-th AP;
[0221] Clustering the target row vector according to the clustering algorithm to obtain a second clustering result;
[0222] The target row vector includes the K+pth row vector corresponding to the first AP and the row vector corresponding to the first STA among the first K row vectors of the sixth intermediate matrix Q, and the second clustering result includes row vectors of multiple clusters, one cluster in the multiple clusters corresponds to one group of the multiple groups;
[0223] The network nodes respectively included in the plurality of groups are determined according to the row vectors of the plurality of clusters.
[0224] In another possible implementation, the grouping module is specifically used for:
[0225] According to the p-th column vector of the channel matrix W, the p+K-th main diagonal element b of the first intermediate matrix D p The first K main diagonal elements in the first intermediate matrix D and the first K main diagonal elements in the first intermediate matrix D determine the pth row vector of the third intermediate matrix R The second intermediate matrix The W T Transpose W;
[0226] According to the stochastic gradient descent algorithm and the pth row vector of the third intermediate matrix R Determine the first matrix X M p row vectors and the second matrix Y M , the first matrix X M The similarity with the third matrix is greater than or equal to the first preset threshold, and the second matrix Y M The degree of similarity with the fourth matrix is greater than or equal to a second preset threshold;
[0227] The third matrix is a matrix obtained by selecting M column vectors from a left singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the left singular matrix has a corresponding singular value, and the M column vectors of the third matrix are M column vectors selected from the left singular matrix in descending order of singular value size;
[0228] The fourth matrix is a matrix obtained by selecting M column vectors from a right singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the right singular matrix has a corresponding singular value, and the M column vectors of the fourth matrix are M column vectors selected from the right singular matrix in descending order of singular value size;
[0229] The M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2(K+L) means finding the logarithm of K+L with base 2, and min(K, L) means taking the minimum value of K and L;
[0230] According to the p+Kth main diagonal element b of the first intermediate matrix D p , the first K main diagonal elements in the first intermediate matrix D, the first matrix X M The p-th row vector and the second matrix Y M Calculate the sixth intermediate matrix The first K row vectors and the K+pth row vector of .
[0231] In another possible implementation, the grouping module is specifically used for:
[0232] Generate a first cluster center according to the first random seed, the first cluster center including cluster centers corresponding to the multiple clusters respectively, and each cluster corresponds to a cluster center;
[0233] Determine, according to the first cluster center, the cluster to which each row vector in the target row vector belongs, and obtain the row vector included in each cluster in the multiple clusters;
[0234] When the first preset condition is met, the network nodes corresponding to the row vectors of the same cluster among the multiple clusters are divided into the same group.
[0235] In another possible implementation, when the first preset condition is not met, the grouping module is further configured to:
[0236] Calculate a first average vector corresponding to the row vector of each cluster in the plurality of clusters according to the gossip protocol;
[0237] Using the first average vector of the row vector of each cluster in the multiple clusters as the cluster centers corresponding to the multiple clusters, respectively, to obtain second cluster centers;
[0238] Determine, according to the second cluster center, the cluster to which each row vector in the target row vector belongs;
[0239] When the second preset condition is met, the network nodes corresponding to the row vectors of the same cluster among the multiple clusters are divided into the same group.
[0240] In another possible implementation, the first preset condition includes:
[0241] The update times of the cluster centers of the clusters corresponding to the multiple groups are greater than or equal to a third preset threshold.
[0242] In another possible implementation, the first preset condition includes:
[0243] The first convergence accuracy is less than or equal to the preset convergence accuracy;
[0244] The first convergence accuracy is a ratio of an absolute value of the first difference to the first global error, and the first difference is a difference between the first global error and a preset initialization error;
[0245] The first global error is the sum of first local errors respectively calculated by all APs in the first network;
[0246] The first local error of the first AP includes a sum of errors corresponding to each of the multiple clusters determined by the first AP;
[0247] The error corresponding to each cluster is the sum of errors between row vectors included in each cluster and a first average vector corresponding to each cluster. The first average vector corresponding to each cluster is calculated by the first AP according to a gossip protocol.
[0248] In another possible implementation, the group management module sends information about the group to which the first AP belongs through the transceiver module, and the information about the group to which the first AP belongs is used by STAs in the first network to determine whether to access the first AP.
[0249] In another possible implementation, the information of the group to which the first AP belongs includes a first group number of the group to which the first AP belongs; the group management module sends a first beacon frame through the transceiver module, and the first beacon frame carries the first group number.
[0250] In another possible implementation, the group management module is specifically used for:
[0251] Determine whether the second STA that has established a connection with the first AP belongs to the same group;
[0252] If yes, keep providing service to the second STA;
[0253] If not, determining the group to which the second STA belongs, and sending first indication information to the third AP;
[0254] The first indication information is used to indicate that the third AP provides a service for the second STA, and the third AP and the second STA belong to the same group.
[0255] In another possible implementation, the transceiver module is further used for:
[0256] receiving first grouping information sent by a fourth AP, the first grouping information including grouping information to which the fourth AP belongs and an identifier of the fourth AP, the fourth AP including APs other than the first AP in the first network;
[0257] The group management module is also used to:
[0258] Determine, according to the first grouping information, that the fifth AP and the first AP belong to the same group, and that the fifth AP is a partial AP in the fourth AP;
[0259] Determining whether the priority of the first AP is higher than the priority of the fifth AP according to the identifier of the fifth AP, the identifier of the first AP and a preset AP priority rule;
[0260] If the priority of the first AP is higher than the priority of the fifth AP, selecting the first channel by scanning channels, and adjusting the channel accessed by the first AP to the first channel;
[0261] The transceiver module is also used for:
[0262] First channel allocation information is sent to the fifth AP, where the first channel allocation information is used to instruct the fifth AP to access the first channel.
[0263] In another possible implementation, if the priority of the first AP is lower than the priority of the fifth AP, the transceiver module is further configured to:
[0264] receiving second channel allocation information sent by a sixth AP, where the second channel allocation information is information of a second channel selected by the sixth AP, and the sixth AP is an AP with the highest priority among the fifth APs;
[0265] The group management module is also used to:
[0266] The channel accessed by the first AP is adjusted to the second channel according to the second channel allocation information.
[0267] An eighth aspect of the embodiments of the present application provides a control device, the control device comprising:
[0268] The transceiver module is used to send first channel parameter information to the server, where the first channel parameter information includes channel parameter information between each AP of all APs in the first network and STAs within the signal coverage of each AP, where the first network is a network controlled by the control device; and receive grouping results of multiple groups from the server, where the multiple groups include all STAs and all APs in the first network;
[0269] A grouping management module, used to determine the AP and STA included in each of the multiple groups according to the grouping result;
[0270] The transceiver module is further used to send second group information, where the second group information includes information about groups to which the APs of the first network respectively belong and information about groups to which the STAs of the first network respectively belong.
[0271] In a possible implementation, the group management module is further used to:
[0272] Allocating a same channel to APs in the same group among the multiple groups;
[0273] The transceiver module is specifically used for:
[0274] The second grouping information is sent, where the second grouping information includes channel allocation information respectively corresponding to the plurality of groups.
[0275] In another possible implementation, the group management module is further used to:
[0276] A corresponding channel is allocated to the APs of each group in the multiple groups according to the first channel parameter information, wherein the control device allocates the same channel to the APs of the same group.
[0277] A ninth aspect of an embodiment of the present application provides a second AP, the second AP including:
[0278] A transceiver module, configured to send second channel parameter information to a control device, the second channel parameter information including channel parameter information between the second AP and STAs within the signal coverage range of the second AP; receive second group information sent by the control device, the second group information including information about groups to which the APs of the first network respectively belong and information about groups to which the STAs of the first network respectively belong;
[0279] The group management module is used to determine the group to which the second AP belongs according to the second group information; and adjust the STA connected to the second AP according to the group to which the second AP belongs.
[0280] In a possible implementation, the second grouping information further includes third channel allocation information, where the third channel allocation information includes information about a third channel allocated by the control device to the second AP; and the grouping management module is further configured to:
[0281] The channel accessed by the second AP is adjusted to the third channel according to the third channel allocation information.
[0282] In another possible implementation, the group management module sends information about the group to which the second AP belongs through the transceiver module. The information about the group to which the second AP belongs is used to instruct the STA in the first network to determine whether to access the second AP. The first network is a network controlled by the control device.
[0283] In another possible implementation, the information of the group to which the second AP belongs includes a second group number of the group to which the second AP belongs;
[0284] The group management module sends a second beacon frame through the transceiver module, and the second beacon frame carries the second group number.
[0285] In another possible implementation, the group management module is specifically used for:
[0286] Determine the STAs included in the group to which the second AP belongs;
[0287] Determine whether the second AP and the third STA belong to the same group, the third STA being a STA that has established a connection with the second AP;
[0288] If yes, keep providing service to the third STA;
[0289] If not, determining a seventh AP according to the group to which the third STA belongs, and sending third indication information to the seventh AP;
[0290] The third indication information is used to indicate that the seventh AP provides services for the third STA, and the seventh AP and the third STA belong to the same group.
[0291] In a tenth aspect, an embodiment of the present application provides a fourth STA, the fourth STA including:
[0292] A transceiver module, used for receiving information of a group to which the second AP belongs, sent by the second AP;
[0293] The group management module is used to determine that the second AP and the fourth STA belong to a group according to the information of the group to which the second AP belongs; and adjust the AP to which it is connected to be the second AP.
[0294] In a possible implementation manner, the information of the group to which the second AP belongs includes a second group number of the group to which the second AP belongs; and the transceiver module is specifically configured to:
[0295] receiving a second beacon frame sent by the second AP, wherein the first beacon frame carries the second group number;
[0296] The group management module is specifically used for:
[0297] If the second group number is consistent with the group number of the group to which the fourth STA belongs, it is determined that the second AP and the fourth STA belong to the same group.
[0298] In another possible implementation, the transceiver module is further used for:
[0299] The information of the group to which the fourth STA belongs is received from an eighth AP, where the eighth AP is the AP to which the fourth STA was connected before the connection adjustment.
[0300] In an eleventh aspect of the present application, a communication device is provided, the communication device comprising: a processor and a memory. The memory stores a computer program or a computer instruction, and the processor is used to call and run the computer program or the computer instruction stored in the memory, so that the processor implements any one of the implementation methods in the first aspect.
[0301] Optionally, the communication device further includes a transceiver, and the processor is used to control the transceiver to send and receive signals.
[0302] The twelfth aspect of the present application provides a control device, the control device comprising: a processor and a memory. The memory stores a computer program or a computer instruction, and the processor is used to call and run the computer program or the computer instruction stored in the memory, so that the processor implements any one of the implementation methods in the second aspect.
[0303] Optionally, the control device further includes a transceiver, and the processor is used to control the transceiver to send and receive signals.
[0304] In a thirteenth aspect of the present application, a second AP is provided, the second AP comprising: a processor and a memory. The memory stores a computer program or a computer instruction, and the processor is used to call and run the computer program or the computer instruction stored in the memory, so that the processor implements any one of the implementation methods in the third aspect.
[0305] Optionally, the second AP further includes a transceiver, and the processor is used to control the transceiver to send and receive signals.
[0306] In a fourteenth aspect of the present application, a fourth STA is provided, the fourth STA comprising: a processor and a memory. The memory stores a computer program or a computer instruction, and the processor is used to call and run the computer program or the computer instruction stored in the memory, so that the processor implements any one of the implementation methods in the fourth aspect.
[0307] Optionally, the fourth STA also includes a transceiver, and the processor is used to control the transceiver to send and receive signals.
[0308] A fifteenth aspect of the present application provides a computer program product comprising instructions, characterized in that when the computer program product is run on a computer, the computer is caused to execute an implementation method as described in any one of the first, second, third and fourth aspects.
[0309] The sixteenth aspect of the present application provides a computer-readable storage medium, comprising computer instructions, which, when executed on a computer, enables the computer to execute any one of the implementation methods of the first, second, third and fourth aspects.
[0310] In the seventeenth aspect of the present application, a chip device is provided, comprising a processor for calling a computer program or computer instruction in the memory so that the processor executes any one of the implementation methods of any one of the above-mentioned first aspect, second aspect, third aspect and fourth aspect.
[0311] Optionally, the chip device further includes a memory, which is used to store computer programs or computer instructions, etc. The chip device is composed of a chip, and may also include a chip and other discrete devices.
[0312] Optionally, the processor is coupled to the memory via an interface.
[0313] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:
[0314] Through the above technical solution, it can be known that the first communication device obtains the first channel parameter information, and the first channel parameter information includes the channel parameter information between the first AP in the first network and the STA within the signal coverage of the first AP; the first communication device groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups, and the first STA includes some or all of the STAs within the signal coverage of the first AP; then, the first communication device sends the grouping results of the multiple groups, and the grouping results are used to indicate the AP and STA included in each of the multiple groups. In this way, the AP or STA receives the grouping result and can adjust the connection in combination with the group to which it belongs, thereby realizing the management and planning of the STA connected to the AP. Since the multiple groups are divided by the first communication device according to the first channel parameter information, the management and planning of the STA connected to the AP through the solution of the embodiment of the present application can reduce interference between user devices, thereby improving the performance of network transmission. BRIEF DESCRIPTION OF THE DRAWINGS
[0315] Figure 1A A schematic diagram of a network management system according to an embodiment of the present application;
[0316] Figure 1B A schematic diagram of the structure of the server in the embodiment of the present application;
[0317] Figure 1C A schematic diagram of the structure of the control device of the embodiment of the present application;
[0318] Figure 1D A schematic diagram of the structure of the AP in the embodiment of the present application;
[0319] Figure 1E This is a schematic diagram of the structure of the STA in the embodiment of the present application;
[0320] Figure 2AThis is another schematic diagram of an embodiment of the network management system of the present application;
[0321] Figure 2B This is a schematic diagram of the structure of the first AP in the embodiment of the present application;
[0322] Figure 2C This is a structural diagram of the second STA in an embodiment of the present application;
[0323] Figure 3 A schematic diagram of an embodiment of the network management method of the present application;
[0324] Figure 4A This is another schematic diagram of an embodiment of the network management method of the present application;
[0325] Figure 4B A schematic diagram of a weighted bipartite graph in an embodiment of the present application;
[0326] Figure 4C-1 It is a schematic diagram of the connection status between the STA and the AP in the first network under the traditional single AP transmission mechanism;
[0327] Figure 4C-2 A schematic diagram of group distribution of multiple groups divided according to regions;
[0328] Figure 4C-3 A schematic diagram of group distribution of multiple groups divided by a server in a centralized network management system in an embodiment of the present application;
[0329] Figure 4D This is a schematic diagram of the frame structure of the second beacon frame in an embodiment of the present application;
[0330] Figure 5A This is another schematic diagram of an embodiment of the network management method of the present application;
[0331] Figure 5B A schematic diagram of group distribution of multiple groups obtained by dividing multiple APs in a collaborative manner in a distributed network management system according to an embodiment of the present application;
[0332] Fig. 6A This is another schematic diagram of an embodiment of the network management method of the present application;
[0333] Figure 6B This is a schematic diagram of the structure of the first broadcast message in an embodiment of the present application;
[0334] Figure 7 This is another schematic diagram of an embodiment of the network management method of the present application;
[0335] Fig. 8A This is another schematic diagram of an embodiment of the network management method of the present application;
[0336] Figure 8B A schematic diagram of the structure of the second broadcast message in an embodiment of the present application;
[0337] Fig. 9 A schematic diagram of the structure of a communication device according to an embodiment of the present application;
[0338] Fig.10 A schematic diagram of the structure of the control device of the embodiment of the present application;
[0339] Fig.11 This is a schematic diagram of the structure of the second AP in the embodiment of the present application;
[0340] Fig.12 This is a structural diagram of the fourth STA in an embodiment of the present application;
[0341] Fig.13 A schematic diagram of a network management system according to an embodiment of the present application;
[0342] Fig.14 This is another schematic diagram of the network management system according to an embodiment of the present application. DETAILED DESCRIPTION
[0343] The embodiments of the present application provide a network management method and related devices, which are used to manage and plan STAs connected to an AP, thereby reducing interference between user devices and improving network transmission performance.
[0344] The technical solution of the embodiment of the present application can be applied to a wireless local area network (WLAN), which is usually called a wireless WiFi network, and the standard adopted is the Institute of Electrical and Electronics Engineers (IEEE) 802.11 series of standards. The network nodes that a WLAN may include are stations (STA), and the stations include access point stations (AP) and non-access point stations (None Access Point Station, Non-AP STA). Hereinafter, access point stations are referred to as APs, and non-access point stations are referred to as STAs.
[0345] Access point type sites are also called wireless access points or hotspots. AP is an access point for mobile users to enter the wired network. It is mainly deployed in homes, buildings and campuses. The typical coverage radius is tens of meters to hundreds of meters. Of course, it can also be deployed outdoors. AP is equivalent to a bridge connecting wired and wireless networks. Its main function is to connect various wireless network clients together and then connect the wireless network to Ethernet. Specifically, AP can be a terminal device or network device with a WiFi chip. Optionally, AP can be a device that supports the 802.11ax standard. Furthermore, optionally, the AP can be a device that supports multiple WLAN standards such as 802.11be, 802.11ac, 802.11n, 802.11g, 802.11b and 802.11a.
[0346] A non-access point station (Non-AP STA) can be a wireless communication chip, a wireless sensor or a wireless communication terminal. For example: a mobile phone that supports WiFi communication function, a tablet computer that supports WiFi communication function, a set-top box that supports WiFi communication function, a smart TV that supports WiFi communication function, a smart wearable device that supports WiFi communication function, a vehicle-mounted communication device that supports WiFi communication function, and a computer that supports WiFi communication function. Specifically, STA can be a terminal device or a network device with a wireless fidelity chip. Optionally, the station can support the 802.11ax standard, and further optionally, the station supports multiple WLAN standards such as 802.11be, 802.11ac, 802.11n, 802.11g, 802.11b and 802.11a.
[0347] It should be noted that the wireless signal interference between devices is caused by the dense deployment of devices in the wireless network. Therefore, the technical solution of the embodiment of the present application is also applicable to the network management and planning of other wireless communication networks with dense deployment of devices.
[0348] In order to achieve the management and planning of WiFi networks, the present application embodiment introduces two possible network management systems, namely, a centralized network management system and a distributed network management system. Figure 1A and Figure 2A Provide a detailed introduction.
[0349] See also Figure 1A , Figure 1A A schematic diagram of a network management system according to an embodiment of the present application. Figure 1A The network management system shown is a centralized network management system. The centralized network management system includes a server, a control device, an AP and a STA in a WiFi network ( Figure 1A Only 5 APs and 3 STAs are shown in the figure).
[0350] The control device obtains the first channel parameter information and sends the first channel parameter information to the server. The first channel parameter information includes channel parameter information between each AP in all APs in the WiFi network and the STA within the signal coverage range of each AP. For example, Figure 1A As shown, the signal coverage of AP2 is Figure 1A The range within the dotted circle shown in the figure includes STA1 and STA2. Then, the control device can obtain the channel parameter information between the AP2 and the STA1 and STA2 respectively. For other APs in the WiFi network, it is similar and will not be described in detail one by one.
[0351] Then, the server groups all APs and STAs in the WiFi network according to the first channel parameter information to obtain multiple groups, and then feeds back the grouping results of the multiple groups to the control device. The control device manages and plans the WiFi network according to the grouping results.
[0352] Specifically, the control device manages and plans the WiFi network mainly as follows: the control device notifies the AP and STA in the WiFi network of the group to which they belong. Then, the AP in the WiFi network adjusts the STA connected to the AP to achieve connection management of the STA and reduce mutual interference between devices. Furthermore, the control device can also allocate the same channel to APs in the same group, so that multiple APs in the same group can achieve multi-AP joint transmission, thereby meeting the requirements of high throughput and low latency of the WiFi network, while also improving spectrum utilization.
[0353] Above Figure 1A In the embodiment, the server is a cloud server, a cloud computing center server, or other types of servers, which are not specifically limited in this application. The control device is a central controller, or a master AP, etc., which are not specifically limited in this application.
[0354] based on Figure 1A The network management system shown below is Figure 1B , Figure 1C , Figure 1D and Figure 1E The structures of the server, control device, AP and STA of the network management system are introduced respectively. Figure 1C The AP shown is any AP in the first network. Figure 1E The STA shown is any STA in the first network.
[0355] First, let's introduce the server in this network management system. Figure 1B , Figure 1B This is a schematic diagram of the structure of the server in the embodiment of the present application. Figure 1B In the example, the server includes a first transceiver module 101 and a first grouping module 102 .
[0356] The transceiver module 101 is used to receive first channel parameter information sent by a control device.
[0357] The grouping module 102 is used to group all APs and all STAs in the first network according to the first channel parameter information to obtain grouping results of multiple groups;
[0358] The transceiver module 101 is further configured to send the grouping result of the multiple groups to the control device.
[0359] Figure 1B The server shown is used to execute the following Figure 3 Some or all of the steps in the embodiments shown in the figure, and the execution of the following Figure 4A In the embodiment shown, the server executes some or all of the steps. For example, the transceiver module 101 is used to execute the following Figure 3 Steps 301 and 303 in the embodiment shown, and Figure 4A Steps 402 and 404 in the embodiment shown in the figure. The grouping module 102 is used to perform the following Figure 3 Step 302 in the illustrated embodiment and Figure 4A Step 403 in the embodiment shown. Figure 3 and Figure 4A Related introduction in the illustrated embodiment.
[0360] See also Figure 1C , Figure 1C 1 is a schematic diagram of a control device according to an embodiment of the present application. The control device includes a transceiver module 103 and a group management module 104 .
[0361] The transceiver module 103 is used to receive the first channel parameter information reported by the AP or STA in the WiFi network, and send the first channel parameter information to the server; receive the grouping results of multiple groups sent by the server;
[0362] The grouping management module 104 is used to determine the AP and STA included in each of the multiple groups according to the grouping results of the multiple groups.
[0363] The transceiver module 103 is further configured to send the grouping information of the multiple groups to an AP or a STA in the WiFi network.
[0364] Optionally, the group management module 104 is further configured to allocate the same channel to APs in the same group. The transceiver module 103 is further configured to send channel allocation information corresponding to each of the multiple groups to the APs or STAs in the WiFi network.
[0365] Figure 1C The control device shown is used to execute the following Figure 4A In the embodiment shown, the control device performs some or all of the steps. For example, the transceiver module 103 is used to perform the following Figure 4A Steps 401 and 405 in the embodiment shown in the figure. The group management module 104 is used to execute the following Figure 4A Step 405a in the embodiment shown in the figure is described later. Figure 4A The relevant introduction in the illustrated embodiment will not be repeated here.
[0366] See also Figure 1D , Figure 1D FIG. 1 is a schematic diagram of the structure of an AP according to an embodiment of the present application. The AP includes a transceiver module 105 and a group management module 106. Optionally, the AP includes a signal measurement module 107.
[0367] The transceiver module 105 is used to receive grouping information of multiple groups sent by a control device.
[0368] The group management module 106 is used to determine the group to which the AP belongs according to the group information; and adjust the STA connected to the AP according to the group to which the AP belongs.
[0369] Optionally, the grouping information also includes channel allocation information allocated by the control device to the AP. The grouping management module 106 is further configured to determine a corresponding channel according to the channel allocation information, and adjust the channel accessed by the first AP to the channel corresponding to the channel allocation information.
[0370] In the WiFi network, the AP can measure the channel parameter information between the AP and the STA within the signal coverage of the AP. Optionally, the signal measurement module 107 is used to measure the channel parameter information between the AP and the STA within the signal coverage of the AP.
[0371] Figure 1D The AP shown is used to perform the following Figure 4A For example, the transceiver module 105 is used in the following Figure 4A Steps 401 and 405 in the embodiment shown in the figure. The group management module 106 is used to execute the following Figure 4A Steps 406 and 407 in the embodiment shown. The signal measurement module 107 is used to perform measurement to obtain Figure 4AThe second channel parameter information of step 401 in the illustrated embodiment.
[0372] See also Figure 1E , Figure 1E This is a schematic diagram of the structure of the STA in the embodiment of the present application. Figure 1E In the embodiment, the STA includes a transceiver module 108 and a group management module 109. Optionally, the STA also includes a signal measurement module 110.
[0373] The transceiver module 108 is used to receive grouping information of multiple groups sent by the control device.
[0374] The group management module 109 is used to determine the group to which the STA belongs according to the group information, and adjust the AP to which the STA is connected according to the group to which the STA belongs.
[0375] In the WiFi network, the STA can measure the channel parameter information between the STA and the AP in the first network. Optionally, the signal measurement module 110 is used to measure the channel parameter information between the STA and the AP in the first network. The transceiver module 108 is also used to send the channel parameter information to the AP connected to the STA, and then the AP sends it to the control device.
[0376] Specifically, the STA is used to execute the following Figure 4A In the embodiment shown, the fourth STA performs some or all of the steps. For example, the transceiver module 108 is used to perform Figure 4A In the embodiment shown, step 407b is performed. The group management module 109 is used to perform the above Figure 4A In the embodiment shown, steps 407c to 407e
[0377] It should be noted that the above Figure 1A In the centralized network management system shown, the control device can also group all APs and STAs in the WiFi network according to the first channel parameter information to obtain multiple groups, and the control device manages and plans the WiFi network according to the grouping results, which is not limited in this application. Figure 1A The server may not be included in the network management system shown.
[0378] See also Figure 2A , Figure 2A This is another schematic diagram of the network management system according to an embodiment of the present application. Figure 2A The network management system shown is a distributed network management system. The distributed network management system includes multiple APs and multiple STAs ( Figure 2A Only 5 APs and 3 STAs are shown in the figure).
[0379] The following takes AP2 as an example to introduce the distributed network management system. AP2 obtains the first channel parameter information, which includes the channel parameter information of the AP2 and the STA within the signal coverage range of the AP2. Then, the AP1 groups the STAs connected to the AP2 within the signal coverage range of the AP2 and the AP2 according to the first channel parameter information to obtain multiple groups. The AP2 performs connection management on the STAs connected to the AP2 and adjusts the channel accessed by the AP2 according to the grouping results of the multiple groups, and performs similar operations on other APs, thereby realizing connection management of the STAs in the WiFi network. The connection management of the STA is realized through the above process, and mutual interference between devices is reduced. Furthermore, multiple APs in the same group can access the same channel, so that multiple APs in the same group can realize multi-AP joint transmission, thereby meeting the requirements of high throughput and low latency of the WiFi network, and also improving spectrum utilization.
[0380] and, Figure 2A In the network management system shown, in the WiFi network, different APs can exchange information, that is, the APs can group the APs and STAs of the WiFi network in a collaborative manner. Specifically, different APs can exchange information through a wired network, such as Ethernet.
[0381] based on Figure 2A The network management system shown below is Figure 2B and Figure 2C The structures of the first AP and the second STA of the network management system are respectively introduced. The second STA is a STA within the signal coverage of the first AP.
[0382] See also Figure 2B , Figure 2B FIG. 1 is a schematic diagram of the structure of the first AP in the embodiment of the present application. Figure 2B In the embodiment, the first AP includes a transceiver module 201, a grouping module 202 and a grouping management module 203. Optionally, the first AP also includes a signal measurement module 204.
[0383] The transceiver module 201 is used to obtain first channel parameter information;
[0384] The grouping module 202 is used to group the first AP and the first STA within the signal coverage of the first AP and having established a connection with the first AP according to the first channel parameter information, to obtain multiple groups, and to obtain grouping results of the multiple groups and the group to which the first AP belongs.
[0385] The group management module 203 is used to adjust the STAs connected to the first AP according to the group to which the first AP belongs.
[0386] The transceiver module 201 is also used to send the grouping results of the multiple groups.
[0387] In the WiFi network, the first AP can measure the channel parameter information between the first AP and the STA within the signal coverage of the first AP. Optionally, the signal measurement module 204 is used to measure the first channel parameter information between the first AP and the STA within the signal coverage of the first AP.
[0388] Specifically, the first AP is used to execute the following Figure 3 Steps 301 to 303 in the embodiment shown in the figure, and execution of the following Figure 5A Figure 6 Figure 7 and Fig. 8A In the embodiment shown, the first AP performs some or all of the steps. For example, the transceiver module 201 is used to perform Figure 5A In the embodiment shown, steps 501, 505 and 509 are shown. The grouping module 201 is used to perform Figure 5A In the embodiment shown, step 502 is performed. The group management module 203 is used to perform Figure 5A Step 504 in the embodiment shown. Figure 3 , Figure 5A Figure 6 Figure 7 and Fig. 8A The relevant introduction in the illustrated embodiment will not be repeated here.
[0389] See also Figure 2C , Figure 2C This is a schematic diagram of the structure of the second STA in the embodiment of the present application. Figure 2C In the embodiment, the second STA includes a transceiver module 205 and a group management module 206. Optionally, the second STA also includes a signal measurement module 207.
[0390] The transceiver module 205 is used to receive the grouping result of multiple groups sent by the first AP.
[0391] The group management module 206 is used to determine the group to which the second STA belongs according to the grouping result, and adjust the AP to which the second STA is connected according to the group to which the second STA belongs.
[0392] In the WiFi network, the second STA can measure the channel parameter information between the second STA and the first AP. Optionally, the signal measurement module 207 is used to measure the channel parameter information between the second STA and the first AP; the transceiver module 205 is used to send the signal parameter information to the first AP.
[0393] The technical solution of the embodiments of the present application is introduced below in conjunction with the embodiments.
[0394] See also Figure 3 , Figure 3 This is a schematic diagram of an embodiment of the network management method of the present application. Figure 3 In the network management method, the network management method includes:
[0395] 301. A first communication device obtains first channel parameter information.
[0396] The first channel parameter information includes channel parameter information between a first AP in the first network and a STA within a signal coverage range of the first AP.
[0397] Optionally, the first channel parameter information includes a signal gain between the first AP and a STA within the signal coverage range of the first AP, where the signal gain is also called received signal strength or interference signal strength. The specific signal strength measurement method includes the following two possible implementation methods:
[0398] 1. A first AP sends a data packet on a channel to which the first AP accesses, and STAs within the signal coverage of the first AP receive the data packet sent by the first AP and respectively measure the received signal strength of the data packet sent by the first AP.
[0399] 2. STAs within the signal coverage of the first AP access the channel accessed by the first AP. The first AP receives data packets sent by STAs within the signal coverage of the first AP; then, the first AP measures the received signal strength of the data packets sent by STAs within the signal coverage of the first AP.
[0400] 1. Based on Figure 1A In the centralized network management system shown, the first communication device may be a server, and the first channel parameter information includes the channel parameter information between each AP among all APs in the first network and the STA within the signal coverage of each AP. That is, the first channel parameter information can be understood as the global interference information of the first network, and the global interference information includes the interference signal strength between each AP among all APs in the first network and the STA within the signal coverage of each AP. The first AP includes all APs in the first network. For the specific process of the server obtaining the first channel parameter information, please refer to the following text Figure 4A Related introduction in the illustrated embodiment.
[0401] 2. Based on Figure 2AIn the distributed network management system shown, the first communication device may be the first AP. Then, the first channel parameter information includes the channel parameter information between the first AP and the STA within the signal coverage range of the first AP. That is, the first channel parameter information can be understood as the local interference information of the first network obtained by the first AP, and the local interference information includes the interference signal strength between the first AP and the STA within the signal coverage range of the first AP. For the specific process of the first AP obtaining the first channel parameter information, please refer to the following text Figure 5A Related introduction in the illustrated embodiment.
[0402] It should be noted that the above shows that the first channel parameter information includes the channel parameter information between the first AP and the STA within the signal coverage range of the first AP. In practical applications, the first channel parameter information may also include the channel parameter information between the first AP and the STA within the preset range of the first AP. The preset range is a range divided according to geographical location; or, the preset range is a range where the STA corresponding to the signal gain greater than or equal to the fourth preset threshold is located.
[0403] Among them, the signal gain greater than or equal to the fourth preset threshold is the channel gain between the first AP measured by the first AP and the STA corresponding to the signal gain greater than or equal to the fourth preset threshold, or the signal gain greater than or equal to the fourth preset threshold is the channel gain between the STA corresponding to the signal gain greater than or equal to the fourth preset threshold and the first AP measured respectively.
[0404] Factors considered in setting the fourth preset threshold include the computing power of the first communication device, the accuracy requirements for the grouping results, the network status and channel of the first network, etc., which are not specifically limited in this application.
[0405] For example, the fourth preset threshold may be 5 dB (decibel) or 10 dB.
[0406] In the following embodiments, only the example in which the first channel parameter information includes the channel parameter information between the first AP and the STA within the signal coverage of the first AP is used for explanation, and the present application does not make any specific limitation.
[0407] 302. The first communication device groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups.
[0408] The first STA includes part or all of the STAs within the signal coverage of the first AP.
[0409] 1. Based on Figure 1AIn the centralized network management system shown, the first STA includes all STAs within the signal coverage range of each AP in the first network and each AP. That is, the first STA includes all STAs in the first network. Specifically, based on the centralized network management system, the specific execution process of the server can be found in Figure 4A The embodiment shown.
[0410] 2. Based on Figure 2A The distributed network management system shown, the first STA includes the STA that has established a connection with the first AP within the signal coverage of the first AP. Specifically, based on the distributed network management system, the specific execution process of the first AP can be found in Figure 5A The embodiment shown.
[0411] 303. The first communication device sends the grouping result of the multiple groups.
[0412] The grouping result is used to indicate the AP and STA included in each of the multiple groups.
[0413] In an embodiment of the present application, a first communication device obtains first channel parameter information, and the first channel parameter information includes channel parameter information between a first AP in a first network and a STA within the signal coverage of the first AP; the first communication device groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups, and the first STA includes some or all of the STAs within the signal coverage of the first AP; then, the first communication device sends a grouping result of the multiple groups, and the grouping result is used to indicate the AP and STA included in each of the multiple groups. In this way, the AP or STA receives the grouping result and can adjust the connection in combination with the group to which it belongs, thereby realizing the management and planning of the STA connected to the AP. In the technical solution of the embodiment of the present application, the multiple groups are divided by the first communication device according to the first channel parameter information, which can reduce interference between user devices and improve the performance of network transmission.
[0414] The following specific Figure 1A The centralized network management system shown introduces the technical solution of the embodiment of the present application.
[0415] See also Figure 4A , Figure 4A This is a schematic diagram of an embodiment of the network management method of the present application. Figure 4A In the network management method, the network management method includes:
[0416] 401. The second AP sends second channel parameter information to a control device.
[0417] The second channel parameter information is channel parameter information between the second AP and the STA within the signal coverage range of the second AP, the second AP is an AP in the first network, and the first network is a network controlled by the control device.
[0418] Specifically, Figure 1A As shown, the first network is Figure 1A The WiFi network controlled by the control device shown. The STA in the WiFi network selects an AP from the WiFi network to connect in an existing manner. For example, the STA can connect to the AP corresponding to the beacon frame with the highest signal strength. The AP in the WiFi network can select a channel and access the channel through existing technology. For example, the AP selects a channel and accesses the channel through LCSS technology. Then, the second AP measures the second channel parameter information between the second AP and the STA within the signal coverage range of the first AP, and sends the second channel parameter information to the control device. Other APs in the WiFi network also perform corresponding measurements and report channel parameter information. Then, the control device can collect the first channel parameter information between each AP in all APs in the WiFi network and the STA within the signal coverage range of each AP.
[0419] 402. The control device sends first channel parameter information to a server.
[0420] The first channel parameter information includes channel parameter information between each AP among all APs in the first network and a STA within a signal coverage range of each AP.
[0421] Optionally, the first channel parameter information includes a channel matrix W, the channel matrix The channel matrix W is a matrix of K rows by L columns, where K is the number of STAs included in the first network and L is the number of APs included in the first network. ij is the channel gain (or the average channel gain) between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, and j is an integer greater than or equal to 1 and less than or equal to L.
[0422] Specifically, the first network includes K STAs and L APs. The first network is modeled as a bipartite graph G = (V, E). The vertex set V can be divided into two non-intersecting vertex subsets V STA and the vertex subset V AP , the vertex subset V STA and the vertex subset V AP The following equations (1) and (2) are satisfied:
[0423] V STA ∪VAP =V Formula (1)
[0424]
[0425] Among them, in the above formula (2), Is an empty set.
[0426] For example, Figure 4B As shown, the weighted bipartite graph includes two types of vertices, a first type of vertex is a STA in the first network, and a second type of vertex is an AP in the first network.
[0427] The edge set E of the weighted bipartite graph represents the channel condition between the STA in the first network and the AP in the first network.
[0428] The edge weight of the weighted bipartite graph represents the channel gain between the STA and the AP in the first network, that is, the weight marked on the edge of the weighted bipartite graph. For example, STA1 is the first STA in the first network, and AP1 is the first AP in the first network, then the edge weight of the edge connecting STA1 and AP1 is w 11 , represents the channel gain between STA1 and AP1. Figure 4B The edge weights of other edges in are similar and will not be explained one by one here.
[0429] Optionally, the channel matrix W is an edge weight matrix of the weighted bipartite graph.
[0430] 403. The server groups all APs and all STAs in the first network according to the first channel parameter information to obtain multiple groups.
[0431] The following introduces step 403 in conjunction with the first channel parameter information including the channel matrix W in step 402. Specifically, step 403 specifically includes steps 403a to 403d.
[0432] Step 403a: The server calculates the first intermediate matrix D according to the channel matrix W.
[0433] Among them, the first intermediate matrix D 1 is a matrix of K rows and K columns, D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All off-diagonal elements of are zero.
[0434] D 2 is a matrix of L rows by L columns, D 2The elements b on the main diagonal of j is the sum of channel gains between the jth AP in the first network and all STAs in the first network.
[0435] 0 K×L is a matrix of K rows by L columns, 0 K×L All elements in are zero. L×K is a matrix of L rows by K columns, 0 L×K The elements in are all zero.
[0436] Optionally, when the channel matrix W is an edge weight matrix W of a weighted bipartite graph, step 403a is specifically as follows: the server generates a degree matrix of the edge weight matrix W according to the edge weight matrix W. Wherein, each diagonal element of the degree matrix of the edge weight matrix W is the degree of the node corresponding to each diagonal element, and the degree of the node corresponding to each diagonal element is Figure 4B The sum of the edge weights of the edges associated with the node in the weighted bipartite graph (i.e., the edges connected to the node in the weighted bipartite graph). Figure 4B As shown in , the nodes connected to STA1 include AP1, AP2, AP3 and AP4, then the degree of STA1 is the sum of the edge weights of the edge connecting STA1 and AP1, the edge weights of the edge connecting STA1 and AP2, the edge weights of the edge connecting STA1 and AP3, and the edge weights of the edge connecting STA1 and AP4. The degree matrix of the edge weight matrix W is the first intermediate matrix D mentioned above.
[0437] Step 403b: The server determines the second intermediate matrix Z according to the channel matrix W and the first intermediate matrix D.
[0438] Each row vector in the second intermediate matrix Z corresponds to a network node, and the network node is a STA or an AP in the first network.
[0439] The following describes the process of the server determining the second intermediate matrix Z in the above steps in combination with steps 1 to 3.
[0440] Step 1: The server determines the third intermediate matrix R according to the channel matrix W and the first intermediate matrix D.
[0441] Among them, the third intermediate matrix W T is the transpose of the channel matrix W.
[0442] Step 2: The server performs singular value decomposition on the third intermediate matrix R to obtain the left singular matrix U and the right singular matrix of the third intermediate matrix R.
[0443] Among them, the third intermediate matrix R = UΣV T, Σ is the diagonal matrix obtained by singular value decomposition of the third intermediate matrix R, V T It is the transpose of the right singular matrix V obtained by singular value decomposition of the third intermediate matrix R.
[0444] Step 3: The server selects the corresponding M column vectors from the left singular matrix U in descending order of the singular values in the diagonal matrix Σ to obtain the fourth intermediate matrix U M , and select the corresponding M column vectors from the right singular matrix V according to the order of the singular values in the diagonal matrix Σ from large to small, and obtain the fifth intermediate matrix V M .
[0445] Where M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L). 2 (K+L) refers to the logarithm of K+L with base 2, and min(K, L) refers to the minimum value between K and L.
[0446] Optionally, the value of M may be the number of the multiple groups. The number of the multiple groups is an integer greater than 1 and less than min(K, L).
[0447] It should be noted that the value of M can also be determined in combination with actual conditions, and this application does not limit it. The above example of the value range of M is determined by experimental data analysis, and the value range of M is in log 2 When the value is an integer between (K+L) and min(K, L), the server has a better grouping effect on the STAs and APs in the first network. For example, the sum of the effective signal strengths in the subnetwork formed by the APs and STAs in each group is large, and the signal interference strength between the network nodes (APs or STAs) in the subnetwork and outside the subnetwork is low.
[0448] Step 4: The server calculates the second intermediate matrix Z according to the fourth intermediate matrix, the fifth intermediate matrix and the first intermediate matrix.
[0449] Among them, the second intermediate matrix
[0450] Specifically, each row vector in the second intermediate matrix Z corresponds to a network node, which is a STA or AP in the first network. Specifically, the i-th row vector of the second intermediate matrix Z is the row vector corresponding to the i-th STA in the first network, and the K+j-th row vector of the second intermediate matrix Z is the row vector corresponding to the j-th AP in the first network.
[0451] Step 403c: The server clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result.
[0452] The first clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group in the multiple groups.
[0453] Specifically, in the above step 403b, the M column vectors selected by the server from the left singular matrix U and the M column vectors selected from the right singular matrix V according to the size of the singular value contain more characteristic information characterizing the weighted bipartite graph, then in step 403c, the server clusters the second intermediate matrix Z by spectral clustering, so that the server can achieve a better grouping effect of STAs and APs in the first network. For example, the sum of the effective signal strengths in the subnetwork formed by the APs and STAs in each group is large, and the signal interference strength between the network nodes (APs or STAs) in the subnetwork and outside the subnetwork is low.
[0454] In this embodiment, the clustering algorithm is a k-means algorithm, a k-means++ algorithm, a mean shift clustering algorithm, an expectation-maximization algorithm (EM) clustering algorithm, a representative density-based spatial clustering of applications with noise (DBSCAN), a graph community detection clustering algorithm, a graph neural network (GCN) clustering algorithm, etc.
[0455] Optionally, the server performs clustering on the second intermediate matrix The K+L row vectors included are clustered to obtain row vectors of multiple clusters.
[0456] It should be noted that the server is in the second intermediate matrix When clustering the K+L row vectors included, the server can use the second intermediate matrix All or part of each row vector in the included K+L row vectors represents the row vector for clustering. That is, the server can select part of the content of each row vector to represent the row vector for clustering.
[0457] Step 403d: The server determines the network nodes included in the multiple groups according to the row vectors of the multiple clusters.
[0458] Specifically, it can be known from the above step 403c that each row vector in the second intermediate matrix Z corresponds to a network node, so the server divides the network nodes corresponding to the row vectors of the same cluster among the row vectors of the multiple clusters into the same group.
[0459] For example, Figure 1A As shown, the AP and STA in the WiFi network are divided into two groups, and the two groups correspond to two clusters. Here, the two clusters include cluster 1 and cluster 2, and the two groups include group 1 and group 2. Among them, cluster 1 corresponds to group 1, and cluster 2 corresponds to group 2.
[0460] Among them, STA1 corresponds to row vector 1 of the second intermediate matrix Z, STA2 corresponds to row vector 2 of the second intermediate matrix Z, and STA3 corresponds to row vector 3 of the second intermediate matrix Z. AP1 corresponds to row vector 4 of the second intermediate matrix Z, AP2 corresponds to row vector 5 of the second intermediate matrix Z, AP3 corresponds to row vector 6 of the second intermediate matrix Z, AP4 corresponds to row vector 7 of the second intermediate matrix Z, and AP5 corresponds to row vector 8 of the second intermediate matrix Z.
[0461] Cluster 1 includes row vector 1, row vector 2, row vector 4, row vector 5 and row vector 6, while cluster 2 includes row vector 3, row vector 7 and row vector 8. Then the server determines that group 1 includes AP1, AP2, AP3, STA1 and STA2 according to the row vectors included in cluster 1. The server determines that group 2 includes AP4, AP5 and STA3 according to the row vectors included in cluster 2.
[0462] As can be seen from the above, for group C of the multiple groups, the sum of the signal strengths of the mutual interference between the network nodes in the sub-network formed by the network nodes in the group C and the network nodes outside the sub-network can be expressed as:
[0463]
[0464] Among them, STA k Indicates the kth STA, AP l Represented as the lth AP.
[0465] The effective strength of the sub-network formed by the network nodes in the group C can be defined as the sum of the signal strengths between the network nodes in the sub-network, specifically expressed as:
[0466]
[0467] In order to reduce the overall interference of the network and improve the throughput, the mutual interference between the subnetworks formed in each group should be minimized as much as possible, and the effective signal strength in the same subnetwork should be as large as possible. Here, the multiple groups are N, and N is an integer greater than or equal to 2. The first network can be divided into N subnetworks corresponding to the N groups. In order to achieve the effect of minimizing the mutual interference between the subnetworks formed in each group and maximizing the effective signal strength in the same subnetwork, the problem can be located as follows:
[0468]
[0469] argmin is an operator for the arguments of the maxima. Pointing to make Minimized {G 1 ,G 2 ,…,G N}.
[0470] st is the operator of the constraint condition. Indicates that it should satisfy and c is constraint 1, is constraint 2, is a universal quantifier, is an empty set, G d ∩G g Point to G d and G g The intersection of G b refers to the b-th cluster, where b is an integer greater than or equal to 1 and less than or equal to N.
[0471] In the above step 401, the control device collects global interference information in the first network, and sends the global interference information in the first network to the server in step 402. In the above step 403, the server abstracts the first network into a weighted bipartite graph to perform spectral graph decomposition, and performs spectral graph decomposition on the weighted bipartite graph through graph theory methods and spectral clustering to solve the above problem. That is, all APs and all STAs of the first network are grouped through graph theory methods and spectral clustering, so that the signal interference between sub-networks formed by each group in the obtained multiple groups is small, and the effective strength within the sub-network is large.
[0472] For example, Figure 4C-1 FIG. 1 is a schematic diagram of the connection status between the AP and the STA in the first network under the traditional single AP transmission mechanism. Figure 4C-1 In the STA, the STA selects the nearest AP to establish a connection, and the AP provides services to the STA. Figure 4C-2 This is a schematic diagram of the distribution of groups divided by region. Figure 4C-3 Schematic diagram of the group distribution of multiple groups under the network management method of the embodiment of the present application, Figure 4C-3 As shown, the APs and STAs in the first network are divided into multiple groups, compared to Figure 4C-2 The division method, Figure 4C-3 In the , network nodes that are close to each other are divided into the same group. Figure 4C-1 The traditional single AP transmission mechanism, Figure 4C-3In this case, multiple APs in the same group can provide services to STAs in the same group at the same time, realizing multi-AP joint transmission and improving network throughput and spectrum utilization.
[0473] 404. The server sends the grouping results of the multiple groups to the control device.
[0474] The grouping results of the multiple groups are used to indicate the AP and STA included in each of the multiple groups.
[0475] For example, the grouping result includes the group number of each group in the multiple groups, and the AP and STA corresponding to each group number (ie, the AP and STA included in each group).
[0476] 405. The control device sends second group information.
[0477] The second group information includes information about groups to which all APs in the first network belong respectively and information about groups to which all STAs in the first network belong respectively.
[0478] For example, the second grouping information includes group numbers of groups to which all APs and all STAs in the first network belong respectively.
[0479] Optionally, this embodiment further includes step 405a, and step 405a is performed before step 405.
[0480] Step 405a: The control device allocates the same channel to APs in the same group among the multiple groups.
[0481] Specifically, the control device selects a channel for the AP in each group of the multiple groups according to the first channel parameter information; and the control device selects the same channel for the AP in each group.
[0482] Based on step 405a, optionally, the second grouping information further includes third channel allocation information. The third channel allocation information is information about the channel allocated by the control device to the AP in the first network.
[0483] It should be noted that step 405a shows the implementation of using a control device to allocate corresponding channels to the multiple groups. In practical applications, in a centralized network management system, APs in the same group may also select channels. The specific selection process is similar to that in the subsequent Figure 5A Steps 507 to 511 in the illustrated embodiment are similar. Please refer to the relevant introduction below for details, which will not be repeated here.
[0484] In step 405, the control device broadcasts the second grouping information of all APs and all STAs in the first network as an example for introduction. In practical applications, the control device can also determine the information of the group to which each network node in the first network belongs from the grouping results of multiple groups, and send the grouping information corresponding to the network node to the corresponding network node. That is, the control device sends the grouping information in a unicast manner.
[0485] It should be noted that the control device may also send the second grouping information only to all APs in the first network. Optionally, the server is transparent to the grouping management of all APs and all STAs in the first network; or, the APs in each group notify the STAs in the same group of the relevant grouping information, which is not limited in this application.
[0486] Since the control device collects the global interference information in the first network in the centralized network management system, it is more reasonable for the control device to allocate a corresponding channel to each group according to the global interference information. Compared with the method of selecting channels through APs in the same group, the method of selecting a corresponding channel for each group by the control device can reduce information interaction between devices, reduce signaling overhead, and avoid unnecessary waste of resources.
[0487] In addition, the control device allocates the same channel to APs in the same group, so that multiple APs in the same group can achieve multi-AP joint transmission, thereby meeting the requirements of high throughput and low latency of the WiFi network, while also improving spectrum utilization.
[0488] 406. The second AP determines the group to which the second AP belongs according to the second group information.
[0489] Specifically, the second AP is an AP in the first network, and the second AP determines the group to which the second AP belongs according to the second group information. Optionally, the second group information includes a group number to which the second AP belongs; and the second AP determines the group to which the second AP belongs according to the group number to which the second AP belongs.
[0490] Optionally, the second grouping information further includes third channel allocation information. Then this embodiment further includes step 406a, and step 406a and step 406 can be performed simultaneously, or step 406 can be performed first, or step 406a can be performed first, which is not limited in this application.
[0491] Step 406a: The second AP determines the third channel selected by the control device for the second AP according to the third channel allocation information.
[0492] For example, the third channel allocation information includes indication information of a frequency band used for the third channel or the frequency band of the third channel. The second AP may determine the frequency band of the third channel according to the indication information of the frequency band used for the third channel and access the frequency band; or the second AP accesses the frequency band of the third channel.
[0493] Step 406b: The second AP adjusts the channel accessed by the second AP to the third channel.
[0494] 407. The second AP adjusts the STAs connected to the second AP according to the group to which the second AP belongs.
[0495] There are multiple ways for the second AP to adjust the STA of the second connection in step 407, and two possible implementation ways are shown below.
[0496] Implementation method 1: The second AP sends information about the group to which the second AP belongs, and the information about the group to which the second AP belongs is used by the STA in the first network to determine whether to access the second AP.
[0497] Based on implementation mode 1, the above step 407 specifically includes step 407a.
[0498] Step 407a: The second AP sends information about the group to which the second AP belongs.
[0499] The information of the group to which the second AP belongs is used by the STA in the first network to determine whether to access the second AP.
[0500] Specifically, the second AP may send the information of the group to which the second AP belongs in a broadcasting manner.
[0501] Optionally, the information of the group to which the second AP belongs may be sent in a message broadcast by the second AP, or carried in a beacon frame, which is not specifically limited in this application.
[0502] The following shows how the second AP sends information about the group to which the second AP belongs through a beacon frame. Optionally, the information about the group to which the second AP belongs includes the second group number of the group to which the second AP belongs, and the above step 407a specifically includes:
[0503] The second AP sends a second beacon frame, where the second beacon frame carries a second group number of the group to which the second AP belongs.
[0504] For example, Figure 4D As shown, the second AP modifies the format of the second beacon frame. Figure 4D In the second beacon frame, the second AP includes a fixed field field and an optional field. The second AP extends the group number field in the optional field and carries the second group number on the bits corresponding to the group number field.
[0505] Based on the first implementation mode, this embodiment further includes step 407b to step 407e, and step 407b to step 407e are executed after step 407a.
[0506] Step 407b: The fourth STA receives the information of the group to which the second AP belongs, which is sent by the second AP.
[0507] Optionally, the information of the group to which the second AP belongs includes the second group number.
[0508] Optionally, the fourth STA receives a second beacon frame sent by the second AP, where the second beacon frame carries the second group number.
[0509] Specifically, the fourth STA scans all channels and receives the second beacon frame sent by the second AP; then, the fourth STA compares the second group number carried in the second beacon frame with the group number of the group to which the fourth STA belongs.
[0510] It should be noted that, before step 407b, the fourth STA receives the information of the group to which the fourth STA belongs, which is sent by the eighth AP. The eighth AP is the AP to which the fourth STA adjusts the connection. That is to say, the fourth STA can determine the group to which the fourth STA belongs through the group information sent by the eighth AP to which it is connected. For example, as shown in the figure, the fourth STA is STA3, and STA3 is connected to AP2, and AP2 sends the information of the group to which the fourth STA belongs to STA3.
[0511] Step 407c: The fourth STA determines whether the second AP and the fourth STA belong to the same group according to the information of the group to which the second AP belongs. If so, execute step 407d; if not, execute step 407e.
[0512] Optionally, the information of the group to which the second AP belongs includes the second group number, and step 407c is specifically that the fourth STA determines whether the group number of the group to which the fourth STA belongs is consistent with the second group number, if so, execute step 407d, if not, execute step 407e.
[0513] Step 407d: The fourth STA adjusts the AP to which the fourth STA is connected to the second AP.
[0514] For example, Figure 2A As shown, the fourth STA is STA3, STA2 is connected to AP2 before adjustment, and the second AP is AP3. Then STA3 can adjust the connection from AP2 to AP3.
[0515] Step 407e: The fourth STA maintains connection with the eighth AP.
[0516] In the technical solution of the above-mentioned implementation method 1, the fourth STA mainly adjusts its associated AP, thereby realizing the connection management of the STA.
[0517] Implementation method 2: Implementation method 2 is introduced below in combination with steps 407f to 407i.
[0518] Step 407f: The second AP determines the STAs included in the second group.
[0519] Specifically, the second AP determines the STAs included in the group to which the second AP belongs according to the second grouping information sent by the control device.
[0520] Step 407g: The second AP determines whether the second AP and the third STA belong to the same group. If so, execute step 407h; if not, execute step 407i.
[0521] The third STA is a STA that has established a connection with the second AP.
[0522] In one possible implementation, the second AP determines the STAs included in the group to which the second AP belongs based on the second grouping information, and determines whether the STAs included in the group to which the second AP belongs include the third STA. If so, execute step 407h; if not, execute step 407i.
[0523] In another possible implementation, the second AP determines the group to which the second AP belongs and the group to which the third STA belongs based on the second group information; then, the second AP determines whether the group to which the second AP belongs and the group to which the third STA belongs are the same, and if so, executes step 407h; if not, executes step 407i.
[0524] Step 407h: The second AP keeps providing services for the third STA.
[0525] If the second AP and the third STA belong to the same group, the second AP keeps providing services for the third STA.
[0526] Step 407i: The second AP determines the seventh AP according to the group to which the third STA belongs, and sends third indication information to the seventh AP; or, the second AP sends fourth indication information to the third STA.
[0527] The third indication information is used to indicate that the seventh AP provides services for the third STA, and the seventh AP and the third STA belong to the same group.
[0528] Specifically, the second AP determines the group to which the third STA belongs according to the second grouping information, and determines the seventh AP from the group to which the third STA belongs. For example, the seventh AP may be an AP with the largest signal strength or a medium signal strength in the group to which the third STA belongs, which may be specifically determined by the signal strength of the beacon frame sent by the seventh AP. Then, the second AP sends the third indication information to the seventh AP to indicate that the seventh AP provides services for the third STA, thereby realizing the connection management of the STA connected to the AP.
[0529] In this implementation, the second AP mainly adjusts the STA connected to it. That is, the connection adjustment between the AP and the STA is transparent to the STA in the first network. The association adjustment mechanism performed by the second AP is compatible with the existing WiFi7 standard, and the association adjustment process is transparent to the STA in the first network, without the need to expand and modify the functions of the STA in the first network.
[0530] The fourth indication information is used to instruct the third STA to request the seventh AP to provide services for the third STA.
[0531] Specifically, the second AP determines the group to which the third STA belongs according to the second grouping information, and determines the seventh AP from the group to which the third STA belongs. For example, the seventh AP may be an AP with the largest signal strength or a medium signal strength in the group to which the third STA belongs, which may be specifically determined by the signal strength of the beacon frame sent by the seventh AP. Then, the second AP sends the fourth indication information to the seventh AP to instruct the third STA to request the seventh AP to provide services for the third STA.
[0532] In an embodiment of the present application, the control device receives the second channel parameter information sent by the second AP; then, the control device sends the first channel parameter information to the server. The server groups all APs and all STAs in the first network according to the first channel parameter information to obtain multiple groups, and sends the grouping results of the multiple groups to the control device. The control device sends the second grouping information; then, the second AP determines the group to which the second AP belongs according to the second grouping information, and adjusts the STA connected to the second AP according to the group to which the second AP belongs, thereby realizing the management and planning of the STA connected to the AP. Secondly, in an embodiment of the present application, all APs and all STAs of the first network are grouped through graph theory methods and spectral clustering to obtain multiple groups, so that the signal interference between the sub-networks corresponding to each group in the multiple groups is small, and the effective strength within the sub-network is large, which can reduce interference between user devices and improve the performance of network transmission.
[0533] See also Figure 5A , Figure 5AThis is another embodiment schematic diagram of the network management method of the present application embodiment. Figure 5A In the network management method, the network management method includes:
[0534] 501. A first AP obtains first channel parameter information.
[0535] The first channel parameter information includes channel parameter information between the first AP and STAs within the signal coverage of the first AP.
[0536] Specifically, Figure 2A As shown, the STA in the WiFi network selects an AP from the WiFi network to connect in an existing manner. For example, the STA can connect to the AP corresponding to the beacon frame with the highest signal strength. The AP in the WiFi network can select a channel and access the channel through existing technology. Here, the first AP is AP2, and AP2 obtains the AP2 and the STAs (including Figure 2A Specifically, AP2 measures the channel parameter information between AP2 and STA1 and STA2 respectively; or STA1 measures the channel parameter information between STA1 and AP2 and reports it to AP2; and STA2 measures the channel parameter information between STA2 and AP2 and reports it to AP2.
[0537] 502. The first AP groups the first AP and the first STA according to the first channel parameter information to obtain multiple groups.
[0538] The first STA is a STA within the signal coverage of the first AP that has established a connection with the first AP.
[0539] The following describes the process of the first AP determining the multiple groups in step 502 in conjunction with steps 502a to 502e.
[0540] Step 502a: The first AP calculates the p+Kth main diagonal element b of the first intermediate matrix D according to the pth column vector of the channel matrix W. p .
[0541] The first network includes L APs and K STAs, the first AP is the pth AP in the first network, L is an integer greater than or equal to 1, K is an integer greater than or equal to 1, and p is an integer greater than or equal to 1 and less than or equal to L.
[0542] For an introduction to the channel matrix W, please refer to the previous Figure 4A The relevant introduction of the channel matrix W in step 402 in the embodiment shown is not repeated here. For the relevant introduction of the first intermediate matrix D, please refer to the aforementioned Figure 4A The introduction related to the first intermediate matrix D in step 403 in the illustrated embodiment will not be repeated here.
[0543] Since the first AP obtains the pth column vector of the channel matrix W, the first AP calculates the p+Kth main diagonal element b of the first intermediate D according to the pth column vector of the channel matrix W. p .
[0544] Step 502b: The first AP obtains the first K main diagonal elements of the first intermediate matrix D.
[0545] The first AP obtains the first K diagonal elements of the first intermediate matrix D from the STAs in the first network and other APs in the first network, that is, a 1 to a K Specifically, the first AP and other APs in the first network may exchange information via a wired network using an inter access point protocol (IAPP) to obtain information sent by other APs.
[0546] Step 502c: The first AP calculates the p+Kth main diagonal element b of the first intermediate matrix D according to the p+Kth main diagonal element b p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined.
[0547] Each row vector of the sixth intermediate matrix Q corresponds to a network node, and the network node is an AP or a STA in the first network.
[0548] Step 502c is described below in conjunction with steps 1 to 3.
[0549] Step 1: The first AP calculates the channel matrix W based on the pth column vector and the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K main diagonal elements in the first intermediate matrix D determine the pth row vector of the third intermediate matrix R
[0550] Among them, the second intermediate matrix The W T Transpose this W.
[0551] Since the first AP only obtains the pth column vector of the channel matrix W, the first K main diagonal elements of the first intermediate matrix and the p+Kth main diagonal element b p , therefore, the first AP can only determine the pth row vector of the first intermediate matrix R
[0552] Step 2: The first AP is based on the stochastic gradient descent algorithm and the pth row vector of the third intermediate matrix R Determine the first matrix X M The p-th row vector and the second matrix Y M .
[0553] Among them, the first matrix X M The similarity between the second matrix Y and the third matrix is greater than or equal to a first preset threshold. M The degree of similarity to the fourth matrix is greater than or equal to a second preset threshold.
[0554] The third matrix is a matrix obtained by selecting M column vectors from the left singular matrix obtained by singular value decomposition of the third intermediate matrix R. Each column vector in the left singular matrix has a corresponding singular value, and the M column vectors of the third matrix are M column vectors selected from the left singular matrix in descending order of singular value size.
[0555] The fourth matrix is a matrix obtained by selecting M column vectors from the right singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the right singular matrix has a corresponding singular value, and the M column vectors of the fourth matrix are M column vectors selected from the right singular matrix in descending order of singular value size.
[0556] M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), log 2 (K+L) refers to the logarithm of K+L with base 2, and min(K, L) refers to the minimum value of K and L.
[0557] It should be noted that the value of M can also be determined in combination with actual conditions, and this application does not limit it. The above example of the value range of M is determined by experimental data analysis, and the value range of M is in log 2 When the value is an integer between (K+L) and min(K, L), the server can achieve a better grouping effect on the STAs and APs in the first network. For example, the sum of the effective signal strengths in the sub-network formed by the APs and STAs in each group is large, and the signal interference strength between the network nodes (APs or STAs) in the sub-network and outside the sub-network is low.
[0558] Optionally, the value of M may be the number of the multiple groups. The number of the multiple groups is an integer greater than 1 and less than min(K, L).
[0559] The magnitudes of the first preset threshold and the second preset threshold can be determined in combination with experimental data, and specifically can be determined in combination with the following formula (5).
[0560] In this embodiment, the APs in the first network implement the singular value decomposition of the third intermediate matrix R and the clustering of the row vectors of the sixth intermediate matrix Q in a distributed network management system in a collaborative manner. In the distributed network management system, each AP measures the channel parameter information between each AP and the STA within the signal coverage range of each AP (the channel parameter information measured by each AP corresponds to a column of the channel matrix W. For example, if the first AP is the pth AP in the first network, then the first AP can calculate the pth row vector of the third intermediate matrix R The distributed stochastic gradient descent algorithm between APs in the first network achieves a result similar to the result obtained by performing singular value decomposition on the third intermediate matrix R. Specifically, the problem can be approximated as:
[0561]
[0562] in, Pointing to make The minimized X M and Y M . To obtain The Frobenius norm of .
[0563] Specific details are provided later Fig. 6A The calculation process of the embodiment shown in the figure solves the problem of the above equation (7), please refer to the following text Fig. 6A Related introduction in .
[0564] Step 3: The first AP is based on the p+Kth main diagonal element b of the first intermediate matrix D p , the first K main diagonal elements in the first intermediate matrix D, the first matrix X M The p-th row vector and the second matrix Y M Calculate the sixth intermediate matrix The first K row vectors and the K+pth row vector of .
[0565] Among them, the first K row vectors of the sixth intermediate matrix Q are:
[0566]
[0567] The K+pth row vector of the sixth intermediate matrix Q is:
[0568]
[0569] The K+pth row vector of the sixth intermediate matrix Q corresponds to the pth AP, and the first K row vectors of the sixth intermediate matrix Q correspond to the K STAs in the first network respectively.
[0570] Step 502d: The first AP clusters the target row vector according to a clustering algorithm to obtain a second clustering result.
[0571] The target row vector includes the K+pth row vector corresponding to the first AP and the row vector corresponding to the first STA in the first K row vectors of the sixth intermediate matrix Q. One cluster in the multiple clusters corresponds to one group in the multiple groups.
[0572] Optionally, the clustering algorithm is a k-means algorithm, a k-means++ algorithm, a mean shift clustering algorithm, a maximum expectation clustering algorithm, a DBSCAN, a graph group detection clustering algorithm, a GCN clustering algorithm, or the like.
[0573] In this embodiment, the first AP is introduced by k-means as an example, and the specific Figure 7 The illustrated embodiment introduces the determination process of the first AP in step 502d.
[0574] Step 502e: The first AP determines the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters.
[0575] Specifically, since the K+pth row vector of the sixth intermediate matrix Q corresponds to the pth AP, and the first K row vectors of the sixth intermediate matrix Q correspond to the K STAs in the first network, the first AP can divide the network nodes corresponding to the row vectors of the same cluster in the multiple clusters into the same group, and the network nodes are the first AP or the first STA.
[0576] 503. The first AP sends a grouping result of multiple groups.
[0577] The grouping result is used to indicate the AP and STA included in each of the multiple groups.
[0578] Since the objects of the first AP grouping include the first AP and the first STA, the multiple groups include the first AP and the first STA. The grouping result may include the group number corresponding to each group, and the AP and STA corresponding to each group number.
[0579] 504. The first AP adjusts the STAs connected to the first AP according to the group to which the first AP belongs.
[0580] The adjustment method of the first AP in step 504 is the same as that described above. Figure 4A In the embodiment shown, the adjustment method of the second AP in step 407 is similar. For details, please refer to the aforementioned Figure 4A The relevant introduction of step 407 in the illustrated embodiment will not be repeated here.
[0581] In a distributed network management system, multiple APs can group all APs and STAs in the first network in a collaborative manner. The specific distribution of multiple groups is as follows: Figure 5B As shown, the grouping effect of multiple groups is the same as the above Figure 4C-3 The grouping effects of multiple groups are similar, the sum of effective signal strengths within the sub-network formed by the APs and STAs in each group is larger, and the signal interference intensity between the network nodes (APs or STAs) within the sub-network and outside the sub-network is lower.
[0582] In an embodiment of the present application, the first AP obtains first channel parameter information; then, the first AP groups the first AP and the STA connected to the first AP within the signal coverage of the first AP according to the first channel parameter information to obtain multiple groups; then, the first AP sends the grouping results of the multiple groups, and the first AP adjusts the STA connected to the first AP according to the group to which the first AP belongs, thereby realizing the management and planning of the STA connected to the AP. Secondly, in an embodiment of the present application, in a distributed network management system, each AP of the first network groups all APs and all STAs of the first network through a collaborative manner, graph theory methods and spectral clustering to obtain multiple groups, so that the signal interference between the sub-networks corresponding to each group in the multiple groups is small, and the effective signal strength within the sub-network is large, which can reduce interference between user devices and improve the performance of network transmission.
[0583] In a distributed network management system, multiple APs group all APs and all STAs in the first network in a collaborative manner. Each AP can group each AP and the STAs that have established connections with each AP. Then, in a distributed network management system, the channel allocation process of APs in the same group can be specifically referred to the relevant introduction of the following steps 505 to 511.
[0584] Then, the above Figure 5A The illustrated embodiment also includes steps 505 to 511. Moreover, steps 505 to 511 are executed before step 502. There is no obvious execution order between steps 505 to 511 and steps 503 to 504. Steps 503 to 504 may be executed first, or steps 505 to 511 may be executed first; or, depending on the situation, steps 505 to 511 and steps 505 to 511 may be executed simultaneously, which is not limited in this application.
[0585] 505. The first AP receives first group information sent by the fourth AP.
[0586] The first grouping information includes grouping information to which the fourth AP belongs and an identifier of the fourth AP, and the fourth AP includes APs other than the first AP in the first network.
[0587] Specifically, the first group includes the group number of the group to which the fourth AP belongs. The first AP receives first group information broadcasted by other APs in the first network to determine the groups to which the other APs belong.
[0588] 506. The first AP determines, according to the first grouping information, that the fifth AP and the first AP belong to the same group.
[0589] The fifth AP is a part of the fourth AP, and the fifth AP and the first AP belong to the same group.
[0590] Specifically, the first AP determines, according to the group number included in the first group information, that the group number of the group to which the first AP belongs is the same as the group number of the group to which the fifth AP belongs, that is, the first AP can determine that the first AP and the fifth AP belong to the same group.
[0591] 507. The first AP determines whether the priority of the first AP is higher than the priority of the fifth AP according to the identifier of the first AP, the identifier of the fifth AP and the preset AP priority rule. If so, execute step 508; if not, execute step 510.
[0592] The preset AP priority rule includes the priority corresponding to the identifier of each AP in the first network. The first AP can determine the priority of the first AP and the priority of the fifth AP according to the preset AP priority rule, and determine whether the priority of the first AP is higher than the priority of the fifth AP. If so, execute step 508; if not, execute step 510.
[0593] Optionally, the preset AP priority rule is pre-configured in the first AP, or is specified by a communication protocol, or is sent to the first AP by other network devices.
[0594] The preset AP priority rule can be set in any of the following ways:
[0595] 1. Set the preset AP priority rule according to the AP hardware configuration. For example, the higher the AP hardware configuration, the higher the priority of the AP.
[0596] 2. The preset AP priority rule is set according to the AP identification (index) order. For example, the larger the AP index is, the higher the priority of the AP is.
[0597] The above setting method is only an example, and the preset AP priority rule can also be set in other ways, which is not limited in this application.
[0598] 508. The first AP selects a first channel by scanning channels and accesses the first channel.
[0599] For example, the first AP may use the LCCS technology to scan channels, select a first channel, and access the first channel.
[0600] 509. The first AP sends first channel allocation information to the fifth AP.
[0601] The first channel allocation information is used to instruct the fifth AP to access the first channel.
[0602] Optionally, the first channel allocation information includes indication information for indicating a frequency band of the first channel or the frequency band of the first channel.
[0603] 510. The first AP receives second channel allocation information sent by the sixth AP.
[0604] The second channel allocation information is information about the second channel selected by the sixth AP, which is the AP with the highest priority among the fifth APs. The second channel allocation information is similar to the first channel allocation information, and please refer to the related introduction of the first channel allocation information for details.
[0605] 511. The first AP adjusts the channel accessed by the first AP to the second channel according to the second channel allocation information.
[0606] For example, the second channel allocation information includes a frequency band of the second channel, and the first AP can access the frequency band of the second channel.
[0607] It can be seen from the above steps 505 to 511 that in the distributed network management system, the same channel is allocated to multiple APs in the same group, so that multiple APs in the same group can realize multi-AP joint transmission, thereby meeting the requirements of high throughput and low latency of the WiFi network, and also improving spectrum utilization.
[0608] It should be noted that the channel allocation method from step 505 to step 511 is only an example. In a distributed network management system, multiple APs in the same group can also select an AP by listening and competing. Then, the AP selects a channel and sends the information of the selected channel to other APs in the group, so that multiple APs in the same group can access the same channel to achieve multi-AP joint transmission; and meet the requirements of high throughput and low latency of WiFi networks, while also improving spectrum utilization.
[0609] See also Fig. 6A , Fig. 6A This is another embodiment schematic diagram of the network management method of the present application embodiment. Fig. 6AThe method comprises:
[0610] 601. The first AP randomly initializes the first matrix X M The p-th row vector x p and the second matrix Y M .
[0611] Among them, x p =[x p1 x p2 … x pM ],x p is a matrix of one row by M columns, x pf For x p The element of the fth column vector in Y M =[y 1 y 2 … y M ] is a matrix of K rows and M columns, y f Y M The element of the f-th column vector in , where f is an integer greater than or equal to 1 and less than or equal to M.
[0612] 602. The first AP initializes the momentum term Δx p and the momentum term ΔY M , so that the momentum term Δx p and the momentum term ΔY M Both are 0.
[0613] This momentum term Δx is used in the stochastic gradient descent process p and the momentum term ΔY M , in order to accelerate the convergence process of the stochastic gradient descent algorithm, so as to solve the minimum problem of the above equation (5). p =[Δx p1 Δx p2 …Δx pM ], Δx pf is the momentum term Δx p The momentum term ΔY is the f-th column vector of M =[Δy 1 Δy 2 … Δy M ], Δy f is the momentum term ΔY M The f-th column vector of .
[0614] 603. The first AP broadcasts the second matrix Y to other APs in the first network except the first AP. M .
[0615] After the first AP performs the initialization process of step a and step b, the first AP periodically broadcasts the second matrix Y to other APs.M , other APs in the first network also periodically broadcast the second matrix Y on the other APs M .
[0616] Specifically, the first AP may broadcast the second matrix Y through a first broadcast message. M , the first broadcast message carries the second matrix Y M , first identifier and grouping round number.
[0617] The first identifier is used to indicate that the first broadcast message is a broadcast message for stochastic gradient descent processing. The grouping round number is used to identify the round number of the current grouping performed by the first AP, that is, the number of times the first AP currently groups the first AP and the first STA. When the first AP receives a first broadcast message with a higher grouping round number, the first AP ends the execution of the stochastic gradient descent corresponding to the grouping round number and enters the next round of stochastic gradient descent process.
[0618] For example, Figure 6B As shown, the first bit of the payload of the first broadcast message is used to identify the first broadcast message as a broadcast message for stochastic gradient descent processing. For example, when the first bit is "0", it means that the first broadcast message is used for stochastic gradient descent processing. The remaining 7 bits of the first byte of the payload of the first broadcast message, except the first bit, are used to identify the grouping round number, which is used to identify the round number of the current grouping of the first AP. The second matrix Y M It is carried in the remaining bytes except the first byte in the payload of the first broadcast message.
[0619] 604. The first AP starts a first timer.
[0620] The first timer is used for the first AP to monitor the second matrix Y broadcast by other APs in the first network within the duration of the first timer. M .
[0621] 605. The first AP determines the interruption type of the first interruption. If the first interruption is caused by the first AP receiving the second matrix Y sent by other APs M If an interruption occurs, step 606 is executed; if the first interruption is caused by the timeout or interruption of the first timer, the first AP returns to step 604 and restarts the first timer.
[0622] When the first AP has a first interrupt, the first AP determines the interrupt type of the first interrupt. If the first AP receives a second matrix Y sent by another AP, MIf an interruption occurs due to the timeout of the first timer or an interruption, the first AP returns to step 604, and the first AP continues to receive and monitor the second matrix Y sent by other APs in the first network. M .
[0623] 606. The first AP updates the first matrix X according to the stochastic gradient descent rule. M The p-th row vector x p and the second matrix Y M .
[0624] Step 606 is described below in conjunction with steps 1 to 3.
[0625] Step 1: The first AP calculates the pth row vector of the third intermediate matrix R
[0626] By the above Figure 5A In the embodiment shown, step 502c knows that the first intermediate matrix The first K main diagonal elements of All the main diagonal elements of and the p+Kth main diagonal element b of the first intermediate matrix D p So we know that is a matrix of one row by K columns, for The element in the i-th column of .
[0627] Step 2: The first AP saves the second matrix Y received from other APs M .
[0628] Step 3: The first AP is based on the p-th row vector of the third intermediate matrix R and received by other APs sending the second matrix Y M Update the first matrix X on the first AP M The p-th row vector x p and the second matrix Y M .
[0629] Next, combine steps a to d to update the x p =[x p1 x p2 … x pM The first element x in ] p1 and Y M =[y 1 y 2 … y M The first element y in ] 1 The process of step 3 is introduced as an example.
[0630] Step a: The first AP calculates the error vector
[0631] Step b: The first AP updates Δx p1 and Δy 1 , the updated Δx p1 =αΔx p1 +ηey 1 , the updated Δy 1 =αΔy 1 +ηe T x p1 , α and η are constant parameters, α is the momentum parameter, η is the learning rate, e T is the transpose of e, for y 1 The transpose of .
[0632] Step c: First AP update Get the updated
[0633] Step d: The first AP updates x p1 and 1 , and get the updated x p1 Equal to x before update p1 Add the updated Δx p1 , the updated y 1 Equal to y before update 1 Add the updated Δy 1 .
[0634] For this x p The update process of other elements in is the same as x p1 The update process is similar to that for Y M The update process of other elements in y is the same as 1 The update process is similar to that of , and it should be noted that when calculating the error vector e, the error vector obtained in the previous update should be used. Calculate the error vector. The previous update obtained It refers to the e obtained by the first AP updating the previous element of the element currently updated. For example, the above steps a to d update x p The first element x in p1 and Y M The first element y in 1 , then the value obtained in step 1 of step 606 is used. If the first AP updates x p The second element x in p2 and Y M The first element y in 2 , when the first AP calculates the error vector, it uses the updated error vector from step c above And so on, I will not explain them one by one here.
[0635] It should be noted that after the first AP executes step 606, the first AP may restart the first timer and continue to monitor the first matrix Y broadcasted by other APs in the first network. M , when the second matrix Y sent by the other AP is monitored M When , the first AP performs an operation similar to step 606, that is, updates the first matrix X obtained in step 606 again M The p-th row vector x p and the second matrix Y M , and so on, each update process is similar, until the number of restarts of the first timer reaches a preset number, then the first AP determines the last updated first matrix X M The p-th row vector x p and the second matrix Y M .
[0636] Above Figure 5A In step 502d of the embodiment shown, the first AP clusters the target row vector according to the clustering algorithm to obtain a second clustering result. Figure 7 The illustrated embodiment introduces this step 502d. Figure 7 , Figure 7 This is another embodiment schematic diagram of the network management method of the present application embodiment. Figure 7 The method comprises:
[0637] 701. A first AP generates a first cluster center according to a first random seed.
[0638] The first cluster center includes cluster centers corresponding to the multiple clusters generated by the first AP, each cluster corresponds to a cluster center, and each cluster in the multiple clusters corresponds to a group in the multiple groups.
[0639] Specifically, all APs in the first network use the first random seed to generate cluster centers corresponding to multiple clusters. For example, the first AP randomly generates first cluster centers G corresponding to the multiple clusters according to the first random seed. 1 , c 2 , c 3 , c 4 ……c N}, c b is the cluster center corresponding to the bth cluster, and b is an integer greater than or equal to 1 and less than or equal to N. That is, the cluster centers corresponding to the multiple randomly initialized clusters are obtained, and cluster G b Corresponding cluster center c b .
[0640] It should be noted that the first random seed may be pre-configured or generated by an AP (eg, a master AP) in the first network and then sent by the AP to other APs in the first network, which is not specifically limited in this application.
[0641] 702. The first AP determines, according to the first cluster center, a cluster to which each row vector in the target row vector belongs, and obtains row vectors included in each cluster in the multiple clusters.
[0642] Specifically, the first AP searches for the cluster center closest to each row vector in the target row vector, and divides each row vector into the cluster corresponding to the cluster center. and the K+pth row vector of the sixth intermediate matrix Q Select the target row vector from The first AP is The row vectors included in find the cluster corresponding to the cluster center closest to each row vector.
[0643] It should be noted that the first AP uses all or part of each row vector in the target row vector as a representative of the row vector for clustering. That is, the first AP can select part of the content of each row vector to represent the row vector for clustering.
[0644] Based on the above steps 701 and 702, the above Figure 5A Step 502e in the illustrated embodiment specifically includes:
[0645] When the first preset condition is met, the first AP determines the network nodes respectively included in the plurality of groups according to the row vectors of the plurality of clusters.
[0646] Specifically, when the first preset condition is met, the first AP divides the network nodes corresponding to the row vectors of the same cluster among the row vectors of the multiple clusters into the same group.
[0647] There are multiple types of the first preset condition, and two possible implementation methods are shown below.
[0648] Implementation method 1: The first preset condition includes that the update times of the cluster centers respectively corresponding to the multiple clusters are greater than or equal to the third preset threshold.
[0649] Specifically, the first AP determines whether the update times of the cluster centers corresponding to the multiple clusters are greater than or equal to the third preset threshold. If so, the first AP determines the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters; if not, executes the following steps 512 to 514.
[0650] For example, the third preset threshold is 100. The third preset threshold is mainly determined based on experimental data, that is, the number of updates of the cluster centers corresponding to the multiple clusters when the clustering process reaches a convergence level.
[0651] Implementation method 2: The first preset condition introduced below in combination with step 1 to step 4 includes: the first convergence accuracy is less than or equal to the preset convergence accuracy.
[0652] The first convergence accuracy is a ratio of an absolute value of the first difference to the first global error.
[0653] The first difference is the difference between the first global error and a preset initialization error.
[0654] The first global error is the sum of first local errors respectively calculated by all APs in the first network.
[0655] The first local error of the first AP includes the sum of errors corresponding to each of the multiple clusters determined by the first AP, and the error corresponding to each cluster is the sum of errors between the row vectors included in each cluster and the first average vector corresponding to each cluster, and the first average vector corresponding to each cluster is calculated by the first AP according to a gossip protocol.
[0656] Then, when the first AP determines that the first convergence accuracy is less than or equal to the preset convergence accuracy, the first AP executes the above step 502e.
[0657] For example, the initialization error is ε, and the first global error is The preset convergence accuracy is τ; then the first AP determines whether If yes, execute step 502e; if no, execute the following steps 512 to 514.
[0658] For example, the preset convergence accuracy is 0.01, and the preset initialization error is infinite.
[0659] The preset convergence precision is determined based on experimental data and experience, that is, when the above clustering process reaches the convergence degree, the preset convergence precision takes a value.
[0660] Optionally, the preset convergence accuracy is set by the first AP, or may be specified by a communication protocol, or sent to the first AP by other devices, which is not specifically limited in this application.
[0661] The calculation process of the first convergence accuracy is introduced below in conjunction with steps 1 to 3.
[0662] Step 1: The first AP calculates a first average vector corresponding to a row vector of each cluster in the multiple clusters according to a gossip protocol (also known as an epidemic protocol, a rumor algorithm, or an epidemic propagation algorithm);
[0663] Specifically, the first AP calculates the number of network nodes size included in each of the multiple clusters determined by the first AP b The sum of the row vectors included in multiple clusters b , b is an integer greater than or equal to 1 and less than or equal to N, and N is the number of groups in the plurality of groups. Then, the first AP uses the gossip protocol to calculate the first average vector corresponding to the row vector of each cluster in the plurality of clusters If the first AP is the pth AP, each AP in the first network stores a local variable Δ p1 and the local control parameter ω p1 . The first AP sets the Δ p1 ={sum 1 ,sum 2 ......sum N}, that is, the sum of the row vectors included in the multiple clusters. The first AP sets the ω on the first AP p1 ={size 1 , size 2 ......size N Then, the first AP uses the gossip protocol to calculate the first average vector corresponding to the row vector of each cluster in the multiple clusters.
[0664] It should be noted that the first AP uses the gossip protocol for related introduction, please refer to the following text Fig. 8A Related introduction of the illustrated embodiment.
[0665] Step 2: The first AP calculates the first local error.
[0666] Specifically, the first AP calculates the first local error according to the row vectors respectively included in the multiple clusters obtained in the above step 702 and the first average vector corresponding to the row vector of each cluster in the multiple clusters.
[0667] The first local error includes the sum of errors corresponding to each cluster in the plurality of clusters determined by the first AP. The error corresponding to each cluster is the sum of errors between row vectors included in each cluster and the first average vector corresponding to each cluster.
[0668] For example, the first local error The first local error represents the sum of root mean square errors corresponding to multiple clusters in the first AP. represents the row vector in the bth cluster among the multiple clusters, c b Indicates the cluster center corresponding to the b-th cluster. Represents the sum of the root mean square errors of the row vectors of the b-th cluster and the cluster centers corresponding to the b-th cluster.
[0669] Step 3: The first AP calculates a first global error according to the gossip protocol and the first local error.
[0670] The first global error is the sum of first local errors respectively calculated by all APs in the first network.
[0671] Specifically, if the first AP is the pth AP, each AP in the first network stores a local variable Δ p2 and the local control parameter ω p2 . The Δ on the first AP p2 Set to the sum of the root mean square errors corresponding to multiple clusters in the first AP, and set ω on the first AP p2 is set to 1, and the local control parameters on other APs are 0. Then, the first AP uses the gossip protocol and the first local error calculation to obtain a first global error, that is, to obtain the sum of the first local errors calculated by all APs in the first network.
[0672] In this embodiment, the other APs of the first network calculate the first local error in a similar manner to the way in which the first AP calculates the first local error on the first AP in step 2, which will not be described in detail here. When each AP calculates the first local error of each AP, it is calculated by the clustering result of the most recent clustering of the row vectors of multiple clusters by each AP and the clustering result of the previous clustering of the row vectors of multiple clusters. The time when other APs in the first network send the first local error can be different or the same. For example, AP1 sends the first local error of AP1 to the first AP at time 1, that is, AP1 should select the clustering result of the most recent clustering of the row vectors of multiple clusters from time 1 and the clustering result of the previous clustering of the row vectors of multiple clusters to calculate the first global error of AP1. And AP2 sends the first local error of AP2 to the first AP at time 2, that is, AP2 should select the clustering result of the most recent clustering of the row vectors of multiple clusters from time 2 and the clustering result of the previous clustering of the row vectors of multiple clusters to calculate the first global error of AP2. The same is true for other APs in the first network, which will not be described one by one here.
[0673] It should be noted that step 3 is different from the following Fig. 8AIn steps 801 to 811, the first AP uses the gossip protocol to determine the first average vector corresponding to the row vector of each cluster in the multiple clusters. Similar, except that the first AP converts the local variable Δ p2 Set to the sum of the root mean square errors corresponding to multiple clusters in the first AP, and set ω p2 Set to 1, so that through similar Fig. 8A The execution flow from step 801 to step 811 shown can obtain the sum of the first local errors calculated by all APs in the first network.
[0674] Step 4: The first AP calculates the first convergence accuracy according to the first global error and a preset initialization error.
[0675] The first convergence accuracy is equal to the absolute value of the quotient of the first difference and the first global error, and the first difference is the difference between the first global error and the initialization error.
[0676] In a possible implementation, if the first preset condition is not met, the above Figure 5A The illustrated embodiment further includes steps 512 to 514 .
[0677] Step 512: The first AP calculates a first average vector corresponding to the row vectors of the multiple clusters according to the gossip protocol;
[0678] Step 512 is similar to step 1 of the second implementation method in the aforementioned first preset condition. For details, please refer to the relevant introduction of step 1 of the second implementation method in the aforementioned first preset condition, which will not be repeated here.
[0679] Step 513: The first AP uses the first average vector corresponding to the row vector of each cluster in the multiple clusters as the cluster centers corresponding to the multiple clusters, respectively, to obtain second cluster centers;
[0680] Step 514: the first AP determines the cluster to which each row vector in the target row vector belongs according to the second cluster center;
[0681] Steps 513 to 514 are the same as those described above. Figure 7 In the embodiment shown, step 701 is similar to step 702. For details, please refer to the aforementioned Figure 7 The related introduction of step 701 and step 702 in the illustrated embodiment will not be repeated here.
[0682] Step 515: When the second preset condition is met, the first AP divides the network nodes corresponding to the row vectors of the same cluster among the multiple clusters into the same group.
[0683] Step 515 is similar to the related introduction of the aforementioned step 502e. Please refer to the related introduction of the aforementioned step 502e for details, and will not be repeated here.
[0684] The second preset condition is similar to the first preset condition, and two possible implementation methods are shown below.
[0685] Implementation method 1: The second preset condition includes that the number of updates of the cluster centers corresponding to the multiple clusters is greater than or equal to a third preset threshold.
[0686] For the relevant introduction of the third preset threshold, please refer to the relevant introduction of the third preset threshold in the first implementation method of the first preset condition, which will not be repeated here.
[0687] For example, steps 512 to 515 are the first update of the cluster centers corresponding to the multiple clusters by the first AP. Therefore, if the third preset threshold is 1, the first AP executes step 515; if the third preset threshold is an integer greater than 1, the second AP performs the third clustering of the target row vector, and the specific clustering process is similar to the aforementioned steps 512 to 515.
[0688] Implementation method 2: The second preset condition includes: the second convergence accuracy is less than or equal to the preset convergence accuracy, and the second convergence accuracy is the ratio of the absolute value of the difference between the second global error and the first global error to the second global error.
[0689] Then, when the first AP determines that the second convergence accuracy is less than or equal to the preset convergence accuracy, the first AP executes the above step 515.
[0690] The calculation process of the second convergence accuracy is introduced below in combination with step a and step d.
[0691] Step a: The first AP calculates a second average vector corresponding to the row vector of each cluster in the multiple clusters according to the gossip protocol;
[0692] Specifically, it can be seen from the above steps 513 and 514 that the row vector of each cluster in the multiple clusters is updated again, so the first AP uses the gossip protocol to calculate the second average vector corresponding to the row vector of each cluster in the multiple clusters. The specific calculation method is similar to step 1 in the second implementation method in the above-mentioned first preset condition, and will not be repeated here.
[0693] Step b: The first AP calculates the second local error.
[0694] The first AP obtains the row vectors respectively included in the multiple clusters and the second average vector corresponding to the row vector of each cluster in the multiple clusters according to the above step 514.
[0695] Step c: The first AP calculates a second global error according to the gossip protocol and the second local error.
[0696] Step b to step c are similar to step 2 to step 3 in implementation method 2 of the aforementioned first preset condition. Please refer to the relevant introduction of step 2 to step 3 in implementation method 2 of the aforementioned first preset condition, and no further details will be given here.
[0697] Step d: The first AP calculates a second convergence accuracy according to the second global error and the first global error.
[0698] The second convergence accuracy is equal to the ratio of the absolute value of the second difference to the quotient of the second global error, and the second difference is the difference between the second global error and the first global error.
[0699] Specifically, since steps 512 to 515 are the second clustering of the target row vector, the second global error should be compared with the first global error calculated by the previous clustering (i.e., the first clustering), that is, the initialization error in the aforementioned step 502e is updated to the first global error. If the first AP needs to cluster the target row vector for the third time, the initialization error is updated to the second global error, and so on.
[0700] Therefore, when the absolute value of the quotient of the second difference and the second global error is less than or equal to the preset convergence accuracy, the first AP executes step 515. When the absolute value of the quotient of the second difference and the second global error is greater than the preset convergence accuracy, the first AP continues to cluster the target row vector for the third time, and the specific clustering process is similar to the aforementioned steps 512 to 515.
[0701] Combine the following Fig. 8A The illustrated embodiment introduces a process in which the first AP calculates the first average vector corresponding to the row vector of each cluster in the plurality of clusters according to the gossip protocol. Fig. 8A , Fig. 8A This is another embodiment of the network management method of the present application. Fig. 8A In the network management method, the network management method includes:
[0702] 801. The first AP sets a local variable Δ p1 and the local control parameter ω p1 .
[0703] Each AP in the first network stores a local variable Δ p1 and the local control parameter ω p1 . The first AP is set to Δ p1 ={sum 1 ,sum 2 ......sum N}, that is, the sum of the row vectors included in the multiple clusters, set ω p1 ={size 1 , size 2 ......size N}, that is, the number of row vectors included in each of the multiple clusters. The first AP calculates the Δ p1 and ω p1 .
[0704] 802. The first AP determines a first node cache set.
[0705] The first node cache set includes the I selected by the first AP from the AP directly connected to the first AP in the first network. max The set of APs is I max is an integer greater than or equal to 1.
[0706] TheI max The size of is related to the cache size or storage space size of the first AP. For example, the larger the storage space of the first AP, the larger the storage space of the first AP. max The bigger.
[0707] 803. The first AP selects a target node from the first node cache set.
[0708] 804. The first AP determines the Δ p1 / 2,ω p1 / 2 and r1.
[0709] Among them, r1 is used to indicate whether the first AP exchanges the Δ p1 and ω p1 For example, when r1 is 1, it indicates that the first AP exchanges the Δ p1 and ω p1 When r1 is 0, it indicates that this is not the first time that the first AP exchanges the Δ p1 and ω p1 .
[0710] 805. The first AP sends a second broadcast message to the target node.
[0711] The second broadcast message carries the group round number, Δ p1 / 2,ω p1 / 2 and r1. For the relevant introduction of the group round number, please refer to the previous Fig. 6A The relevant introduction in the illustrated embodiment will not be repeated here.
[0712] For example, Figure 8B As shown, the second broadcast message carries the group round number, Δ p1 / 2,ωp1 / 2, r1 and the identifiers of the nodes included in the first cache set.
[0713] It should be noted that Figure 8B The second broadcast message shown is the first broadcast message sent by the first AP to the target node for the first time, so the second broadcast message carries Δ p1 / 2,ω p1 / 2 and r1 is 1; if the first AP is not sending the second broadcast message to the target node for the first time, the second broadcast message should carry the Δ p1 ,ω p1 and r1=0. For example, when the first AP sends the second broadcast message to the target node for the second time, the updated Δ p1 and ω p1 The first AP changes the Δ p1 / 2 as the updated Δ on the first AP p1 , replace ω in step 801 p1 / 2 as the updated ω on the first AP p1 Then, the first AP carries the updated Δ p1 ,ω p1 And r1=0.
[0714] 806. The first AP starts a second timer.
[0715] The second timer is used for the first AP to monitor the second broadcast message sent by other APs except the first AP in the first network within the duration of the second timer.
[0716] 807. The first AP determines the interruption type of the second interruption. If the second interruption is caused by the first AP receiving a second broadcast message sent by other APs in the first network except the first AP, execute step 808; if the second interruption is caused by the second timer timing out or interruption, the first AP returns to step 806.
[0717] If the second interruption is caused by the timeout or interruption of the second timer, the first AP returns to step 806, ie, restarts the second timer, and then executes steps 806 to 810 again.
[0718] It should be noted that, if the second broadcast message of the target node is not received after restarting the second timer for multiple times, the first AP can reselect other nodes, and then perform information exchange between the first AP and the selected other nodes according to a process similar to step 804 to step 810. Specifically, the first AP can be configured with the number of times the first AP restarts the second timer for the selected node. When the second broadcast message sent to the node is not received after reaching the number of times, the first AP selects other nodes, and then performs information exchange between the first AP and the selected other nodes according to a process similar to step 804 to step 810.
[0719] 808. The first AP determines whether r1 in the second broadcast message received from other APs in the first network except the first AP is 1. If so, execute steps 809 to 810; if not, execute step 810.
[0720] If the first AP receives the second broadcast message sent by other APs in the first network except the first AP, the first AP determines whether the reply message r1 is 1, and if so, executes steps 809 and 810; if not, executes step 810.
[0721] 809. The first AP sends the Δ before the update of the first AP to the first node. p1 / 2 and ω p1 / 2.
[0722] The first node is the node that sends the second broadcast message to the first AP within the second timer in the above step 807.
[0723] 810. The first AP updates Δ according to the second broadcast message sent by other APs in the first network. p1 / 2 and ω p1 / 2.
[0724] If r1 in the second broadcast message sent by the first node is 1, the first AP will p1 / 2 plus Δ carried in the second broadcast message sent by the first node p1 / 2, and the ω on the first AP p1 / 2 plus ω carried in the second broadcast message sent by the first node p1 / 2.
[0725] If r1 in the second broadcast message sent by the first node is 0, the first AP does not need to perform the above step 809. p1 / 2 plus Δ carried in the second broadcast message sent by the first node p1 / 2, and the ω on the first AP p1 / 2 plus ω carried in the second broadcast message sent by the first node p1 / 2.
[0726] 811. The first AP updates the first node cache set.
[0727] Specifically, the first AP selects I from the second node cache set carried in the second broadcast message sent by the first node, the first node, and the first node cache set. max APs are used as APs included in the first node cache set.
[0728] The above shows the process of executing the gossip protocol between the first AP and the target node, and realizes the Δ p1 and ω p1 Then similarly for other nodes in the first node cache set, the first AP performs the process from step 803 to step 811 to achieve Δ between the first AP and all nodes in the first node cache set. p1 and ω p1 The first AP uses the gossip protocol to calculate the first average vector corresponding to the row vector of each cluster in the multiple clusters.
[0729] The present application also provides a communication device. Fig. 9 , another structural diagram of a communication device 900 in an embodiment of the present application, the communication device 900 can be used to perform Figure 3 The steps performed by the first communication device in the embodiment shown, Figure 4A The steps performed by the server in the embodiment shown, and Figure 5A For the steps performed by the first AP in the illustrated embodiment, reference may be made to the relevant descriptions in the above method embodiments.
[0730] The communication device 900 includes: a processor 901 , a memory 902 , and a transceiver 903 .
[0731] The processor 901, the memory 902 and the transceiver 903 are connected via buses respectively, and computer instructions are stored in the memory.
[0732] In a possible implementation, when the communication device 900 is used to execute Figure 4A When the server in the embodiment shown executes the steps, the aforementioned Figure 1B The grouping module 102 in the embodiment may specifically be the processor 901 in this embodiment, so the specific implementation of the processor 901 is not described in detail. Figure 1BThe transceiver module 101 may specifically be the transceiver 903 in this embodiment, so the specific implementation of the transceiver 903 is not repeated any more.
[0733] In another possible implementation, when the communication device 900 is used to execute Figure 5A In the embodiment shown, the first AP performs the steps described above. Figure 2B The grouping module 202, the grouping management module 203 and the signal measurement module 204 in the embodiment may be the processor 901 in this embodiment, so the specific implementation of the processor 901 is not repeated here. Figure 1B The transceiver module 201 in the embodiment may specifically be the transceiver 903 in this embodiment, so the specific implementation of the transceiver 903 is not described in detail.
[0734] The present application also provides a control device. Fig.10 , another structural diagram of the control device 1000 in the embodiment of the present application, the control device 1000 can be used to perform Figure 4A For the steps executed by the control device in the illustrated embodiment, reference may be made to the relevant description in the above method embodiment.
[0735] The control device 1000 includes: a processor 1001 , a memory 1002 , and a transceiver 1003 .
[0736] The processor 1001, the memory 1002 and the transceiver 1003 are connected via buses respectively, and computer instructions are stored in the memory.
[0737] The foregoing Figure 1C The group management module 104 in the embodiment may specifically be the processor 1001 in this embodiment, so the specific implementation of the processor 1001 is not described in detail. Figure 1C The transceiver module 103 may specifically be the transceiver 1003 in this embodiment, so the specific implementation of the transceiver 1003 is not described in detail any more.
[0738] The present application embodiment also provides a second AP, see Fig.11 , another structural diagram of the second AP1100 in the embodiment of the present application, the second AP1100 can be used to perform Figure 4A For the steps performed by the second AP in the illustrated embodiment, reference may be made to the relevant descriptions in the above method embodiment.
[0739] The second AP 1100 includes a processor 1101 , a memory 1102 , and a transceiver 1103 .
[0740] The processor 1101, the memory 1102 and the transceiver 1103 are connected via buses respectively, and computer instructions are stored in the memory.
[0741] The foregoing Figure 1D The group management module 106 and the signal measurement module 107 in the embodiment may specifically be the processor 1101 in this embodiment, so the specific implementation of the processor 1101 is not described in detail. Figure 1D The transceiver module 105 may specifically be the transceiver 1103 in this embodiment, so the specific implementation of the transceiver 1103 is not described in detail any more.
[0742] The present application embodiment also provides a fourth STA, see Fig.12 , another structural diagram of the fourth STA1200 in an embodiment of the present application.
[0743] The fourth STA 1200 includes: a processor 1201 , a memory 1202 , and a transceiver 1203 .
[0744] The processor 1201, the memory 1202 and the transceiver 1203 are connected via buses respectively, and computer instructions are stored in the memory.
[0745] The foregoing Figure 1E The processing module 109 and the signal measuring module 110 in the embodiment may specifically be the processor 1201 in this embodiment, so the specific implementation of the processor 1201 is not described in detail. Figure 1E The transceiver module 108 may specifically be the transceiver 1203 in this embodiment, so the specific implementation of the transceiver 1203 is not described in detail.
[0746] The present application also provides a network management system. Fig.13 The network management system includes Figure 1B The server shown, Figure 1C The control devices shown, Figure 1D The AP and Figure 1E STA shown.
[0747] Figure 1B The server shown is used to perform the above Figure 3 In the embodiment shown, some or all of the steps performed by the first communication device are used to perform the above Figure 4A The server in the illustrated embodiment executes some or all of the steps.
[0748] Figure 1C The control device shown is used to perform the above Figure 4A In the illustrated embodiments, the control device executes part or all of the steps.
[0749] Figure 1D The AP shown is used to perform the above Figure 4A Part or all of the steps performed by the second AP in the described embodiments.
[0750] Figure 1E The STA shown is used to perform the above Figure 4A In the embodiment shown, the fourth STA performs steps 407b to 407e. Figure 3 and Figure 4A The relevant introduction of the illustrated embodiment will not be repeated here.
[0751] The present application also provides a network management system. Fig.14 The network management system includes Figure 2B The first AP and Figure 2C The second STA is shown.
[0752] Figure 2B The first AP shown is used to perform the above Figure 5A , Fig. 6A , Figure 7 and Fig. 8A In the illustrated embodiment, the first AP executes part or all of the steps.
[0753] Optionally, the network management system further includes a fourth AP, which is used to perform the above Figure 5A Some or all of the steps are performed by the fourth AP in the illustrated embodiment.
[0754] The fourth AP includes a fifth AP and a sixth AP. The fifth AP is used to perform the above Figure 5A In the embodiment shown in the figure, the fifth AP performs some or all of the steps, and the sixth AP is used to perform the above steps. Figure 5A In the illustrated embodiment, the sixth AP performs some or all of the steps.
[0755] Please refer to the above Figure 5A , Fig. 6A , Figure 7 and Fig. 8A The relevant introduction of the illustrated embodiment will not be repeated here.
[0756] The present application provides a computer program product including instructions, characterized in that when the computer program product is run on a computer, the computer is caused to execute the following Figure 3 , Figure 4A , Figure 5A , Fig. 6A , Figure 7 and Fig. 8A Any one of the implementation methods.
[0757] The present application provides a computer-readable storage medium including computer instructions. When the instructions are executed on a computer, the computer executes the following Figure 3 , Figure 4A , Figure 5A , Fig. 6A and Figure 7 Any implementation method.
[0758] The present application provides a chip device, including a processor, for calling a computer program or a computer instruction in the memory so that the processor executes the above Figure 3 , Figure 4A , Figure 5A , Fig. 6A , Figure 7 and Fig. 8A Any one of the implementations in .
[0759] Optionally, the chip device further includes a memory, which is used to store computer programs or computer instructions, etc. The chip device is composed of a chip, and may also include a chip and other discrete devices.
[0760] Optionally, the processor is coupled to the memory via an interface.
[0761] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0762] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.
[0763] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0764] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0765] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.
[0766] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims
1. A network management method, It is characterized in that The method comprises: The first communication device acquires first channel parameter information, where the first channel parameter information includes channel parameter information between a first AP in a first network and a STA within a signal coverage range of the first AP; The first communication device groups the first AP and the first STA according to the first channel parameter information to obtain a plurality of groups, where the first STA includes some or all of the STAs within the signal coverage of the first AP; The first communication device sends a grouping result of the multiple groups, where the grouping result is used to indicate the AP and STA included in each group of the multiple groups, and the grouping result is used by the first AP to adjust the connected STAs.
2. The method according to claim 1, It is characterized in that The first communication device is a server, the first AP includes all APs of the first network, and the first STA includes all STAs of the first network; the first communication device acquires first channel parameter information, including: The server receives the first channel parameter information sent by the control device, where the first channel parameter information includes channel parameter information between each AP of all APs in the first network and a STA within a signal coverage range of each AP, and the first network is a network controlled by the control device; The first communication device sends the grouping result of the plurality of groups, including: The server sends grouping results of the plurality of groups to the control device.
3. The method according to claim 2, It is characterized in that The first channel parameter information includes a channel matrix W, where the channel matrix W is a matrix of K rows by L columns, where K is the number of STAs included in the first network, and L is the number of APs included in the first network. w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, and j is an integer greater than or equal to 1 and less than or equal to L; The first communication device groups the first AP and the first STA according to the first channel parameter information to obtain a plurality of groups, including: The server calculates a first intermediate matrix D according to the channel matrix W. The first intermediate matrix D D 1 is a K-row by K-column matrix, D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All the off-diagonal elements of D are zero; 2 is a matrix of L rows by L columns, D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix of K rows by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero; The server determines a second intermediate matrix according to the channel matrix W and the first intermediate matrix D, each row vector in the second intermediate matrix corresponds to a network node, and the network node is a STA or an AP in the first network; The server clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result, where the first clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group in the multiple groups; The server determines the network nodes respectively included in the plurality of groups according to the row vectors of the plurality of clusters.
4. The method according to claim 3, It is characterized in that The channel matrix W is an edge weight matrix W of a weighted bipartite graph, the weighted bipartite graph includes first type vertices and second type vertices, the first type vertices represent STAs in the first network, the second type vertices represent APs in the first network, the elements of the edge weight matrix W are edge weights of the weighted bipartite graph, the edge weights of the weighted bipartite graph are channel gains between STAs and APs in the first network, and the first intermediate matrix is the degree matrix of the weighted bipartite graph.
5. The method according to claim 3 or 4, It is characterized in that The server determines a second intermediate matrix according to the channel matrix W and the first intermediate matrix D, including: The server determines a third intermediate matrix R according to the channel matrix W and the first intermediate matrix D, W T is the transpose of W; The server performs singular value decomposition on the third intermediate matrix R to obtain a left singular matrix U of the third intermediate matrix R and a right singular matrix V of the third intermediate matrix R; wherein the third intermediate matrix R=UΣV T , the Σ is a diagonal matrix, the V T is the transpose of the right singular matrix V; The main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the left singular matrix U, and the main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the right singular matrix V, where r is an integer greater than or equal to 1 and less than or equal to min(K, L), and min(K, L) refers to the minimum value of K and L; The server selects corresponding M column vectors from the left singular matrix U in descending order of the singular values in the diagonal matrix Σ to obtain a fourth intermediate matrix U M , and select corresponding M column vectors from the right singular matrix V in descending order of the singular values in the diagonal matrix Σ, to obtain the fifth intermediate matrix V M , where M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2 (K+L) means taking the logarithm of K+L with base 2; The server calculates the second intermediate matrix Z according to the fourth intermediate matrix, the fifth intermediate matrix and the first intermediate matrix, 6. The method according to claim 5, It is characterized in that The first communication device clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result, including: The server performs clustering on the second intermediate matrix Clustering the K+L row vectors to obtain row vectors of multiple clusters, where one cluster in the multiple clusters corresponds to one group in the multiple groups; The server determines the network nodes respectively included in the plurality of groups according to the row vectors of the plurality of clusters, including: The server divides the network nodes corresponding to the row vectors of the same cluster among the row vectors of the multiple clusters into the same group.
7. The method according to claim 1, It is characterized in that The first communication device is the first AP, and the first STA includes a STA that has established a connection with the first AP within a signal coverage range of the first AP; the method further includes: The first AP adjusts the STAs connected to the first AP according to the group to which the first AP belongs.
8. The method according to claim 7, It is characterized in that The first AP is the pth AP in the first network, the first channel parameter information includes the pth column vector of the channel matrix W, the channel matrix w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, j is an integer greater than or equal to 1 and less than or equal to L, p is an integer less than or equal to L, L is the number of APs included in the first network, and K is the number of STAs included in the first network; The first communication device groups the first AP and the first STA according to the first channel parameter information to obtain a plurality of groups, including: The first AP calculates the p+Kth main diagonal element b of the first intermediate matrix D according to the pth column vector of the channel matrix W. p ; Among them, the first intermediate matrix The D 1 is a matrix of K rows by K columns, The D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, and the D 1 The off-diagonal elements of are all zero; 2 is a matrix of L rows by L columns, The D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix multiplied by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero; The first AP obtains the first K main diagonal elements of the first intermediate matrix D; The first AP calculates the channel matrix W based on the pth column vector, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, each row vector of the sixth intermediate matrix Q corresponds to a network node, and the network node is an AP or STA in the first network. The i-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the i-th STA, and the K+j-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the j-th AP; The first AP clusters the target row vector according to a clustering algorithm to obtain a second clustering result; The target row vector includes the K+pth row vector corresponding to the first AP and the row vector corresponding to the first STA in the first K row vectors of the sixth intermediate matrix Q, and the second clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group of the multiple groups; The first AP determines the network nodes respectively included in the plurality of groups according to the row vectors of the plurality of clusters.
9. The method according to claim 8, It is characterized in that The first AP is based on the pth row vector of the channel matrix W, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, including: The first AP calculates the channel matrix W based on the pth column vector, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K main diagonal elements in the first intermediate matrix D determine the pth row vector of the third intermediate matrix R The second intermediate matrix The W T is the transpose of W; The first AP is based on a stochastic gradient descent algorithm and the pth row vector of the third intermediate matrix R. Determine the first matrix X M p row vectors and the second matrix Y M , the first matrix X M The similarity with the third matrix is greater than or equal to the first preset threshold, and the second matrix Y M The degree of similarity with the fourth matrix is greater than or equal to a second preset threshold; The third matrix is a matrix obtained by selecting M column vectors from a left singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the left singular matrix has a corresponding singular value, and the M column vectors of the third matrix are M column vectors selected from the left singular matrix in descending order of singular value size; The fourth matrix is a matrix obtained by selecting M column vectors from a right singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the right singular matrix has a corresponding singular value, and the M column vectors of the fourth matrix are M column vectors selected from the right singular matrix in descending order of singular value size; The M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2 (K+L) means finding the logarithm of K+L with base 2, and min(K, L) means taking the minimum value of K and L; The first AP is based on the p+Kth main diagonal element b of the first intermediate matrix D. p , the first K main diagonal elements in the first intermediate matrix D, the first matrix X M The pth row vector and the second matrix Y M Calculate the sixth intermediate matrix The first K row vectors and the K+pth row vector of .
10. The method according to claim 9, It is characterized in that The first AP clusters the target row vector according to a clustering algorithm to obtain a second clustering result, including: The first AP generates a first cluster center according to the first random seed, wherein the first cluster center includes cluster centers corresponding to the multiple clusters respectively, and each cluster corresponds to a cluster center; The first AP determines, according to the first cluster center, a cluster to which each row vector in the target row vector belongs, and obtains a row vector included in each cluster in the multiple clusters; The first AP determines the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters, including: When a first preset condition is met, the first AP divides the network nodes corresponding to the row vectors of the same cluster among the multiple clusters into the same group.
11. The method according to claim 10, It is characterized in that When the first preset condition is not met, the method further includes: The first AP calculates a first average vector corresponding to a row vector of each cluster in the plurality of clusters according to a gossip protocol; The first AP uses the first average vector of the row vector of each cluster in the multiple clusters as the cluster centers corresponding to the multiple clusters respectively, to obtain second cluster centers; The first AP determines, according to the second cluster center, a cluster to which each row vector in the target row vector belongs; When the second preset condition is met, the first AP divides the network nodes corresponding to the row vectors of the same cluster among the multiple clusters into the same group.
12. The method according to claim 10 or 11, It is characterized in that The first preset condition includes: The update times of the cluster centers of the clusters corresponding to the multiple groups are greater than or equal to a third preset threshold.
13. The method according to claim 10 or 11, It is characterized in that The first preset condition includes: The first convergence accuracy is less than or equal to the preset convergence accuracy; The first convergence accuracy is a ratio of an absolute value of a first difference to a first global error, and the first difference is a difference between the first global error and a preset initialization error; The first global error is the sum of first local errors respectively calculated by all APs in the first network; The first local error of the first AP includes a sum of errors corresponding to each of the multiple clusters determined by the first AP; The error corresponding to each cluster is the sum of errors between the row vectors included in each cluster and the first average vector corresponding to each cluster; the first average vector corresponding to each cluster is calculated by the first AP according to the gossip protocol.
14. The method according to claim 7, It is characterized in that The first AP adjusting, according to the group to which the first AP belongs, the STA connected to the first AP, includes: The first AP sends information about the group to which the first AP belongs, and the information about the group to which the first AP belongs is used by STAs in the first network to determine whether to access the first AP.
15. The method according to claim 14, It is characterized in that The information of the group to which the first AP belongs includes a first group number of the group to which the first AP belongs; The first AP sends information about a group to which the first AP belongs, including: The first AP sends a first beacon frame, where the first beacon frame carries the first group number.
16. The method according to claim 7, It is characterized in that The first AP adjusting, according to the group to which the first AP belongs, the STA connected to the first AP, includes: The first AP determines whether the second STA that has established a connection with the first AP belongs to the same group; If yes, the first AP continues to provide services for the second STA; If not, the first AP determines the group to which the second STA belongs, and sends first indication information to the third AP; The first indication information is used to indicate that the third AP provides a service for the second STA, and the third AP and the second STA belong to the same group.
17. The method according to claim 7, It is characterized in that The method further comprises: The first AP receives first grouping information sent by a fourth AP, where the first grouping information includes grouping information to which the fourth AP belongs and an identifier of the fourth AP, and the fourth AP includes an AP other than the first AP in the first network; The first AP determines, according to the first grouping information, that a fifth AP and the first AP belong to the same group, and the fifth AP is a partial AP in the fourth AP; The first AP determines, according to the identifier of the fifth AP, the identifier of the first AP and a preset AP priority rule, whether the priority of the first AP is higher than the priority of the fifth AP; If the priority of the first AP is higher than the priority of the fifth AP, the first AP selects a first channel by scanning channels, and adjusts the channel accessed by the first AP to the first channel; The first AP sends first channel allocation information to the fifth AP, where the first channel allocation information is used to instruct the fifth AP to access the first channel.
18. The method according to claim 17, It is characterized in that If the priority of the first AP is lower than the priority of the fifth AP, the method further includes: The first AP receives second channel allocation information sent by a sixth AP, where the second channel allocation information is information of a second channel selected by the sixth AP, and the sixth AP is an AP with the highest priority among the fifth APs; The first AP adjusts the channel accessed by the first AP to the second channel according to the second channel allocation information.
19. A first communication device, It is characterized in that The first communication device comprises: A transceiver module, configured to obtain first channel parameter information, where the first channel parameter information includes channel parameter information between a first AP in a first network and a STA within a signal coverage range of the first AP; a grouping module, configured to group the first AP and the first STA according to the first channel parameter information to obtain a plurality of groups, wherein the first STA includes some or all of the STAs within the signal coverage of the first AP; The transceiver module is used to send the grouping results of the multiple groups, where the grouping results are used to indicate the AP and STA included in each group of the multiple groups, and the grouping results are used by the first AP to adjust the connected STAs.
20. The first communication device according to claim 19, It is characterized in that The first communication device is a server, the first AP includes all APs of the first network, and the first STA includes all STAs of the first network; the first communication device acquires first channel parameter information, including: The server receives the first channel parameter information sent by the control device, where the first channel parameter information includes channel parameter information between each AP of all APs in the first network and a STA within a signal coverage range of each AP, and the first network is a network controlled by the control device; The first communication device sends the grouping result of the plurality of groups, including: The server sends grouping results of the plurality of groups to the control device.
21. The first communication device according to claim 20, It is characterized in that The first channel parameter information includes a channel matrix W, where the channel matrix W is a matrix of K rows by L columns, where K is the number of STAs included in the first network, and L is the number of APs included in the first network. w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, and j is an integer greater than or equal to 1 and less than or equal to L; The first communication device groups the first AP and the first STA according to the first channel parameter information to obtain a plurality of groups, including: The server calculates a first intermediate matrix D according to the channel matrix W. The first intermediate matrix D D 1 is a K-row by K-column matrix, D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, D 1 All the off-diagonal elements of D are zero; 2 is a matrix of L rows by L columns, D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix of K rows by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero; The server determines a second intermediate matrix according to the channel matrix W and the first intermediate matrix D, each row vector in the second intermediate matrix corresponds to a network node, and the network node is a STA or an AP in the first network; The server clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result, where the first clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group in the multiple groups; The server determines the network nodes respectively included in the plurality of groups according to the row vectors of the plurality of clusters.
22. The first communication device according to claim 21, It is characterized in that The channel matrix W is an edge weight matrix W of a weighted bipartite graph, the weighted bipartite graph includes first type vertices and second type vertices, the first type vertices represent STAs in the first network, the second type vertices represent APs in the first network, the elements of the edge weight matrix W are edge weights of the weighted bipartite graph, the edge weights of the weighted bipartite graph are channel gains between STAs and APs in the first network, and the first intermediate matrix is the degree matrix of the weighted bipartite graph.
23. The first communication device according to claim 21 or 22, It is characterized in that The server determines a second intermediate matrix according to the channel matrix W and the first intermediate matrix D, including: The server determines a third intermediate matrix R according to the channel matrix W and the first intermediate matrix D, W T is the transpose of W; The server performs singular value decomposition on the third intermediate matrix R to obtain a left singular matrix U of the third intermediate matrix R and a right singular matrix V of the third intermediate matrix R; wherein the third intermediate matrix R=UΣV T , the Σ is a diagonal matrix, the V T is the transpose of the right singular matrix V; The main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the left singular matrix U, and the main diagonal elements corresponding to the r-th column vector of the diagonal matrix Σ are the singular values corresponding to the r-th column vector of the right singular matrix V, where r is an integer greater than or equal to 1 and less than or equal to min(K, L), and min(K, L) refers to the minimum value of K and L; The server selects corresponding M column vectors from the left singular matrix U in descending order of the singular values in the diagonal matrix Σ to obtain a fourth intermediate matrix U M , and select corresponding M column vectors from the right singular matrix V in descending order of the singular values in the diagonal matrix Σ, to obtain the fifth intermediate matrix V M , where M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2 (K+L) means taking the logarithm of K+L with base 2; The server calculates the second intermediate matrix Z according to the fourth intermediate matrix, the fifth intermediate matrix and the first intermediate matrix, 24. The first communication device according to claim 21, It is characterized in that The first communication device clusters the row vectors of the second intermediate matrix according to a clustering algorithm to obtain a first clustering result, including: The server performs clustering on the second intermediate matrix Clustering the K+L row vectors to obtain row vectors of multiple clusters, where one cluster in the multiple clusters corresponds to one group in the multiple groups; The server determines the network nodes respectively included in the plurality of groups according to the row vectors of the plurality of clusters, including: The server divides the network nodes corresponding to the row vectors of the same cluster among the row vectors of the multiple clusters into the same group.
25. The first communication device according to claim 19, It is characterized in that The first communication device is the first AP, and the first STA includes STAs that have established connections with the first AP within the signal coverage of the first AP; the first AP adjusts the STAs connected to the first AP according to the group to which the first AP belongs.
26. The first communication device according to claim 25, It is characterized in that The first AP is the pth AP in the first network, the first channel parameter information includes the pth column vector of the channel matrix W, the channel matrix w ij is the channel gain between the i-th STA and the j-th AP, i is an integer greater than or equal to 1 and less than or equal to K, j is an integer greater than or equal to 1 and less than or equal to L, p is an integer less than or equal to L, L is the number of APs included in the first network, and K is the number of STAs included in the first network; The first communication device groups the first AP and the first STA according to the first channel parameter information to obtain a plurality of groups, including: The first AP calculates the p+Kth main diagonal element b of the first intermediate matrix D according to the pth column vector of the channel matrix W. p ; Among them, the first intermediate matrix The D 1 is a matrix of K rows by K columns, The D 1 The elements a on the main diagonal of i is the sum of the channel gains between the i-th STA in the first network and all APs in the first network, and the D 1 The off-diagonal elements of are all zero; 2 is a matrix of L rows by L columns, The D 2 The elements b on the main diagonal of j is the sum of the channel gains between the jth AP in the first network and all STAs in the first network, 0 K×L is a matrix multiplied by L columns, 0 K×L All elements in are zero, 0 L×K is a matrix of L rows by K columns, 0 L×K All elements in are zero; The first AP obtains the first K main diagonal elements of the first intermediate matrix D; The first AP calculates the channel matrix W based on the pth column vector, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, each row vector of the sixth intermediate matrix Q corresponds to a network node, and the network node is an AP or STA in the first network. The i-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the i-th STA, and the K+j-th row vector of the sixth intermediate matrix Q is the row vector corresponding to the j-th AP; The first AP clusters the target row vector according to a clustering algorithm to obtain a second clustering result; The target row vector includes the K+pth row vector corresponding to the first AP and the row vector corresponding to the first STA in the first K row vectors of the sixth intermediate matrix Q, and the second clustering result includes row vectors of multiple clusters, and one cluster in the multiple clusters corresponds to one group of the multiple groups; The first AP determines the network nodes respectively included in the plurality of groups according to the row vectors of the plurality of clusters.
27. The first communication device according to claim 26, It is characterized in that The first AP is based on the pth row vector of the channel matrix W, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K row vectors and the K+pth row vector of the sixth intermediate matrix Q are determined, including: The first AP calculates the channel matrix W based on the pth column vector, the p+Kth main diagonal element b of the first intermediate matrix D. p The first K main diagonal elements in the first intermediate matrix D and the first K main diagonal elements in the first intermediate matrix D determine the pth row vector of the third intermediate matrix R The second intermediate matrix The W T is the transpose of W; The first AP is based on a stochastic gradient descent algorithm and the pth row vector of the third intermediate matrix R. Determine the first matrix X M p row vectors and the second matrix Y M , the first matrix X M The similarity with the third matrix is greater than or equal to the first preset threshold, and the second matrix Y M The degree of similarity to the fourth matrix is greater than or equal to a second preset threshold; The third matrix is a matrix obtained by selecting M column vectors from a left singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the left singular matrix has a corresponding singular value, and the M column vectors of the third matrix are M column vectors selected from the left singular matrix in descending order of singular value size; The fourth matrix is a matrix obtained by selecting M column vectors from a right singular matrix obtained by singular value decomposition of the third intermediate matrix R, each column vector in the right singular matrix has a corresponding singular value, and the M column vectors of the fourth matrix are M column vectors selected from the right singular matrix in descending order of singular value size; The M is greater than or equal to log 2 (K+L) and an integer less than or equal to min(K, L), the log 2 (K+L) means finding the logarithm of K+L with base 2, and min(K, L) means taking the minimum value of K and L; The first AP is based on the p+Kth main diagonal element b of the first intermediate matrix D. p , the first K main diagonal elements in the first intermediate matrix D, the first matrix X M The pth row vector and the second matrix Y M Calculate the sixth intermediate matrix The first K row vectors and the K+pth row vector of .
28. The first communication device according to claim 27, It is characterized in that The first AP clusters the target row vector according to a clustering algorithm to obtain a second clustering result, including: The first AP generates a first cluster center according to the first random seed, wherein the first cluster center includes cluster centers corresponding to the multiple clusters respectively, and each cluster corresponds to a cluster center; The first AP determines, according to the first cluster center, a cluster to which each row vector in the target row vector belongs, and obtains a row vector included in each cluster in the multiple clusters; The first AP determines the network nodes respectively included in the multiple groups according to the row vectors of the multiple clusters, including: When a first preset condition is met, the first AP divides the network nodes corresponding to the row vectors of the same cluster among the multiple clusters into the same group.
29. The first communication device according to claim 28, It is characterized in that When the first preset condition is not met, the first AP calculates a first average vector corresponding to a row vector of each cluster in the multiple clusters according to a gossip protocol; The first AP uses the first average vector of the row vector of each cluster in the multiple clusters as the cluster centers corresponding to the multiple clusters respectively, to obtain second cluster centers; The first AP determines, according to the second cluster center, a cluster to which each row vector in the target row vector belongs; When the second preset condition is met, the first AP divides the network nodes corresponding to the row vectors of the same cluster among the multiple clusters into the same group.
30. The first communication device according to claim 28 or 29, It is characterized in that The first preset condition includes: The update times of the cluster centers of the clusters corresponding to the multiple groups are greater than or equal to a third preset threshold.
31. The first communication device according to claim 28 or 29, It is characterized in that The first preset condition includes: The first convergence accuracy is less than or equal to the preset convergence accuracy; The first convergence accuracy is a ratio of an absolute value of a first difference to a first global error, and the first difference is a difference between the first global error and a preset initialization error; The first global error is the sum of first local errors respectively calculated by all APs in the first network; The first local error of the first AP includes a sum of errors corresponding to each of the multiple clusters determined by the first AP; The error corresponding to each cluster is the sum of errors between the row vectors included in each cluster and the first average vector corresponding to each cluster; the first average vector corresponding to each cluster is calculated by the first AP according to the gossip protocol.
32. The first communication device according to claim 25, It is characterized in that The first AP adjusting, according to the group to which the first AP belongs, the STA connected to the first AP, includes: The first AP sends information about the group to which the first AP belongs, and the information about the group to which the first AP belongs is used by STAs in the first network to determine whether to access the first AP.
33. The first communication device according to claim 32, It is characterized in that The information of the group to which the first AP belongs includes a first group number of the group to which the first AP belongs; The first AP sends information about a group to which the first AP belongs, including: The first AP sends a first beacon frame, where the first beacon frame carries the first group number.
34. The first communication device according to claim 25, It is characterized in that The first AP adjusting, according to the group to which the first AP belongs, the STA connected to the first AP, includes: The first AP determines whether the second STA that has established a connection with the first AP belongs to the same group; If yes, the first AP continues to provide services for the second STA; If not, the first AP determines the group to which the second STA belongs, and sends first indication information to the third AP; The first indication information is used to indicate that the third AP provides a service for the second STA, and the third AP and the second STA belong to the same group.
35. The first communication device according to claim 25, It is characterized in that The first AP receives first grouping information sent by a fourth AP, where the first grouping information includes grouping information to which the fourth AP belongs and an identifier of the fourth AP, and the fourth AP includes an AP other than the first AP in the first network; The first AP determines, according to the first grouping information, that a fifth AP and the first AP belong to the same group, and the fifth AP is a partial AP in the fourth AP; The first AP determines, according to the identifier of the fifth AP, the identifier of the first AP and a preset AP priority rule, whether the priority of the first AP is higher than the priority of the fifth AP; If the priority of the first AP is higher than the priority of the fifth AP, the first AP selects a first channel by scanning channels, and adjusts the channel accessed by the first AP to the first channel; The first AP sends first channel allocation information to the fifth AP, where the first channel allocation information is used to instruct the fifth AP to access the first channel.
36. The first communication device according to claim 35, It is characterized in that If the priority of the first AP is lower than the priority of the fifth AP, the method further includes: The first AP receives second channel allocation information sent by a sixth AP, where the second channel allocation information is information of a second channel selected by the sixth AP, and the sixth AP is an AP with the highest priority among the fifth APs; The first AP adjusts the channel accessed by the first AP to the second channel according to the second channel allocation information.
37. A network management method, It is characterized in that The method comprises: The control device sends first channel parameter information to the server, where the first channel parameter information includes channel parameter information between each AP of all APs in the first network and a STA within the signal coverage range of each AP, where the first network is a network controlled by the control device; The control device receives a grouping result of a plurality of groups from a server, wherein the plurality of groups include all STAs and all APs in the first network; The control device determines, according to the grouping result, an AP and a STA included in each of the plurality of groups; The control device sends second grouping information, where the second grouping information includes information about groups to which the APs of the first network respectively belong and information about groups to which the STAs of the first network respectively belong.
38. The method according to claim 37, It is characterized in that The method further comprises: The control device allocates the same channel to APs in the same group among the multiple groups; The control device sends the second group information, including: The control device sends the second grouping information, where the second grouping information includes channel allocation information corresponding to the plurality of groups respectively.
39. The method according to claim 38, It is characterized in that The control device allocates the same channel to APs in the same group among the multiple groups, including: The control device allocates a corresponding channel to the APs in each of the multiple groups according to the first channel parameter information, wherein the control device allocates the same channel to the APs in the same group.
40. A control device, It is characterized in that The control device comprises: A transceiver module is used to send first channel parameter information to a server, wherein the first channel parameter information includes channel parameter information between each AP of all APs in a first network and a STA within a signal coverage range of each AP, wherein the first network is a network controlled by the control device; and receive grouping results of multiple groups from a server, wherein the multiple groups include all STAs and all APs in the first network; A grouping management module, configured to determine the AP and STA included in each of the plurality of groups according to the grouping result; The transceiver module is used to send second group information, where the second group information includes information about groups to which the APs of the first network respectively belong and information about groups to which the STAs of the first network respectively belong.
41. The control device according to claim 40, It is characterized in that The control device allocates the same channel to APs in the same group among the multiple groups, including: The control device allocates a corresponding channel to the APs in each of the multiple groups according to the first channel parameter information, wherein the control device allocates the same channel to the APs in the same group.
42. A network management method, It is characterized in that The method comprises: The second access point AP sends second channel parameter information to the control device, where the second channel parameter information includes channel parameter information between the second AP and the STA within the signal coverage range of the second AP; The second AP receives second grouping information sent by the control device, where the second grouping information includes information about groups to which the APs of the first network respectively belong and information about groups to which the STAs of the first network respectively belong; The second AP determines, according to the second group information, the group to which the second AP belongs; The second AP adjusts the STAs connected to the second AP according to the group to which the second AP belongs.
43. The method according to claim 42, It is characterized in that The second grouping information further includes third channel allocation information, and the third channel allocation information includes information about a third channel allocated by the control device to the second AP; the method further includes: The second AP adjusts the channel accessed by the second AP to the third channel according to the third channel allocation information.
44. The method according to claim 42 or 43, It is characterized in that The second AP adjusts the STA connected to the second AP according to the group to which the second AP belongs, including: The second AP sends information about the group to which the second AP belongs, where the information about the group to which the second AP belongs is used to instruct a STA in a first network to determine whether to access the second AP, where the first network is a network controlled by the control device.
45. The method according to claim 44, It is characterized in that The information of the group to which the second AP belongs includes a second group number of the group to which the second AP belongs; The second AP sends information about the group to which the second AP belongs, including: The second AP sends a second beacon frame, where the second beacon frame carries the second group number.
46. The method according to claim 42 or 43, It is characterized in that The second AP adjusts the STA connected to the second AP according to the group to which the second AP belongs, including: The second AP determines the STAs included in the group to which the second AP belongs; The second AP determines whether the second AP and a third STA belong to the same group, the third STA being a STA that has established a connection with the second AP; If yes, the second AP keeps providing services for the third STA; If not, the second AP determines a seventh AP according to the group to which the third STA belongs, and sends third indication information to the seventh AP; The third indication information is used to indicate that the seventh AP provides services for the third STA, and the seventh AP and the third STA belong to the same group.
47. A second access point AP, It is characterized in that The second AP includes: A transceiver module, configured to send second channel parameter information to a control device, wherein the second channel parameter information includes channel parameter information between the second AP and STAs within a signal coverage range of the second AP; receive second grouping information sent by the control device, wherein the second grouping information includes information about groups to which the APs of the first network respectively belong and information about groups to which the STAs of the first network respectively belong; A group management module is used to determine the group to which the second AP belongs according to the second group information; and adjust the STA connected to the second AP according to the group to which the second AP belongs.
48. The second AP according to claim 47, It is characterized in that The second grouping information also includes third channel allocation information, and the third channel allocation information includes information of a third channel allocated by the control device to the second AP; the second AP adjusts the channel accessed by the second AP to the third channel according to the third channel allocation information.
49. The second AP according to claim 47 or 48, It is characterized in that The second AP adjusts the STA connected to the second AP according to the group to which the second AP belongs, including: The second AP sends information about the group to which the second AP belongs, where the information about the group to which the second AP belongs is used to instruct a STA in a first network to determine whether to access the second AP, where the first network is a network controlled by the control device.
50. The second AP according to claim 49, It is characterized in that The information of the group to which the second AP belongs includes a second group number of the group to which the second AP belongs; The second AP sends information about the group to which the second AP belongs, including: The second AP sends a second beacon frame, where the second beacon frame carries the second group number.
51. The second AP according to claim 47 or 48, It is characterized in that The second AP adjusts the STA connected to the second AP according to the group to which the second AP belongs, including: The second AP determines the STAs included in the group to which the second AP belongs; The second AP determines whether the second AP and a third STA belong to the same group, the third STA being a STA that has established a connection with the second AP; If yes, the second AP keeps providing services for the third STA; If not, the second AP determines a seventh AP according to the group to which the third STA belongs, and sends third indication information to the seventh AP; The third indication information is used to indicate that the seventh AP provides services for the third STA, and the seventh AP and the third STA belong to the same group.
52. A computer-readable storage medium, It is characterized in that The method comprises computer instructions, which, when executed on a computer, cause the computer to execute the method as described in any one of claims 1 to 18, or cause the computer to execute the method as described in any one of claims 37 and claim 39, or cause the computer to execute the method as described in any one of claims 42 to 46.
