A method and device for rapid network reconstruction and routing maintenance based on link status prediction

By predicting the orientation and attitude information of nodes in a wireless network, predicting link information is formed, and the problem that the network topology reconstruction method in the prior art cannot accurately reflect the topology connection status is solved, and link stability and routing are achieved. promote.

CN118590959BActive Publication Date: 2024-11-15BEIJING INST OF REMOTE SENSING EQUIP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411003454.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-25
Publication Date
2024-11-15
Estimated Expiration
2044-07-25

AI Technical Summary

Technical Problem

The existing network topology reconstruction method simply uses orientation information and effective communication distance to cannot accurately reflect the topological connection status between nodes, resulting in inaccurate link status prediction.

Method used

In a wireless network with time division multiple access, by predicting the orientation information and attitude information of local nodes and neighbor nodes, the accurate probability of predicting attitude information is determined, and the link information of predicted neighbor nodes is formed, and the neighbor table and routing table are maintained based on this. , perform network topology prediction and routing maintenance.

Benefits of technology

It significantly enhances link stability, reduces the impact of node movement on network topology stability, ensures that the network always maintains efficient and stable operation, and supports more accurate routing and reliable maintenance and reconstruction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118590959B_ABST
    Figure CN118590959B_ABST
Patent Text Reader

Abstract

This specification provides a method and device for rapid network reconstruction and routing maintenance of link status prediction, which relates to the field of wireless communication networking. Predict the position information, posture information and position information of the local node and the neighboring node, and determine the accuracy probability of the posture information prediction; based on the local node prediction information and the posture information prediction accuracy probability, form the predicted neighboring node link information; judge the link building capability between the local node and the neighboring node; the local node and the neighboring node exchange position information, routing information and predicted neighboring node link information, and maintain the neighbor table and routing table of the local node; in the clustered network, the super cluster head controls the cluster heads of each cluster, and the cluster head of each cluster controls each node in the cluster to perform a priori information backup and full network routing maintenance. It solves the problem that the existing network topology reconstruction method simply uses the position information combined with the effective communication distance to determine the topological relationship, and cannot accurately reflect the topological connection status between nodes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This document relates to the field of wireless communication networking technology, and in particular to a method and device for rapid network reconstruction and routing maintenance based on link status prediction. Background Art

[0002] Due to their high mobility, aircraft such as drones and high-altitude platforms will have an impact on wireless communication links, resulting in rapid changes in network topology during networking. In order to ensure reliable communication between nodes, routing maintenance and network reconstruction capabilities are required. In view of the impact of high node mobility on network topology, the prediction results of link status are often used to predict changes in network topology, and quickly implement network maintenance and on-demand reconstruction. The link status characteristics can reflect the ability of a node to communicate wirelessly with other nodes, which is mainly related to the distance between the nodes. Therefore, in existing research, since the positional relationship between nodes can directly reflect the link status characteristics, the known node motion state information can be combined with the trajectory prediction algorithm to predict its orientation information, and then the topological connection between nodes can be determined based on the predicted orientation information and the communication distance to obtain the link status information.

[0003] However, dynamic nodes such as aircraft do not have the ability to communicate in the entire airspace due to constraints on antenna installation, that is, they only have communication capabilities in certain specific directions. Therefore, simply using the azimuth information combined with the topological relationship determined by the effective communication distance cannot accurately reflect the topological connection status between nodes, and thus cannot truly reflect the link status information.

[0004] Therefore, a network reconstruction and routing maintenance method is urgently needed to solve the problem that the existing network topology reconstruction method simply uses the topological relationship determined by the location information combined with the effective communication distance, and cannot accurately reflect the topological connection status between nodes. Summary of the invention

[0005] The present specification provides a method and device for rapid network reconstruction and routing maintenance based on link status prediction, which is used to solve the problem that the existing network topology reconstruction method simply uses the topological relationship determined by the position information combined with the effective communication distance, and cannot accurately reflect the topological connection status between nodes.

[0006] In a first aspect, this specification provides a method for rapid network reconstruction and routing maintenance for link state prediction, comprising:

[0007] In a wireless network using time division multiple access, a local node transmits and receives communications with neighboring nodes in a specific time slot, and predicts the location information, attitude information of the local node and the location information of the neighboring nodes, and determines the accuracy probability of the attitude information prediction;

[0008] Based on the local node's location information, predicted location information, attitude information, and the probability of accurate attitude information prediction, network topology prediction is performed to form predicted neighbor node link information;

[0009] According to the predicted local node position information, local node posture information, and neighbor node position information, the link building capability between the local node and the neighbor node is determined;

[0010] The local node exchanges location information, routing information, and predicted neighbor node link information with neighbor nodes, and maintains the neighbor table and routing table of the local node;

[0011] Based on the maintained neighbor table and routing table of the local node, in the clustered network, the super cluster head controls the cluster heads of each cluster, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and network-wide routing maintenance.

[0012] In the second aspect, the present specification provides a fast network reconstruction and routing maintenance device for link state prediction, including: a link state information prediction module, a predicted neighbor node link information formation module, a link building capability judgment module, an information interaction module, and a clustered full network update module, wherein:

[0013] The link state information prediction module is used in a wireless network using a time division multiple access method, where a node transmits and receives communications with a neighboring node in a specific time slot, and predicts the position information and posture information of the local node and the position information of the neighboring node, and determines the accuracy probability of the posture information prediction;

[0014] The predicted neighbor node link information forming module is used to perform network topology prediction based on the local node's position information, predicted position information, posture information and posture information prediction accuracy probability, and form predicted neighbor node link information;

[0015] The link building capability judgment module is used to judge the link building capability between the local node and the neighboring node based on the predicted position information and posture information of the local node and the position information of the neighboring node;

[0016] The information exchange module is used for exchanging position information, routing information and predicted neighbor node link information between the local node and the neighbor node, and maintaining the neighbor table and routing table of the local node;

[0017] The clustered full network update module is used to perform a priori information backup and full network routing maintenance based on the neighbor table and routing table of the maintained local node. In the clustered network, the super cluster head controls the cluster heads of each cluster, and the cluster head of each cluster controls each node in the cluster to perform a priori information backup and full network routing maintenance.

[0018] The beneficial effects of the present invention are as follows:

[0019] The present specification provides a method and device for rapid network reconstruction and routing maintenance of link state prediction, the method comprising: predicting the position information, posture information and position information of the local node and the neighboring node, determining the accuracy probability of the posture information prediction; forming the predicted neighboring node link information based on the local node prediction information and the posture information prediction accuracy probability; judging the link building capability between the local node and the neighboring node; exchanging the position information, routing information and the predicted neighboring node link information between the local node and the neighboring node, maintaining the neighbor table and routing table of the local node; in the clustered network, the super cluster head controls the cluster heads of each cluster based on the neighbor table and the routing table, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and full network routing maintenance. The method can accurately predict the direction and its own posture information when communicating with the neighboring node through the prediction of its own posture and position and the prediction of the position of the neighboring node, effectively predict the success rate of link building with the neighboring node, and then make control adjustments such as the antenna beam width in advance based on its own antenna communication capability, significantly enhance the stability of the link, greatly reduce the potential impact of node movement on the stability of the network topology, and ensure that the network always maintains efficient and stable operation. At the same time, a route maintenance and reconstruction method is proposed under a hierarchical network architecture, which uses the beam pointing and position information obtained by link state prediction obtained across layers to achieve rapid determination of new cluster heads when the network topology changes. This route maintenance and reconstruction method uses prediction accuracy information to quantify the accuracy of the route, supports more accurate route selection, and can also achieve reliable route maintenance and reconstruction under external navigation constraints. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings described herein are used to provide a further understanding of this specification and constitute a part of this specification. The illustrative embodiments and descriptions of this specification are used to explain this specification and do not constitute an improper limitation on this specification. In the drawings:

[0021] Figure 1 It is a schematic diagram of a fast network reconstruction and routing maintenance method for link state prediction provided in an embodiment of this specification;

[0022] Figure 2 is a schematic diagram of a beam width provided in an embodiment of this specification;

[0023] Figure 3 It is a schematic diagram of a fast network reconstruction and routing maintenance device for link state prediction provided in an embodiment of this specification. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solutions and advantages of this specification clearer, the technical solutions of this application will be clearly and completely described below in conjunction with the specific embodiments of this application and their corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this document.

[0025] The technical solutions provided by the embodiments of this specification are described in detail below in conjunction with the accompanying drawings. Specific embodiment 1

[0026] This embodiment provides a fast network reconstruction and routing maintenance method for link state prediction, see Figure 1 ;

[0027] First of all, it should be noted that this embodiment proposes to expand the correlation elements of link state information from position information to position information and attitude information for TDMA (Time Division Multiple Access) systems, and to predict the link state through the prediction results of position and attitude information, thereby supporting reliable routing maintenance and reconstruction. The routing maintenance and reconstruction process is applicable to hierarchical networks in clustering form, that is, there are cluster heads in the network, and the cluster heads are responsible for managing the entire network.

[0028] In a wireless network using TDMA, local nodes communicate with neighboring nodes in specific time slots, and exchange location information in different time slot resources while performing business interaction. After the nodes are networked, each node is in a high-mobility state. The fast and reliable network reconstruction and routing maintenance method based on link state prediction is used to achieve network reconstruction and routing maintenance. The specific processing steps are as follows:

[0029] Step 102: In a wireless network using a time division multiple access method, a local node performs transceiver communication with a neighboring node in a specific time slot, and predicts the position information and attitude information of the local node and the position information of the neighboring node, and determines the accuracy probability of the attitude information prediction;

[0030] The local node refers to any one of the communication nodes in the network, and the neighboring node refers to an adjacent node within one hop range of the local node;

[0031] Specifically, a specific implementation of step 102 may be:

[0032] S21, predicting the position information of the local node and neighboring nodes using a trajectory prediction algorithm based on the real-time measurement information;

[0033] S22, predicting the posture information of the local node;

[0034] S23, determining the accuracy probability of the predicted posture information according to the preset posture prediction method and prediction time;

[0035] The accuracy probability of the predicted posture information is the accuracy probability of the posture information prediction.

[0036] Specific example:

[0037] (1) Use trajectory prediction algorithms to predict the location information of the local node and its neighboring nodes;

[0038] This embodiment adopts the classic Kalman filtering method, and uses the kinematic equation and known measurement information to predict the node position. The known measurement information for predicting its own position comes from the navigation information, and the known measurement information for predicting the position of the neighboring node comes from the neighboring node position information sent by the neighboring node in the past.

[0039] (2) Predict the posture information of the local node itself

[0040] The prediction of posture information can be obtained by using the dynamic equations and kinematic equations of the local node, or by extrapolation based on the current posture information.

[0041] (3) Determination of parameters for accuracy of self-attitude prediction

[0042] According to the prediction accuracy of the selected posture prediction method, determine the accuracy probability of the predicted posture information The accuracy probability depends on the accuracy of the prediction algorithm and the prediction time. The accuracy probability definition and design can be carried out according to factors such as the actual prediction algorithm and application requirements. Here is a definition method for the accuracy probability. Assume that the maximum change rate of the known posture is , for the current time The time is calculated using the prediction algorithm. The posture changes after , the accurate probability of posture information can be expressed as .

[0043] It should be noted that in a highly dynamic network, nodes are highly mobile and network topology changes rapidly. By predicting its own posture and orientation and the orientation of neighboring nodes, it is possible to accurately predict the direction and its own posture information when communicating with neighboring nodes. This embodiment expands the prediction. In addition to the prediction of node position, there is also the prediction of node direction, which provides a basis for the subsequent prediction of the actual link connection status.

[0044] Step 104: Based on the position information of the local node, the predicted position information, the posture information and the probability of accurate posture information prediction, network topology prediction is performed to form predicted neighbor node link information;

[0045] The predicted neighbor node link information includes: the accuracy probability of posture information prediction, the topology effective time, the destination node identification address, and the last hop node address to the destination node.

[0046] Specifically, this embodiment provides a technical solution for forming predicted neighbor node link information, which can be specifically implemented as follows:

[0047] S41, determining the accuracy probability of posture information prediction based on its own posture information prediction result;

[0048] S42, using the prediction results of the position information of the local node and the neighboring nodes, as well as the accuracy probability of the local node's own posture information prediction, to form the predicted neighboring node link information;

[0049] It should be noted that each node will locally store neighbor tables and routing table entries containing information related to neighbor nodes, and the node maintains the neighbor table and routing table by exchanging control information such as location and routing. In this embodiment, each node will perform the following additional work while normally exchanging control information such as current location and routing information:

[0050] (1) Predictive information interaction

[0051] The node updates the topology information based on its own prediction information, and forms the updated topology to predict the neighboring node link information. The predicted neighboring node link information also includes its own posture information prediction accuracy probability , to characterize the accuracy of the prediction information;

[0052] Nodes communicate with each other periodically during their allocated network management time slots. The network topology is predicted by the neighboring nodes. The size can be determined based on the network scale and the interactive information carrying capacity. The specific information content is as follows:

[0053] Each node sends If the number of predicted neighbor node link information generated by itself is Greater than or equal to , select the one with the longest validity period If the number of predicted neighbor node link information generated by itself is Less than , among the predicted neighboring node link information sent by other nodes, select the one with the largest P_self The predicted neighbor node link information is randomly sent to the neighbor node.

[0054] The predicted neighbor node link information received by each node will be used to update the locally stored neighbor table and routing table entries.

[0055] In a hierarchical clustering network, the super cluster head, cluster head and ordinary nodes in the cluster send the predicted neighbor node link information at different frequencies. At the same time, the super cluster head, cluster head and ordinary nodes receive the predicted neighbor node link information sent by other nodes to complete the interaction of predicted neighbor node link information.

[0056] Among them, the frequency of sending the predicted neighbor node link information by ordinary nodes is the highest, the frequency of the cluster head is the lowest, and the frequency of sending the predicted neighbor node link information by the super cluster head is the lowest;

[0057] If the sending frequency of the super cluster head is , the maximum number of nodes in each level of cluster is , the transmission frequency of the cluster head is , the sending frequency of ordinary nodes is ,in Indicates approaching is less than integer, satisfying .frequency Determine based on actual network wireless resource conditions.

[0058] Based on this, by predicting neighboring node link information through interaction between nodes and nearby nodes, each node can effectively predict the future network topology based on this information and avoid some unnecessary topology adjustments.

[0059] Step 106: judging the link building capability between the local node and the neighboring node according to the predicted position information of the local node, the posture information of the local node, and the position information of the neighboring node;

[0060] The link building capability refers to that the local node and the neighboring node at least meet the communication distance constraint and the beam pointing constraint at the same time;

[0061] Communication capability refers to the ability of a node to transmit and receive signals with another node through its own antenna;

[0062] Specifically, a specific implementation of step 106 may be:

[0063] S61, predicting the communication distance between the local node and the neighboring node according to the predicted position information of the local node and the neighboring node, if the communication distance is less than or equal to the maximum transmission distance of the beam, the communication distance constraint between the local node and the neighboring node is satisfied;

[0064] S62, for the neighboring node that meets the communication distance constraint, determine the angle between the antenna normal and the neighboring node's orientation according to the predicted orientation information of the local node, the attitude information of the local node, and the position information of the neighboring node; if the angle of the angle is less than the maximum scanning angle of the antenna, it is determined that the communication between the local node and the neighboring node meets the beam pointing constraint;

[0065] S63. For neighboring nodes that meet the beam pointing constraint, the local node adjusts the beam width according to the neighboring node position information, the local node posture information, and the local node antenna transmission capability, and selects a beam with the best width for communication.

[0066] Furthermore, if the specific neighboring node does not have the ability to establish a link with the local node, the local node will no longer attempt to establish a link with the specific neighboring node at the next moment. At the same time, the local node will update the neighbor table and routing table entries, and realize service transmission with the specific neighboring node by cooperating with other neighboring nodes.

[0067] The specific neighboring node refers to any one of all neighboring nodes within one hop range of the local node.

[0068] Specific example:

[0069] The node will also evaluate its own communication capabilities based on the predicted orientation information and its own posture information to determine whether it has the ability to establish an efficient and stable wireless communication link with other neighboring nodes in the future. If a neighboring node meets the ability to establish a communication link with the local node, the local node will adjust the beam width based on factors such as location information, posture information, and transmission capacity to select a beam with the optimal width to ensure communication quality and reduce interference. If a neighboring node cannot meet the link establishment requirements with the local node, the local node will no longer attempt to establish a link with the neighboring node at the next moment. At the same time, the local node will update the neighbor table and routing table entries to find other available communication paths and meet the transmission requirements through collaboration and cooperation with other neighboring nodes. The method for determining the ability to establish a link with a neighboring node is as follows:

[0070] A node periodically evaluates the link stability of its neighboring nodes within one hop range.

[0071] 1) Communication distance determination

[0072] The node will sequentially consult the locally stored neighbor table, which records the predicted results of the one-hop node's position information, including longitude, latitude, and altitude. Then, the node calculates the communication distance with the node based on the haversine formula, which is as follows: (1)

[0073] in, Indicates the predicted longitude and latitude of the node. Represents the predicted longitude and latitude of the neighboring node, is the difference in latitude between two points, is the difference in longitude between two points, and 6378.137 is the equatorial radius of the earth, in kilometers.

[0074] Assume that the maximum transmission distance corresponding to the narrowest optional beam width is b. If S is greater than b, it means that the distance between them has exceeded the maximum communication range of the node itself and cannot meet the basic requirements of link establishment. If S is less than or equal to b, it means that the communication distance between the two nodes is within an acceptable range and has the potential to establish a link. In this case, the node will further judge the posture to comprehensively evaluate the feasibility of link establishment.

[0075] 2) Beam pointing determination

[0076] After meeting the communication distance limit, the node will further judge the beam pointing based on the attitude information to determine whether the node can configure the beam to accurately align the antenna array received by the neighboring node. According to the predicted own position, attitude and position of the neighboring node, the angle between the antenna normal and the azimuth of the neighboring node is determined. If the angle is less than the maximum scanning angle of the antenna, it is considered that communication with the node can be achieved, that is, at a certain moment in the future, the antenna can correctly point to the neighboring node. Otherwise, it is considered that communication with the node cannot be achieved.

[0077] 3) Beam width adjustment

[0078] After completing the communication capability evaluation based on azimuth and attitude information, if the two nodes have the ability to establish a communication link, the node will comprehensively consider the communication distance, attitude information, and the distribution of surrounding neighboring nodes to adaptively adjust the beam width. The beam width modulation strategy is designed to ensure that the signal can accurately cover the neighboring nodes while minimizing potential interference to other nodes.

[0079] See also Figure 2 , assuming the maximum beam coverage is First, the node determines the feasible domain of the adjustable beam width range, the lower limit of which is the narrowest beam width. . It is approximately assumed that there is a linear relationship between arc length and radius, that is, , it is approximately assumed that the maximum beam coverage for: ;

[0080] Therefore, the upper bound of the beam width is It can be approximately calculated by the following formula: , where L is the straight-line distance between nodes;

[0081] Assume that the shortest distance between the receiving node and the remaining nodes is At this time, it should be ensured that when the beam reaches the receiving node, its coverage range satisfy , which corresponds to the beam width Therefore, the actual adjustable bandwidth range of the node is .

[0082] According to different anti-interference performance requirements, by quantifying the required indicators, intelligent search algorithms such as particle swarm algorithm and ant colony algorithm can be used to conduct rapid search and determine the specific beam width.

[0083] Based on this, by setting constraints on the communication distance and beam pointing between the node and its neighboring nodes, the prediction of the link connection status between the node and its neighboring nodes is more accurate; based on its own communication capabilities, control adjustments such as antenna beam width are made in advance, which significantly enhances the stability of the link and greatly reduces the potential impact of node movement on the stability of the network topology, ensuring that the network always maintains efficient and stable operation.

[0084] Step 108: The local node exchanges position information, routing information, and predicted neighbor node link information with neighbor nodes to maintain the neighbor table and routing table of the local node.

[0085] Step 110: Based on the maintained neighbor table and routing table of the local node, in the clustered network, the super cluster head controls the cluster heads of each cluster, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and network-wide routing maintenance.

[0086] It should be noted that in a hierarchical network, a node is selected as a super cluster head to manage the entire network. The entire network is divided into several clusters, each cluster has a cluster head, and the cluster head is used to manage the member nodes in each cluster;

[0087] The super cluster head, the cluster head of each cluster, and the nodes of each cluster update the topology table at the MAC (Media Access Control) layer based on the link information. The MAC layer sends the lower layer topology update identifier to the network layer. The network layer triggers the routing update mechanism and updates the routing table based on the topology table.

[0088] Specifically, a specific implementation of step 110 may be:

[0089] S101, according to a preset accuracy probability threshold, the super cluster head and the cluster head of each cluster select a node whose prediction accuracy probability is higher than the accuracy probability threshold to perform a priori information backup;

[0090] For the clustered network under the hierarchical topology architecture, each cluster is equipped with a backup cluster head, and the cluster head exchanges information with the backup cluster head in a specific time slot.

[0091] Prior information backup includes the prior information backup between the cluster head and the backup cluster head and the prior information backup between nodes that exchange information. The prior information backup between the cluster head and the backup cluster head helps to achieve rapid reconstruction of the network, and the prior information backup between any nodes helps to improve the trajectory prediction accuracy of the nodes.

[0092] The cluster head and the selected backup cluster head communicate in a fixed communication time slot. Regardless of whether there is application layer data transmission, data packets will be packaged and sent according to a fixed format. In this communication time slot, the cluster head will also send network topology prediction information to the backup cluster head, including the position, speed, acceleration and other information of other members in the cluster connected to the current cluster head, other cluster heads in the network and the super cluster head obtained through interactive prediction information.

[0093] When any node in the network communicates in the allocated communication time slot, it will send its current location information and its predicted location information. The receiving node can improve the location prediction accuracy of the node based on this information.

[0094] When a node communicates with its neighboring nodes, it needs to predict the accurate probability based on its own posture information. Determine the antenna control method. , then the existing antenna control method is used to directly control the antenna. , then recalculate the antenna control mode according to the current attitude and adjust the beam pointing and beam width; is the preset accuracy probability threshold;

[0095] The position acquisition between different node pairs is asynchronous. The calculation conditions that the asynchronous acquisition time should meet are:

[0096] ① Calculate the escape time based on the node spacing and the maximum relative movement speed of the nodes as the upper bound of the maintenance cycle ;

[0097] ② Determine the maximum time for asynchronously obtaining local link messages based on the fixed communication time slot arrangement between nodes (The longest interaction interval between a pair of nodes).

[0098] The above time should meet ,according to Sure The time slot arrangement rules.

[0099] Based on this, the accuracy probability of the posture information prediction of each node is judged against the accuracy probability threshold. For nodes whose accuracy probability is lower than the threshold, no backup is performed, which improves the backup efficiency of super cluster heads and cluster heads and greatly saves overhead.

[0100] After executing step S101, this embodiment further provides a technical solution for determining the existence of a cluster head, which may specifically be:

[0101] According to the time slot arrangement rules, if the backup cluster head does not receive a network layer message packet from the cluster head during a periodic interaction until the end of the current communication time slot with the cluster head, the backup cluster head will determine that the cluster head is damaged and further set the identification bit to drive the physical layer and MAC layer to complete the re-establishment of the link.

[0102] Furthermore, this embodiment also provides a technical solution for establishing a new cluster head link, which may specifically be:

[0103] The MAC layer link establishment process is as follows:

[0104] ① The standby cluster head modifies the identity field in its own attributes and becomes the new cluster head;

[0105] ② According to the time slot arrangement rules, the backup cluster head sends a backup cluster head upgrade message to the nodes in the cluster in the time slot when the node communicates with other cluster members, informing them of the cluster head update. Because this message is sent and received in the corresponding time slot, the beam pointing mode is directional reception and directional transmission. The backup cluster head upgrade message mainly contains the address of the new cluster head after the upgrade and the message identification bit;

[0106] ③Complete the link communication according to the super cluster head location information obtained in the backup phase, establish a link with the super cluster head in the corresponding time slot and send the above backup cluster head upgrade message to inform the cluster head of the update.

[0107] Based on this, when a cluster head is damaged, a new cluster head needs to be enabled to build a link, ensuring the stability of the grid structure.

[0108] S102, after the topology set is initialized, according to the preset update and maintenance constraints, the nodes of each cluster update and maintain local topology items, generate and send local link messages, receive and forward local link messages, and update the routing table.

[0109] The update and maintenance constraints include:

[0110] If the validity time of the current topology item is within the preset time threshold range, and the predicted probability of the posture information in the neighboring node link information is greater than the preset accuracy threshold, the current topology item is updated; otherwise, the current topology item is not updated;

[0111] It should be noted that the data items included in the topology item constitute a topology table, which includes: the destination node identification address, the last hop node address to the destination node, the topology accuracy probability, the link support rate, and the topology validity time; wherein the topology accuracy probability is derived from the accuracy probability of the posture information prediction of the current node;

[0112] The routing table includes: a destination node identification address, a next hop address, a link support rate, and a link reliability probability;

[0113] The link reliability probability depends on the topology accuracy probability of each hop stored in the topology table.

[0114] Specific example:

[0115] After the link is established, the super cluster head, new cluster head and cluster members will each update the contents of the topology table in the MAC layer according to the new link information. The MAC layer will then send an identification message to the network layer, indicating that the lower layer topology has been updated. The network layer will then trigger the routing update mechanism to update the routing table according to the topology table.

[0116] The specific topology update method is as follows:

[0117] 1) Topology set initialization

[0118] The node maintains a topology set locally, which consists of several topology items, each of which represents a link determined by T_dest and T_last. The data items contained in the topology item are as follows:

[0119]

[0120] When at least one link exists, the topology set is initialized and the topology item corresponding to the local link is established. The values ​​are as follows:

[0121] a) T_dest: the address of the neighboring node connected to this node in one hop;

[0122] b) T_last: the address of this node;

[0123] c) P_accu: link accuracy probability, derived from the accuracy probability of posture information prediction of T_last node. If the value of P_accu is 0, it means that the current topology information is not derived from the predicted neighbor node link information;

[0124] d) T_speed: link supported rate;

[0125] f) T_seq: topology serial number, initially 0, incremented by 1 each time it is updated;

[0126] g) T_time: effective time, set to current time + periodic update time.

[0127] 2) Topology set maintenance

[0128] a) Update and maintain local topology items

[0129] The update and maintenance of local topology items are derived from the monitoring of neighboring nodes and the predicted neighboring node link information of the interaction. During the operation of the network, the node maintains the topology of T_last = the address of this node based on the monitored changes in the local link and the predicted neighboring node link information stored in itself:

[0130] The network management resource slice periodically checks the status of the neighboring nodes of the node at different rates.

[0131] If there is such a neighbor node, set the current time to , determine the current time and predicting neighbor node link information Size, where is a time increment. And predict the neighbor node link information When , the topology item update comes from the predicted neighbor node link information, and in other cases, the predicted neighbor node link information is not used (in this case, P_accu takes the value of 0). It can be determined based on network characteristics and should be greater than the update cycle;

[0132] When the neighboring node cannot be detected, if there is corresponding predicted neighboring node link information, the topology item is updated, otherwise the corresponding topology item is deleted;

[0133] When a new neighboring node is detected, a corresponding topological item is added, and the value is obtained as above;

[0134] The link supported rate changes: calculate and update T_speed;

[0135] T_seq: Each time an update is performed, the topology serial number of all local topology items is increased by 1;

[0136] In addition, the node will also use the acquired predicted neighbor node link information to maintain the topology of T_last = other nodes.

[0137] b) Generation and transmission of local link messages

[0138] During network operation, nodes need to periodically generate local link messages and send them to all one-hop neighbors at different rates.

[0139] Periodic mode: In the scanning frame, local link messages are sent at different rates to update the topology serial number and link rate. The format of the local link message packet is as follows:

[0140] 3) Receiving and forwarding local link messages

[0141] After receiving the local link message at different speeds, determine whether it is an "intra-cluster local link message" or an "inter-cluster local link message" based on the "routing type". Then compare the "source MAC address" of the local link message with the T_last of all topology items in the topology set of the corresponding speed:

[0142] a) If there is no topology item with T_last = "source MAC address", the message is forwarded to all one-hop neighboring nodes at the same rate gear, the message content remains unchanged, and a new topology is established and placed in the corresponding rate gear topology set:

[0143] ①T_dest = "one-hop neighbor address" of the local link message obtained;

[0144] ②T_last = the "source MAC address" of the local link message obtained;

[0145] ③P_accu=acquired local link accuracy probability;

[0146] ④T_speed=the "link rate" of the local link message obtained;

[0147] ⑤T_seq = the "serial number" of the local link message obtained;

[0148] ⑥T_time=current time+update cycle;

[0149] b) If there is a topology item with T_last = "source MAC address", compare the topology serial number in the local link message with that in the topology item:

[0150] ① If T_seq is less than the "serial number" of the local link message, the packet is discarded and not processed;

[0151] ② Otherwise, forward the message to all one-hop neighboring nodes and update the topology items at the same time. The update method is:

[0152] If T_dest = "one-hop neighbor address" of the local link message, then update P_accu, T_speed and T_seq according to the local link message, and T_time = current time + update period;

[0153] If T_dest is not equal to the "one-hop neighbor address" of the local link message, delete the item;

[0154] If there is no topology item with T_dest = "one-hop neighbor address" of the local link message, the method for creating a new one is the same as a).

[0155] 4) Routing table update and route selection

[0156] After completing the above topology table update process, the routing table should be updated accordingly. The specific routing update method is as follows:

[0157] First, the routing table is calculated, and then the path transmission is selected according to the table items corresponding to the different rates in the routing table. The routing table entry structure includes:

[0158]

[0159] The establishment of a routing table is to calculate the path to each node in the network. The node calculates the path to each node in the topology set by finding the spanning tree with the root node of the node. The algorithm for calculating the routing table of all other nodes is as follows:

[0160] a) Generate a routing entry for each topology item (neighbor link) in the topology set whose T_last is equal to the address of this node, and set the topology item to calculated. The value of the data item is as follows:

[0161] ①R_dest= T_dest;

[0162] ②R_next= T_dest;

[0163] ③R_speed= T_speed;

[0164] ④P_R depends on the P_accu value of each hop stored in the topology table. If the route goes through w hops to reach the destination node, the calculation method of P_R is:

[0165] b) Determine whether all routing items have been calculated. If so, the calculation process ends; otherwise, continue to execute a);

[0166] c) For the newly created routing item (called the original routing item) in the previous step, find the topological items whose T_last is R_dest and T_dest is not equal to the current node (with T_last as the relay), establish routes for the T_dest of these topological items, and set these topological items as calculated. The values ​​are as follows:

[0167] ①R_dest=T_dest;

[0168] ②R_next=R_next of the original routing item;

[0169] ③R_speed = min(original routing item R_speed, T_speed);

[0170] d) Go back to c) and iterate until all topological terms have been calculated;

[0171] e) Recalculate the routing table when a new topology item is added or T_speed changes;

[0172] f) Recalculate the routing table when a topology entry times out or is deleted.

[0173] Based on this, by adding the accurate probability constraint of posture information prediction in the constraint conditions for updating the current topology items, the accuracy of maintenance updates is further improved, and the topology information interaction overhead of highly dynamic networks is greatly saved, ensuring fast and stable updates of maintenance routes.

[0174] Furthermore, this embodiment also provides a technical solution for selecting an optimal route, which may specifically be:

[0175] When selecting routing for service, routing service can be provided according to the P_R of the routing of different services.

[0176] First, match the service type, and then determine the service carrying order. According to the transmission requirements of the service, it is divided into two categories: "high priority" and "low priority". For each type of service, it is arranged from large to small according to the P_R value of its corresponding route, and sent in sequence. After completing n high-priority service services, a low-priority service is provided. When the route is updated, the order of the two queues is updated at the same time.

[0177] When the route is updated, if there are multiple routing information for the same destination node, the route with the largest link reliability probability value is selected as the selected route;

[0178] The link reliability probability is used to describe the impact of the posture of the nodes in the routing table on the link reliability. A route with a high link reliability probability indicates that the reliability of the communication link is high when the information transmission passes through the node in the future when the posture changes, and the transmission success rate of the information transmitted through this route is higher.

[0179] In summary, this embodiment determines the accuracy probability of posture information prediction by predicting the local node's position information, posture information and the position information of the neighboring nodes; based on the local node prediction information and the accuracy probability of posture information prediction, the predicted neighboring node link information is formed; the link building capability between the local node and the neighboring nodes is judged; the local node and the neighboring nodes exchange position information, routing information and predicted neighboring node link information, and maintain the neighbor table and routing table of the local node; in the clustered network, the super cluster head controls the cluster heads of each cluster based on the neighbor table and routing table, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and full network routing maintenance. This method can accurately predict the direction and its own posture information when communicating with neighboring nodes through the prediction of its own posture and position and the prediction of the position of neighboring nodes, effectively predict the success rate of link building with neighboring nodes, and then make control adjustments such as antenna beam width in advance based on its own communication capability, significantly enhance the stability of the link, greatly reduce the potential impact of node movement on the stability of network topology, and ensure that the network always maintains efficient and stable operation. At the same time, a route maintenance and reconstruction method is proposed under a hierarchical network architecture, which uses the beam pointing and position information obtained by link state prediction obtained across layers to achieve rapid determination of new cluster heads when the network topology changes. This route maintenance and reconstruction method uses prediction accuracy information to quantify the accuracy of the route, supports more accurate route selection, and can also achieve reliable route maintenance and reconstruction under external navigation constraints. Specific embodiment 2

[0180] This embodiment provides a fast network reconstruction and routing maintenance method for link state prediction, which specifically includes:

[0181] This embodiment aims at the TDMA (Time Division Multiple Access) system, and proposes to expand the correlation elements of link state information from position information to position information and attitude information, and to predict the link state through the prediction results of position and attitude information, so as to support reliable route maintenance and reconstruction. The route maintenance and reconstruction process is applicable to a hierarchical network in the form of clustering, that is, there is a cluster head in the network, and the cluster head is responsible for managing the entire network.

[0182] In a wireless network using TDMA, nodes communicate with neighboring nodes in a specific time slot, and exchange location information in different time slot resources while performing business interaction. After the nodes are networked, each node is in a high-mobility state. The fast and reliable network reconstruction and routing maintenance method based on link state prediction is used to achieve network reconstruction and routing maintenance. The specific processing steps are as follows:

[0183] Step 1: Prediction of link status related information

[0184] (1) Use the prediction algorithm to predict the location information of itself and its neighboring nodes

[0185] Classical Kalman filtering and other methods can be used to predict the node's position using kinematic equations and known measurement information. When predicting its own position, the known measurement information comes from navigation information, and when predicting the position of a neighboring node, the known measurement information comes from the neighboring node's position information sent by the neighboring node in the past.

[0186] (2) Predict its own posture information

[0187] The prediction of posture information can be obtained by using the node's dynamic equations and kinematic equations, or by extrapolation based on the current posture information.

[0188] (3) Determination of parameters for accuracy of self-attitude prediction

[0189] According to the prediction accuracy of the selected posture prediction method, determine the accuracy probability of the predicted posture information The accuracy probability depends on the accuracy of the prediction algorithm and the prediction time. The accuracy probability definition and design can be carried out according to factors such as the actual prediction algorithm and application requirements. Here is a definition method for the accuracy probability. Assume that the maximum change rate of the known posture is , for the current time The time is calculated using the prediction algorithm. The posture changes after , The accurate probability of posture information can be expressed as .

[0190] Step 2: Predict information interaction and determine neighbor node link building capabilities

[0191] Similar to conventional routing methods, each node will store neighbor tables and routing table entries containing neighbor-related information locally, and nodes maintain neighbor tables and routing tables by exchanging control information such as location and routing. In the proposed method, each node will perform the following additional work while exchanging control information such as current location and routing information.

[0192] (1) Predictive information interaction

[0193] Nodes communicate with each other periodically during their allocated network management time slots. The network topology is predicted by the neighboring nodes. The size of can be determined according to the network scale and the carrying capacity of interactive information. The node updates the topology information based on its own prediction information, and forms the updated topology to predict the link information of neighboring nodes. The predicted link information of neighboring nodes also includes its own posture information prediction accuracy probability , to characterize the accuracy of the prediction information, the specific information content is as follows:

[0194]

[0195] Each node sends If the number of predicted neighbor node link information generated by itself is Greater than or equal to , select the one with the longest validity period If the number of predicted neighbor node link information generated by itself is Less than , among the predicted neighboring node link information sent by other nodes, select the one with the largest P_self The predicted neighbor node link information is randomly sent to the neighbor node.

[0196] The predicted neighbor node link information received by each node will be used to update the locally stored neighbor table and routing table entries.

[0197] (2) Determination of neighbor node link building capability

[0198] The node will also evaluate its own communication capabilities based on the predicted orientation information and its own posture information to determine whether it has the ability to establish an efficient and stable wireless communication link with other neighboring nodes in the future. If the neighboring node meets the ability to establish a communication link with the local node, the local node will adjust the beam width based on factors such as location information, posture information, and transmission capacity to select a beam with the optimal width to ensure communication quality and reduce interference. If the neighboring node cannot meet the link establishment requirements with the local node, the local node will no longer attempt to establish a link with the neighboring node at the next moment. At the same time, the local node will update the neighbor table and routing table entries to find other available communication paths and meet the transmission requirements through collaboration and cooperation with other nodes. The method for determining the link establishment capability with a neighboring node is as follows:

[0199] A node periodically evaluates the link stability of its neighboring nodes within one hop range.

[0200] 1) Communication distance determination

[0201] The node will sequentially consult the locally stored neighbor table, which records the predicted results of the one-hop node's position information, including longitude, latitude, and altitude. Then, the node calculates the communication distance with the node based on the haversine formula, which is as follows:

[0202] (1)

[0203] in, Indicates the longitude and latitude of the predicted node, Represents the predicted longitude and latitude of the neighboring node, is the difference in latitude between two points, is the difference in longitude between two points, and 6378.137 is the equatorial radius of the earth, in kilometers.

[0204] Assume that the maximum transmission distance corresponding to the narrowest optional beam width is b. If S is greater than b, it means that the distance between them has exceeded the maximum communication range of the node itself and cannot meet the basic requirements of link establishment. If S is less than or equal to b, it means that the communication distance between the two nodes is within an acceptable range and has the potential to establish a link. In this case, the node will further judge the posture to comprehensively evaluate the feasibility of link establishment.

[0205] 2) Beam pointing determination

[0206] After meeting the communication distance limit, the node will further determine the beam pointing based on the attitude information to determine whether the local node can configure the beam to accurately align the antenna array received by the neighboring node. According to the predicted own position, attitude and position of the neighboring node, the angle between the antenna normal and the azimuth of the neighboring node is determined. If the angle is less than the maximum scanning angle of the antenna, it is considered that communication with the node can be achieved, that is, at a certain moment in the future, the antenna can correctly point to the neighboring node. Otherwise, it is considered that communication with the node cannot be achieved.

[0207] 3) Beam width adjustment

[0208] After completing the communication capability evaluation based on azimuth and attitude information, if the two nodes have the ability to establish a communication link, the node will comprehensively consider the communication distance, attitude information, and the distribution of surrounding neighboring nodes to adaptively adjust the beam width. The beam width modulation strategy is designed to ensure that the signal can accurately cover the neighboring nodes while minimizing potential interference to other nodes.

[0209] See also Figure 2 , assuming that the maximum coverage of the beam is R. First, the node determines the feasible domain of the adjustable beam width range, and its lower bound is the narrowest beam width α . It is approximately assumed that there is a linear relationship between arc length and radius, that is, , it is approximately assumed that the maximum beam coverage for: ;

[0210] Therefore, the upper bound of the beam width is It can be approximately calculated by the following formula: ; Where L is the straight-line distance between two nodes.

[0211] Assume that the shortest distance between the receiving node and the remaining nodes is At this time, it should be ensured that when the beam reaches the receiving node, its coverage range satisfy , which corresponds to the beam width Therefore, the actual adjustable bandwidth range of the node is .

[0212] According to different anti-interference performance requirements, by quantifying the required indicators, intelligent search algorithms such as particle swarm algorithm and ant colony algorithm can be used to conduct rapid search and determine the specific beam width.

[0213] Step 3: Route maintenance and network reconstruction

[0214] For clustered networks under hierarchical topology, each cluster has a backup cluster head, and the cluster head exchanges information with the backup cluster head at a specific time slot. The maintenance and reconstruction of the network mainly includes four stages: prior information backup, cluster head existence judgment, new cluster head link establishment, and network-wide routing update. In addition, based on the maintained routing information, service-centric routing selection can be performed. In the prior information backup and network-wide routing update stages, the accuracy probability of posture information will be considered. The impact of the link is quantified to describe the accuracy of the link, thereby affecting the route construction and selection.

[0215] (1) Prior information backup

[0216] Prior information backup includes the prior information backup between the cluster head and the backup cluster head and the prior information backup between nodes that exchange information. The prior information backup between the cluster head and the backup cluster head helps to achieve rapid reconstruction of the network, and the prior information backup between any nodes helps to improve the trajectory prediction accuracy of the nodes.

[0217] The cluster head and the selected backup cluster head communicate in a fixed communication time slot. Regardless of whether there is application layer data transmission, data packets will be packaged and sent according to a fixed format. In this communication time slot, the cluster head will also send network topology prediction information to the backup cluster head, including the position, speed, acceleration and other information of other members in the cluster connected to the current cluster head, other cluster heads in the network and the super cluster head obtained through interactive prediction information.

[0218] When any node in the network communicates in the allocated communication time slot, it will send its current position information and the predicted position information of itself. The receiving node can improve the trajectory prediction accuracy of the node based on this information.

[0219] When a node communicates with its neighboring nodes, it needs to predict the correct probability based on its own posture information. Determine the antenna control method. , then use the antenna control method in step 2 to directly control the antenna. , then recalculate the antenna control mode according to the current attitude and adjust the beam pointing and beam width;

[0220] The position acquisition between different node pairs is asynchronous. The calculation conditions that the asynchronous acquisition time should meet are:

[0221] ① Calculate the escape time based on the node spacing and the maximum relative movement speed of the nodes as the upper bound of the maintenance cycle ;

[0222] ② Determine the maximum time for asynchronously obtaining local link messages based on the fixed communication time slot arrangement between nodes (The longest interaction interval between a pair of nodes).

[0223] The above time should meet ,according to Sure The time slot arrangement rules.

[0224] (2) Cluster head existence judgment

[0225] According to the time slot arrangement rules, if the backup cluster head does not receive a network layer message packet from the cluster head during a periodic interaction until the end of the current communication time slot with the cluster head, the backup cluster head will determine that the cluster head is damaged and further set the identification bit to drive the physical layer and MAC layer to complete the re-establishment of the link.

[0226] (3) New cluster head link establishment

[0227] The MAC layer link establishment process is as follows:

[0228] ① The standby cluster head modifies the identity field in its own attributes and becomes the new cluster head;

[0229] ② According to the time slot arrangement rules, the backup cluster head sends a backup cluster head upgrade message to the nodes in the cluster in the time slot when the node communicates with other cluster members, informing them of the cluster head update. Because this message is sent and received in the corresponding time slot, the beam pointing mode is directional reception and directional transmission. The backup cluster head upgrade message mainly contains the address of the new cluster head after the upgrade and the message identification bit;

[0230] ③Complete the link communication according to the super cluster head location information obtained in the backup phase, establish a link with the super cluster head in the corresponding time slot and send the above backup cluster head upgrade message to inform the cluster head of the update.

[0231] (4) Network-wide routing update

[0232] After the link is established, the super cluster head, new cluster head and cluster members will each update the contents of the topology table in the MAC layer according to the new link information. The MAC layer will then send an identification message to the network layer, indicating that the lower layer topology has been updated. The network layer will then trigger the routing update mechanism to update the routing table according to the topology table.

[0233] The specific topology update method is as follows:

[0234] 1) Topology set initialization

[0235] The node maintains a topology set locally, which consists of several topology items, each of which represents a link determined by T_dest and T_last. The data items contained in the topology item are as follows:

[0236]

[0237] When at least one link exists, the topology set is initialized and the topology item corresponding to the local link is established. The values ​​are as follows:

[0238] a) T_dest: the address of the neighboring node connected to this node in one hop;

[0239] b) T_last: the address of this node;

[0240] c) P_accu: link accuracy probability, derived from the accuracy probability of posture information prediction of T_last node. If the value of P_accu is 0, it means that the current topology information is not derived from the predicted neighbor node link information;

[0241] d) T_speed: link supported rate;

[0242] f) T_seq: topology serial number, initially 0, incremented by 1 each time it is updated;

[0243] g) T_time: effective time, set to current time + periodic update time.

[0244] 2) Topology set maintenance

[0245] a) Update and maintain local topology items

[0246] The update and maintenance of local topology items are derived from the monitoring of neighboring nodes and the predicted neighboring node link information of the interaction. During the operation of the network, the node maintains the topology of T_last = the address of this node based on the monitored changes in the local link and the predicted neighboring node link information stored in itself:

[0247] The network management resource slice periodically checks the status of the neighboring nodes of the node at different rates.

[0248] If there is such a neighboring node, let the current time be T_n, and determine the current and the T_time size of the predicted neighbor node link information, where is a time increment. And predict the neighbor node link information When , the topology item update comes from the predicted neighbor node link information, and in other cases, the predicted neighbor node link information is not used (in this case, P_accu takes the value of 0). It can be determined based on network characteristics and should be greater than the update cycle;

[0249] When the neighboring node cannot be detected, if there is corresponding predicted neighboring node link information, the topology item is updated, otherwise the corresponding topology item is deleted;

[0250] When a new neighboring node is detected, a corresponding topological item is added, and the value is obtained as above;

[0251] The link supported rate changes: calculate and update T_speed;

[0252] T_seq: Each time an update is performed, the topology serial number of all local topology items is increased by 1;

[0253] In addition, the node will also use the acquired predicted neighbor node link information to maintain the topology of T_last = other nodes.

[0254] b) Generation and transmission of local link messages

[0255] During network operation, nodes need to periodically generate local link messages and send them to all one-hop neighbors at different rates.

[0256] Periodic mode: In the scanning frame, local link messages are sent at different rates to update the topology serial number and link rate. The format of the local link message packet is as follows:

[0257]

[0258] 3) Receiving and forwarding local link messages

[0259] After receiving the local link message at different speeds, determine whether it is an "intra-cluster local link message" or an "inter-cluster local link message" based on the "routing type". Then compare the "source MAC address" of the local link message with the T_last of all topology items in the topology set of the corresponding speed:

[0260] a) If there is no topology item with T_last = "source MAC address", the message is forwarded to all one-hop neighboring nodes at the same rate gear, the message content remains unchanged, and a new topology is established and placed in the corresponding rate gear topology set:

[0261] ①T_dest = "one-hop neighbor address" of the local link message obtained;

[0262] ②T_last = the "source MAC address" of the local link message obtained;

[0263] ③P_accu=acquired local link accuracy probability;

[0264] ④T_speed=the "link rate" of the local link message obtained;

[0265] ⑤T_seq = the "serial number" of the local link message obtained;

[0266] ⑥T_time=current time+update cycle;

[0267] b) If there is a topology item with T_last = "source MAC address", compare the topology serial number in the local link message with that in the topology item:

[0268] ① If T_seq is less than the "serial number" of the local link message, the packet is discarded and not processed;

[0269] ② Otherwise, forward the message to all one-hop neighboring nodes and update the topology items at the same time. The update method is:

[0270] If T_dest = "one-hop neighbor address" of the local link message, then update P_accu, T_speed and T_seq according to the local link message, and T_time = current time + update period;

[0271] If T_dest is not equal to the "one-hop neighbor address" of the local link message, delete the item;

[0272] If there is no topology item with T_dest = "one-hop neighbor address" of the local link message, the method for creating a new one is the same as a).

[0273] 4) Routing table update and route selection

[0274] After completing the above topology table update process, the routing table should be updated accordingly. The specific routing update method is as follows:

[0275] First, the routing table is calculated, and then the path transmission is selected according to the table items corresponding to the different rates in the routing table. The routing table entry structure includes:

[0276]

[0277] The establishment of a routing table is to calculate the path to each node in the network. The node calculates the path to each node in the topology set by finding the spanning tree with the root node of the node. The algorithm for calculating the routing table of all other nodes is as follows:

[0278] a) Generate a routing entry for each topology item (neighbor link) in the topology set whose T_last is equal to the address of this node, and set the topology item to calculated. The value of the data item is as follows:

[0279] ①R_dest= T_dest;

[0280] ②R_next= T_dest;

[0281] ③R_speed= T_speed;

[0282] ④P_R depends on the P_accu value of each hop stored in the topology table. If the route goes through w hops to reach the destination node, the calculation method of P_R is: ;

[0283] b) Determine whether all routing items have been calculated. If so, the calculation process ends; otherwise, continue to execute a);

[0284] c) For the newly created routing item (called the original routing item) in the previous step, find the topological items whose T_last is R_dest and T_dest is not equal to the current node (with T_last as the relay), establish routes for the T_dest of these topological items, and set these topological items as calculated. The values ​​are as follows:

[0285] ①R_dest=T_dest;

[0286] ②R_next=R_next of the original routing item;

[0287] ③R_speed = min(original routing item R_speed, T_speed);

[0288] d) Go back to c) and iterate until all topological terms have been calculated;

[0289] e) Recalculate the routing table when a new topology item is added or T_speed changes;

[0290] f) Recalculate the routing table when a topology entry times out or is deleted.

[0291] Furthermore, this embodiment provides a method for selecting a route, which specifically includes:

[0292] When selecting routing for service, routing service can be provided according to the P_R of the routing of different services.

[0293] First, match the service type, and then determine the service carrying order. According to the transmission requirements of the service, it is divided into two categories: "high priority" and "low priority". For each type of service, arrange them from large to small according to the P_R value of the corresponding route, and send them in sequence. After completing n high-priority service services, a low-priority service is performed. When the route is updated, the order of the two queues is updated at the same time. For the same destination node, there may be multiple routing information, so the route with a large P_R value is preferentially selected as the selected route.

[0294] P_R describes the impact of the posture of the nodes in the routing table on the link reliability. A route with a large P_R indicates that the reliability of the communication link is high when the posture of the node changes in the future, so the transmission success rate of the information transmitted through this route is higher.

[0295] In summary, this embodiment determines the accuracy probability of posture information prediction by predicting the local node's position information, posture information and the position information of the neighboring nodes; based on the local node prediction information and the accuracy probability of posture information prediction, the predicted neighboring node link information is formed; the link building capability between the local node and the neighboring nodes is judged; the local node and the neighboring nodes exchange position information, routing information and predicted neighboring node link information, and maintain the neighbor table and routing table of the local node; in the clustered network, the super cluster head controls the cluster heads of each cluster based on the neighbor table and routing table, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and full network routing maintenance. This method can accurately predict the direction and its own posture information when communicating with neighboring nodes through the prediction of its own posture and position and the prediction of the position of neighboring nodes, effectively predict the success rate of link building with neighboring nodes, and then make control adjustments such as antenna beam width in advance based on its own communication capability, significantly enhance the stability of the link, greatly reduce the potential impact of node movement on the stability of network topology, and ensure that the network always maintains efficient and stable operation. At the same time, a route maintenance and reconstruction method is proposed under a hierarchical network architecture, which uses the beam pointing and position information obtained by link state prediction obtained across layers to achieve rapid determination of new cluster heads when the network topology changes. This route maintenance and reconstruction method uses prediction accuracy information to quantify the accuracy of the route, supports more accurate route selection, and can also achieve reliable route maintenance and reconstruction under external navigation constraints. Specific embodiment three

[0296] This embodiment provides a fast network reconstruction and routing maintenance device for link state prediction, see Figure 3 , including: a link state information prediction module 301, a predicted neighbor node link information formation module 302, a link building capability judgment module 303, an information interaction module 304, and a clustering network-wide update module 305, wherein:

[0297] The link state information prediction module 301 is used in a wireless network using a time division multiple access method, where a node transmits and receives communications with a neighboring node in a specific time slot, and predicts the position information and posture information of the local node and the position information of the neighboring node, and determines the probability of accurate prediction of the posture information;

[0298] The predicted neighbor node link information forming module 302 is used to perform network topology prediction based on the local node's position information, predicted position information, posture information and posture information prediction accuracy probability, and form predicted neighbor node link information;

[0299] The link building capability judgment module 303 is used to judge the communication link building capability between the local node and its neighboring node based on the predicted local node position information, the local node posture information, and the neighboring node position information;

[0300] The information interaction module 304 is used for exchanging position information, routing information and predicted neighbor node link information between the local node and neighbor nodes, and maintaining the neighbor table and routing table of the current node;

[0301] The clustering network-wide update module 305 is used in the clustering network. The super cluster head controls the cluster heads of each cluster based on the neighbor table and the routing table. The cluster head of each cluster controls each node in the cluster to perform prior information backup and network-wide routing maintenance.

[0302] Optionally, the link state information prediction module 301 is specifically used to:

[0303] Based on real-time measurement information, the trajectory prediction algorithm is used to predict the location information of the local node and neighboring nodes;

[0304] Predict the posture information of the local node;

[0305] According to the preset posture prediction method and prediction time, the accuracy probability of the predicted posture information is determined.

[0306] Optionally, the link building capability determination module 303 is specifically configured to:

[0307] Predict the communication distance between the local node and the neighboring node based on the predicted position information of the local node and the neighboring node. If the communication distance is less than or equal to the maximum transmission distance of the beam, the communication distance constraint is satisfied between the local node and the neighboring node. Determine the neighboring nodes that satisfy the communication distance constraint with the local node in turn.

[0308] For the neighboring node that meets the communication distance constraint, the angle between the antenna normal and the neighboring node's orientation is determined based on the predicted local node's orientation information, the local node's attitude information, and the neighbor's position information. If the angle is less than the maximum scanning angle of the current antenna, the neighboring node is determined to meet the beam pointing constraint.

[0309] For neighboring nodes that meet the beam pointing constraints, the local node adjusts the beam width based on the neighboring node location information, the local node attitude information, and the local node antenna transmission capability, and selects the beam with the optimal width for communication.

[0310] Optionally, the device further comprises: a link establishment requirement not being met module, wherein the link establishment requirement not being met module is configured to:

[0311] If a specific neighboring node does not have the ability to establish a communication link between local nodes, the local node will no longer attempt to establish a link with the specific neighboring node at the next moment. At the same time, the local node will update the neighbor table and routing table entries, and realize service transmission with the specific neighboring node by cooperating with other nodes.

[0312] Optionally, the predicted neighbor node link information includes: the accuracy probability of posture information prediction, the topology validity time, the destination node identification address, and the last hop node address to the destination node.

[0313] Optionally, in a hierarchical clustering network, the super cluster head, cluster head and common nodes in the cluster send predicted neighbor node link information at different frequencies, and at the same time, the super cluster head, cluster head and common nodes receive the predicted neighbor node link information sent by other nodes to complete the interaction of predicted neighbor node link information;

[0314] Among them, the frequency of sending the predicted neighbor node link information by ordinary nodes is the highest, and the frequency of sending the predicted neighbor node link information by the super cluster head is the lowest;

[0315] If the sending frequency of the super cluster head is , the maximum number of nodes in each level of cluster is , the transmission frequency of the cluster head is , the sending frequency of ordinary nodes is ,in Indicates approaching is less than integer, satisfying .

[0316] Optionally, the clustering network-wide updating module 305 is specifically used for:

[0317] According to a preset accuracy probability threshold, the super cluster head and the cluster head of each cluster select nodes with a predicted accuracy probability higher than the accuracy probability threshold to perform prior information backup;

[0318] After the topology set is initialized, according to the preset update and maintenance constraints, the nodes of each cluster update and maintain local topology items, generate and send local link messages, receive and forward local link messages, and update the routing table.

[0319] Optionally, the update and maintenance constraints include:

[0320] If the validity time of the current topology item is within the preset time threshold range, and the predicted probability of the posture information in the neighboring node link information is greater than the preset accuracy threshold, then the current topology item is updated;

[0321] The data items included in the topology items constitute a topology table, which includes: the destination node identification address, the last hop node address to the destination node, the topology accuracy probability, the link support rate, and the topology validity time; wherein the topology accuracy probability is derived from the accuracy probability of the posture information prediction of the current node;

[0322] The routing table includes: a destination node identification address, a next hop address, a link support rate, and a link reliability probability;

[0323] The link reliability probability depends on the topology accuracy probability of each hop stored in the topology table.

[0324] Optionally, when the route is updated, if there are multiple routing information for the same destination node, a route with a larger link reliability probability value is selected as the selected route;

[0325] The link reliability probability is used to describe the impact of the posture of the nodes in the routing table on the link reliability. A route with a high link reliability probability indicates that the reliability of the communication link is high when the information transmission passes through the node in the future when the posture changes, and the transmission success rate of the information transmitted through this route is higher.

[0326] In summary, this embodiment determines the accuracy probability of posture information prediction by predicting the local node's position information, posture information and the position information of the neighboring nodes; forms the predicted neighboring node link information based on the local node prediction information and the accurate probability of posture information prediction; determines the link building capability between the local node and the neighboring nodes; the local node and the neighboring nodes exchange position information, routing information and predicted neighboring node link information, and maintain the neighbor table and routing table of the local node; in the clustered network, the super cluster head controls the cluster heads of each cluster based on the neighbor table and routing table, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and full network routing maintenance. This method can accurately predict the direction and its own posture information when communicating with neighboring nodes through the prediction of its own posture and position and the prediction of the position of neighboring nodes, effectively predict the success rate of link building with neighboring nodes, and then make control adjustments such as antenna beam width in advance based on its own communication capability, significantly enhance the stability of the link, greatly reduce the potential impact of node movement on the stability of network topology, and ensure that the network always maintains efficient and stable operation. At the same time, a route maintenance and reconstruction method is proposed under a hierarchical network architecture, which uses the beam pointing and position information obtained by link state prediction obtained across layers to achieve rapid determination of new cluster heads when the network topology changes. This route maintenance and reconstruction method uses prediction accuracy information to quantify the accuracy of the route, supports more accurate route selection, and can also achieve reliable route maintenance and reconstruction under external navigation constraints.

[0327] The above description is only a preferred embodiment of this specification and is not intended to limit this specification. For those skilled in the art, this specification may have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification shall be included in the protection scope of this specification.

Claims

1. A fast network reconstruction and routing maintenance method for link state prediction, characterized in that: include: In a wireless network using time division multiple access, a local node transmits and receives communications with neighboring nodes in a specific time slot, and predicts the location information, attitude information of the local node and the location information of the neighboring nodes, and determines the accuracy probability of the attitude information prediction; Based on the local node's location information, predicted location information, attitude information, and the probability of accurate attitude information prediction, network topology prediction is performed to form predicted neighbor node link information; Based on the predicted position information, posture information of the local node and the position information of the neighboring nodes, the link establishment capability between the local node and the neighboring nodes is determined; If the local node and the neighboring node have the ability to establish a link, the local node and the neighboring node exchange location information, routing information and predicted neighboring node link information to maintain the local node's neighbor table and routing table; Based on the maintained neighbor table and routing table of the local node, in the clustered network, the super cluster head controls the cluster heads of each cluster, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and network-wide routing maintenance; The step of judging the link establishment capability between the local node and the neighboring node based on the predicted position information, posture information of the local node, and position information of the neighboring node includes: The communication distance between the local node and the neighboring node is predicted based on the predicted position information of the local node and the neighboring node. If the communication distance is less than or equal to the maximum transmission distance of the beam, the communication distance constraint between the local node and the neighboring node is satisfied. For neighboring nodes whose communication meets the communication distance constraint, the angle between the antenna normal and the neighboring node direction is determined based on the predicted local node position information, the local node posture information and the neighboring node position information. If the angle is less than the maximum scanning angle of the antenna, it is determined that the neighboring node meets the beam pointing constraint. For neighboring nodes whose communications meet the beam pointing constraints, the local node adjusts the beam width based on the predicted neighboring node's position information, the local node's posture information, and the local node's antenna transmission capability, and selects the beam with the optimal width for communication.

2. The method according to claim 1, characterized in that The predicting the position information, attitude information of the local node and the position information of the neighboring nodes, and determining the accuracy probability of the attitude information prediction, includes: Based on real-time measurement information, the trajectory prediction algorithm is used to predict the location information of the local node and neighboring nodes; Predict the posture information of the local node; According to the preset posture prediction method and prediction time, the accuracy probability of the predicted posture information is determined.

3. The method according to claim 1, characterized in that If the specific neighboring node does not have the ability to establish a link with the local node, the local node will no longer attempt to establish a link with the specific neighboring node at the next moment. At the same time, the local node will update the neighbor table and routing table entries, and realize service transmission with the specific neighboring node by cooperating with other neighboring nodes.

4. The method according to claim 1, characterized in that: The predicted neighbor node link information includes: the accuracy probability of posture information prediction, the topology effective time, the destination node identification address, and the last hop node address to the destination node.

5. The method according to claim 1, characterized in that In a hierarchical clustering network, the super cluster head, cluster head and ordinary nodes in the cluster send the predicted neighbor node link information at different frequencies. At the same time, the super cluster head, cluster head and ordinary nodes receive the predicted neighbor node link information sent by other nodes to complete the interaction of predicted neighbor node link information. Among them, the frequency of sending the predicted neighbor node link information by ordinary nodes is the highest, and the frequency of sending the predicted neighbor node link information by the super cluster head is the lowest; If the transmission frequency of the super cluster head is f max , the maximum number of nodes in each level of cluster is K, and the sending frequency of the cluster head is The sending frequency of ordinary nodes is 6. The method according to claim 1, characterized in that Based on the neighbor table and routing table of the maintained local node, in the clustered network, the super cluster head controls the cluster heads of each cluster, and the cluster head of each cluster controls each node in the cluster to perform prior information backup and network-wide routing maintenance, including: According to a preset accuracy probability threshold, the super cluster head and the cluster head of each cluster select nodes with a predicted accuracy probability higher than the accuracy probability threshold to perform prior information backup; After the topology set is initialized, according to the preset update and maintenance constraints, the nodes of each cluster update and maintain local topology items, generate and send local link messages, receive and forward local link messages, and update the routing table.

7. The method according to claim 6, characterized in that The update and maintenance constraints include: If the validity time of the current topology item is within the preset time threshold range, and the predicted probability of the posture information in the neighboring node link information is greater than the preset accuracy threshold, then the current topology item is updated; The data items included in the topology items constitute a topology table, which includes: the destination node identification address, the last hop node address to the destination node, the topology accuracy probability, the link support rate, and the topology validity time; wherein the topology accuracy probability is derived from the local node posture information prediction accuracy probability; The routing table includes: a destination node identification address, a next hop address, a link support rate, and a link reliability probability; The link reliability probability depends on the topology accuracy probability of each hop stored in the topology table.

8. The method according to claim 1, characterized in that When the route is updated, if there are multiple routing information for the same destination node, the route with the largest link reliability probability value is selected as the selected route; The link reliability probability is used to describe the impact of the posture of the nodes in the routing table on the link reliability. A route with a high link reliability probability indicates that the reliability of the communication link is high when the information transmission passes through the node in the future when the posture changes, and the transmission success rate of the information transmitted through this route is higher.

9. A fast network reconstruction and routing maintenance device for link state prediction, used to implement the method according to any one of claims 1 to 8, characterized in that: include: Link status information prediction module, predicted neighbor node link information formation module, link building capability judgment module, information interaction module, clustering full network update module, among which: The link state information prediction module is used in a wireless network using a time division multiple access method, where a node performs transceiver communication with a neighboring node in a specific time slot, and predicts the position information and posture information of the local node and the position information of the neighboring node, and determines the probability of accurate prediction of the posture information; The predicted neighbor node link information forming module is used to perform network topology prediction based on the local node's position information, predicted position information, posture information and posture information prediction accuracy probability, and form predicted neighbor node link information; The link building capability judgment module is used to judge the link building capability between the local node and the neighboring node based on the predicted position information and posture information of the local node and the position information of the neighboring node; The information exchange module is used for exchanging position information, routing information and predicted neighbor node link information between the local node and the neighbor node, and maintaining the neighbor table and routing table of the local node; The clustered full network update module is used to perform a priori information backup and full network routing maintenance based on the neighbor table and routing table of the maintained local node. In the clustered network, the super cluster head controls the cluster heads of each cluster, and the cluster head of each cluster controls each node in the cluster to perform a priori information backup and full network routing maintenance.