Method and system for link handover of vehicle-mounted access terminal based on wireless map

Abstract

The application discloses a wireless map-based link switching method and system for a vehicle-mounted access terminal, and the method comprises the following steps: acquiring LTE-M wireless characteristic information and position information of the vehicle-mounted access terminal, and constructing historical wireless link big data; constructing a wireless link map based on the historical wireless link big data of the vehicle-mounted access terminal by using a spectral clustering algorithm, and acquiring an output result of the spectral clustering algorithm; comparing and analyzing the output result of the spectral clustering algorithm with LTE-M network indexes, and performing link switching of the vehicle-mounted access terminal according to a comparison and analysis result. The system comprises an acquisition module, a clustering module and a switching module. By using the application, intelligent link switching of a rail transit vehicle-mounted access terminal can be performed according to a comparison result of real-time wireless link data and a wireless link map. The application can be widely applied to the technical field of wireless link switching.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless link switching, in particular to a link switching method and system for a vehicle-mounted access terminal based on a wireless map. BACKGROUND

[0002] The vehicle-mounted access terminal of rail transit can realize data communication under the LTE-M network, and can independently bear three train services of CBTC (metro train control), CCTV (video monitoring) and PIS (passenger information system). The vehicle-mounted access terminal is arranged at the front and rear of the train, and the two terminals realize hot backup and redundancy through the VRRP protocol, meet the redundancy requirements of signal system communication personnel, and under normal circumstances, both devices operate, one is the main device, and the other is the auxiliary device. The two devices use different and fixed wireless links for data transmission, data is forwarded by the main device to realize train-ground communication, when the main device fails or the link is disconnected, data is immediately forwarded by the auxiliary device to ensure that data will not be interrupted. The 1.8G frequency band of the current LTE-M network is an industry frequency band, which is not special for the subway and is easy to be interfered by the adjacent frequency of the operator base station. When the wireless link of the main device of the vehicle-mounted access terminal of rail transit cannot automatically switch to the wireless link of the auxiliary device due to external interference, the data transmission performance is reduced, which affects the train-ground communication and thus affects the safety of train operation. Moreover, the two vehicle-mounted access terminals on the train work in a hot backup and redundancy mode, and can only automatically switch the wireless link when the main device fails or the link is disconnected. When the data transmission performance of the wireless link of the main device is reduced, the wireless link cannot be automatically switched. SUMMARY

[0003] In order to solve the above technical problems, the purpose of the present application is to provide a link switching method and system for a vehicle-mounted access terminal based on a wireless map, which can intelligently switch the link of the vehicle-mounted access terminal of rail transit according to the comparison result of real-time wireless link data and wireless link map.

[0004] The first technical solution adopted by the present application is a link switching method for a vehicle-mounted access terminal based on a wireless map, comprising the following steps:

[0005] Obtain the LTE-M wireless feature information and position information of the vehicle-mounted access terminal, and construct historical wireless link big data;

[0006] Based on the historical wireless link big data of the vehicle-mounted access terminal, construct a wireless link map through a spectral clustering algorithm, and obtain the output result of the spectral clustering algorithm;

[0007] Compare and analyze the output result of the spectral clustering algorithm with the LTE-M network index, and switch the link of the vehicle-mounted access terminal according to the comparison and analysis result.

[0008] Further, the step of acquiring LTE-M wireless feature information and position information of the vehicle-mounted access terminal and constructing historical wireless link big data specifically includes:

[0009] Based on the pre-set center frequency and sampling rate, the vehicle-mounted access terminal is subjected to information acquisition and processing by the radio frequency acquisition module to obtain LTE-M wireless signals;

[0010] The LTE-M wireless signals are subjected to data conversion processing to obtain baseband data;

[0011] Based on the baseband decoding module, the baseband data is subjected to analysis processing according to the characteristics of the LTE-M standard to obtain LTE-M wireless feature information;

[0012] The position information of the vehicle-mounted access terminal is acquired by the inertial navigation module;

[0013] The LTE-M wireless feature information and the position information of the vehicle-mounted access terminal are integrated to construct historical wireless link big data.

[0014] Further, the LTE-M wireless feature information includes the reference signal receiving power of the SSS channel and the signal-to-interference-plus-noise ratio of the SSS channel, and the position information of the vehicle-mounted access terminal includes the longitude information of the vehicle-mounted access terminal and the latitude information of the vehicle-mounted access terminal.

[0015] Further, the input parameters of the wireless link graph include LTE-M wireless feature information , sample set size , generation method of similarity matrix, clustering method, dimension reduction dimension and clustering dimension , and the output parameter of the wireless link graph is a clustering set , i.e., the output result of the spectral clustering algorithm.

[0016] Further, the step of constructing a wireless link graph based on the historical wireless link big data of the vehicle-mounted access terminal by the spectral clustering algorithm and acquiring the output result of the spectral clustering algorithm specifically includes:

[0017] Based on any pair of data set points in the wireless link graph, the corresponding similarity value is acquired and a similarity matrix is generated according to the generation method of the similarity matrix;

[0018] According to the similarity matrix, an adjacency matrix and a degree matrix are constructed;

[0019] The adjacency matrix and the degree matrix are subjected to difference calculation processing to construct a Laplacian matrix;

[0020] The Laplacian matrix is subjected to standardization processing to obtain a standardized Laplacian matrix;

[0021] According to a feature vector calculation formula, a feature vector corresponding to the smallest eigenvalue of the normalized Laplacian matrix is calculated.

[0022] The feature vector is processed by row normalization to obtain a normalized feature vector.

[0023] The k-means clustering method is used to cluster each row in the normalized feature vector to obtain the output result of the spectral clustering algorithm.

[0024] Further, the step of constructing an adjacency matrix and a degree matrix according to the similarity matrix specifically includes:

[0025] The edge connection between the vertices in the wireless link graph is determined, and the weight of the edge connection of the vertices is defined.

[0026] The weight of the edge connection of the vertices with a similarity value greater than or equal to 0 is selected and integrated to construct the adjacency matrix.

[0027] Based on each point in the wireless link graph, the weight of all connected edges of each point is defined.

[0028] The weight of all connected edges of each point is integrated to construct the degree matrix.

[0029] Further, the expression of the feature vector calculation formula is specifically as follows:

[0030]

[0031] In the above formula, denotes the feature vector, denotes the normalized Laplacian matrix, denotes the weight of the edge connection of the vertices, , denotes the feature vector of the vertices, , denotes the degree of the vertices.

[0032] Further, the step of comparing and analyzing the output result of the spectral clustering algorithm with the LTE-M network index and switching the link of the vehicle-mounted access terminal according to the comparison and analysis result specifically includes:

[0033] The output result of the spectral clustering algorithm is compared and analyzed with the LTE-M network index to obtain a comparison and analysis result.

[0034] The comparison and analysis result is judged.

[0035] If the value of the comparison and analysis result is greater than or equal to a preset value, the link of the vehicle-mounted access terminal is switched. ​

[0036] If the value of the comparison analysis result is less than a preset value, it is determined that the link of the vehicle-mounted access terminal at the last moment is maintained.

[0037] Further, the requirements of the LTE-M network index are as follows:

[0038] The reference signal received power of the SSS channel is greater than -95dbm;

[0039] The signal-to-interference-plus-noise ratio of the SSS channel is greater than 3db.

[0040] The second technical solution adopted by the present application is: a vehicle-mounted access terminal link switching system based on a wireless map, comprising:

[0041] An acquisition module is configured to acquire LTE-M wireless feature information and position information of a vehicle-mounted access terminal, and construct historical wireless link big data;

[0042] A clustering module is configured to construct a wireless link map based on the historical wireless link big data of the vehicle-mounted access terminal, and acquire an output result of a spectral clustering algorithm;

[0043] A switching module is configured to compare and analyze the output result of the spectral clustering algorithm with the LTE-M network index, and perform vehicle-mounted access terminal link switching according to the comparison analysis result.

[0044] The method and system of the present application have the following advantages: the present application acquires LTE-M wireless feature information and position information of a vehicle-mounted access terminal in a subway wireless link, further constructs a wireless link map, compares real-time wireless link data with the wireless link map, analyzes whether the wireless link of the main device is normal, and performs intelligent switching of the rail transit vehicle-mounted access terminal link according to the analysis result, thereby making up for the deficiency of the existing switching condition of two main and standby vehicle-mounted access terminals, effectively improving the availability of the rail transit vehicle-mounted access terminal, and fully ensuring the safety of the vehicle-ground communication. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 is a step flow chart of the vehicle-mounted access terminal link switching method based on the wireless map of the present application;

[0046] Figure 2 is a structural block diagram of the vehicle-mounted access terminal link switching system based on the wireless map of the present application;

[0047] Figure 3 is a step flow chart of the intelligent switching of the rail transit vehicle-mounted access terminal link of the present application;

[0048] Figure 4 is a collection flow chart of the LTE-M wireless feature information and position information of the wireless link of the present application. DETAILED DESCRIPTION

[0049] The application will be described in further detail below with reference to the drawings and specific embodiments. For the step numbers in the following embodiments, they are only set for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0050] Referring to Figure 1 and Figure 3 , the application provides a wireless map-based vehicle-mounted access terminal link switching method, which comprises the following steps:

[0051] S1, acquiring LTE-M wireless feature information and position information of a vehicle-mounted access terminal, and constructing historical wireless link big data;

[0052] Specifically, referring to Figure 4 , two vehicle-mounted access terminals collect LTE-M wireless feature information and position information of their respective wireless links through a wireless detection module, accumulate historical wireless link big data, and the wireless detection module collects the LTE-M wireless feature information by a radio frequency collection module, collects LTE-M wireless signals according to a preset center frequency and sampling rate, and converts to obtain baseband data, and the baseband data is analyzed according to the characteristics of the LTE-M mode by a baseband decoding module to obtain the LTE-M wireless feature information.

[0053] The LTE-M wireless feature information comprises:

[0054] (1) SSS_RSRP: reference signal received power of the SSS channel;

[0055] (2) SSS_SINR: signal-to-interference-plus-noise ratio of the SSS channel;

[0056] The LTE-M wireless feature information is mainly used as a judgment condition for wireless link switching.

[0057] S2, constructing a wireless link map based on the historical wireless link big data of the vehicle-mounted access terminal by a spectral clustering algorithm, and obtaining an output result of the spectral clustering algorithm;

[0058] S21, constructing a wireless link map;

[0059] Specifically, based on the historical wireless link big data of the two vehicle-mounted access terminals, a spectral clustering algorithm is used to construct a wireless link map, and the input of the wireless link map is to divide the subway track line into a grid according to a length of 20 meters, and the LTE-M wireless feature information collected in the grid , the size of the sample set , the generation method of the similarity matrix, the clustering method, and the dimension reduction dimension Clustering Dimension Its output is a cluster set. ;

[0060] S22. Based on the historical wireless link big data of the vehicle access terminal, obtain the output results of the spectral clustering algorithm;

[0061] Specifically, the similarity matrix of the samples is constructed based on the similarity matrix generation method of the input. ;

[0062] (1) For any pair of data points ,remember For their similarity, Similarity matrix The structure is as follows:

[0063]

[0064] (2) Based on the similarity matrix Constructing an adjacency matrix Construct the degree matrix ;

[0065] In a similarity graph, each vertex represents a data point. Two vertices , There are edges connecting them and when Define the weight of the edge. , each When combined, we can obtain adjacency matrix The adjacency matrix It is expressed as follows:

[0066]

[0067] For any point in the graph Its degree Defined as the sum of the weights of all edges connected to it, its calculation formula is as follows:

[0068]

[0069] In the above formula, Represent any point The degree, Indicates the weight of the vertex edge connection;

[0070] each point When combined, we can obtain degree matrix It is a diagonal matrix with values ​​only on the main diagonal, and its expression is:

[0071]

[0072] (3) Further calculate the Laplacian matrix, the calculation formula of the Laplacian matrix is as follows:

[0073]

[0074] In the above formula, represents the Laplacian matrix;

[0075] (4) Construct the normalized Laplacian matrix, the calculation formula of the normalized Laplacian matrix is as follows:

[0076]

[0077] In the above formula, represents the normalized Laplacian matrix;

[0078] (5) Calculate the smallest eigenvalue respectively corresponding to the eigenvector , the calculation formula is as follows:

[0079]

[0080] In the above formula, represents the eigenvector, represents the normalized Laplacian matrix, represents the weight of the vertex edge connection, , represents the eigenvector of the vertex, , represents the degree of the vertex;

[0081] For the smallest eigenvalue respectively corresponding to the eigenvector, wherein the smallest Laplacian matrix represents that the clustering calculation requirement is met, then the spatial characteristics of the clustering can be found through the output eigenvector, and the number of corresponding clusters is determined, and the subsequent K-means clustering model is applied to calculate the corresponding clusters;

[0082] (6) The matrix composed of each eigenvector is normalized by row to obtain an eigenvector matrix of dimension;

[0083] (7) Each row in is taken as a dimensional sample, and there are ​a sample is clustered using k-means, the clustering dimension is the input of ;

[0084] (8) get cluster partition .

[0085] S3, compare and analyze the spectrum clustering algorithm output result with the LTE-M network index, and perform vehicle access terminal link switching according to the comparison and analysis result.

[0086] Specifically, the spectrum clustering algorithm output result is compared with the LTE-M network index requirement, whether the main device wireless link feature information is normal is analyzed, and the vehicle access terminal link intelligent switching is performed according to the analysis result;

[0087] If the number of values in the cluster set that do not meet the LTE-M network index requirement >= 70%, it is determined that the feature information of the main device wireless link of the track grid is abnormal, and the wireless link of the auxiliary device is directly switched when the next train passes;

[0088] If the number of values in the cluster set that do not meet the LTE-M network index requirement < 70%, it is determined that the feature information of the main device wireless link of the track grid is normal, and the wireless link of the main device is continued to be used when the next train passes.

[0089] The requirements of the LTE-M network index include SSS_RSRP: >-95dbm and SSS_SINR: >3db.

[0090] Referring to Figure 2 , the vehicle access terminal link switching system based on a wireless atlas comprises:

[0091] An acquisition module is configured to acquire LTE-M wireless feature information and position information of a vehicle access terminal, and construct historical wireless link big data.

[0092] A clustering module is configured to construct a wireless link atlas based on the historical wireless link big data of the vehicle access terminal by using a spectrum clustering algorithm, and acquire a spectrum clustering algorithm output result.

[0093] A switching module is configured to compare and analyze the spectrum clustering algorithm output result with the LTE-M network index, and perform vehicle access terminal link switching according to the comparison and analysis result.

[0094] The contents in the method embodiments are applicable to the system embodiments, the system embodiments specifically realize the same functions as the method embodiments, and achieve the same beneficial effects as the method embodiments.

[0095] The above is a specific description of the preferred embodiment of the application, but the application is not limited to the described embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the application, and these equivalent modifications or replacements are all included in the scope defined by the claims of the present application.

Claims

1. A method for link handover of a vehicle-mounted access terminal based on wireless graphs, characterized in that, The method comprises the following steps: acquiring LTE-M wireless feature information and position information of the vehicle-mounted access terminal, and constructing historical wireless link big data; constructing a wireless link graph based on the historical wireless link big data of the vehicle-mounted access terminal through a spectral clustering algorithm, and acquiring an output result of the spectral clustering algorithm; comparing and analyzing the output result of the spectral clustering algorithm with LTE-M network indexes, and performing vehicle-mounted access terminal link switching according to a comparison and analysis result; the step of acquiring the LTE-M wireless feature information and the position information of the vehicle-mounted access terminal, and constructing the historical wireless link big data specifically comprises: acquiring LTE-M wireless signal through a radio frequency acquisition module based on a pre-set center frequency and sampling rate; performing data conversion processing on the LTE-M wireless signal to obtain baseband data; performing analysis processing on the baseband data based on a baseband decoding module according to the characteristics of the LTE-M mode to obtain the LTE-M wireless feature information; acquiring the position information of the vehicle-mounted access terminal through an inertial navigation module; integrating the LTE-M wireless feature information and the position information of the vehicle-mounted access terminal to construct the historical wireless link big data; the step of constructing the wireless link graph based on the historical wireless link big data of the vehicle-mounted access terminal through the spectral clustering algorithm, and acquiring the output result of the spectral clustering algorithm specifically comprises: acquiring a corresponding similarity value based on any pair of data set points in the wireless link graph, and generating a similarity matrix according to a generation mode of the similarity matrix; Similarity matrix The expression of the similarity matrix S is given by: ； constructing an adjacency matrix and a degree matrix according to the similarity matrix; performing difference calculation processing on the adjacency matrix and the degree matrix to construct a Laplacian matrix; performing standardization processing on the Laplacian matrix to obtain a standardized Laplacian matrix; According to the eigenvector calculation formula, an eigenvector corresponding to the smallest eigenvalue of the normalized Laplacian matrix is calculated ​ performing row-wise standardization processing on a feature vector to obtain a standardized feature vector; performing clustering processing on each row in the standardized feature vector through a k-means clustering method to obtain the output result of the spectral clustering algorithm.

2. The method of claim 1, wherein, The LTE-M wireless feature information comprises reference signal received power of an SSS channel and signal-to-interference-plus-noise ratio of the SSS channel, and the position information of the vehicle-mounted access terminal comprises longitude information of the vehicle-mounted access terminal and latitude information of the vehicle-mounted access terminal.

3. The method of claim 2, wherein, The input parameters of the wireless link graph include LTE-M wireless feature information , sample set size , generation mode of the similarity matrix, clustering method, dimension reduction dimension and clustering dimension , and the output parameters of the wireless link graph are a clustering set , that is, the output result of the spectral clustering algorithm.

4. The method of claim 3, wherein, The step of constructing the adjacency matrix and the degree matrix according to the similarity matrix specifically comprises: determining edge connections between vertices in the wireless link graph, and defining weights of the vertex edge connections; selecting and integrating the weights of the vertex edge connections with a similarity value greater than or equal to 0 to construct the adjacency matrix; defining weights of all edges connected to each point in the wireless link graph based on each point; integrating the weights of all edges connected to each point to construct the degree matrix.

5. The method of claim 4, wherein, The expression of the feature vector calculation formula is specifically as follows: In the above formula, denotes the eigenvector, denotes the normalized Laplacian matrix, denotes the weight of the vertex edge connection, denotes the eigenvector of the vertex, denotes the degree of the vertex.​​ 6. The method of claim 5, wherein the method further comprises: The step of comparing and analyzing the output result of the spectral clustering algorithm with the LTE-M network indexes, and performing vehicle-mounted access terminal link switching according to the comparison and analysis result specifically comprises: comparing and analyzing the output result of the spectral clustering algorithm with the LTE-M network indexes to obtain a comparison and analysis result; judging the comparison and analysis result; It is judged that the value of the comparison analysis result is greater than or equal to a preset value, and the vehicle-mounted access terminal link is switched; It is judged that the value of the comparison analysis result is less than the preset value, and the vehicle-mounted access terminal link at the last moment is maintained.

7. The method of claim 6, wherein the method further comprises: The requirements of the LTE-M network index are as follows: The reference signal receiving power of the SSS channel is greater than -95dbm; The signal-to-interference-plus-noise ratio of the SSS channel is greater than 3db.

8. A vehicle-mounted access terminal link switching system based on wireless graphs, characterized in that, The method for performing the vehicle-mounted access terminal link switching based on the wireless map according to claim 1 comprises the following modules: An acquisition module is configured to acquire LTE-M wireless feature information and position information of a vehicle-mounted access terminal, and construct historical wireless link big data; A clustering module is configured to construct a wireless link map based on the historical wireless link big data of the vehicle-mounted access terminal through a spectral clustering algorithm, and acquire an output result of the spectral clustering algorithm; A switching module is configured to compare and analyze the output result of the spectral clustering algorithm with the LTE-M network index, and perform vehicle-mounted access terminal link switching according to the comparison analysis result.

Images