Interference elimination algorithm in railway communication
By adopting a nested interference cancellation algorithm of regular multivariate decomposition and a regular multivariate alternating least squares algorithm of blind decomposition, the signal loss problem caused by strong interference in railway loop communication is solved, and higher signal reception quality and communication system performance are achieved.
Patent Information
- Application Number
- CN202510423096.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-07
- Publication Date
- 2025-06-20
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In railway loop communication, strong interference signals can easily cause the receiving equipment to lose useful signal information, and may even cause train collision accidents.
The nested interference cancellation algorithm of regular multivariate decomposition and the regular multivariate alternating least squares algorithm of blind decomposition are used to jointly estimate user symbols and channel state information to eliminate fixed interference and intentional interference through the multivariate decomposition interference cancellation algorithm and singular value decomposition algorithm.
Effectively eliminate interference in complex communication environments, improve signal reception quality and reliability, reduce bit error rates, and improve the overall performance of the communication system.
Smart Images
Figure CN120185975A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of information acquisition, and particularly relates to an interference cancellation algorithm in railway communication. The algorithm includes a nested interference cancellation algorithm based on regular multi-factor decomposition, a multi-factor decomposition interference cancellation algorithm, and a regular multi-factor alternating least squares algorithm based on blind decomposition. Background Art
[0002] Railway loop communication relies on the processing gain of spread spectrum signals to suppress interference. When a strong interference signal acts on the vehicle-mounted receiving device, the receiving device is easily affected by interference and loses useful signal information. In severe cases, it may even cause a train collision accident. To address the above problems, the regular multi-factor alternating least squares algorithm based on blind decomposition can jointly estimate user symbols and channel state information without relying on pilot sequences. The nested interference cancellation algorithm based on regular multi-factor decomposition can avoid the cumulative error effect caused by training sequences for channel estimation. Summary of the Invention
[0003] The purpose of the present invention is to provide an interference cancellation algorithm in railway communication to solve problems such as difficult information acquisition in the above background art.
[0004] The present invention provides an interference cancellation algorithm in railway communication, including the following steps:
[0005] Step 1, randomly initialize and
[0006] Step 2, set the rank of the tensor to M, substitute the matrix factors initialized in Step 1 and M into the multi-factor decomposition interference cancellation algorithm, and obtain
[0007] Step 3, expand the obtained by the previous iteration into a matrix to get
[0008] Step 4, randomly initialize and
[0009] Step 5, set the rank of the tensor Y sr to R, substitute the matrix factors initialized in Step 4 and R into the multi-factor decomposition interference cancellation algorithm, and obtain
[0010] Step 6, obtain
[0011] by the previous iteration
[0012] Among them, the steps of the multivariate decomposition interference elimination algorithm are as follows:
[0013] Step 1: Use interference cancellation method to obtain the received signal tensor
[0014] Step 2: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the signal tensor The corresponding rank-one matrix factorization factor
[0015] Sub: {h 1 ,a 1 ,s 1};
[0016] Step 3, by {h 1 ,a 1 ,s 1}Reconstruct tensor signal Receive signal tensor Subtract the reconstructed tensor signal Get the tensor residual signal
[0018] Step 4: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the tensor residual signal The corresponding rank-one matrix decomposition factor: {h′, a′, s′};
[0019] Step 5: Receive the signal tensor The rank-one matrix factor and tensor residual signal The rank-one matrix factors of the received signal tensor y are combined to obtain the rank-two initialization matrix factors of the received signal tensor y;
[0020] Step 6: Use the regularized multivariate alternating least squares algorithm of blind decomposition to find the rank r matrix decomposition factor {H r ,A r ,S r}, where the initialization matrix factors are {H * ,A * ,S *};
[0021] Step 7, through {H r ,A r ,S r}Reconstruct the tensor signal y′, and subtract the reconstructed tensor signal y′ from the received signal tensor y to obtain the tensor residual signal
[0022] Step 8: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the tensor residual signal The corresponding rank-one matrix decomposition factor: {h′, a′, s′}; the received signal tensor Rank-One Matrix Factor and Tensor Residual Signal Merge the rank-r matrix factors to obtain the received signal tensor The rank-(r + 1) initialized matrix factor; return a series of solutions: {h 1 , a 1 , s 1}, {H 2 , A 2 , S 2}, … {H\ R , A R , S R}.
[0023] Among them, the regular multi-way alternating least squares algorithm process for blind decomposition is as follows:
[0024] Step 1, initialization
[0025] Step 2, when H, A, and S do not reach convergence, let
[0026]
[0027] The condition for algorithm convergence is that the mean square value of the corresponding matrix error in two consecutive iterations is less than 10 -5 .
[0028] Compared with the existing inventions, the beneficial effects of the present invention are:
[0029] The nested interference cancellation algorithm for regular multi-way decomposition can effectively eliminate fixed interference and intentional interference. Especially in a complex communication environment, it shows stronger anti-interference ability to ensure reliable signal transmission; compared with traditional receivers (such as two-stage training receivers), the nested interference cancellation for regular multi-way decomposition can significantly improve the signal reception quality, reduce the bit error rate, and enhance the overall performance of the communication system in the case of two-hop communication. This algorithm speeds up the convergence rate and reduces the number of iterations through the rank-recursive initialization method, and has an obvious advantage in processing speed compared with other algorithms (such as the regular multi-way alternating least squares algorithm for blind decomposition). Brief Description of the Drawings
[0030] Figure 1 is a two-hop multi-antenna loop communication system;
[0031] Figure 2 is the comparison of two-hop and single-hop BER performance. Detailed Embodiment
[0032] The present invention will be described in detail below in conjunction with the embodiments shown in the drawings. However, it should be noted that these embodiments do not limit the present invention. Any equivalent transformation or substitution in function, method, or structure made by those of ordinary skill in the art based on these embodiments shall fall within the protection scope of the present invention.
[0033] The present invention provides an interference cancellation algorithm in railway communication, including the following steps:
[0034] Step 1, randomly initialize and
[0035] Step 2, let the rank of the tensor be M, substitute the matrix factors initialized in Step 1 and M into the multi-way decomposition interference cancellation algorithm, and obtain
[0036] Step 3, expand the obtained by the previous iteration into a matrix to get
[0037] Step 4, randomly initialize and
[0038] Step 5, let the rank of the tensor Y sr be R, substitute the matrix factors initialized in Step 4 and R into the multi-way decomposition interference cancellation algorithm, and obtain
[0039] Step 6, obtain
[0040] by the previous iteration Step 7, use the singular value decomposition algorithm to estimate the user information
[0041] Among them, the steps of the multi-way decomposition interference cancellation algorithm are as follows:
[0042] Step 1, use the interference cancellation method to obtain the received signal tensor y;
[0043] Step 2, use the regular multi-way alternating least squares algorithm for blind decomposition to find the rank-one matrix decomposition factors corresponding to the signal tensor : {h 1 , a 1 , s 1};
[0044] Step 3, reconstruct the tensor signal by {h 1 , a 1 , s 1 The received signal tensor subtract the reconstructed tensor signal Obtain the tensor residual
[0046] Step 4: Use the regular multi - way alternating least squares algorithm for blind decomposition to find the tensor residual signal The corresponding rank - one matrix decomposition factors: {h′, a′, s′};
[0047] Step 5: Combine the rank - one matrix factors of the received signal tensor and the rank - one matrix factors of the tensor residual signal to obtain the rank - two initialization matrix factors of the received signal tensor ;
[0048] Step 6: Use the regular multi - way alternating least squares algorithm for blind decomposition to find the rank - r matrix decomposition factors {H , A r , S r} of the signal tensor r , where the initialization matrix factors are respectively {H * , A * , S *};
[0049] Step 7: Reconstruct the tensor signal r using {H r , A r} and subtract the reconstructed tensor signal from the received signal tensor to obtain the tensor residual signal ;
[0050] Step 8: Use the regular multi - way alternating least squares algorithm for blind decomposition to find the rank - one matrix decomposition factors: {h′, a′, s′} of the tensor residual signal ; Combine the rank - one matrix factors of the received signal tensor and the rank - r matrix factors of the tensor residual signal to obtain the rank - r + 1 initialization matrix factors of the received signal tensor ; Return a series of solutions: {h 1 , a 1 , s 1}, {H 2 , A 2 , S 2}, … {H R , A R , S R}.
[0051] The flow of the regular multi - way alternating least squares algorithm for blind decomposition is as follows:
[0052] Step 1: Initialize
[0053] Step 2, when H, A, and S do not converge, let
[0054]
[0055] The condition for the algorithm to converge is that the mean square value of the corresponding matrix error in two consecutive iterations is less than 10 -5 .
[0056] The present invention has obvious advantages compared with other inventions, including: the interference cancellation system proposed by the present invention can effectively solve the impact of interference on the communication system. The nested interference cancellation algorithm of regular multi - decomposition is specially designed to eliminate fixed interference and intentional interference, which are very common in the railway loop communication system. Through effective interference cancellation, the algorithm can improve the reception quality and reliability of the signal. The algorithm adopts the method of rank - recursive initialization, which helps to accelerate the convergence speed of the algorithm, reduce the number of iterations, and thus improve the operation efficiency of the algorithm. The algorithm takes into account the interference in the non - Gaussian environment during design, so that the algorithm can still maintain stable performance when facing complex and changing interference environments.
[0057] To verify the performance of the proposed algorithm, a system model is established to analyze this method. It should be noted that: once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.
[0058] Please refer to Figure 1 , Figure 1 which is a two - hop multi - antenna loop communication system.
[0059] Step 1, randomly initialize and
[0060] Step 2, let the rank of the tensor be M, substitute the matrix factors initialized in Step 1 and M into the multi - decomposition interference cancellation algorithm to obtain
[0061] Step 3, expand the obtained in the previous iteration into a matrix to get
[0062] Step 4, randomly initialize and
[0063] Step 5, let the rank of the tensor Y sr be R, substitute the matrix factors initialized in Step 4 and R into the multi - decomposition interference cancellation algorithm to obtain
[0064] Step 6, iterate through the previous step to find
[0065] Step 7: Estimate user information using singular value decomposition algorithm
[0066] Among them, the steps of the multivariate decomposition interference elimination algorithm are as follows:
[0067] Step 1: Use interference cancellation method to obtain the received signal tensor
[0068] Step 2: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the signal tensor The corresponding rank-one matrix decomposition factor: {h 1 ,a 1 ,s 1};
[0069] Step 3, by {h 1 ,a 1 ,s 1}Reconstruct tensor signal Receive signal tensor Subtract the reconstructed tensor signal Get tensor residual
[0071] Step 4: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the tensor residual signal The corresponding rank-one matrix decomposition factor: {h′, a′, s′};
[0072] Step 5: Receive the signal tensor The rank-one matrix factor and tensor residual signal The rank-one matrix factor is combined to obtain the received signal tensor The rank-two initialization matrix factor of ;
[0073] Step 6: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the signal tensor The corresponding rank r matrix decomposition factor {H r ,A r ,S r}, where the initialization matrix factors are {H * ,A * ,S *};
[0074] Step 7, through {H r ,A r ,S r}Reconstruct tensor signal Receive signal tensor Subtract the reconstructed tensor signal Obtain the tensor residual signal
[0075] Step 8, use the regular multi - way alternating least squares algorithm for blind decomposition to find the tensor residual signal The corresponding rank - one matrix decomposition factors: {h′, a′, s′}; combine the rank - one matrix factors of the received signal tensor and the rank - r matrix factors of the tensor residual signal to obtain the rank - r + 1 initialization matrix factors of the received signal tensor Return a series of solutions: {h 1 , a 1 , s 1}, {H 2 , A 2 , S 2}, … {H R , A R , S R}.
[0076] The process of the regular multi - way alternating least squares algorithm for blind decomposition is as follows:
[0077] Step 1, initialization
[0078] Step 2, when H, A, and S do not reach convergence, let
[0079]
[0080] The condition for algorithm convergence is that for two consecutive iterations, the mean square value of the corresponding matrix error is less than 10 -5 .
[0081] Figure 2 is for the performance comparison between two - hop and single - hop BER. The results show that the nested interference cancellation receiver with regular multi - way decomposition has no cumulative channel estimation error and can achieve better signal reception performance compared with the two - stage training receiver.
[0082] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above - mentioned exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non - restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, it is intended to embrace all changes within the meaning and scope of the equivalent elements of the claims in the present invention.
Claims
1. An interference elimination algorithm in railway communication, whose system model is: The system model is based on the structure of a two-hop multi-antenna loop communication system. In this system, the train control command center acts as the transmitter and sends signals to the train onboard receiver through the trackside electronic equipment. The system considers two groups of receiving antenna arrays, each with M antennas, and two groups of onboard receiving equipment, each with F antennas. First, it is necessary to obtain the received signal tensor after relay processing. Then, the algorithm estimates the loading matrix factor of the received signal tensor, expands it, decomposes it, and obtains the symbol information. Since the algorithm utilizes the nested characteristics of the signal, it can avoid the cumulative error caused by the training sequence estimation channel, thereby improving the signal reception performance.
2. According to the system model of claim 1, its algorithm features include the following: The regular multivariate decomposition nested interference cancellation algorithm is designed for two-hop communication links, taking into account the transmission of signals between relay nodes and destination nodes, and is suitable for complex communication scenarios. It can effectively eliminate fixed interference and intentional interference, especially in railway loop signal transmission in non-Gaussian environments, and improve the signal reception quality. By comparing with the TST receiver, the regular multivariate decomposition nested interference cancellation algorithm shows better bit error performance.
3. The interference elimination algorithm based on railway communication as claimed in claim 2, comprising the steps of: Step 1: Random initialization and Step 2: Let the tensor The rank is M. Substitute the matrix factors initialized in step 1 and M into the multivariate decomposition interference elimination algorithm to obtain Step 3: Iterate the previous step to find Expand the matrix to get Step 4: Random Initialization and Step 5, let the tensor Y sr The rank is R. Substitute the matrix factors initialized in step 4 and R into the multivariate decomposition interference elimination algorithm to obtain Step 6, iterate through the previous step to find Step 7: Estimate user information using singular value decomposition algorithm 4. The multivariate decomposition interference elimination algorithm mentioned in claim 3, wherein the steps are as follows: Step 1: Use interference cancellation method to obtain the received signal tensor Step 2: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the signal tensor The corresponding rank-one matrix decomposition factor: {h 1 ,a 1 ,s 1 }; Step 3, by {h 1 ,a 1 ,s 1 }Reconstruct tensor signal Receive signal tensor Subtract the reconstructed tensor signal Get the tensor residual signal Step 4: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the tensor residual signal The corresponding rank-one matrix decomposition factor: {h′, a′, s′}; Step 5: Receive the signal tensor The rank-one matrix factor and tensor residual signal The rank-one matrix factor is combined to obtain the received signal tensor The rank-two initialization matrix factor of ; Step 6: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the signal tensor The corresponding rank r matrix decomposition factor {H r ,A r ,S r }, where the initialization matrix factors are {H * ,A * ,S * }; Step 7, through {H r ,A r ,S r }Reconstruct tensor signal Receive signal tensor Subtract the reconstructed tensor signal Get the tensor residual signal Step 8: Use the regularized multivariate alternating least squares algorithm with blind decomposition to find the tensor residual signal The corresponding rank-one matrix decomposition factor: {h′, a′, s′}; the received signal tensor The rank-one matrix factor and tensor residual signal The rank r matrix factor is combined to obtain the received signal tensor Initialize the matrix factors of rank r+1; return a series of solutions: {h 1 ,a 1 ,s 1 },{H 2 ,A 2 ,S 2 },…{H R ,A R ,S R }.
5. The regular multivariate alternating least squares algorithm for blind decomposition mentioned in claim 4, wherein the steps are as follows: Step 1: Estimate the matrix of rank r Step 2, set p = 0, and randomly initialize H, A, and S; Step 3, let p=p+1; Step 4: Use formulas (1) to (3) to obtain estimated values of H, A, and S Step 5, repeat steps 2 to 3 until convergence is reached; Step 6: Eliminate the scale ambiguity and column ambiguity of the estimated values H, A, and S.
6. A two-hop multi-antenna loop communication system, characterized in that The system includes a train control center, trackside electronic equipment, and train onboard receivers, all of which are equipped with multi-antenna arrays. This configuration enhances the system's reception capability and anti-interference performance, and the system is designed with interference cancellation processing to deal with non-Gaussian interference, such as fixed interference and intentional interference. This processing helps improve the signal reception quality and system robustness.
7. An interference elimination algorithm in railway communication, comprising a computer program, which implements the system model in claim 1 when executed.
8. A wireless communication information acquisition algorithm program product, comprising a computer program, which implements the information acquisition algorithm in claims 3, 4, and 5 when executed.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the information acquisition algorithm as claimed in claim 8 is implemented.
10. A non-transitory computer-readable storage medium storing a computer program, characterized in that When the computer program is executed by a processor, it implements the channel estimation algorithm as claimed in claims 3, 4, 5 or the receiver simulation program as claimed in claim 7.