Research On Multipath Time Delay Estimation Algorithm Of Spread Spectrum Signal Based On Convolutional Neural Network

The spread spectrum sequence used in the spread spectrum system has great autocorrelation property,and the receiver can easily extract its signal propagation delay.Thence,the spread spectrum signal is a good TOA/TDOA positioning signal source.However,when the signal propagates in an environment with obstacles such as cities or valleys,the multipath effect will affect the accuracy of the spread spectrum signal delay estimation,thus affecting its positioning performance.Convolutional neural networks have excellent adaptive and generalization capabilities,and have achieved excellent performance in many fields such as image recognition and signal processing.In order to solve the problem of spread spectrum signal delay estimation in multipath environment,this thesis proposes a multipath time delay estimation algorithm based on convolutional neural network.The specific work of this paper is as follows:First,a convolutional neural network structure for multipath time delay estimation is designed in this paper,and excellent delay estimation accuracy is achieved by increasing the depth of the network.After studying the data model of spread spectrum signal in multipath environment,the multi-path signal cross-correlation function after windowing is selected as the input sequence of convolutional neural network.It overcomes the defects that the multipath signal waveform is not obvious and susceptible to noise.Secondly,this paper uses the Pytorch framework to build the above-mentioned delay estimation network,and tests its time delay estimation performance.The effects of convolution kernel size,input sequence length,and training set signal-to-noise ratio on final delay estimation performance are discussed.According to the above test results,two delay estimation networks were selected,one for high SNR scenarios and the other for low SNR scenarios,and both were compared to existing algorithms.The simulation results show that the proposed algorithm overcomes the defects of WRELAX algorithm and MUSIC algorithm that need to predict the number of multipath.Compared with the MUSIC algorithm,the proposed algorithm has higher delay estimation accuracy in the lower SNR and higher SNR scenarios.The performance of the proposed algorithm is similar to that of the MSUIC algorithm.Compared with the WRELAX algorithm,the proposed algorithm has better stability and lower complexity,and the delay estimation accuracy is higher in the lower SNR scenario.