Research On Electronic Reconnaissance Technology Based On Compress Sensing

Digital wideband receiver plays a very important role in electronic reconnaissance.In order to satisfy the requirements of total probability reconnaissance receiving in the situation of modern electronic warfare,it is necessary to be more accurate and wider instant bandwidth for the reconnaissance receiver.But the traditional high speed sampling equipment,which works under the requirement of Nyquist theory,could not meet the demands.The compress sensing theory indicates that if the signal is sparse,we could get the sampler below the Nyquist rate and recover it accurately.In other words,the sampling is based on the information but not the data.Thus,the hardware can be easier realized and needs less data to be transmitted and saved.This kind of compress sensing based analog signal sampling method is called Analog to Information Conversion.This paper studies the application of compress sensing in electronic reconnaissance.The main work is as follows:Firstly,a kind of digital wideband receiver based on MWC has been studied.The digital channelized receiver which is widely used has been studied.Although the digital channelized receiver has good performance,the sampling rate of ADC is still an important factor which restricts the instant bandwidth.So we consider using AIC into the design of electronic reconnaissance receiver.The sparsity of the received signal by electronic reconnaissance receiver has been studied and the characters of two different AIC,i.e.,RD and MWC,have been compared.The MWC has been used as the sampling structure in the electronic reconnaissance receiver.Simulation experiments of the LFM signals which arrive in a certain period prove its performance.Secondly,the design of measurement matrix in compress sensing has been studied.Based on the study of chaotic measurement matrix,a Cat-Toeplitz measurement matrix has been studied.Besides,the Cat chaotic sequence is pseudorandom and easy to produce while the Toeplitz matrix has small degree of freedom and only need a few elements.We combine the advantages of them to construct a new measurement matrix,whose inner related coefficient is calculated and compared with the random Gaussian matrix in order to prove that it satisfies the relative criterion.The simulation experiments of unidimensional time domain signals show that it has good performance compared with random Gaussian matrix or random Bernoulli matrix.The DOA estimation algorithm based on CS has been studied and the proposed matrix has been used into the DOA estimation of electronic reconnaissance.The simulations demonstrate that it has good performance when the receiving channels of array elements decrease.Thirdly,a kind of sampling scheme called C-AIC has been studied and the Hardware In the Loop Simulation system of CS wideband receiver has been constructed.The CT measurement matrix has been used as frequency mixing sequences to improve the performance of MWC,which reduces the complexity of the hardware.The simulation experiments show that the scheme can realize almost the same successful reconstruction rate as MWC when using less sampling channels,and can achieve better performance than MWC when the SNR is small.We design the hardware of the sampling board and the upper computer software.We analyze every part of the circuit and optimize it to make the board have better performance and lower power dissipation.Numerical experiments show that it can reconstruct the original signals accurately when the sampling rate is far below the Nyquist rate.