Research On TDOA Estimation Method Based On Compressive Sampling For Digital Modulated Signals

Study on digital modulated signal compressive sampling time difference of arrival(TDOA)estimation is a current area of research in the realm of signal processing and involves multi-node signal-level fusion processing.How to achieve a high-precision TDOA estimate with minimal data transfer when the transmission bandwidth between nodes is constrained is a research challenge.In order to achieve the goal of accurate TDOA estimation with less data transmission under constrained communication bandwidth,this thesis combines the compressive sampling,edge computing computation offload strategy,and TDOA estimation problem and studies the TDOA estimation method of digital modulation signal compressive sampling.The primary work contains the following:Firstly,a compressive sampling time difference estimation model for digital modulated signals is established,and the problems of directly using compressed data to estimate time difference are discussed.The modified Cramer Rao Lower Bound(MCRLB)of symbol period and time delay under computational offload is obtained from the issue description and signal model of TDOA estimation under the edge computing process.Second,a partial Fourier matrix is employed as the measurement matrix based on the compressive sampling time difference estimation model,and the compressed data is used for cross correlation to produce the time difference estimate.After compressing the partial Fourier matrix,the compressed data discards the amplitude information,further reducing the amount of data transmission.The time difference estimation is then obtained through the phase method in the frequency domain using the characteristics of the signal time domain delay reflected as the corresponding frequency phase shift in the frequency domain.Experimental simulation demonstrates that the correlation method performs better for the TDOA estimate than the phase method.Finally,the computation offload strategy of edge computing is applied to the TDOA estimation problem for the model of a linear digital modulation signal in the scenario of two nodes transmitting compressed data to the processing center,and the TDOA estimation problem is transformed into three parameter estimation problems of signal symbol period,time delay,and symbol period interval.The processing center separates the estimated time difference into two components: the delay that occurs during the symbol period and the pause between symbol periods.When the signal oversampling rate is large,experimental simulation demonstrates that the algorithm exceeds the correlation approach of compressive sampling in TDOA estimation performance.