Joint Estimation Of Spatial And Frequency Domain Parameters Of Multi-band Signal Based On Compressed Sampling

Multi-band signals are widely used in radar detection,ultra-wideband wireless communication,spectrum detection,and cognitive radio.Simultaneous detection of various parameter information in the airspace and frequency domain is always a hot spot in the field of array signal processing.The spectrum of multi-band signals is wider and wider,but the effective occupied bandwidth is narrow,and the carrier frequency of the target signal is unknown in most applications.The Shannon-Nyquist sampling theorem states that the minimum sampling rate of the signal is at least twice the highest frequency of the signal.The excessive sampling rate brings great pressure to the hardware.In recent years,the Modulated Wideband Converter(MWC)system based on Compressed Sensing(CS)theory has solved the sampling problem of multi-band analog signals.Therefore,in view of the high sampling rate required for the joint estimation method of the spatial frequency domain parameters of multi-band signals in array signal processing,this paper studies the joint space-frequency domain parameters estimation method of multi-band signal based on array MWC structure.In this paper,we first study the existing joint one-dimensional Direction-Of-Arrival(DOA)and carrier frequency estimation method based on L-shaped array MWC structure,and mainly study the ESPRIT method and the CP decomposition method.Since the signal in reality exists in three-dimensional space,it is more accurate to use a two-dimensional DOA to describe the direction of the target signal.In view of the simple structure of Lshaped array and excellent estimation performance,based on the joint one-dimensional DOA and carrier frequency estimation method based on L-shaped array MWC structure,the ESPRIT method and CP decomposition method based on double-L shaped array MWC structure are proposed.By transforming the three-dimensional parameter estimation problem into two two-dimensional parameter estimation problems and the corresponding pairing method is given.Then,in order to omit the complex pairing steps and reduce the complexity of the algorithm,this paper further proposes the ESPRIT method and CP decomposition method of the joint two-dimensional DOA and frequency estimation based on L-shaped delay array MWC.The carrier frequency parameters are estimated by adding a fixed delay channel without parameter pairing.The details are as follows:1.Study the existing joint one-dimensional DOA and carrier frequency estimation method based on L-shaped array MWC structure.The array multi-band signal model,Larray MWC sampling structure and reconstruction algorithm are studied.Study the joint one-dimensional DOA and carrier frequency estimation method of Estimating Signal Parameter via Rotational Invariance Techniques(ESPRIT)method and canonical decomposition/ Parallel factor analysis(CANDECOMP/PARAFAC,CP)based on Lshaped array MWC structure.The conditions needed to ensure accurate recovery under ideal conditions were given.The performance of each method was compared and analyzed by MATLAB simulation.At the same time,the influence of parameters and noise on the estimation accuracy of each method.2.Aiming at the problem that one-dimensional DOA can not accurately describe the direction of target signal in space,in order to estimate the two-dimensional DOA parameters of target signal,the L-shaped array MWC structure is extended,and a joint of space-frequency domain parameters estimation method based on double L-shaped array MWC structure is proposed.The estimated parameter dimension is increased by transforming the three-dimensional parameter estimation problem into two twodimensional parameter estimation problems.Since the three axes receive information relatively independently,additional parameter pairing is required,and we also give the corresponding pairing method.Then through MATLAB simulation,the ESPRIT method and CP decomposition method based on double L-shaped array MWC structure are simulated and compared,and the proposed two methods are compared with the existing methods.The influence of the number of system parameters and noise on the estimation accuracy of each method.3.In order to omit the complicated pairing operation,reduce the complexity of the algorithm,we also extend the L-shaped structure,and propose a joint spatial-frequency domain parameters estimation method based on the L-shaped delay array MWC structure.By directly estimating the carrier frequency using the sampled values of the delay channel,the three-bit parameter estimation problem can be directly calculated without additional pairing operations,which reduces the complexity of the algorithm.Then the simulation experiments and comparative analysis of ESPRIT method and CP decomposition method based on L-shaped delay array MWC structure are carried out by MATLAB simulation.The influences of system parameters such as array spacing and channel number and noise on the estimation accuracy of each method are also explored.Based on the existing one-dimensional DOA and carrier frequency joint estimation method based on L-shaped array MWC structure,ESPRIT method and CP decomposition method algorithm are simultaneously improved.Firstly,the ESPRIT method and CP decomposition method based on double-L array MWC structure are proposed.The threedimensional parameter estimation problem is transformed into two two-dimensional parameter estimation problems,and the corresponding pairing method is given.Then,in order to omit the complicated pairing step,this paper further proposes the ESPRIT method and CP decomposition method based on the joint estimation of two-dimensional DOA and frequency of L-shaped delay array MWC structure.No additional pairing is required,reducing algorithm complexity.Finally,we summarize the characteristics and advantages and disadvantages of the double L-shaped array structure and the L-shaped delay array structure,as well as the characteristics and advantages and disadvantages of the ESPRIT algorithm and the CP algorithm method.