Study On Parameter Estimation For DCS-MIMO Radar

Distributed Compressed Sensing (DCS) extends sparsity from single signal to multi-signalensembles. It also enables joint recovery under low-speed sampling that exploits both intra-andinter-signal correlation structures. MIMO radar mainly adopts the diversity technology to improvesystem performance; its multiple received signals have the same sparsity in the target parameter space.Thus DCS will bring new innovations to the signal processing of MIMO radar system. Combinedwith the national natural fund“Study On Joint Sparsity Presentation and Parameter Estimation forDCS-MIMO Radar, this paper mainly focuses on Joint Sparsity Models and Reconstruction algorithmof Jointly Sparse Signals as well as Signal model and target parameter estimation in DCS-MMO radarsystem. The completed major tasks are as follows.Given the common sparse support of the second joint sparsity model, a fast iterative algorithmhas been proposed. In each iteration, it selects the index of the base vector that maximizes the value ofthe sum of the magnitudes of the projections of the residual, and updates the estimate of thecoefficients for all the selected base vectors. Simulation results show that the method can reduce thetime of reconstructing signal ensembles.This paper studies incoherent DCS-MIMO radar. Firstly, we propose signal model and reveal thejoint sparsity presentation of all received signals in the range space. Then, the segmented One StepGreedy and JOMP algorithm are introduced to implement the estimation of delay time and reflectionamplitude. Simulation results show that DCS-MIMO radars can effectively overcome target RCSfluctuations and increases the parameter estimation accuracy.This paper studies the parameter estimation of coherent DCS-MIMO radar, and derives the jointsparsity presentation of all received signals in the angle space. For the monostatic DCS-MIMO radar,1-D spectrum estimation algorithm based on difference is proposed, which takes the second orderdifference of correlation between signals and based vectors as spectral estimator to estimate DOA.Simulation results show that the higher angle resolution is achieved. For bistatic DCS-MIMO radar,2-D spectrum estimation algorithm based on OSGA is proposed, which takes the value of the sum ofthe magnitude of the inner products between the based vector with the same index and thecorresponding signal as spectral estimator to estimate DOA and DOD. Simulation results show thatthe computation is reduced significantly.