Font Size: a A A

Target Detection Methods In MIMO Radar Based On Polarization Diversity Technology And Random Matrix Theory

Posted on:2015-03-25Degree:MasterType:Thesis
Country:ChinaCandidate:X H TangFull Text:PDF
GTID:2268330428498047Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
Multiple-input multiple-output (MIMO) radar has many advantages in targetdetection, parameter estimation and target recognition and other fields. Due to the useof transmitted signal diversity, the target detection performance of the MIMO radarhas been significantly improved.Firstly, MIMO radar target detection based on polarization diversity technologyis conducted in-depth research in this essay. Compared to traditional MIMO radarwhich mainly adopts spatial diversity technique, MIMO radar has the stronger radardetection capability in the background of small target or clutter, due to the use ofpolarization diversity. Different polarizations will cause huge differences in detectionperformance, so the detection performance of MIMO radar can be significantlyimproved by optimizing the design of transmitted polarization waveforms.Furthermore, the essay conducts a further study on MIMO radar target detection basedon random matrix theory (RMT). Currently, most of MIMO radar target detection iscarried out under the assumption that the number of samples is much larger than thenumber of array elements, resulting in reduced performance when the sample size isnot sufficient. Random matrix theory provides a convenient tool for MIMO radarsignal processing, and the method of asymptotic spectral theory (AST) based onrandom matrix theory can conduct the blind detection of bistatic MIMO radar withoutthe knowledge of noise variance and target scattering matrix. The research of thisessay is supported by the National Natural Science Foundation Project of “Jointparameter estimation and target localization based on polarization diversity in MIMOradar”(No.61071140)and “Robust target detection and estimation for MIMO radarsbased on large dimensional random matrix theory”(No.61371158). The following isthe main research of this essay:On the base of the traditional model for distributed MIMO radar, this essaystudys MIMO radar target detection algorithm based on polarization diversity in thebackground of clutter. A polarimetric MIMO radar signal model for target detection isestablished and a polarimetric MIMO radar detection algorithm based on Jones vectoris proposed. This method achieves the optimization of the polarization waveforms by utilizing multiple searches of transmitted antenna waveforms. Simulation results showthe effectiveness of the algorithm, and compared with horizontal, vertical andorthogonal polarization, the algorithm improves target detection performance.To reduce the complexity of multiple searches, this essay furtherly proposes aMIMO radar target detection algorithm based on glowworm swarm optimization(GSO). This method uses maximizing detection probability as the objective functionto select transmitted polarization waveform, conducts multi-dimensional parallelsearch through the use of GSO, and optimizes simultaneously multiple polarizationparameters through parallel processing data, so it solves the elusive problem ofmultiple nested search. The simulation results show that the MIMO radar targetdetection algorithm based on GSO improves the detection performance and reducesthe amount of computation.Despite the above MIMO radar detection method improves target detectionperformance, the performance will still be degraded when the number of samples isinsufficient or the number of samples and array elements are close. For this reason,from the perspective of bistatic MIMO radar model, this essay proposes a MIMOradar target detection algorithm based on random matrix theory in the case of thenumber of samples is close to the number of transmit array elements. This methodachieves blind detection by using random matrix theory, with the target scattering,location information and the noise variance unknown. Since this algorithm reducedgreatly the dependence on priors and was not sensitive to noise, it achieves robustMIMO radar target detection in the large dimensional regime.
Keywords/Search Tags:MIMO radar, target detection, polarization diversity, glowworm swarmoptimization, random matrix theory
PDF Full Text Request
Related items