Joint Target Detection Based On Passive MIMO Radar System

Different from the traditional active radar,the passive Multi-Input Multi-Output(MIMO)radar(PMR)system can work in radio silence and effectively resist external interference.In this thesis,based on the advantage of spatial diversity gain of distributed passive MIMO radar systems,the following studies are conducted on the joint target detection problem.1.In order to provide theoretical support for the subsequent detection content,this thesis introduces the compound hypothesis testing criterion in PMR systems,and provides a detailed description of the constant false alarm detection technique.Meanwhile,the principles related to adaptive filtering are introduced,and the applicability of the Least Mean Square(LMS)algorithm is verified by simulation.In addition,multi-sensor data fusion techniques are introduced,and several common network structures for multisensor signal detection are given.2.In this thesis,the multi-sensor data fusion technology is introduced into the signal detection of PMR system.The distance-weighted and the inspection-weighted novel Generalized Likelihood Ratio Test(GLRT)detection algorithm are proposed without the direct-path interference residual.In the former,the correlation between all observed data is utilized.In the data fusion center,the distance factor prior information is obtained through grid preprocessing to weight the signal and test quantity,and then the constant false alarm threshold judgment is made.The former is suitable for situations where communication bandwidth resources are sufficient,environmental factors are single,and performance requirements are high.The latter uses the Modified Deflection Cofficient(MDC)to assign corresponding weights to each receiver's inspection quantity,and then sends the weighted local inspection quantities to the data fusion center for global decision-making,which is suitable for situations where communication bandwidth resources are limited and the surrounding environment is complex.Finally,it is proved by performance analysis and simulation that the proposed algorithms based on weighted data fusion are better than the traditional detectors and have wider application range.3.When the direct-path residuals in the surveillance channel cannot be completely eliminated,an improved signal model with Direct-Path Interference(DPI)residue is given for the passive MIMO radar dual channel system.The unknown parameters were estimated jointly by GLRT criterion and Rao criterion respectively,and then the corresponding detectors were derived.The computational complexity and performance of the detector are analyzed,and the calculation method of the constant false alarm threshold is given.Meanwhile,in order to facilitate the subsequent verification of the effectiveness of the proposed passive detectors,the active DPI-GLRT and Rao detectors were derived respectively,which can be used as upper bounds on the performance of passive detectors.In the case of DPI residue in the surveillance channel,the experimental results demonstrate that the two detectors can achieve better detection under certain conditions,and the detection performance of the Rao detector is close to that of the GLRT detector in the case of a relatively large sample size.