Track-Before-Detect Algorithms For Target Detection In Passive Radar

Passive radar,which is based on the opportunity illuminators,is a new type of radar.The main feature is that it does not emit electromagnetic wave signals externally,but generally uses commercial broadcast signals as the illumination source,and only receives the reflected signals of the interesting targets to realize the detection and tracking of targets.Among the advantages,the detection and localization of targets is covert,continuous,and also inexpensive,since the transmitter needs neither frequency allocation nor extra hardware.Therefore,passive radar has become an important way to detect targets effectively.However,due to the use of external radiation sources,the problem of low signal-to-noise ratio will inevitably occur,a method adapted to low signal-to-noise ratio should be adopted in the detection and tracking of targets.The classical method will generate thresholds for the original detection data generated by cross-correlation,which will cause problems such as target miss detection and false alarms.The Track-Before-Detect(TBD)with multi-frame joint processing and accumulation,makes it that performance of the weak target detection and tracking compared to the traditional track-after-detect improved significantly.However,at present,the idea of TBD is used for passive radar target detection,and there are few applications for further accumulation of data,and there are fewer applications for using the TBD idea to process the raw data accumulated after mutual accumulation.In view of the above problems,this thesis studied the passive radar target detection and tracking technology based on TBD technology,and mainly did the following work:1.According to the working principle of passive radar,the passive radar observation data is simulated and modeled.Based on the lucubration of particle filter algorithm,an improved particle filter method is proposed,which optimizes particle position initialization and improves the detection performance of particle filter algorithm.The particle filter algorithm is applied to the passive radar,and the multi-frame passive radar observation data is combined to obtain the target delay and Doppler information.2.Lucubrate the dynamic programming algorithm.The dynamic programming algorithm is applied to the passive radar scene,and the performance of the algorithm is simulated.By studying a method of phase weighting for measurement data,the performance of dynamic programming algorithm in passive radar applications is improved.Finally,the dynamic programming algorithm and the particle filter algorithm are simulated and compared,and the differences between the two algorithms are analyzed.3.In view of the target crossing multiple beams,the specific requirements of the application scenario are analyzed.The existing cross-beam accumulation method is studied.The multi-beam non-coherent accumulation method according to the above scenario is proposed,including the determination of initial accumulation time and accumulation method.The simulation experiment is carried out on the method to analyze the performance of the method.The above performance analysis and theoretical derivation and application were verified by simulation experiments,which solved the problem of passive radar target detection and tracking at low SNR,and optimized the particle filter theory application method,which solved the problem of target accumulation across multiple beams.