Research On Real-time Signal Processing For FM-based Passive Radar

Passive bistatic radar utilizes electromagnetic signals emitted by the third party to achieve the purpose of target location and detection,without radiating signals itself.It has many advantages to compare with conventional active radar,such as low cost,strong in anti-interference and harmless to the environment.The FM radio signal possesses the feature of extensive distribution and high emission power,which is of great compatibility to be used as opportunity illuminator of passive radar.FM-based passive radar signal processing has important academic significance and broad application.Due to the reflection from ground and buildings,strong direct path signal and multipath clutter exist in passive receive channel.The power of those disturbance signals is much stronger than the target signals,which shield the target signals in the Rang-Doppler map to resist the extraction of target information.Clutter removal is one of the important chains in passive radar signal processing.This paper deduces the basis of common filter algorithms,and highlights the advantages and disadvantages of the extensive cancellation algorithm(ECA)and its batch version ECA-B,which are mostly used in real signal processing.The sliding extensive cancellation algorithm(ECA-S)is able to work around the limitation on the batch number to meet the requirement in integration time and doppler resolution of radar system at the same time.The cost of using sliding window is to increase the computational complexity,which challenges the demand of real-time signal processing.The Compute Unified Device Architecture(CUDA)proposed by the NVIDIA company is greatly suitable to solve the computational problems with high parallelism.This paper successfully realizes ECA-S parallel speedup with CUD A,and parallel realization about the passive radar signal processing chains of match filtering and target detection are also done to accomplish the whole signal processing procedure speedup.The results of real data processing verify the superiority of ECA-S to compare with ECA-B,and the performance of CUD A parallel speedup meets the demand of the real-time target detection.