GPU-accelerated mismatched filter for analog TV-based passive radar

Passive radar exploits third-party commercial transmitters such as frequency modulated signals, television (TV) signals and navigation signals to detect, locate and track targets. In this paper, for the considerations of both large transmitting power and wide coverage area, analog TV signal is selected as the transmitter of the passive radar. Analog TV signal consists of audio signal and visual signal. Audio signal with the ambiguity function like a pushpin can be viewed as a perfect transmitter, whereas video signal which is periodical in time-domain, resulting in the existence of high ambiguous range side-lobes, is not the case. In general, traditional mismatched filter based on only visual signal addresses the undesired problem at the cost of the loss of signal-noise-ratio (SNR) of target. In this paper, to suppress ambiguous range side-lobes and reduce the loss of SNR to the most degree, a modified mismatched filter based on visual signal and audio signal instead of single visual signal is presented. Furthermore, the detailed parallel implementations based on graphic processing unit (GPU) of the proposed method are described for the purpose of real-time processing. The performance of the proposed method is demonstrated via several simulations.