Multistatic Passive Radar Sensing Algorithms With Calibrated Receivers

Abstract

This paper addresses a multistatic target sensing problem with a calibrated receiver with an unknown but deterministic ambient signal. To do this, we formulate two different target detection problems as composite hypothesis testing problems, where they differ based on whether the reference channel (RC) is available or not. Indeed, the detection problem 1 (P1) is formulated for the case in which each receiver utilizes an RC to collect the ambient signal and a surveillance channel (SC) for sensing potential target echo, while the second detection problem (P2) is for the case without RCs. Then, we resort to the likelihood ratio test (LRT) principle to devise two detection methods, namely P1-LRT and P2-LRT detectors. For the proposed P1-LRT method, we show that any uncertainty in the value of averaged direct signal-to-noise-ratio of the available reference channels, abbreviated as DNR <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">avg</sub> , can result in excessive false alarm probability in low DNR <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">avg</sub> regime. To facilitate efficient operation under uncertainty in DNR <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">avg</sub> , we propose a new threshold-setting strategy to adjust the level of the proposed P1-LRT detector. Through extensive Monte-Carlo (MC) simulations, we examine the above challenge and investigate the efficiency of the proposed threshold-setting strategy. Although the P1-LRT has a fixed level and the P2-LRT is a fixed size test, the P1-LRT detector significantly outperforms the P2-LRT, as indicated by the MC simulation results.