Estimation of radar signals using passive sensor network

The problem of signal estimation in radar using passive receiver networks is addressed. A new method for estimating the radar signal is proposed, based on augmenting the set of signals received at the sensors with a signal chosen from among one of several appropriate signal classes. The choice within each signal class is optimized by maximizing the principal component of the augmented signal matrix, and the best among the signal classes selected as the final estimate. This approach is compared with the signal estimate obtained through unconstrained principal component analysis of the signal matrix. Simulations are presented that demonstrate improved performance using the proposed method and that show good estimates of the signal with signal to noise ratios as low as -20dB.