A New Approach to Non-Stationary Signals Detection Using Adaptive Matched Filter

In this work, we propose a non-stationary signal detection algorithm based on matched filtering to significantly improve the processing performance of high-speed rendezvous target. We transform the target detection problem into a multivariate hypothesis test problem of target signal duration. We derive the test statistics, and prove that the test statistics can be calculated by adaptive matched filter. By searching for the maximum test statistics under each hypothesis test, we can simultaneously make hypothesis test decisions and estimate target trajectory parameters. The simulation results show that the performance of the constructed detector is consistent with the theoretical analysis results. Compared to the traditional two-step method, the proposed algorithm can reduce the estimation error of the signal duration by two orders of magnitude. And it also brings a significant improvement in the performance of the trajectory parameter estimation.