A method of single channel blind source separation of co-frequency 16QAM signals

Abstract

With the rapid development of wireless communication, the receiver equipped one antenna receives more than one signal at the same frequency band simultaneously. For the above-mentioned reasons, the single channel blind source separation (BSS) becomes a hot topic in signal processing field. Nevertheless, the study of single channel blind source separation have great challenges since this problem is usually ill-posed one. Many scholars make hard efforts to solve this ill-posed problem. In this paper, one method is proposed to separate two co-frequency16QAM overlapped signals based maximum likelihood method (ML) since communication signals have finite set of symbols and some preamble symbols we already know. The performance of the method through computer simulations shows its reliability.