Mixed Signal Detection Based on Second-Order Cyclostationary Features

Cyclostationary analysis has heen widely used in signal detection, RF parameter estimation and modulation detection. However, in most of existing work, the target signal is a single communication signal without overlap with other signals in the frequency domain. Hence, it is feasible to use a filter to first distinguish the target signal out and perform these tasks next to avoid the interferences from the spectrum environment. However, in a spectrally congested environment such as cognitive radio and dynamic spectrum access network, or in a spectrally contested environment such as a hattle field, multiple signals are often mixed together with significant overlap in spectrum. It is highly desired to find effective methods to perform signal detection, RF parameter estimation and modulation detection for mixed signals. In this paper, we employ second-order cycloslationary feature, namely the spectral correlation function (SCF), to detect the components of mixed signal. It is shown that by performing second-order cyclostationary analysis, we can successfully detect the existence of multiple signal components in mixed signal, and estimate their carrier frequencies accurately. Simulations over fading channels at different signal to noise ratios validate the effectiveness of the proposed method. In future works, we will investigate the feasibility and performance of signal classification/modulation detection of mixed signals in complex environment using cyclostationary analysis.