Time-frequency peak filtering for the recognition of communication signals

Most existing classification methods cannot work in low signal-to-noise ratio (SNR) environments. This limitation motivates the signal filtering before the classification process. In this paper, a general framework that links the time-frequency peak filtering (TFPF) and traditional feature-based signal classification is explored. As the name suggests, TFPF is a filtering approach to encode the received signal as the instantaneous frequency (IF) of an analytic signal, and then the filtered signal is obtained by estimating the peak in the time-frequency domain of the encoded signal. The proposed framework is tested on the recognition of some communication signals. Numerical results demonstrate the effectiveness of this classification scheme for heavily noise corrupted signals. The TFPF based signal classification method exhibits a much better classification performance than the cases where the filtering process is not used.