A New Polynomial Phase Signal Detection Algorithm Based on the Combination of Filter Bank and Short-Time Radon-Wigner Transform

In this paper, we are interested in the problem of detection polynomial phase signals (PPS) affected by additive Gaussian white noise. A new algorithm for PPS detection is proposed. It is based on the time domain filter-banks method combined with the short time Radon-Wigner transform method. Because of the band-limited character of PPS, the subband signal´s signal-to-noise ratio (SNR) can be enhanced by using filter bank method. On the other hand, the local PPS can be approximated by linear frequency modulated (LFM) signal, so short time Radon-Wigner transform is used to detect and estimate the local signal´s instantaneous frequency. The merit of this new algorithm is validated by illustrative examples in low (SNR) condition.