Novel Numerical Methods for Automatic Identification of Digital Signal in Dispersive Channels

Signal identification plays an important role for various applications and purposes. In this paper, we addressed a problem that did not receive great attention in the literature: numerical classification of signals that propagated through fading channels. We concentrated on the mobile application, and on fast time-selective dispersive channel with no pre-channel estimation. We present two novel systems for the same problem. The first one is based on the phase of the signal, and the approach taken is decentralization that named phase based signal type classifier (PBMC). The second classifier is based on the phase difference of the signal (for eliminating a constant phase offset) named phase difference based signal type classifier (PDBMC). Numerical simulations shows both structures suggest easy and very good solutions for the problem of automatic MPSK signal classification in fast time-selective dispersive channel even in low SNR.