Adaptive soft-decision algorithms for mobile fading channels

This paper addresses the task of joint symbol-by-symbol decisions and channel estimation in a coded/interleaved/frequency-selective fading channel. Based on two popular soft-decision algorithms, namely OSA and SSA, which have an inherent trellis structure, several adaptive derivatives are proposed. In particular. by extending the concept of Conventional Adaptive MLSE (CA-MLSE) receiver to the soft-decision case, the CA-OSA and CA-SSA algorithms are derived. Moreover, using the Per-Survivor Processing (PSP) concept, two new adaptive extensions are presented, namely PSP-OSA and PSP-SSA. The latter of these algorithms (PSP-SSA) has a simplified form with complexity comparable to PSP-MLSE. Simulation results show a significant improvement of the suggested adaptive soft-decision algorithms over the standard configuration, which consists of a hard-decision channel equalizer followed by a convolutional decoder.