Low-complexity iterative demodulation for noncoherent coded transmission over Ricean-fading channels

Power and bandwidth efficient noncoherent transmission over frequency nonse lective Ricean-fading channels is studied. We propose a remarkably low-comp lexity receiver structure, which is very well suited to mobile communicatio n scenarios with time-variant and nonstationary transmission channels. Appl ying bit-interleaved coded modulation with standard convolutional codes, su bstantial gains of several decibels in power efficiency compared to convent ional differential detection are achieved. To obtain the novel noncoherent reception scheme, ideas of iterative decoding with hard-decision feedback a nd prediction-based branch metric calculation are combined and extended. Fu rthermore, the incorporation of combined phase and amplitude modulation for high bandwidth efficiency is focused on. The theoretical analysis of both the convergence and the achievable performance of iterative decoding are gi ven by evaluating the corresponding prediction-error variance and the assoc iated cutoff rate, respectively. The results from information theory are we ll confirmed by simulation results presented for different channel scenario s.