Performance of coded residual arithmetic differential MPSK modulation

A major attraction for differential modulation is in non-coherent reception for a communication system propagating over fast fading channels. However, conventional differential modulation (CDM) leads to a 3 dB loss over coherent modulation. To overcome this with minimal complexity, iterative decoding employing hard decision feedback and bit-interleaved coded modulation have been suggested. The channel encoder often employs convolutional codes which may require low rates or high constraint length to reduce the flooring effect in CDM. In this paper we propose residual arithmetic codes for channel coding. We show through simulated results that the arithmetic codes can achieve similar or better bit error rate (BER) performance than convolutional codes of the same rate and asymptotic coding gain at t = 1, 2 error correcting capability with minimal complexity. We invoke the Chase algorithm to provide the soft input to the hard decision algebraic decoder for the arithmetic codes