Multiple-Symbol Differential Detection for Differentially Encoded LDPC Coded Systems

In this paper, multiple-symbol differential detection (MSDD) is applied to differentially encoded LDPC (DE-LDPC) coded systems, which can avoid phase tracking and channel estimation and compensate the performance loss of conventional differential detection. In the studied systems, the outer iterative decoding is performed between the MSDD softinput soft-output demodulator (SISOD) and the LDPC decoder. To make MSDD suit for iterative decoding, a metric computation algorithm which output soft information for MSDD is derived, and transfer characteristics and performances of the MSDD SISOD are analyzed by the extrinsic information transfer (EXIT) chart and computer simulations. To solve the problem of high complexity of the MSDD SISOD, we propose an improved soft-output M-algorithm (ISOMA) by combining the features of the existing SOMA approaches, and propose an adaptive ISOMA (AISOMA) to further reduce the complexity of the iterative decoding with ISOMA. By computer simulations, it is shown that the computational complexity of the MSDD SISOD as well as the iterative decoding complexity of DELDPC coded systems with MSDD can be significantly reduced by the two approaches, especially for high order modulation schemes. 