Abstract :
The authors show that the decision feedback multiple-symbol differential detection (DF-DD) schemes proposed by Leib and Pasupathy (see IEEE Trans., vol.IT-34, p.1491-1501, 1988) and by Edbauer (see IEEE Trans., vol.COM-40, see p.457-60, 1992) are equivalent, and that DF-DD is an approximate realisation of maximum likelihood differential detection (ML-DD) proposed by Divsalar and Simon (see IEEE Trans. vol.COM-38, p.300-8, 1990). A theoretical BER analysis shows that DF-DD can achieve a BER performance almost the same as ML-DD.
