Multiple-symbol differential unitary space-time demodulation with reduced-complexity

We derive a new decision rule for multiple-symbol detection (MSD) of differential unitary space-time modulation (DUSTM). It is valid for both diagonal and non-diagonal constellations, and for quasi-static fading channels. We then present a fast algorithm based on the extended Euclidean algorithm and bound principles for single symbol detection (SSD) with diagonal constellations. We call it bound-intersection detection (BID) and it is exact maximum likelihood (ML). In high SNR, the complexity of our algorithm is much less than that of the brute-force ML search. We also develop two BID variants for MSD. They are ML but with significantly reduced complexity using branch-and-bound (BnB).