A computationally efficient algorithm for space-time block decoding

This paper presents a computationally efficient, low-power algorithm for the decoding of space-time block codes using a previously reported maximum likelihood algorithm (Tarokh et al. 1998). The new algorithm arrives at the same decision with less computation, and therefore reduces power dissipation for an actual implementation. Based on the previous algorithm, we develop a more efficient algorithm by utilizing the properties of the constellation signals and reported decision metrics. In most cases, the new algorithm results in more than 50% reduction in the computation complexity. The algorithm is illustrated for BPSK, QPSK, PSK or QAM.