Iterative Detection for BLAST Systems with Arbitrary Number of Receive Antennas

Turbo coded Vertical Bell Labs Layered Space-Time Architecture (V-BLAST) is a multiple antenna system employing turbo code and bit-interleaving to offer flexible rate and high speed wireless data communication. The conventional detection techniques based on one-shot demodulation and decoding algorithms, can suffer poor performance or simply fail when there are fewer receive antennas than transmit antennas or when the multiple- input-multiple-output (MIMO) channel exhibits strong spatial correlation. We propose employing an iterative demodulation and decoding algorithm that solves these problems and greatly extends the applicability of V-BLAST. Simulations demonstrate that iterative V-BLAST offers the best-known performance in a wide range of settings. Through complexity analysis, it is shown that a) the incremental computing complexity is small but additional memory is required compared with conventional ML decoding, and b) an iterative algorithm employing generalized sphere decoding will be a low complexity, reduced memory, and high performance solution. Extensions to the technique beyond simple V-BLAST are given. When there exists feedback from the receiver to transmitter, we illustrate how to adapt code rate, modulation, and power distribution to take advantage of the channel knowledge. With a combination of "best in class" receiver performance and the ability to decode V-BLAST transmissions with fewer antennas at the receiver than at the transmitter, the techniques described here can be viewed as key "technology enablers" for BLAST/MIMO in next generation systems.