Data detection in MIMO systems with cochannel interference

The Bell Labs layered space-time (BLAST) architecture has been proposed to achieve high spectral efficiency on multi-input multi-output (MIMO) channels. Most studies of the BLAST algorithm consider spatially and temporally white noise and interference at the receiver. We study channel estimation and data detection of a MIMO system under both spatially and temporally colored interference. We derive maximum likelihood (ML) estimates of channel and spatial interference correlation matrices. By exploiting known temporal interference correlation, we extend one-time-slot ordered minimum mean-squared error (MMSE) ing detection to a multi-time-slot version. We evaluate the symbol error rate of an uncoded QPSK MIMO system under independent Rayleigh fading. The results show that by exploiting both spatial and temporal interference correlation, we achieve about 2dB gain in SIR for a 4×4 MIMO system.