For multiple-input multiple-output (MIMO) wireless communications systems, we propose a new zero-forcing (ZF) detection approach that explicitly accounts for instantaneous channel state information (ICSI) estimation error and spatial correlation. For this ZF approach, we derive an average error probability (AEP) expression for transmit-correlated Rician fading. Our AEP derivation exploits the effective signal-to-noise ratio that results by compounding ICSI estimation error and receiver noise. The derived AEP expression is then applied to evaluate MIMO ZF performance in Rayleigh and Rician fading for samples from recently-measured lognormal azimuth spread (AS) and Rician K-factor distributions, for pilot-based ICSI estimation. Numerical results depict the dependence of the AEP averaged over the AS and K distributions on fading type, rank of the deterministic component of the channel matrix, and AS-K correlation, for realistic scenarios.
展开▼