Soft parallel decoding of per-antenna-coded MIMO systems

In this paper, we consider a soft parallel decoding approach using probabilistic data association (PDA) algorithm for per-antenna-coded (PAC) transmission on multiple-input multiple-output (MIMO) channels. The proposed algorithm allows for decoding of layered signals in a variety of contexts: downlink and uplink in multi-user systems, single user transmission for both real-time and data traffic, systems with fewer receive than transmit antennas and with or without channel state information at the transmitter. We investigate performance of the proposed algorithm in turbo-coded MIMO systems using 16, 64, 256-QAM and spectral efficiencies ranging from 2 to 6 bpcu/layer. We present frame error rates for selected scenarios with the corresponding lower bounds on channel outage probabilities. We also briefly discuss the convergence properties of the PDA algorithm with turbo encoded signals.