Iterative Detection Aided DL SDMA Systems Using Quantized Channel Impulse Response

An iterative detection aided down-link (DL) space division multiple access (SDMA) system is proposed. The base station (BS) requires DL channel state information (CSI) estimated by and fed back from the mobile station (MS) for separating the DL signals to be transmitted to the MSs from the BS. We investigate the achievable system performance subject to finite-accuracy CSI feedback. The achievable receiver performance is investigated with the aid of extrinsic information transfer (EXIT) charts. Furthermore, a novel concept, referred to as the EXIT-chart optimized CSI quantization (ECO-CQ) is proposed. The ECO-CQ scheme assists the system in maintaining the lowest possible CSI feedback overhead, while ensuring that an open EXIT-tunnel is still attainable, which implies maintaining an infinitesimally low BER. Furthermore, we demonstrate that the proposed ECO-CQ may reduce the normalized feedback overhead compared to the conventional CQ. For instance, the ECO-CQ aided iterative DL-SDMA system using an average of q = 2.7 quantization bits per CIR coefficient achieves a 10% normalized overhead reduction at E_b/N₀ = 5 dB, compared to the conventional CQ aided benchmarker system.