FRFD MIMO Systems: Precoded V-BLAST with Limited Feedback Versus Non-orthogonal STBC MIMO

Abstract

Full-rate (FR) and full-diversity (FD) are attractive features in MIMO systems. We refer to systems which achieve both FR and FD simultaneously as FRFD systems. Non-orthogonal STBCs can achieve FRFD without feedback, but their ML decoding complexities are high. V-BLAST without precoding achieves FR but not FD. FRFD can be achieved in V-BLAST through precoding given full channel state information at the transmitter (CSIT). However, with limited feedback precoding, V-BLAST achieves FD, but with some rate loss. Our contribution in this paper is two-fold: $i)$ we propose a limited feedback (LFB) precoding scheme which achieves FRFD in $2\times 2$, $3\times 3$ and $4\times 4$ V-BLAST systems (we refer to this scheme as FRFD-VBLAST-LFB scheme), and $ii)$ comparing the performances of the FRFD-VBLAST-LFB scheme and non-orthogonal STBCs without feedback (e.g., Golden code, perfect codes) under ML decoding, we show that in $2\times 2$ MIMO system with 4-QAM/16-QAM, FRFD-VBLAST-LFB scheme outperforms the Golden code by about 0.6 dB; in $3\times 3$ and $4\times 4$ MIMO systems, the performance of FRFD-VBLAST-LFB scheme is comparable to the performance of perfect codes. The FRFD-VBLAST-LFB scheme is attractive because 1) ML decoding becomes less complex compared to that of non-orthogonal STBCs, 2) the number of feedback bits required to achieve the above performance is small, 3) in slow-fading, it is adequate to send feedback bits only occasionally, and 4) in most practical wireless systems feedback channel is often available (e.g., for adaptive modulation, rate/power control).