Efficient optimal joint channel estimation and data detection for massive MIMO systems

Abstract

Abstract-In this paper, we propose an efficient optimal joint channel estimation and data detection algorithm for massive MIMO wireless systems. Our algorithm is optimal in terms of the generalized likelihood ratio test (GLRT). For massive MIMO systems, we show that the expected complexity of our algorithm grows polynomially in the channel coherence time. Simulation results demonstrate significant performance gains of our algorithm compared with suboptimal non-coherent detection algorithms. To the best of our knowledge, this is the first algorithm which efficiently achieves GLRT-optimal non-coherent detections for massive MIMO systems with general constellations.

VLSI Designs for Joint Channel Estimation and Data Detection in Large SIMO Wireless Systems

“…Because the iterates remain bounded and f is continuous, the left-hand side of ( 25) is bounded from above. We can conclude that s (t−1) − s (t) → 0, and from (21), we see that the residuals converge as well.…”

“…While a considerable number of algorithms and VLSI designs have been proposed for small-and large-scale multi-antenna wireless systems that separate channel estimation and data detection (see, e.g., [8], [17], [19] and the references therein), only a few of results have been proposed for JED. Spheredecoding (SD) algorithms have been proposed to perform exact maximum-likelihood (ML) JED in SIMO and MIMO systems that use a small number of time slots [9]- [12], [20], [21]. Unfortunately, the complexity of SD methods quickly becomes prohibitive for larger dimensional problems [22], [23] and approximate linear methods, which are widely used for coherent data detection in massive MIMO systems [8], cannot be used for JED (see Section II-B for the reasons).…”

“…Furthermore, a theoretical error-rate performance analysis of PrOX is a challenging open research problem. Finally, while some algorithms exist for JED in the MIMO case and for general (non-constant modulus) constellation sets [12], [21], they are either computationally complex or exhibit a considerable loss with respect to the optimal ML-JED performance. The development of efficient JED algorithms that can be implemented as high-throughput VLSI designs for large (multi-user) MIMO systems with arbitrary constellations is part of ongoing work.…”

“…The assumption of having constant-modulus constellations limits our results to low-rate modulation schemes, such as BPSK and PSK constellations. While the (multi-user) MIMO case and higher-order QAM modulation schemes would be of interest in more general deployment scenarios, the associated ML-JED problem is significantly more challenging to solve, and requires different, more complex, and complicated algorithms; see, e.g., [12], [21] for more details-we are planning to address this scenario in the future.…”

Efficient DSP and Circuit Architectures for Massive MIMO: State of the Art and Future Directions

“…Cholesky-decomposition (Z = LL * ) has also been studied for Massive MIMO precoding and detection implementation [46], [47]. It has lower computational complexity than the Neumann series expansion method (with L ≥ 4) [39] and provides accurate processing independent of M and K. More importantly, the Cholesky decomposition imposes lower memory requirements, since only the lower triangular matrix L needs to be stored.…”