Near-Optimal Low-Complexity MIMO Detection via Structured Reduced-Search Enumeration

TL;DR

It is demonstrated that for practical MIMO dimensions and modulation orders, near-ML hard-decision performance can be achieved using a structured reduced-search strategy with complexity linear in constellation size.

Abstract

Maximum-likelihood (ML) detection in high-order MIMO systems is computationally prohibitive due to exponential complexity in the number of transmit layers and constellation size. In this white paper, we demonstrate that for practical MIMO dimensions (up to 8x8) and modulation orders, near-ML hard-decision performance can be achieved using a structured reduced-search strategy with complexity linear in constellation size. Extensive simulations over i.i.d. Rayleigh fading channels show that list sizes of 3|X| for 3x3, 4|X| for 4x4, and 8|X| for 8x8 systems closely match full ML performance, even under high channel condition numbers, |X| being the constellation size. In addition, we provide a trellis based interpretation of the method. We further discuss implications for soft LLR generation and FEC interaction.

Figures (8)

• Figure 1: Spatial Trellis Interpretation of 3x3 MIMO (64QAM). Arc represents select min ( survivor) out of $|\mathcal{X}|$ outgoing paths
• Figure 2: Multi-Pivot in 3x3 MIMO 64QAM: Search one pivoting on $x_3$
• Figure 5: BER performance for 2×2 MIMO 256QAM: Full ML vs Reduced-Search ($2 |\mathcal{X}|$) vs ZF-QR
• Figure 6: BER performance for 3x3 MIMO 256QAM: Full ML vs Reduced-Search ($6 |\mathcal{X}|$ and $3 |\mathcal{X}|$) vs ZF-QR
• Figure 7: BER performance for 4×4 MIMO 64-QAM: Full ML vs Reduced-Search ($8 |\mathcal{X}|$ and $4 |\mathcal{X}|$) vs ZF-QR