Soft MIMO detection through sphere decoding and box optimization

Abstract

Achieving optimal detection performance with low complexity is one of the major challenges, mainly in multiple-input multiple-output (MIMO) detection. This paper presents three low-complexity Soft-Output MIMO detection algorithms that are based mainly on Box Optimization (BO) techniques. The proposed methods provide good performance with low computational cost using continuous constrained optimization techniques. The first proposed algorithm is a non-optimal Soft-Output detector of reduced complexity. This algorithm has been compared with the Soft-Output Fixed Complexity (SFSD) algorithm, obtaining lower complexity and similar performance. The two remaining algorithms are employed in a turbo receiver, achieving the max-log Maximum a Posteriori (MAP) performance. The two Soft-Input Soft-Output (SISO) algorithms were proposed in a previous work for soft-output MIMO detection. This work presents its extension for iterative decoding. The SISO algorithms presented are developed and compared with the SISO Single Tree Search algorithm (STS), in terms of efficiency and computational cost. The results show that the proposed algorithms are more efficient for high order constellation than the STS

Review on various detectors in massive MIMO technology: a performance analysis

“…This further found the simulation of the tree policy and default policy and backpropagation steps related to the MCTS. In 2018, Simarro et al (Ángeles Simarro et al , 2018) laid the foundation to the new optimization technique called BO techniques for the three low-complexity Soft-Output MIMO detection algorithms. Then, the continuous constrained optimization techniques were used to get better performance with a low cost of computation.…”