Reduced Complexity Tree Search Algorithms for MIMO Decoding

Published online: Dec 22, 2015 Full Text: PDF (1.91 MiB) DOI: 10.24138/jcomss.v10i4.119
Cite this paper
Authors:
Gajanan R Patil, Vishwanath K Kokate

Abstract

Maximum Likelihood Decoding (MLD) is computationally complex technique for decoding received information in multiple input multiple output (MIMO) systems. Tree search algorithms such as sphere decoding (SD) and QR decomposition with M survivals (QRD-M) are used to reduce the complexity keeping the performance near ML. This paper presents two techniques for reducing the computational complexities of the tree search algorithms further. The first technique is based on selecting the initial radius for sphere decoding. The main contribution of this paper is that the greedy best first search is used to compute initial radius, instead of Babai estimate. The second contribution is, QRD-M algorithm is modified to prune the nodes in the current layer based on maximum metric of child nodes of smallest surviving node. The performance of the proposed techniques is tested for different MIMO systems in terms of bit error rates (BER) and average number of nodes visited. The proposed schemes have improved computational complexity with no degradation of performance.

Keywords

Best First Search, Maximum Likelihood Decoding, MIMO, QRD-M, Sphere decoding
Creative Commons License 4.0
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.