Open AccessLattice reduction‐based iterative receivers: using partial bit‐wise MMSE filter with randomised sampling and MAP‐aided integer perturbation
Author(s) -
Bai Lin,
Li Tian,
Zhao Lewen,
Choi Jinho
Publication year - 2016
Publication title -
iet communications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.355
H-Index - 62
eISSN - 1751-8636
pISSN - 1751-8628
DOI - 10.1049/iet-com.2015.1062
Subject(s) - lattice reduction , algorithm , minimum mean square error , mathematics , maximum a posteriori estimation , detector , decoding methods , computational complexity theory , performance improvement , mathematical optimization , computer science , statistics , telecommunications , mimo , beamforming , maximum likelihood , estimator , operations management , economics
For iterative detection and decoding (IDD) in multiple‐input multiple‐output systems, the maximum a posteriori probability (MAP) detector would be ideal in terms of the performance. However, due to its high computational complexity, various suboptimal low‐complexity approximate MAP detectors have been studied. In this study, a lattice reduction (LR)‐based detector is considered for a near‐optimal performance for IDD. The authors improve further the performance by employing a partial bit‐wise minimum mean square error (MMSE) approach with randomised sampling, which has a lower complexity than that of the full bit‐wise MMSE method. Moreover, the list of candidate vectors obtained by randomised sampling is extended using a MAP‐aided integer perturbation algorithm for a better performance with low additional complexity. Through simulation results, it is shown that a near‐optimal performance can be obtained which is better than that of the LR‐based randomised successive interference cancellation and the full bit‐wise MMSE methods.