Reserch Of MIMO Detection Algorithms In Wireless Telecommunication System

The technology of MIMO is capable of greatly improving data rate of the communication system,which is generally treated as a key technology of the communication system in the next generation and has been a hot topic in academia and industry.This paper gives a research of several MIMO detection algorithms in wireless communication system,and some imporvement based on existing algorithms.The complexity of nonlinear detection algorithms can increase nonlinearly as the number of antenna increases,such as ML,whose complexity increases exponentially with the increasing of the number of antenna.The performance of ML is the best in all detection algorithms,but it is hard to be implemented industrially because of its complexity.While linear detection algorithms have relatively low complexities with poor performance.Sphere detection algorithm is a kind of fast search algorithm,which make its complexity lower by shrinking its searching areas.Compared with ML's exponential complexity,the complexity of sphere detection is polynomial.Sphere detection transforms MIMO detection into the problem of searching points in the sphere by controlling the changes of its radii.This paper researches the technology of detection algorithms in MIMO communication system.After learning the principle and relative knowledge and many existing detection algorithms,this paper gives a deep analysis of MIMO detection and has proposed three improved algorithms.First,with the characteristic of ML and OSIC,this paper proposes an improved OSIC algorithm based on ML.This algorithm employs ML to the layers of lower SNR,which guarantees accuracy rate of the remaining layers.It makes the performance of OSIC better without much improvement of its complexity,and even lower in some situations.Second,with the characteristic of QR decomposition detection,this paper proposes an improved OSIC algorithm based on QR,which employs QR decomposition detection and ML to OSIC.This algorithm guarantees not only layers of lower SNR,but also layers of higher SNR.Third,this paper proposes a new original radius selection strategy base on SNR threshold,after the research of radius choosing strategies in sphere detection.The proposed algorithm improved the problem of higher complexity because of improper selection of original radius.