Mimo Signal Detection Algorithm

In the past decade, multiple-input multiple-output (MIMO) technology has been widely concerned and studied because it can significantly increase channel capacity and spectrum efficiency without occupying any more bandwidth. Therefore, MIMO has been viewed as one of the key technologies of future wireless communications.The following aspects of signal detection in MIMO systems are analyzed and studied in this paper:1. To improve the performance of traditional detectors such as zero forcing algorithm, minimum mean square error algorithm, successive interference cancellation algorithm, a lattice reduction (LR) aided algorithm has been investigated in this paper. The simulation result shows that LR-aided detector makes the decision region being closer to ML, and the performance of the traditional detectors can be significantly improved.2. To reduce the complexity of sphere decoding (SD) algorithm achieving the performance of ML, a two-step scheme has been proposed. In this framework, a linear detector is firstly adopted to realize a coarse detection to obtain a coarse detected transmitting symbol vector (SDTSV), then, SD is used to search within a small hyper-sphere considering the probability that the exact transmitting symbol vector is far from the SDTSV is small. Simulation result shows that this detection framework takes full advantage of linear detectors and SD algorithm, so it has a near-SD performance with lower complexity.3. Due to high complexity of maximum likelihood (ML) algorithm, the semi-definite relaxation (SDR) algorithm and a near ML decoding algorithm which is based on semi-definite programming (SDP) have been studied in this paper. These two algorithms relax the constraints of ML algorithm and transform it into a convex problem which can be efficiently solved with a polynomial time. Simulation result shows that SDR algorithm reduces the complexity effectively with only a little performance loss, and the SDP shows a near-ML performance because it has a tighter bound.