الفهرس 
Survey on MIMO Systems
Spatial MultiplexingOnly 14 pages are availabe for public view
 |  | from | 130 |  |  |
| | | from | 130 | | |
Abstract
MIMO systems attract a lot of attention nowadays and are expected to be good candidates in the future as they can increase the capacity of mobile communication systems. They adopt spatial multiplexing and eliminate multipath propagation errors through diversity gain. In all communication systems, the receiver has to jointly detect and decode the channel and the code, respectively, which is a time-consuming task even for the SISO case. However, in the MIMO scenario even trying to implement just the ideal channel detector is often very complex, because an ML decision on the transmitted vector will have to account for all the concurrent transmitted symbols possibilities. Such ML judgment will therefore be exponential both in the number of bits per symbol and the number of transmitting antennas. Sphere decoding algorithms try to apply the ML and MAP decision by establishing a search radius close to an initial estimate. This search can be visualized as finding the nearest lattice point to a noisy lattice observation. For these reasons, our thesis is dedicated to use a hybrid combination between the ML and SD with linear detectors such as ZF, MMSE, and V-BLAST.In this thesis, chapter 1: Introduce a general review of the thesis.Chapter 2: shows an overview about MIMO systems. This chapter introduce MIMO System model, MIMO system definitions, the features of using multiple antennas, the models of MIMO receiving and transmitting systems, MIMO system decoding techniques, and a review on channel estimation.Chapter 3 and chapter 4: low complexity and fast techniques are proposed to enhance the performance of MIMO detectors. These algorithms are ML/ZF, ML/MMSE, ML/V-BLAST/ (ZF or MMSE), SD/ZF, and SD/MMSE. These algorithms provide fast, low complexity, and near optimal performance as the ML. The main idea is based on dividing the transmitted symbols vector and the channel matrix into two equal size subsets. The ML or SD is used to detect the first subset and the other detection algorithms are used for the remaining symbols set. Chapter 5: discusses the conclusions and the future work.