Study On Precoding Algorithm For MIMO Channel Based On GMD

In recent years, the wireless communication is developing rapidly. MIMOcan achieve space division multiplexing and improve the diversity gain throughincreasing the number of antennas at the transmitting and receiving, it can alsoimprove the spectrum efficiency and the reliability of communications withoutadditional bandwidth and transmission power. The wireless cooperative relayingtechnique can support higher data transmission speed, improve the spectrumefficiency, expand the coverage, resist fading and increase system capacity.MIMO cooperative relaying can be viewed as a new collaboration whichincreases the number of relay nodes. Therefor it increases the complexity of thecooperative transmission system, at the same time, it can provide highermulti-antenna gain.The research on the relay precoding design for MIMO cooperative relayingsystem becomes one of the important issues. The relay precoding can improvethe channel capacity and the link quality. However, some existing precoderdesigns adopt the linear precoding based on the zero forcing (ZF) criterion or theminimum mean square error (MMSE) criterion, assume a linear receiver at thedestination and a precoder at the relay. Some works even ignore the transmissionof the direct link (channel link from source to destination). Therefor theperformance enhancement of such a system is then limited.Fortunately, a kind of source-relay precoders design that adopts nonlinearreceiver at the destination appears. The most typical system is the source-relayprecoders design based on GMD, which has been adopted in two-hop system.This method significantly enhances the performance of MIMO cooperativerelaying system as compared to other precoder design methods.In this paper, we apply the source-relay precoders design based on GMD toMIMO cooperative relaying tree-hop system, fully exploring informationprovided by direct and relay links, transforming this design problem to a scalaroptimization problem with the help of matrix analysis theory. Finally, we solvethe optimal solution through mixed programming with the optimization softwareLINGO and C++. Simulations show that the proposed design can significantlyenhance the performance of a MIMO relay three-hop system. Because thethree-hop system has more advantages than two-hop system, such as, it has largercoverage and it can reduce the communication dead region, therefor the reseachon three-hop system has a greater value.