Fisher Linear Discrimination Based Multiuser Detector
Wen He · Wireless Communication Technology · 2007
This paper proposes a new strategy for multiuser detection which is based oo Fisher linear discriminant analysis.New detector imply the time variety channel characteristic in the optimal projection direction vector that obetained from training process.It is suitable for various channel entironment in spite of any type of noise or interference.Besides training samples the detector don't need signature sequence of the other users in the system.It was shown via simulations that these algorithms can serve as means to reduce the computational load while giving excellent performance compare with MMSE and SVM detector.