Multi-user detector for MC-CDMA using nonorthogonal wavelet neural networks

Abstract

Multi-user detection in a fading channel environment is a challenging problem in mobile communication systems like multicarrier code-division multiple access (MC-CDMA). In this paper we propose a multi-user detector using a nonorthogonal wavelet neural network. The proposed method operates in frequency domain to detect multi-users in a frequency selective fading channel environment. Kalman filter algorithm is used to estimate the parameters of the neural network adaptively. Nonorthogonal bases have a good resolution when constructing a wavelet network than orthogonal basis. It is due to the difficulty of obtaining an analytical solution to generate orthogonal basis to constrict a wavelet network with better resolution. The performance of the proposed detector shows the ability of adaptive multi-user detection in a Rayleigh fading channel