Performance analysis of DS-CDMA in Nakagami fading multipath channel

Abstract

With the growth of wireless personal communication systems and wireless local area networks, the demand for reliable voice and data transmissions has become evident. The code division multiple access (CDMA) system meets the demand to a large number of users. A transmitted signal is affected severely by fading, and the reliable transmission is dominated by the bit error rate (BER) in wireless channels. In wireless channels with CDMA systems, the signals are corrupted severely by the interference due to other users who are not a desired user as well as the fading phenomenon. They are also polluted by background Gaussian noise somewhat. The standard CDMA adopts convolutional coding as encoding scheme, however the BER performance could be increased by using a trellis coded modulation (TCM) scheme with no parallel paths that has the shortest error event path in a trellis diagram as long as possible and the branch distance product of the path as large as can be. In particular, the line-of-sight (direct) component of a signal is more dominant than the scattered component in a satellite channel. In this case, Nakagami````s m-distribution is appropriate to model the channel. The Nakagami fading channel also includes Rayleigh, and approximates Rician distributions for a given value of m in comparatively small areas. In this dissertation, an analytical approach to improve the BER and throughput performance on direct sequence CDMA (DS-CDMA) in a Nakagami fading mulitpath channel is presented. To obtain this object, convolutional coding or TCM scheme is considered. The binary phase shift keying (BPSK) is also considered. It is assumed that the fading parameters and average signal powers per diversity branch are identical. The distance between diversity branches is so small that the correlation between them may occur. The occurrence of correlation deteriorates the BER performance. It is assumed to be equally correlated between diversity branches. In case of correlation occu...