Title

New Schemes of MUD for Synchronous DS-CDMA & its Overloaded Systems

Abstract

The demand for better techniques in the field of multiuser detection for CDMA and WCDMA systems has grown especially in the last decade. This is due to the exponential growth in the number of subscribers and high data rate requirements due to multimedia and wireless internet services. However, there are many formidable limiting factors on the way to achieving this objective. One of the most challenging issue is the multiple access interference (MAI) in multiuser detection (MUD). Another issue is the computational complexity of the multiuser detector, which grows exponentially with the number of users for maximum likelihood detector (MLD). This will prohibit the use of MLD as multiuser detector unless its computational complexity is decreased substantially. These two major issues have been addressed in this dissertation.

The problem of MAI has been tackled in a simple manner in this dissertation by using the concept of an additional redundant user, named as ‘Pseudo-user’, for DSCDMA systems. In this proposed scheme, the matched filters followed by minor linear processing have resulted in the removal of MAI completely. This scheme is theoretically independent of the number of users. However, this has been achieved at the expense of a minor percentage loss in bandwidth and noise enhancement. The percentage loss of bandwidth can be decreased by increasing the number of users

The complexity issue of MLD has been addressed by using particle swarm optimization (PSO) technique. Two soft versions of PSO have been proposed. The results achieved are almost optimal, but at a much lower computational complexity as compared to that of conventional approach for MLD.

The above two schemes, i.e. the pseudo-user scheme and PSO scheme along with multicarrier modulation, have been applied to WCDMA overloaded systems in order to assert their practical applicability. The proposed schemes were tried on two types of channels ─ simple additive white Gaussian noise (AWGN) channel and slowly flat fading channel. The results have been compared with some standard techniques as well as recently reported techniques in the literature.

Download full paper