Research On Multiuser Detection In Spread Spectrum Communications Systems

Mobile communications systems have been developed very fast in the past decades. The second generation mobile communications systems are gradually replaced by the third generation mobile communications systems. The third generation mobile communications systems have the advantage of large capacity, low consumed power, fine voice quality and confidentiality. CDMA is the optimal multiple access method of the third generation mobile communications systems, and the three major standards of IMT-2000, that is, WCDMA, cdma2000 and TD-SCDMA, use CDMA technique. But because CDMA is a self-interference system, there exist severe MAI and near-far problem which reduce the performance and capacity of CDMA. Great importance has been continuously attached to multiuser detection, one of the key techniques for wideband CDMA communications systems. In this thesis, we do research on multiuser detection in the third mobile communications systems. Specific contributions are as follows:1. The minimum bit-error rate linear multiuser detection problem can be transformed to a bit-error rate cost function problem. Firstly, the cost function is analyzed and a mathematical map is proposed, and the coefficient vector of linear multiuser detectors is mapped into the interior and surface of a hyperellipsoid in K-dimensional Euclidean space, that is, the problem can be studied from new solution space and a new cost function is given. By the mapping, the minimum bit-error rate linear multiuser detection problem can be converted into an optimization problem for the new cost function which is subjected to the constraint of the hyperellipsoid. Then the constrained optimization problem is analyzed in detail. It is proved that all the local minimizers of the constrained optimization problem lie on the surface of the hyperellipsoid. The eigen set of equations of the constrained optimization problem is deduced. It is proved that any local minimizer is a solution of the egien set of equations and any solution of the eigen set of equations satisfies the necessary conditions as a local minimizer, which means the minimum bit-error rate linear multiuser detection problem is equivalent to the eigen set of equations.2. The new expression of the of conventional matched filter detector, the decorrelating detector and the MMSE detector are given based on the mapping and it is proved all these...