Performance of code division multiple access on multipath channels - an exact analysis (1994)
Type of ContentTheses / Dissertations
Thesis DisciplineElectrical Engineering
Degree NameDoctor of Philosophy
PublisherUniversity of Canterbury. Electrical and Electronic Engineering
AuthorsCowl, David J.show all
This thesis presents an exact analysis approach to the investigation of the average probability of error performance for a code division multiple access (CDMA) system operating over multipath fading channels. The system in question has K users transmitting direct sequence spread spectrum signals over the same bandwidth. The receiver employs a selection diversity algorithm to select the path with the strongest desired signal at any given time. Performance results are presented for three classes of channel model, those being the single path non-fading channel, the single path fading channel and the multiple path fading channel. The performance is also presented for a system where ideal power control is applied. The channel model for the single path fading channel is a single path whose gain has a Rayleigh distribution. The channel model for the multiple path fading channel is a tapped delay line model, where each tap is equally spaced in time, and has a gain with a Rayleigh distribution. The average gain for each tap is specified by an average delay profile. The exact analysis approach involves finding the probability density functions (pdf's) for the per-user multi-user interference and the inter-symbol interference from the desired user. These probability density functions are derived from the pdf's of each contributing factor to the interference, including the empirical cross-correlation function pdf. The characteristic functions of the per-user multi-user interference and the inter-symbol interference are found from the pdf's, and the characteristic function of the total interference is given by the product of the inter-symbol interference characteristic function and the K - 1 multi-user interference characteristic functions. The average probability of error is calculated using the characteristic function of the total interference. Conclusions on the performance of the selection diversity algorithm and the power control algorithm are drawn, based on the results.