Abstract
In wireless communication, a signal is impaired by interference, multipath fading and channel noise. Diversity techniques can be used to reduce these impairments. In this thesis, an asynchronous multicarrier direct sequence code division multiple access (MC-DS-CDMA) system is addressed, and the minimum mean square error (MMSE) combiner is adopted as the diversity combining technique. As the combiner is dependent on a number of factors, including the spreading codes, channel fading coefficients, and relative delays among users, it needs to be computed from symbol to symbol. Therefore, it is desirable to have a reduced-complexity method for obtaining the combiner. A reduced-rank approach is employed in this thesis. The received signal is projected to the Krylov subspace, and the projection on the subspace is used as the observation to obtain a reduced-rank MMSE combiner. It is found that this reduced-rank combiner converges to a deterministic filter in a large system. The deterministic combiner obtained in the large system assumption is used as the approximation for the combiner in a finite-size system. Thus, there is no need to re-compute the combiner at every symbol. From the simulation results, it is found that the performance loss due to this approximated combiner is negligible.