| This dissertation studies blind adaptive interference cancellation and multiuser detection schemes for Code Division Multiple Access (CDMA) systems. The idea is to use the tremendous capabilities of modern digital signal processing (DSP) algorithms to solve problems in CDMA wireless communications. The major contributions of this dissertation are summarized below. (1) Present a blind subspace tracking algorithm that does not require knowledge of spreading codes, signal energies and time delays for all but the user of interest. Simulation results showed comparable performance in interference cancellation between a receiver that had other users' information and a receiver that did not have other users' information, but instead, used the blind algorithm. The blind algorithm is especially useful in downlink situations, where a mobile user receives interfering signals from many other users, but does not know anything about the other users. (2) Develop a projection receiver that incorporates branch processing. As shown theoretically and demonstrated by experimental results, a projection receiver with branch processing results in lower bit-error-rate, i.e., better interference cancellation, than the decorrelating receiver in all operating conditions or the MMSE receiver in a wide range of operating conditions. (3) Present an adaptive algorithm for multipath fading channel estimation. This algorithm allows the receiver to track the channel variations based only on the received signals and the spreading code of the user of interest. As will be seen in its derivation, the adaptive channel estimator cancels the interference among the replica signals of the user of interest. Therefore, it is superior to the matched filter based estimator used in current CDMA systems. (4) Finally, this dissertation incorporates blind subspace tracking, branch processing, and adaptive channel estimation techniques into the projection receivers with forward error control coding. Computer simulations illustrate that this receiver can adaptively and effectively operate in severe communication environments. |
