The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	Interference suppression and diversity for CDMA systems
Author:	Cai, Xiaodong
View Online:	njit-etd2001-058 (xiv, 111 pages ~ 6.0 MB pdf)
Department:	Department of Electrical and Computer Engineering
Degree:	Doctor of Philosophy
Program:	Electrical Engineering
Document Type:	Dissertation
Advisory Committee:	Akansu, Ali N. (Committee chair) Ge, Hongya (Committee member) Haddad, Richard A. (Committee member) Haimovich, Alexander (Committee member) Winters, Jack H. (Committee member)
Date:	2001-01
Keywords:	CDMA Reduced Rank Packet Combining Transmit Diversity
Availability:	Unrestricted
Abstract:	In code-division multiple-access (CDMA) systems, due to non-orthogonality of the spreading codes and multipath channels, the desired signal suffers interference from other users. Signal fading due to multipath propagation is another source of impairment in wireless CDMA systems, often severely impacting performance. In this dissertation, reduced-rank minimum mean square error (MMSE) receiver and reduced-rank minimum variance receiver are investigated to suppress interference; transmit diversity is applied to multicarrier CDMA (MC-CDMA) systems to combat fading; packet combing is studied to provide both interference suppression and diversity for CDMA random access systems. The reduced-rank MMSE receiver that uses a reduced-rank estimated covariance matrix is studied to improve the performance of MMSE receiver in CDMA systems. It is shown that the reduced-rank MMSE receiver has much better performance than the full-rank MMSE receiver when the covariance matrix is estimated by using a finite number of data samples and the desired signal is in a low dimensional subspace. It is also demonstrated that the reduced-rank minimum variance receiver outperforms the full-rank minimum variance receiver. The probability density function of the output SNR of the full-rank and reduced-rank linear MMSE estimators is derived for a general linear signal model under the assumption that the signals and noise are Gaussian distributed. Space-time coding that is originally proposed for narrow band systems is applied to an MC-CDMA system in order to get transmit diversity for such a wideband system. Some techniques to jointly decode the space-time code and suppress interference are developed. The channel estimation using either pilot channels or pilot symbols is studied for MC-CDMA systems with space-time coding. Performance of CDMA random access systems with packet combining in fading channels is analyzed. By combining the current retransmitted packet with all its previous transmitted copies, the receiver obtains a diversity gain plus an increased interference and noise suppression gain. Therefore, the bit error rate dramatically decreases with the number of transmissions increasing, which in turn improves the system throughput and reduces the average delay.