Parameter estimation and receiver design for overlay communication systems.

摘要

The design and implementation of wireless communication systems have seen tremendous growth recently. This will continue in the years to come as wireless multimedia, such as voice, video and data, becomes pervasive. To address the demands, higher rate services need to be provided which indicates larger bandwidth. There are many new problems in the design of such wideband systems. First of all, due to limited band resource, the new systems may have to be overlaid upon the existing narrowband systems, this usually implies a multirate signal structure since the target services are different. Secondly, to enable the provision of multimedia services, multirate signal detection needs to be tackled within the new systems. Also, channel characteristics will change with respect to the increased bandwidth. All these make new receiver designs imperative.; We propose two solutions to overlay applications. Specifically, the minimum mean square error (MMSE) and the parallel interference cancellation (PIC) receivers are developed. We address the problem from a group-blind perspective since information such as the number of active users and band occupancies usually is not exchanged between different systems. Simulations demonstrate the effectiveness of the methods and indicate that for the N-CDMA system, the proposed MMSE receiver performs better than the proposed PIC receiver; for the W-CDMA system, the opposite is true when the system is lightly loaded, when there are many users in the system, it is always better to exploit the MMSE receiver.; For new wideband system designs alone, we first focus on multirate CDMA which has been standardized for the third generation cellular systems. Then attention is given to single rate systems. Among others, channel estimation is a very important issue since it is typical of wideband systems that channels tend to be frequency selective, in clear contrast to narrowband systems. We thus adopt the tapped-delay-line model and study the estimation of the tap coefficients. In simulations, we combine channel estimates with existing signal detection strategies and obtain satisfactory system performance.