In the past, the design of PC algorithms for CDMA systems has remained at the physical layer to compensate for slow and fast channel impairments (known as fast PC and slow PC). The TDMA/FDMA manages inter-cell interference at the beginning of the radio planning process. In SS technology, real time adaptive PC and power management algorithms would need to work coherently to ensure reliable multi-media services, and the need for this real-time hybrid structure of PC and power management has only been shown recently. The emphasis in this dissertation is therefore on the design of a QoS-based PC structure in W-CDMA applications, the ultimate goal being to evaluate the new QoS-based PC structure by means of a Monte Carlo computer simulation; a multi-user, multimedia W-CDMA simulation package. Before the design of the QoS-based PC structure, this dissertation examines and proposes a new power-sensitive model that addresses factors affecting the W-CDMA system capacity. Consequently, PC problems are put into a framework for various optimization criteria. Finally the design of a QoS-based PC structure by means of Monte Carlo computer simulation is described and evaluate. The first problem is closely related to the fact that W-CDMA is a design of a power management network architecture. The power management can co-exist in every layer of operation with different specific time scale and optimization objectives. The solution to this problem is therefore to introduce a general and mathematically tractable power-sensitive model to identify factors that influence the capacity of W-CDMA cellular systems and then articulate the general power sensitive model to form a PC framework aimed at finding a common systematic treatment for different schools of thought on PC algorithms. This dissertation proves the benefits of layered PC operation for guaranteed QoS transmission and also shows that this research coincides with and extend the literature on PC management by categorizing PC algorithms according to various optimization objectives and time scales. The second problem is to evaluate the new QoS-based PC structure in a channel coded and RAKE combining uplink UMTSIUTRA cellular environment using the Monte Carlo simulation package. The UMTS radio channel models are described in terms of frequency-selective Rayleigh fading: Indoor-Office, Outdoor and Pedestrian and Vehicular environments. The package is simulated in Matlab. The influence of the number of multipath components, of Doppler Spread, the number of received antenna, the coding scheme and multi-access interference are discussed in the dissertation. The performance evaluation criteria for utility-based PC structures are Bit-Error-Rate (BER) performance (robustness), outage performance (tracking ability) and rate of convergence. The first test shows that the new proposed unbalanced step-size closed-loop FPC schemes can provide better SINR tracking ability and better BER performance than conventional balanced step-size PC schemes. The unbalanced FPCs have better PC error distribution in all scenarios. The second test shows that the proposed BER-prediction distributed OPC schemes can provide better BER tracking ability. This scheme converges iteratively to an optimal SINR level under current network settings with no excessive interference to other active users. / Dissertation (M Eng (Electronics))--University of Pretoria, 2006. / Electrical, Electronic and Computer Engineering / Unrestricted
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:up/oai:repository.up.ac.za:2263/28078 |
| Date | 21 September 2005 |
| Creators | Song, Ting-Chen Tom |
| Contributors | Prof L P Linde, Xia, Xiaohua |
| Source Sets | South African National ETD Portal |
| Detected Language | English |
| Type | Dissertation |
| Rights | © 2002, University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. |
Page generated in 0.0137 seconds