Date of Award


Document Type

Master Thesis

Degree Name

Master of Engineering (Research)


Electronic Engineering

First Advisor

Dr. Dirk Pesch


Recent years have seen a dramatic increase in demand for mobile communication services and with the introduction of 2.5G services such as general packet radio service, this trend is expected to increase further. These new services will introduce highly dynamic tele-traffic variations due to the inherent variability in resource usage of the new data services. The net effect will be considerable increase in load on the available radio resources. The currently employed fixed channel allocation scheme lacks the flexibility to support such traffic variations. A plethora of dynamic channel allocation schemes have been proposed in the past in an effort to increase network capacity by managing the available radio resources in a more flexible manner. Most of these schemes adopt a reactive approach, which makes their incorporation into current cellular networks impractical, due to their high levels of complexity and signalling load. In contrast, this study takes a more pro-active approach to radio resource management in which the cellular network frequency plan adapts itself to the varying traffic conditions by predicting future radio resource requirements for new and handover circuit-switched GSM call traffic and packet-switched GPRS sessions using neural networks. The predicted resources are then assigned to cell- sites in the network using a genetic algorithm. As each cell-site receives just the required number of frequencies for the next hour, a much more flexible allocation of resources throughout the network is obtainable. This study also investigates the effects of increased load in the network and makes direct comparisons in the resource gain achievable between the currently employed fixed channel allocation and proposed adaptive radio resource management schemes for different load scenarios. A dynamic channel allocation scheme is also implemented so as to make comparisons of the resource gain achievable with the proposed adaptive scheme and that of dynamic channel allocation. The non-invasive implementation of the proposed scheme makes it an attractive strategy for the provision of increased radio resources and their management in the evolution toward 2.5G systems.

Access Level