Abstract
The mobile communications industry has undergone rapid developments in the last decade. Consumers and businesses have become increasingly technology savvy and are likely to expect creative wireless access to many services. Consequently, future mobile systems are envisioned to support a wide variety of applications. The support of these diverse services (e.g. data and multimedia) has led to different bandwidth requirements, inevitably rendering many of the traditional radio resource management (RRM) mechanisms deficient. For one, the present radio resource dimensioning strategy that dimensions radio resource based on "busy hour" traffic does not allow operators to efficiently utilised the radio resource, which is scarce and limited. In line with the future demand driven resource allocation, two novel Scalable Resource Allocation (ScRA) techniques are proposed in this thesis to optimise the utilisation of radio resources within multiple services. The proposed novel ScRA techniques performances are evaluated on a typical GSM/GPRS mobile network through our developed dynamic system level simulator. Based on the simulation results, significant radio resource utilization gain, higher data throughput, lesser delay for data connections can be achieved with the use of our ScRA technique over the conventional StRA technique. For example, when the WWW arrival rate reaches approximate 35 Kbits/sec, ScRA technique can attain the throughput difference as high as 27.663%. On the other hand, the system throughput gain can yield as high as 27.44% when the Email arrival rate reached 1769 bits/sec for ScRA technique. Above all, we have also verified that the ScRA techniques perform well within the consideration of desirable voice QoSs at all times. The better performance of this technique is credited to its flexible and adaptive radio channel allocation (or re-allocation) between the multiple services during each sampled time. Key words; Radio Resource Management, Call Admission Control, Scalable Resource Allocation.