AU-KBC RESEARCH CENTRE

Wireless Communication Systems have evolved from ones that were primarily used for carrying voice traffic to the present day systems that primarily carry data. In fact, data services are expected to dominate the wireless communications scenario as people expect wireless systems to provide services akin to that of the wired netowrk. But the cost of wireless access is still very high compared to that of wires access and this has been a major hindrance to the widespread use of wireless data services. Spectrum contributes to a major chunk of the cost of wireless systems and hence future networks must utilize spectrum as efficiently as possible. In our thesis, we examine the issue of resource allocation and we propose a ‘Distributed Dynamic Packet Assignment’ method that uses the available spectrum more efficiently compared to conventional resource allocation methods. Supporting physical layer techniques like multi-carrier modulation schemes are expected to play a vital role in the implementation of our scheme. Besides improving spectrum efficiently, our scheme also supports high data rates. Data rates up to 1mbps have been demonstrated in our simulation and the technique can be extended to support even higher data rates.

Another important factor in data networks is the ability to provide different qualities of service to different types of traffic. This provides flexibility to the service provider and also paves the way for differential pricing policies that utilize the quality of service guaranteed to a user to determine pricing. In our work, we have implemented a scheduling scheme that minimizes the queuing delay suffered by real-time traffic, thus enabling aggressive QoS guarantees. We have demonstrated that efficacy of this scheme by comparing it with the performance of a First Come First Serve scheduler.