Channel management schemes for the RNC system in wireless communications network

Abstract

This thesis was concerned with RNC system, which is one of the subsystems in the next generation wireless networks, and its main role is to perform channel assignment at the request of base stations (BS``s) and mobile switching center (MSC). In CHAPTER 2, we dealt with channel management of the RNC system that is capable of supporting multimedia traffic. A basic model based on the Markov chain is introduced to evaluate the performance of a simplified RNC system, which has one THP. According to the schemes of resource partitioning for two types of service calls and channel assignment policies for non-voice call, we propose four channel allocation strategies, ACAP-SC, RCAP-SC, ACAP-DC and RCAP-DC along with the associated Markovian model. The validity of the models is examined in terms of the blocking probability of the system by comparing the simulation results. Through sensitivity studies, it is shown that RCAP-00 generally outperforms ACAP-00 due to the efficiency of random allocation policy on sever utilization. However, we find that ACAP-00 has better performance than RCAP-00 at high traffic intensity with low proportion of Type2 calls. Considering the advantage of random assignment policy over adjacent assignment policy in resource utilization point of view, the observation seems unexpected one. The principle cause comes from the nature of the current performance measures ($P_B, T_S$), in that they do not fully reflect the difference in the resource consumption level of each type of traffic calls. CHAPTER 3 dealt with channel management of the RNC system with multiple Traffic Handling Processor (THP) that is capable of supporting multimedia traffic. A basic model based on the Markov chain is introduced to evaluate the performance of a RNC system with multiple THPs. According to the schemes of resource partitioning for two types of service calls, we propose two channel allocation strategies, RSS and RPS along with the associated Markovian model. The validity...