Improvement in the grade-of-service in a cooperative overlay heterogeneous network

Abstract

We present a analytical platform for investigating the achievable performance gain when multiple networks open up their resources for sharing. We assume that an additional adaptation layer is possible to facilitate the sharing of resources. This paper examines the enhancement in the service capacity measured in terms of grade-of-service (GoS) and studies the tradeoffs incurred in terms of the vertical handoff probability in a cooperative overlay heterogeneous networks environment. We consider two separate networks having different amount of available radio resources and each providing a default service. The services also take up different amount of radio resources. If there are insufficient radio resources in the default network to serve the new call, we assume that the mobile terminals can resume their service requests in the other network. In our model, a particular service request is first served by the default network. When the default network has exhausted its resources, the terminal performs a vertical handoff and reconfigures itself so as to be served by the cooperating network. If that cooperating network is also unable to serve the request, the call will be blocked and cleared. All calls are served on a first-come-first- service basis. Using a four dimension Markov chain, we are able to model and study the performance and tradeoffs of the cooperative scheme. The call blocking probabilities and vertical handoff probabilities are computed for the respective networks. The results show a significant increase in service capacity and resource utilization with respect to the ratio of amount of radio resource available in the two networks. The analytical results derived from the Markov chain model show good agreement with the simulation results.