Vertical handoff decision and network merit for integrated wireless and mobile networks

Abstract

Next generation wireless networks must be able to coordinate services between heterogeneous networks using a mobile terminal with multiple access interfaces. One of the challenges for continuous seamless mobility is vertical handoff which is the decision for a mobile node to switch for one network to another different type of networks. Moreover, how to select a desired target network is an important issue to balance against the network condition and user preferences. In this paper, the policies of multiple metrics to trigger a handoff while a mobile node is located in UMTS and WLAN/WiMAX are designed. The received signal strength (RSS) of neighbor networks predicted by the back propagation neural network is beneficial to perform handoff early. We propose a dwell time calculation depending to the user speed and moving pattern. If the RSS conditions are consistently true during dwell time, then unnecessary handoffs are avoidable. To select an optimal target network, policies in the merit function is presented. The weights of each handoff metric are dynamic to several networks. The results indicate that the proposed policy-enabled vertical handoff decision algorithm and network selection outperforms the other two approaches in reducing the number of vertical handoffs, Grade of Service (GoS) while increasing the average utilization per call of WLAN/WiMAX networks.