Resource Sharing in a WiFi-WiMAX Integrated Network

Abstract

Resource allocation is a challenging issue in integrated wireless environments due to factors like difference in participating network protocols, design of hierarchical network architecture, user mobility and multiple traffic types. In order to ensure fair access and efficiency of bandwidth usage among the component networks in an integrated network, resource allocation algorithms should be properly designed. The resource considered in this thesis refers to the wireless channel bandwidth.The design of such resource allocation algorithms becomes complex in wireless networks when compared to wired networks because of the error prone nature of the wireless medium. However, when a widespread Wi-Fi network is integrated with a next generation network such as WiMAX, both the subscribers and the service providers benefit significantly. This thesis describes two different resource allocation algorithms in an integrated network of Wi-Fi and WiMAX. The first algorithm is an improvement over an existing two threshold mechanism suggested by Yahiya et. al. for an integrated Wi-Fi and WiMAX network. In the existing model, the channels are classified as reserved and shared. Reserved channels are to be accessed only by specific users to which they are assigned and shared channels can be accessed by users of the component networks i.e. Wi-Fi or WiMAX based on availability. The proposed algorithm assigns priorities to the users and employs hybrid resource sharing. Complete sharing, complete partitioning and hybrid sharing approaches and their shortcomings in WiMax-Wi-Fi integrated networks are described. To overcome some of the shortcomings, a second algorithm called Prioritized Resource Sharing algorithm is proposed. In this algorithm, the channels are prioritized for different traffic classes rather than strict reservation or open access. The prioritized resource sharing algorithm’s objective is to improve channel utilization. Channel utilization is improved because prioritized sharing allows the users to access any of the channels if they are not being used, unlike in complete partitioning where users are restricted to use only allocated channels even when channels restricted to other users in the integrated network are not being used. Prioritized sharing method is an improvement over complete sharing as it guarantees a minimum number of channels due to the prioritization, unlike in complete sharing where no reservation is provided which could lead to users of one network capturing all the channels instead of fairly sharing them. Finally, the proposed algorithms are analyzed using two dimensional continuous time Markov chains. It is observed via simulations that using the forced termination model, in unbalanced loads both Wi-Fi and WiMAX component networks get unbiased service and neither network is starved. The simulation results also indicate that the Prioritized Sharing model achieves the best system utilization compared to other algorithms.