Secure Application-Aware Service Differentiation in Public AreaWireless Networks

We are witnessing the increasing demand forpervasive Internet access from public area wireless networks(PAWNs). As their popularity grows, the inherent untrustednature of public places and the diverse service requirementsof end users are two key issues that need to be addressed. We haveproposed two approaches to address these issues. First, the Home-based Authentication Protocol (HAP) that provides a frameworkby which to establish trust between a nomadic client and a serviceprovider using a trusted third party (home). Second, we argue thatthe best-effort-based service model provided by many access pointsis not enough to satisfy the end user fairness and to maximizethe wireless link utilization for a diverse user population.We have proposed an application-aware service differentiation (AASD)mechanism that takes both application semantics and userrequirements into consideration. Our analysis of this frameworkshows several fruitful results. The total authentication latencyincreases with the number of clients but at a rate that is much lessthan linear increasing latency. Also, in comparison with two otherbandwidth allocation approaches, the best effort and static accesscontrol, our proposed application-aware service differentiationmethod, outperforms them in terms of the client fairness andwireless bandwidth utilization.