Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
452309 | Computer Networks | 2009 | 17 Pages |
The dynamic nature of mobile ad hoc networks poses fundamental challenges to the design of service composition schemes that can satisfy the end-to-end quality of service requirements and minimize the effect of service disruptions caused by dynamic link and node failures. Although existing research on mobile ad hoc networks has focused on improving reliability, little existing work has considered service deliveries spanning multiple components. Moreover, service composition strategies proposed for wireline networks (such as the Internet) are poorly suited for highly dynamic wireless ad hoc networks.This paper proposes a new service composition and recovery framework designed to achieve minimum service disruptions for mobile ad hoc networks. The framework consists of two tiers: service routing, which selects the service components that support the service path, and network routing, which finds the optimal network path that connects these service components. Our framework is based on the disruption index, which is a novel concept that characterizes different service disruption aspects, such as frequency and duration, that are not captured adequately by conventional metrics, such as reliability and availability.Using the definition of disruption index, we formulate the problem of minimum-disruption service composition and recovery (MDSCR) as a dynamic programming problem and analyze the properties of its optimal solution for ad hoc networks with known mobility plan. Based on the derived analytical insights, we present our MDSCR heuristic algorithm for ad hoc networks with uncertain node mobility. This heuristic algorithm approximates the optimal solution with one-step lookahead prediction, where service link lifetime is predicted based on node location and velocity using linear regression. We use simulations to evaluate the results of our algorithm in various network environments. The results validate that our algorithm can achieve better performance than conventional methods.