Multi-constrained quality of service aware routing in mobile ad hoc wireless networks

Mobile ad hoc networks (MANETs) usually consist of battery operated mobile nodes. Conservation of battery life is important to increase the lifetime of the energy constrained ad hoc network. Energy aware routing brings about energy conservation by selecting energy efficient, routes. Real time applications running on MANETs impose another set of additional quality of service (QoS) constraints for routing packets. It is essential for these types of packets to arrive at the destination within a certain delay bound. Furthermore, real time flows require a minimum available bandwidth on the selected route as well as a bound on its end-to-end loss rate. The multi-constrained QoS routing problem for a mobile ad hoc network is to find a route in the ad hoc network that satisfies all the QoS requirements of the incoming traffic and at the same time reduces constrained resources consumption as much as possible. This thesis focuses on developing a framework for supporting multi-constrained QoS aware routing for mobile ad hoc networks; denoted as MCQAR. MCQAR enables the wireless mesh network nodes to gather QoS and constrained resource information of the network. Based on this information, a source node employs a heuristic algorithm to find a route that minimizes constrained resources consumption and satisfies incoming traffic QoS requirement bounds. MCQAR supports multiple classes of traffic, each with varying QoS requirements. MCQAR is implemented as an extension to the Dynamic Source Routing (DSR) protocol. Experiments on a real life testbed show MCQAR improves the delivery ratio of real time flows, at the same time reducing the per packet energy consumption as compared to DSR.