Optimized Service Selection for MANETs using an AODV-based Service Discovery Protocol *

networks, which operate without the need of an infrastructure. In such networks, mobile nodes must rely on their own capabilities and on other peers' capabilities to perform their tasks. In this context, many service discovery protocols have been proposed in order for mobile peers to be able to discover and take advantage of each other's services. A basic building block for service discovery protocols is their service selection strategy. Especially in service-rich environments with many (possibly resource poor) mobile nodes, the service selection strategy employed is of major importance. In this paper we investigate the impact of two basic and easy to implement service selection strategies on the lifetime of mobile servers. The first strategy takes into account hop-based server proximity and promotes the selection of the nearest mobile server, while the second strategy takes into account the remaining energy of service providers and promotes the selection of the mobile server with the maximum remaining energy. Through extensive simulations, we show that the provider's remaining energy yields the best performance when used as a service selection criterion under most situations, and that the shortest path selection criterion presents very competitive performance. The performance metrics used for evaluating both service selection strategies are service and network lifetimes, service success and service discoverability ratios.