Performance evaluation of power-aware routing algorithms in personal networks

Abstract¿Personal Networks (PN) is a concept in the area of modern communication networks in which devices belonging to a person cooperate to offer services to the person regardless of their geographical locations. A PN may consist of several adhoc sub-networks (or clusters) which are interconnected through infrastructure networks such as the Internet. Failure of personal devices due to battery shortage is one of the main threads for service availability in PNs. Power-aware routing (PAR) has been proposed as an effective scheme to prolong the lifetime of battery-operated nodes in wireless ad-hoc networks. This scheme, however, could be utilized in PNs as well, since a PN cluster can be modeled as a wireless ad-hoc network. Aiming at prolonging the lifetime of nodes in PNs, in this paper we propose a PAR algorithm for PNs and we evaluate and compare its performance with the performance of other PAR algorithms proposed so-far for wireless ad-hoc networks. We analyze the effect of node density, routing overhead, heterogeneity of PN nodes in terms of their power supplies, and communication pattern within PN clusters on the performance of the PAR algorithms. Taking into account various parameters, we show that in a PN environment, our proposed scheme can profoundly extend the lifetime of PN nodes compared to the other algorithms. We also show that there can be an optimal value for route refresh interval, as a parameter which affects the generated routing overhead, and the node density of PN clusters, in such a way that the network lifetime is maximized.