Energy conservation for wireless ad hoc routing

Self-configuring wireless ad hoc networks have attracted considerable attention in the last few years due to their valuable civil and military applications. One aspect of such networks that has been studied insufficiently is the energy efficiency. Energy efficiency is crucial to prolong the network lifetime and thus make the network more survivable.;Nodes in wireless ad hoc networks are most likely to be driven by battery and hence operate on an extremely frugal energy budget. Conventional ad hoc routing protocols are focused on handling the mobility instead of energy efficiency. Energy efficient routing strategies proposed in literature either do not take advantage of sleep modes to conserve energy more efficiently, or incur much overhead in terms of control message and computing complexity to schedule sleep modes and thus are not scalable.;In this dissertation, a novel strategy is proposed to manage the sleep of the nodes in the network so that energy can be conserved and network connectivity can be kept. The novelty of the strategy is its extreme simplicity. The idea is derived from the results of the percolation theory, typically called gossiping. Gossiping is a convenient and effective approach and has been successfully applied to several areas of the networking. In the proposed work, we will develop a sleep management protocol from gossiping for both static and mobile wireless ad hoc networks. Then the protocol will be extended to the asynchronous network, where nodes manage their own states independently. Analysis and simulations will be conducted to show the correctness, effectiveness and efficiency of the proposed work. The comparison between analytical and simulation results will justify them for each other. We will investigate the most important performance aspects concerning the proposed strategy, including the effect of parameter tuning and the impacts of routing protocols. Furthermore, multiple extensions will be developed to improve the performance and make the proposed strategy apply to different network scenarios.