Design And Implementation Of An Energy-Efficient Routing Protocol For Wireless Sensor Networks

The resource limitations of wireless sensor networks (WSNs), in terms of energy, computing power, bandwidth and memory storage, pose significant challenges for the design of routing protocols. Moreover, the amount of sensor nodes in deployment is very huge and each node only knows its own local topology information. In addition, the topology of a WSN is highly dynamic caused by frequent node mobility, node failure and unreliable wireless medium. It makes that conventional routing protocols for wireless ad hoc networks are not feasible for WSNs.In this report, we first presented an energy-efficient on-demand routing protocol (ESR) for WSNs. The routing protocol enables nodes to reduce energy consumption, simplify computational complexity and only store small-length routing information. Moreover, ESR works collaboratively with the underlying MAC layer, which helps to provide local topology information of nodes, so that ESR recovers the broken links efficiently, and also finds out the optimal route during routing maintanence. In route reestablishment, a limted directional flooding alleviates costly power consumption. Then, we simulated ESR on the OMNeT++ simulator and presented the performance evaluation results. Finally, we implemented ESR on a 12-node multihop WSN, using cross-layer information from the underlying LMAC protocol on top of AmbientRT operating system. Both simulation results and implementation results show that ESR achieves improved throughput performance over conventional ad hoc routing algorithms (DSR, AODV), while greatly reducing power consumption on the routing control overhead. The main contribution of this paper is as following.Firstly, we presented an energy-efficient on-demand routing algorithm for WSNs, which has a fast recovery mechasnims, relying on MAC layer feed back to counter the mobility and unreliability of node. In the route reestablishment, a geographically restricted directional flooding scheme based on the previous knowledge of the location of the destination is deviced.Secondly, we simulated the routing protocol in the simulation environment, and further implemented it in a WSN consisting of sensor node prototypes using LMAC on top of AmbientRT operating system. The results of performance evaluation show that ESR dramatically reduces power consumption and handles the dynamic topology of WSNs, especially when the density or the size of the network increases.