Energy-efficient return-path routing in wireless sensor networks

Abstract

As sensor nodes are powered by battery of limited lifespan, energy consumption has always been an important issue in sensor networks. New generation of sensor network technology bring about the possibility of constructing a spatial index over a sensor network such that sensor queries are forwarded only to the relevant nodes. Existing spatial indices share the common feature of using the same node path for propagating queries and routing back results. Hence, intermediate nodes along the path incur heavier workload than non-intermediate nodes and drain their battery more quickly. In this paper, we propose a multi-parent method to share the workload of intermediate nodes so that sensor results have alternative return paths to the sink node. As a result, the lifespan of the entire sensor network can be prolonged. We study the performance and effects of this method under three different kinds of spatial index structures. Our experiments show that the proposed approach is able to prolong the sensor network lifespan especially when the sensor network is dense. The experiments also yield various interesting information on the performance of the proposed approach under various conditions of the sensor network.

Power-aware routing in wireless sensor networks

“…To demonstrate improvements arising from the use of virtual nodes, we first show performance without virtual node, as shown in Figure 4.6. Three spatial index approaches are adopted; namely, OMSI-tree, Nearest Parent Tree(NPT) and Hybrid Tree [64]. From Figure 4.6, we note that as density increases, performance of all three spatial indices decreases.…”

“…As the objective of this work is to adapt spatial indices to large scale and high-density sensor network, we adopt the spatial index approach instead of ad hoc protocols.We adopt three spatial indices to evaluate the performance of our virtual node idea. The three spatial indices were first proposed by Zha et al[64]. We make modifications to these indices as follows: We introduce a new characteristic for virtual node: strength.…”

“…Hybrid Tree. The Strongest Parent Tree used here is basically the same as the Nearest Parent Tree in an earlier work[64]. However, due to a uniformly partitioned network, a virtual node is quite likely to have two or more neighbors being the nearest parents.…”