Research On Energy-Density Routing Algorithm For Wireless Sensor Networks Based On Boundary Recognition

Wireless sensor network is a kind of self-organized temporary wireless networkconstructed by sensor nodes under the environment of no fixed infrastructure. Theenergy supply, computing and communication power of the nodes are seriouslylimited, so how to extend the lifetime of the networks and/or increase networkefficiency are the key to realize them.Because of these features and requirements, the network routing algorithm andtopology recognition have become important research issues. Here boundaryrecognition is the core of topology recognition in wireless sensor networks. A moreimportant issue is, the information collecting, processing and transmission in wirelesssensor network require effective routing algorithm as the basic guarantee, while thecombination of routing algorithm with boundary recognition can provide betterrouting efficiency.In this paper, considering the existing boundary recognition and classic routingalgorithm for wireless sensor networks, a novel state parameter, energy density, isdefined for each node based on the physical concept density. Energy density isdesigned to be positive proportion to the remaining energy of node and fall off as thesquare of the distance between itself and the sink node. Accordingly, a new routingalgorithm, boundary recognition-based energy density routing algorithm, is proposedfor wireless sensor networks.In boundary recognition-based energy density routing algorithm, the networkfirstly recognize the boundary of the network and mark the boundary nodes. Eachnode in the network transmits data to only one of their adjacent nodes in the commontransmittance radius. The adjacent node is selected by two conditions. One is that itsdistance to sink node becomes smaller, the other is that its energy density is largest ofthe adjacent nodes that satisfy the first condition except the marked nodes. When allthe nodes satisfy the first condition are marked nodes, the recent node transmits itsdata back to its last node and mark itself as boundary.On the basis of the above algorithm, this paper discusses another type of priorboundary recognition-based energy density routing algorithm, in which after thenodes in the boundary are marked, they would start a self-detecting process. Thenodes around the boundary will automatically detect whether the nodes satisfying therouting conditions can be found. If the nodes can be found, then nothing is disposed. If the nodes can not be found, then the recent node marks itself as the boundary nodesand informs the adjacent nodes.Next, this paper introduces the simulation process by using the Java language torealize the routing algorithm and analyzes the results of the simulation. Furthermore,the paper explores the influence of some key parameters used in the boundaryrecognition-based energy density routing algorithm and gets the regularity of theparameters’ change impacting on the performance of algorithm.Finally, the paper makes a summary and draws the prospect of research.Especially, it points out the focus and direction in the future research.