Font Size: a A A

Research On Localization Algorithm For Wireless Sensor Networks

Posted on:2012-12-03Degree:MasterType:Thesis
Country:ChinaCandidate:T YuFull Text:PDF
GTID:2178330335478364Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
As a new network of information acquiring and transmitting, wireless sensor network(WSN) is made up some micro-sensor nodes. It can be used in Real-time monitoring and collecting various environmental information, then forward the information to ours for being analyzed and used. WSN has been applied widely in environment science, industry control, military defense, medical health and intelligent transport. Node localization technology is most important supporting technique and location information has a key role for application of WSN.In this paper, the basic notion, system architecture, key technology and application field of the WSN are first introduced. Then node localization technology of WSN is studied thoroughly. The representative algorithms of the WSN in recent years are summarized, and we analyze the advantages and disadvantages of various algorithms. The extensions of localization algorithm and facing new challenges in WSN are noted.Secondly this paper proposes an algorithm based on the RBF neural network for wireless sensor networks. We collect signal strength at different positions in wireless sensor networks covering areas and employ Radial Basis Function(RBF)neural network to construct a mapping model between signal strength and the coordinate of a node. Then train such a neural network with the collected signal strength value(an input vector) before use the trained neural network to localize an unknown node. Experimental results show that this algorithm can obtain more accurate and more reliable coordinate position estimation.Finally this paper proposes a DV-Hop localization algorithm based on modifying weighted average hop distance. Aiming to problem with the big average hop distance error of DV-Hop localization algorithm, by computing weight coefficient correct the average one-hop distance of network, this not only reduce the distance error but also make the estimation of average one-hop distance more reasonable. Meanwhile, using the Taylor expansion based the weighted least-squares method to compute node self-coordinate position. Simulation shows the effectiveness of improvement algorithm by comparing with each performance index.
Keywords/Search Tags:wireless sensor network, DV-Hop algorithm, RBF, neural network, location algorithm
PDF Full Text Request
Related items