ments for nodes in WSNs is one to obtain their positions.So node localization of key problems and is the basis on and prerequisite for practical applications of WSNs.Therefore,studying theoretically node localization is of great importance both and practically for WSNs. on Firstly,the characteristics and some novel research WSNs is analysed in depth.Then localization the beacon selection optimization strategy in distributive multi-energy node node localization algorithm is studied in this paper.This on paper also studies the problem in mobile paper WSNs based Monte Carlo method.Finally,all the theories and algorithms are The main work of the simulated witll discrete event simulator are called OMNeT++. summarized as and research contributions briefly follows: 1)In order to overcome the high computational complexity anchor concentric anchor beacon localization algorithm,an problem selection scheme in is proposed.The Based proposed scheme carl not only reduce algorithm complexity,but also improve the localization accuracy.The main idea of the algorithm is as following: on the network physical topology characteristics,only three anchors selected among all the neighboring.The selected anchors are used for node localization. What’S more,the three beacons between should be selected SO tllat the slim of the distances any two of thcrn is the farthest among all possible combinations of neighbor centroid beacons.The estimation location can be calculated based on 2)The which multi—energy idea is further introduced to algorithm. problem of node on address the localization in mobile sensor network.A localization algorithm based power node Monte Carlo, is suitable for mobile wireless sensor network is proposed in this paper.Each at Mobile anchor emits or beacon information different levels.From the determine which is information received by each unknown nodes,the particular ring sensor Can inner circle it lies within from the correspondent anchor,which lI 武汉理工大学硕士学位论文 called constraint region.The positions of unknown a nodes on are able to be estimated by bunch of particles which ale sample