The Research On Self-Localization Of Node In Wireless Sensor Network Based On Genetic Algorithm

It is an intersectional technique of multiple subjects such as sensing, computing, communicating, information processing that the technology of Wireless Sensor Network has wide application in both military and civil areas. The technology of node self-localization is crucial but basic one of Wireless Sensor Network that the node location should be known first among numerous study on application.Centroid algorithm localization depends upon connexity of network and node density and the error of node orientation is high. To finite node energy, a simple localizable arithmetic of node is its preponderance. To keep the advantage, this thesis brings forward ameliorated model based on computing model of maximum likelihood estimation and revises centroid algorithm value of coordinates and then optimizes the model based on centroid algorithm. Ordinary arithmetic on node self-localization is capable of saving energy of node and prolonging natural life of network.The main research work and results are as follows:1. This topic studies centroid algorithm idea and mathematics former and discusses the infection about the precision of node orientation by node density, connectedness etc.2. The technology of node self-localization based on centroid algorithm takes maximum likelihood estimation model which is the common distance measuring in practice because of the error among practical value and computing one, maximum likelihood estimation model is used on computering the error value of node location. The errorless value is modified on the node self-localization coordinates value which based on centroid algorithm idea.3. It optimizes the computing on maximum likelihood estimation measuring model by genetic algorithm and establish the emulational former and designs the arithmatic flow and also elicits emulational conclusion of genetic algorithm. The outcome shows the more node density and connectivity, the less error furthermore iterative time better and the error less.It has a better conclusion that the maximum likelihood estimation measuring model optimized by genetic algorithm on optimzeing of the error value of node location. But, genetic algorithm need much more iterative calculating and to the finite node energy it doesn't fit odd node because of exact hardware, so the optimizing model fit central localization only.