Research And Implementation Of Wireless Sensor Network Node Localization Algorithm

Localization, as a support technology in wireless sensor network (WSN), is thebasis of many applications. If the data collected by nodes is not relevant with location,that is meaningless. Thus, the research of nodes location has important theoreticalsignificance and application value. The thesis has made a thorough research onlocalization technology and the main work is as follows:1. Describe the background, significance and development status in domestic andinternational of localization technology on the basis of a large number of relevantliteratures. Introduce the basic concepts and methods of localization. The thesissummarizes the current range-based and range-free localization algorithms and analyzesthe advantages and disadvantages of each algorithm.2. The thesis has made a depth study of the localization algorithm based onmultidimensional scaling (MDS) technology and improved its ineffective in thelarge-scale networks. The improved algorithm employed FCM to split the sub-networkbased on the idea of partitioned sub-networks.3. By the influence of non-line-of-sight, WSN has poor localization effect forindoor application. For this defect, the thesis promoted the database-based localizationidea, the weighted centroid and BP neural network used in algorithm. The twolocalization algorithms avoid the use of the signal propagation attenuation model toestimate distance and the information collected in the database already contains theenvironment and other factors. Thus, it has higher positioning accuracy for indoorlocalization than common algorithm (Trilateration, etc.), which has been confirmed bysimulation results.4. Based on the CC2430hardware platform, we design a simple localizationsystem to verify the two localization algorithms based on database and achieved goodresults laying the foundation for further study.Through this research of the thesis, the classical MDS algorithm has improved toa large network with a good accuracy; proposed two types of database-basedlocalization algorithms and validation. The work of the thesis is of practical significance and provides a reference for further research and practical application.