Study Of Localization Algorithm For Wireless Sensor Network

Wireless Sensor Network was firstly proposed by the US，which was mainly used inmilitary technology. Afterwards, due to its extensive application, besides using in“Cooperative Engagement Capability” research and “Unattended Ground SensorsGroup” in military, it can also perceive the surrounding environment that it is hard forhuman beings to reach by temperature sensor, pressure sensor, gas sensitive transducer,etc. It can be applied to many fields such as environmental monitoring, fire prevention,biomedicine, and city management. As a matter of fact, Wireless Sensor Network hasbeen a hot area of research at home and abroad.Node localization of wireless sensor is one of the key technologies in sensor networkapplication. Without the specific location of sensor node, all parameters measured aremeaningless. So the thesis takes the node localization technology of sensor as its researchobject. The main work is:(1)On the basis of the three types of sensor networklocalization——centralized and distributed localization algorithms, ranging andrange-free localization algorithms, relative and absolute localization algorithms,coarse-grained and fine-grained localization algorithms, this thesis mainly introducesseveral classical algorithms. Then, by comparing the region-based superposed APIT withROCRSSI algorithm, the result shows that the loop superposed ROCRSSI algorithm hasmore advantages than the delta superposed APIT, such as more accuracy, higher rate ofcoverage, less communication expenses.(2)Graded Circle was applied to ROCRSSIalgorithm, and it was improved. Thus a new algorithm was proposed, namedIROCRSSIGC (improved ROCRSSI based on Graded Circle). This new algorithm is toupgrade the high accurate localized nodes (not less than Grade4) to reference nodes,which will make more unknown nodes join into the location. It retains the advantages of ROCRSSI, increases the rate of coverage, and reduces the number of nodes which areunable to be localized. Through simulation analysis, it shows that the improvedIROCRSSIGC algorithm has higher positioning accuracy and wider coverage thanROCRSSI.