The Research Of Localization Techniques In Wireless Sensor Networks

Wireless sensor networks (WSN) which has wide application prospects is considered as one of the most important technologies in 21 centuries. The localization of sensor-nodes is one of the key technologies of WSN, the information collected by the node must be combined with its location, and the location information of the node is precondition of functions and service-applications of WSN.The paper discusses self-localization technology of node in WSN in depth. Wer introduce the typical localization algorithms, and compare them in accuracy and power consumption. We propose and design the parameter-identification of location algorithm based on Received Signal Strength Indicator (RSSI), according to the characteristic of radio propagation and the former technical literature.The localization algorithm based on RSSI estimates the distance between nodes according to the pass-loss of the signal in the transmission process, then it can estimate the location of the unknown nodes. Because of the uncertainty of RF signal transmission and the diversity of the environment, the algorithm should be self-adaptive, and it could have a well positioning result in different network environment. We do a large number of experiments and analysis the signal data, we propose and design the parameter-identification of location algorithm based on RSSI. It can identify the parameters of environment. The algorithm is universal to many applications. We modify the theoretical model of RF signal attenuation using system identification and Weighted Least Squares. Using ranging errors as a weighted feedback coefficient, we modify the parameters of the environment on line. Then we use the triangular-centroid method to estimate the location of the unknown nodes. The simulation result shows that the algorithm could reflect the rapid changes of the network environment rapidly, and it could identify the parameters of the environment online and reduce the time witch was spent on the parameters configuration. Meanwhile, the algorithm could reduce the distance-estimated error and improve the positioning accuracy. Then we discuss the probabilistic model of localization algorithm, and educe the distance probability density function between the nodes, and design a probability model. We establish a Maximum Likelihood Estimator to estimate the location of the unknown nodes. In the end we analyse the positioning accuracy and variance by Fisher information matrix. At last, the paper summarizes the previous work, and makes a scheme for the future work.