Research On Centroid Localization Algorithm Based On RSSI In Wireless Sensor Network

Wireless sensor network(WSN)is a blend of multidisciplinary technology such as network communication,micro sensor.WSN has a broad application prospect in many fields including military,disaster prevention,transportation,medical.With the popularity of the concept of IOT,WSN technology attracts more enthusiastic attention of international academia and industry.In the practical application of wireless sensor network,there is no attached node location information data is meaningless,and for each node is equipped with GPS to locate the high cost.Using a few position known the connectivity between nodes and the network,locating the unknown node by localization algorithm is the general method of node localization in WSN.The centroid localization algorithm based on RSSI combines classic RSSI localization algorithm and centroid localization algorithm,it is a kind of low cost,low consumption and good localization algorithm accuracy,but we still have some room to improve on the positioning accuracy and coverage.Based on the comprehensive analysis on the basis of the centroid localization algorithm based on RSSI,according to the algorithm that may have caused the error link,from gaussian filtering,error correction,the more power value weighted,unknown nodes reuse four aspects,such as optimizing the centroid localization algorithm based on RSSI,and two improved algorithms are put forward.In this paper,the main work includes:1.Through the analysis of the centroid localization algorithm based on RSSI may produce error of link,this paper proposes a difference correction centroid localization algorithm based on RSSI.First of all,to obey the normal distribution of RSSI values for gaussian filtering processing,using gauss filter method to eliminate error of RSSI values after calculating mean,reduce volatility;Second,the RSSI value by converting the distance value correction model,introducing differential method with beacon node measurement error correction of the coordinates of the unknown node;The simulation results show that the proposed algorithm is compared with the original algorithm on positioning accuracy improved about 20%.2.Through the analysis based on RSSI difference correction centroid localization on the coverage is insufficient,proposed a multiple weight centroid localization algorithm based on RSSI.Although difference method modified centroid localization algorithm in a single node positioning accuracy is improved,but there is no choice of beacon nodes is optimized,rely too much on the uniform distribution of beacon nodes,in the condition of uneven distribution of poor performance on the coverage of positioning of the entire network.To solve these problems,the introduction of multiple weighting factors to the unknown nodes around the beacon nodes are weighted,select beacon nodes,and position has completed the unknown node as auxiliary positioning beacon nodes.Experiments prove that this method is compared with the other two algorithms on coverage increased about 20%,on a single node location accuracy is about 8%.