Indoor WSN Region Localization Based on Time Window Statistics

In recent years, indoor localization is getting more and more attention. The uncertainty of data commonly exists due to the inaccuracy of measurement instrument and the irregularity (multipath, diffraction and obstacle) of wireless signal transmission model in indoor environment. This uncertainty will affect positioning precision. In order to solve such issue, a distributed localization algorithm RLTWS (Region Localization Based on Time Window Statistics) is proposed based on pattern matching principles. RLTWS is a two phase algorithm. First, an unknown node can determine its own sub-region by means of RSSI probability distribution Second, its position in sub-region can be calculated more accurately on the basis of optimal reference node. The distributed algorithm can make full use of the calculating capability of nodes and greatly reduce the amount of network traffic. The field experiments show that more stable position information is achieved compared with traditional methods and that outliers are reduced effectively, as long as the time window matches the sub-region accurately.