An improved sequence-based indoor localization algorithm in WSNs

Sequence-based localization is a novel RF localization technique. The algorithm is achieved by constituting RSSI-based constraint tables and comparing data between two tables. But, the definitions of the constraint relation and the centroid in the algorithm are imperfect. In this paper, we present a new sequence localization method that involves with correlation metric and centroid. First, we use rank order correlation coefficient instead of constraint tables. It simplifies the algorithm implementation. Furthermore, the definition of centroid is amended according to the nearest location regions. It makes the algorithm more reasonable, especially in the edge of the localization area. The simulation shows that the time-consumption and the localization accuracy of the new algorithm are improved.