Iterative Positioning Algorithm of Reference Point Based on DBSCAN

Abstract

There are the problems of low ranging accuracy caused by noise interference during the process of received signal strength indication (RSSI) transmitting and low positioning accuracy of traditional positioning algorithm in the wireless sensor network (WSN), hence, an iterative positioning algorithm of reference point based on the density-based spatial clustering of applications with noise (DBSCAN) is proposed in this paper. In the proposed algorithm, DBSCAN clustering algorithm is firstly used to cluster RSSI signals, eliminate the outliers and obtain the clustered RSSI. Then, the distances between the target node and the beacon nodes are calculated, via which the coordinates of three reference points on the overlapping arc of three positioning circles are calculated, and the centroid of three reference points is taken as the initial position coordinate of target node to calculate the distance deviation coefficient. Finally, the ratio of distance measured by the log-range path loss model to the median of distance deviation coefficient is taken as the new distance between the target node and the beacon node. Based on the above new distances, the reference point coordinates are iterated repeatedly to update the positioning coordinate of target node until the criteria of iteration is met. In the simulation experiment, the mean filtering algorithm and DBSCAN clustering algorithm were adopted respectively to preprocess RSSI signal. The trilateral centroid positioning algorithm, weighted centroid positioning algorithm, and the proposed iterative positioning algorithm of reference point (IPAoRP) were adopted respectively to locate the target node and calculate target node coordinate under the condition of same node distribution. Experimental results show that the iterative positioning algorithm of reference point based on DBSCAN has improved greatly the ranging accuracy and the positioning accuracy of target nodes.