Localization in Wireless Networks Using Decision Trees and K-Means Clustering

Node localization is employed in many wireless networks as it can be used to improve routing and enhance security. In this paper, we propose a new algorithm based on decision tree classification and K-means clustering which are well known techniques in data mining. Several performance measures are used to compare the K-means localization algorithm with those using linear least squares (LLS) and weighted linear least squares based on singular value decomposition (WLS-SVD). It is shown that the proposed algorithm performs better than the LLS and WLS-SVD algorithms even when the geometric anchor distribution about an unlocalized node is poor.