A Hierarchical Clustering Technique for Radio Map Compression in Location Fingerprinting Systems

Abstract

Location Fingerprinting is one of the radio positioning techniques which has been proposed in the field of Location Based Services (LBS). Considering the actual trends towards energy efficient systems and green networking, reducing the energy consumption has become a challenging issue in the context of fingerprinting systems. In this paper we present a clustering technique which aims to compress the radio database and hence to reduce the online processing load of the system. We propose a hierarchical clustering method which is applied in a concatenated location-radio signal space. %The clustering technique tries to provides high quality signal matching in radio signal space and low positioning error in geographical domain. Computer simulations have been conducted to evaluate the performance of the proposed technique in environments with different shadowing configurations. The results show that the proposed clustering technique outperforms the conventional griding method, and besides allows us to reduce the size of the database significantly while keeping an acceptable level of performance for the positioning system.