An approach for near-optimal distributed data fusion in wireless sensor networks Damianos Gavalas Æ Aristides Mpitziopoulos Æ Grammati Pantziou Æ Charalampos Konstantopoulos Ó Springer Science+Business Media, LLC 2009 Abstract In wireless sensor networks (WSNs), a lot of sensory traffic with redundancy is produced due to massive node density and their diverse placement. This causes the decline of scarce network resources such as bandwidth and energy, thus decreasing the lifetime of sensor network. Recently, the mobile agent (MA) paradigm has been pro- posed as a solution to overcome these problems. The MA approach accounts for performing data processing and making data aggregation decisions at nodes rather than bring data back to a central processor (sink). Using this approach, redundant sensory data is eliminated. In this article, we consider the problem of calculating near- optimal routes for MAs that incrementally fuse the data as they visit the nodes in a WSN. The order of visited nodes (the agent’s itinerary) affects not only the quality but also the overall cost of data fusion. Our proposed heuristic algorithm adapts methods usually applied in network design problems in the specific requirements of sensor networks. It computes an approximate solution to the problem by suggesting an appropriate number of MAs that minimizes the overall data fusion cost and constructs near- optimal itineraries for each of them. The performance gain of our algorithm over alternative approaches both in terms of cost and task completion latency is demonstrated by a quantitative evaluation and also in simulated environments through a Java-based tool. Keywords Data fusion  Wireless sensor networks  Mobile agents  Itinerary optimization  Heuristic 1 Introduction Data fusion is the process of combining data and knowl- edge from different sources with the aim of maximizing the useful information content. It improves reliability while offering the opportunity to minimize the data retained. Multiple sensor data fusion is an evolving technology, concerning the problem of how to fuse data from multiple sensors in order to make a more accurate estimation of the environment [29]. Applications of data fusion cross a wide spectrum, including environment monitoring, automatic target detection and tracking, battlefield surveillance, remote sensing, global awareness, etc. [1]. They are usually time-critical, cover a large geographical area and require reliable delivery of accurate information for their com- pletion. Most energy-efficient proposals are based on the traditional client/server computing model to handle mult- isensor data fusion in wireless sensor networks (WSNs); in that model, each sensor sends its sensory data to a back-end processing element (PE) or sink. However, as advances in sensor technology and computer networking allow the deployment of large amount of smaller and cheaper sensors, huge volumes of data need to be processed in real-time. D. Gavalas (&)  A. Mpitziopoulos Department of Cultural Technology and Communication, University of the Aegean, Lesvos, Greece e-mail: dgavalas@aegean.gr A. Mpitziopoulos e-mail: crmaris@aegean.gr G. Pantziou Department of Informatics, Technological Educational Institution of Athens, Athens, Greece e-mail: pantziou@teiath.gr C. Konstantopoulos Department of Informatics, University of Piraeus, Piraeus, Greece e-mail: konstant@unipi.gr 123 Wireless Netw DOI 10.1007/s11276-009-0211-0