A wireless sensor network (WSN) typically consists of a large number of small, low-cost sensor nodes distributed over a large area. The sensor nodes are integrated with sensing, processing and wireless communication capabilities.
Each node is usually equipped with a wireless radio transceiver, a small microcontroller, a power source and multi-type sensors.
Although, WSN was originally considered as a monitoring platform, recent advances and achievements have made them suitable candidate also for event detection and actuation. In this end, event detection using the WSN has recently attracted much attention. Interesting applications vary from industrial safety and security to meteorological hazard, earthquake, and fire detection.
Resource constraints of the sensor nodes, dynamic and often volatile nature of the deployment area and the network itself introduce unique challenges for researchers in the field. The designed event detection technique should therefore be light-weight because of limited computational capability of the sensor nodes.
It should be distributed to reduce communication and transmission overhead as well as to increase the robustness by overcoming the problem of a single point failure.
Studying related work in the field of event detection reveals two major trends, i.e., centralized and decentralized. In centralized approaches, in which event detection is conducted in a base station, the focus has mostly been on data-aggregation and reducing communication overhead.
In decentralized approaches, in which event detection is carried out inside the network and even inside every individual node, the focus has often been on networking aspects and consensus. What is greatly missing in the field of event detection for the WSN, however, is the issue of online and distributed data mining, feature extraction, pattern recognition and matching, and data/sensor fusion.
Although, monitoring was the initial application of wireless sensor networks, in-network data processing and (near) real-time actuation capability have made wireless sensor networks suitable candidate for event detection and alarming applications as well. Unreliability and dynamic (e.g. in terms of deployment area, network resources, and topology) are normal practices in the field of WSN.
Therefore, effective and trustworthy event detection techniques for the WSN require robust and intelligent methods of mining hidden patterns in the sensor data, while supporting various kinds of dynamicity. Due to the fact that events are often functions of more than one attribute, data fusion and use of more features can help increasing event detection rate and reducing false alarm rate.
In addition, sensor fusion can lead to more accurate and robust event detection by eliminating outliers and erroneous readings of individual sensor nodes and combining individual event detection decisions.
In this paper, we propose a two-level sensor fusion-based event detection technique for the WSN. The first level of event detection in our proposed approach is conducted locally inside the sensor nodes, while the second level is carried out in a level higher (e.g., in a cluster head or gateway).
It incorporates a fusion algorithm to reach a consensus among individual detection decisions made by sensor nodes. By considering fire as an event, we evaluate our approach through several experiments and illustrate impact of sensor fusion on achieving better results.
To read this external content in full, download the complete paper from the author archives at Twente University.