In-Network Data Reduction and Coverage-Based Mechanisms for Generating Association Rules in Wireless Sensor Networks

Sensor association rules, which are a kind of behavioral pattern that aims to capture the temporal relations between sensor nodes, has proven to be a promising tool for improving wireless sensor network (WSN) performance and its quality of service (QoS) by participating in the resource management process and by compensating for the undesired effects of wireless communication. To prepare the data needed for generating sensor association rules, each sensor node should monitor its activity over time and inform the sink about the time in which events are detected. However, without an efficient extraction mechanism, this process is costly, giving the limited resources of sensor nodes. In this paper, we propose an in-network data reduction mechanism to reduce the amount of data (about sensors' behaviors) by removing some of the data's redundancies. In addition, we propose a relaxed version of sensor association rules that emphasizes the correlation between a set of locations (areas) rather than individual sensor nodes. We refer to the new rules proposed by coverage-based sensor association rules.