Self-configuring Node Clusters, Data Aggregation, and Security in Microsensor Networks

Microsensors operate under severe energy constraints and should be deployed in large numbers without any pre-configuration. We construct a generalized self-clustering algorithm, called Low-energy Localized Clustering (LLC). It integrates the ideas of two self-configuring clustering algorithms: the Localized algorithm and the Low Energy Adaptive Clustering Hierarchy algorithm. LLC covers a range of behaviors from the better-clustering performance of the Localized method to the energy-efficient operation of the LEACH method. The main advantage of LLC is that it can be energy-efficient while maintaining localization. Data aggregation techniques such as summarization, finding representative data items, and pattern matching are proposed. Data aggregation is a necessity in microsensor networks, since transmitting huge volumes of raw data is an energy-intensive operation. Finally, security issues are discussed and an energyefficient Randomized Data Authentication algorithm is designed specifically for microsensor applications.