Energy optimized fault tolerance for pervasive communication spaces

In a pervasive environment, wireless sensor networks are one backbone for the interaction with the real world. Therefore, pervasive computing relies on a networked infrastructure with an inherent high dynamics due to mobility of nodes, unreliable communication channels and node failures. Additionally, wireless sensor networks are a class of networks with a limited energy budget. To fullfill a given task they are required to tradeoff the important QoS requirements with energy. The most energy consuming operation in a wireless sensor network is communication. This proposal illustrates an approach for exploiting application specific query workloads to find the best tradeoff between query communication costs and availability in a data storage centric sensor network.