Exploiting semantic quorum-based data replication in wireless sensor networks

Most current wireless sensor networks (WSN) are connected to a base station that provides persistent data storage. In application scenarios without a constantly connected base station, in-network storage of the data is necessary. Usually, individual sensor nodes of WSNs are unreliable. To ensure dependable persistent storage within the WSN, redundancy of sensor data is required. Generally, data replication is used for increasing availability of data in distributed systems. The downside of data replication is increased energy consumption, which reduces the lifetime of WSNs. In this paper, we present the adaption of a semantic data replication strategy for WSNs. The approach guarantees increased availability of data in WSNs while reducing the communication costs by the exploitation of semantic properties of the application scenario. A probabilistic version of the approach is considered for further reducing the energy consumption. The example of a distributed FIFO queue is used to illustrate the approach.