ZeDDS - Fault-Tolerant Data Management in Wireless Sensor Networks with Mobile Users

Kamenik, Jens and Peuser, Christoph and Gollücke, Volker and Lorenz, Daniel and Piechocki, Roland and Wasmann, Merlin and Oliver, Theel
Ubiquitous wireless sensor networks (WSNs) consist of sensor nodes which may communicate with each other via unreliable communication links. Furthermore, the sensor nodes themselves may fail. Ubiquitous WSNs may be used in application scenarios where they autonomously monitor the environment and are only sporadically visited by the mobile user for harvesting the collected sensor data. Thus, high availability of the measured data is of very high priority. But how can the mobile user formulate this QoS requirement and how can a WSN – honoring such a QoS requirement – be efficiently implemented? We propose ZeDDS a middleware and control framework for providing high available data storage in WSNs. In ZeDDS, we assume that the WSN is meant for collecting and reliably storing measured data until the mobile user contacts the WSN for data harvesting. ZeDDS enables the mobile user to explicitly specify a particular replication strategy exhibiting a certain data availability and energy consumption. At run-time, ZeDDS is appropriately configured and replicates the measured sensor data according to the replication strategy given. We evaluate our ZeDDS implementation in terms of write operation availability measurements of a WSN consisting of TelosB sensor nodes using three different well-known replication strategies.
11 / 2011
Cognitive Logistiknetzwerke