On using temporal consistency for parallel execution of real-time queries in wireless sensor systems

In this paper, we study the important issues for execution of real-time queries in a wireless sensor system in which sensor nodes are distributed to monitor the events that have occurred in the environment. Three important objectives in processing the real-time queries are: (1) to minimize the number of missed deadlines, (2) to minimize the processing costs, especially in data communication; and (3) to provide temporally consistent sensor data values for query execution. To reduce the data transmission cost and delay time in gathering the right versions of data items for a query, we propose the Parallel Data Shipping with Priority Transmission (PAST) scheme to determine where and how to execute a real-time query. To meet the deadlines of the queries, a deadline-driven priority policy is adopted to schedule the transmission of sensor data versions to the coordinator node. © Springer-Verlag Berlin Heidelberg 2004.