Adaptive Aggregation Scheduling Using Aggregation-degree Control in Sensor Network

Data stream processing applications have recently increased. In sensor network these applications use in-network processing or aggregation to process continuous and unbounded data stream. However, it may lead to delaying total execution time and increasing data loss. In previous work, total number of transmitting messages increases, since data aggregator should send refresh messages to acquire the results that satisfy tolerable error bound the client specified. Thus, it needs aggregation scheduling method to minimize reduced accuracy and to reduce the number of transmitting messages. We propose an AASM (Aggressive Aggregates-Scheduling Method) method using the reliable prediction interval to filter just related data and the aggregation-degree control to reset window size according to satisfying with answering client’s requirements to tolerable error bounds. Our proposed algorithm simulation resulted in the better performance in terms of reducing total number of transmitting messages and minimizing data accuracy loss.