Complex Event Processing over Distributed Uncertain Event Streams

Abstract

In the 21st century, as technologies of perceptual recognition develops, devices of information generation begin to accurately sense, measure and monitor the physical world in real time.Complex Event Processing(CEP), which can be used to extract user level information from raw data, becomes the key part of the IoT middleware.Most of the current study of complex event processing has not focus on distributed uncertain event streams.In this paper, a high performance method over distributed uncertain event streams is proposed.This method uses improved matched tree that include node buffer list and probability calculation to detect complex event in single uncertain event stream.Distributed stream processing platform Storm also is brought in to improve the performance and stability.Based on Storm, two level parallel methods are proposed to increase the throughput.The experiment study shows that this method is efficient to process complex events over distributed uncertain event streams.