A Queuing-based Analytical Model for Complex Event Processing on Stream Platform

Abstract

In recent years, streaming data are everywhere. For example, many sensors in mobile devices produce various kinds of massive data. These data are streaming type. Some important implicit information is hidden in streaming data. To process streaming data, we need a data-driven system to which the inputs are streaming data instead of queries. Complex event processing(CEP) is related to data-driven services. CEP can analyze streaming data from different resources to identify the special event and respond to the event. According to these streaming data, lots of real-time services are needed to satisfy different Quality of Service(QoS) requirements. In this thesis, we introduce a queuing model on CEP application for different QoS requirements. Based on our model, the computing resources can be more effectively and flexibly allocated for various computing environments.

In recent years, streaming data are everywhere. For example, many sensors in mobile devices produce various kinds of massive data. These data are streaming type. Some important implicit information is hidden in streaming data. To process streaming data, we need a data-driven system to which the inputs are streaming data instead of queries. Complex event processing(CEP) is related to data-driven services. CEP can analyze streaming data from different resources to identify the special event and respond to the event. According to these streaming data, lots of real-time services are needed to satisfy different Quality of Service(QoS) requirements. In this thesis, we introduce a queuing model on CEP application for different QoS requirements. Based on our model, the computing resources can be more effectively and flexibly allocated for various computing environments.