Real-time Processing of Rule-based Complex Event Queries for Tactical Moving Objects

Abstract

Target data for tactical moving objects are streaming data collected in real time via radar, sonar, and other sensors. A system of continuous complex event query with dynamic rule definitions and high performance is needed to process that target data. We develop a continuous complex event query system with rule-based layered architecture. A continuous processing flow is decomposed into four modules hierarchically, which are event filtering, event capture, Continuous Queries (CQ) and Complex Event Processing (CEP). Each module has its responsibility but works together for a completed continuous processing flow. This paper shows that it is possible to dynamically insert, update, delete and search rule specifications of each layered modules through the decomposition of the whole system. Many rules are registered in the system for processing input event data continuously in real time. To improve the performance of getting matching CQs for each incoming event, CQ index is developed. Finally, experimentations and performance evaluations are