基于事件共享机制的物联网复杂事件处理方法

Method of IOT complex event processing based on event sharing mechanism

针对物联网(IOT)复杂事件查询处理过程中的重复查询、存储和处理的问题,提出了事件共享机制(ESM)。首先,为了实现复杂事件的查询与检测,给出了物联网语义事件定义及事件操作符的语义描述;其次,从公共子查询的定义、公共内部查询结构的设计以及事件资源的共享三个角度对物联网事件共享机制展开研究,通过查询表达式的重写、有向无环图(DAG)的构建,以及在结点上使用改进的Continuous参数上下文对事件流进行处理,实现公共子事件查询、存储和处理的共享;最后,构建了基于事件共享机制的语义形式化查询计划处理模型(SFQPM),该模型可自动对查询表达式和查询谓词进行处理,实现复杂事件检测和处理的自动化。仿真结果表明,与基于二叉树(BTree)的处理方法进行对比,所提出的SFQPM具有较高的处理效率和可靠性,实现了复杂事件检测与中间结果共享机制的有机统一,提高了系统的处理效率。最后通过案例研究验证了所提出算法的有效性和可行性。

Concerning the problems including repetitive query, storage and processing in the process of complex event query processing in Internet of Things （IOT）, an Event Sharing Mechanism （ESM） was proposed. Firstly, in order to realize the query and detection of complex events, a semantic event definition about IOT and semantic descriptions of event operators were presented. Secondly, research on the IOT ESM was done from the following three aspects： the definition of public subqueries, the design of public internal query structure and the sharing of event resources. Through rewriting the query expression, building the Directed Acyclic Graph （DAG） related to the query expression and using the improved Continuous, one of the parameter contexts, on each node to handle event streams, the sharing of public events＇ query, storage and processing was implemented. Finally, a Semantics Formal Query-plan Processing Model （SFQPM） based on ESM was also designed, which could process query expressions and predicates automatically, and fulfill the automation of complex event detection and processing. The simulation results show that, compared with the method based on BTree （ Binary Tree）, the proposed SFQPM has high efficiency and reliability in processing, and can process massive and real-time IOT event streams timely and efficiently. In addition, a case study was given to verify the effectiveness and feasibility of the proposed SFQPM.