Enhancing Query Efficiency in CEPList Systems Through Node Reordering

Complex Event Processing (CEP) is a technique for identifying target matching sequences from real time data streams, playing a critical role in application domains such as financial trading and aircraft flight testing. Despite substantial advances in event stream pattern matching research, event stream-based Complex Event Processing still confronts fundamental challenges in achieving efficient and adaptive processing under operational conditions. Existing approaches typically employ nondeterministic finite automata (NFA) for query matching. However, automaton-based methods predominantly rely on sequential matching mechanisms, leading to suboptimal efficiency. Moreover, current research primarily focuses on optimizing the overall matching process rather than specifically addressing sequential matching constraints. To overcome these limitations, this paper proposes a tree-structured monitoring model that enables flexible complex event matching through reordering of underlying query nodes. The node reordering mechanism operates based on the selectivity of events in the data stream, allowing the model to transcend the sequential matching restrictions inherent in traditional automata. Experimental results demonstrate that our method achieves superior query performance across diverse datasets and varying time window configurations, effectively breaking through the efficiency bottlenecks of conventional sequential matching approaches.