A fast and anti-matchability matching algorithm for content-based publish/subscribe systems

Abstract

The content-based publish/subscribe system is a flexible many-to-many communication middleware that meets the demands of many large-scale distributed applications. It is well known that event matching is a fundamental component of the content-based publish/subscribe system. When designing matching algorithms, matching speed is a major objective being pursued. Moreover, through theoretical analysis and experimental verification, we discover that the matching speed of most existing matching algorithms is affected by the subscriptions’ matchability which is defined as the matching probability of subscriptions with events. Nevertheless, this problem has not been considered in existing matching algorithms. To address this problem, we propose REIN (REctangle INtersection), a fast and anti-matchability matching algorithm for content-based publish/subscribe systems. REIN is a fast matching algorithm, following the conventional design objective of pursuing a high matching speed. Furthermore, due to the utilization of a negative searching strategy that aims to filter out unmatching subscriptions in the matching process, the matching speed of REIN is not affected by the subscriptions’ matchability, but rather is improved. To evaluate the performance of REIN, comprehensive experiments are conducted. The experiment results show that REIN not only has an excellent matching performance, but also possesses a beneficial anti-matchability feature.