A grouping matching algorithm for large-scale Message-Oriented Middleware with Pub/Sub model on TT&C computer

According to the Problems that predicate counting algorithm used for Message-oriented with Publish/Subscribe model on TT&C computer has incomplete preprocessing and complex counting when processing large-scale information exchange. A grouping matching algorithm was put forward, it grouped subscription by binary tree according to message name and deleted subscription group whose message name set was not subset of the event. As for remaining subscription group, it constructed different index structure according to covered relationship between different predicates and filtered corresponding attributes. The result show that time efficiency of grouping matching algorithm is higher than traditional predicate counting algorithm.