Abstract:
In the widely-used map-based content publishing/subscription systems (CPS), the system always match each constraint in the subscription with the corresponding event. However, the large number of constraints in each subscription of the system call for higher efficiency of event matching. Based on the coverage pruning and predicate index with precise location properties in the matching network algorithm, this paper proposes a matching algorithm based on matching feature fusion. The data structure, subscription processing flow and matching processing flow of the algorithm are given. Theoretical analysis and typical experimental comparisons show that the algorithm can provide more efficient matching compared with the simple coverage algorithm and indexing algorithm.