Filter-based publish/subscribe systems suffer from high subscription maintenance cost for each broker in the system stores a large number of subscriptions. Advertisement and covering are not sufficient to conquer such problem. Thus, subscription merging is proposed. However, current researches lack of an efficient and practical merging mechanism. In this paper, we propose a novel subscription merging mechanism. The mechanism is both time and space efficient, and can flexibly control the merging granularity. The merging mechanism has been verified through both theoretical and simulation-based evaluation.
展开▼