Sluggish Ladder Structure for Query Processing over Sliding Windows

The notion of update pattern awareness in the context of continuous queries over relations, streams and sliding windows was introduced in previous researches. This investigation concerns the access costs of data structures for the intermediate results or final results of continuous query processing over sliding windows by update pattern awareness.The ladder structure was employed to store intermediate results or final results and the trunk list and the branch list were designed into ladder structure. The branch lists were used to gather the result tuples with the identical expiration time , when the result tuples expired through the three components of the ladder structure. Our experiments show that the new data structure outperforms the partioned structure(calendar queue) and other data structures by an order of magnitude.