PEJA:Progressive Energy-Efficient Join Processing for Sensor Networks  

作　　者：赖永炫 陈毅隆 陈红 

机构地区：[1]School of Information,Renmin University of China [2]Key Laboratory of Data Engineering and Knowledge Engineering,Ministry of Education

出　　处：《Journal of Computer Science & Technology》2008年第6期957-972,共16页计算机科学技术学报（英文版）

基　　金：partly supported by the National High Technology Research and Development 863 Program of China under Grant No. 2008AA01Z133;the National Natural Science Foundation of China under Grant Nos. 60673138, 60603046;the Science Technology Research Program of MOE under Grant No. 106006, and the Program for New Century Excellent Talents in University.

摘　　要：Sensor networks are widely used in many applications to collaboratively collect information from the physical environment. In these applications, the exploration of the relationship and linkage of sensing data within multiple regions can be naturally expressed by joining tuples in these regions. However, the highly distributed and resource-constraint nature of the network makes join a challenging query. In this paper, we address the problem of processing join query among different regions progressively and energy-efficiently in sensor networks. The proposed algorithm PEJA （Progressive Energy-efficient Join Algorithm） adopts an event-driven strategy to output the joining results as soon as possible, and alleviates the storage shortage problem in the in-network nodes. It also installs filters in the joining regions to prune unmatchable tuples in the early processing phase, saving lots of unnecessary transmissions. Extensive experiments on both synthetic and real world data sets indicate that the PEJA scheme outperforms other join algorithms, and it is effective in reducing the number of transmissions and the delay of query results during the join processing.Sensor networks are widely used in many applications to collaboratively collect information from the physical environment. In these applications, the exploration of the relationship and linkage of sensing data within multiple regions can be naturally expressed by joining tuples in these regions. However, the highly distributed and resource-constraint nature of the network makes join a challenging query. In this paper, we address the problem of processing join query among different regions progressively and energy-efficiently in sensor networks. The proposed algorithm PEJA （Progressive Energy-efficient Join Algorithm） adopts an event-driven strategy to output the joining results as soon as possible, and alleviates the storage shortage problem in the in-network nodes. It also installs filters in the joining regions to prune unmatchable tuples in the early processing phase, saving lots of unnecessary transmissions. Extensive experiments on both synthetic and real world data sets indicate that the PEJA scheme outperforms other join algorithms, and it is effective in reducing the number of transmissions and the delay of query results during the join processing.

关 键 词：progressive join minimal join set in-network processing sensor network 

分 类 号：TN929.5[电子电信—通信与信息系统] TP212.9[电子电信—信息与通信工程]