Title :
Maximizing Throughput for Queries over Streaming Sensor Data
Author :
Gomes, Joseph ; Choi, Hyeong-Ah
Author_Institution :
Dept. of Comput. Sci., George Washington Univ., DC
Abstract :
Sensors are becoming ubiquitous, and increasingly integrated with our lives. Sensors usually send sampled data periodically using wireless connections to server machines. The servers perform various operations (e.g. filter, aggregate, join etc) on this data in real-time according to predefined queries or rules. In this paper, we address the problem of finding an optimal join tree that maximizes throughput for sliding window based multi-join queries over continuous sensor data streams. We develop a dynamic programming algorithm OptDP, that produces an optimal tree but runs in an exponential time in the number of input streams. We then present a polynomial time greedy algorithm XGreedyJoin. Our experiments in ARES show that for almost all instances, trees from XGreedyJoin perform close to the optimal trees from OptDP, and significantly better than existing XJoin based heuristic algorithms
Keywords :
dynamic programming; greedy algorithms; query processing; trees (mathematics); wireless sensor networks; OptDP algorithm; XGreedyJoin algorithm; dynamic programming algorithm; exponential time; optimal join tree; polynomial time; sensor data streaming; sliding window based multijoin queries; Aggregates; Biosensors; Computer science; Dynamic programming; Filters; Heuristic algorithms; Intelligent sensors; Sensor systems; Throughput; Wireless sensor networks;
Conference_Titel :
Mobile Adhoc and Sensor Systems (MASS), 2006 IEEE International Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
1-4244-0507-6
Electronic_ISBN :
1-4244-0507-6
DOI :
10.1109/MOBHOC.2006.278617