Power Saving Algorithm for Monitoring Extreme Values in Sensor Networks

Abstract

Monitoring extreme values (maximum or minimum) is important to many applications in wireless sensor networks. A previous work, called Hierarchy Adaptive Threshold (HAT), proposed a tree-based structure to distribute queries efficiently and filter out the unnecessary data updates that are not extreme values. In this paper, a data reduction algorithm is presented to reduce energy consumption of the HAT due to network transmission. The proposed method utilizes historical information of extreme values and their corresponding node ID to adjust the reporting rate of sensors properly and eases the burden of the parent of extreme nodes by balancing the packets from extreme nodes to all their possible parents. We evaluate the performance of the proposed algorithm by NS-2 network simulator and real-world data traces. The results indicate that the overall network packets are reduced to 80% with 1% data error in comparison with HAT.

Keywords

• Sensor networks
• Extreme values
• Data reduction
• Power saving