Home > Published Issues > 2016 > Volume 11, No. 1, January 2016 >

A Novel Event Detection Framework for Wireless Sensor and Actor Networks

Haoyun Wang 1,2, Yuhan Fei 1, Jiaojiao Liu 1, Yingjun Xiong 1,2, Shougang Ren 1,2, and Huanliang Xu1,2
1. College of Information Science and Technology, Nanjing Agricultural University Nanjing, China
2. Collaborative Innovation Center of Meat Production and Processing

Abstract—Wireless Sensor and Actor Network (WSAN) can observe the physical world using event detection methods by sensor nodes and implement appropriate operation by actor nodes. Compared with Wireless Sensor Network (WSN), WSAN set high requests for real-time, collaboration, mobility, etc. Therefore, the event detection methods applied to WSN cannot be adopted for WSAN. In this paper, we present a three-layer Cooperative Event Detection Framework (3-CEDF) for WSAN. Firstly, sensor nodes cooperate with each other to affirm the occurrence of events via voting mechanism. Then, cluster-head nodes use a modified K-means algorithm and a proposed type matching mechanism to detect the event types with the boundary information of event data. Finally, actor nodes get the priority of events processing using a modified k-nearest neighbor algorithm with the slope similarity measure. The performances of effectiveness, real-time, mobility and communication traffic for 3-CEDF are obtained through analysis and simulations, when 3-CEDF is compared to the centralized event detection method adopted for WSN.

Index Terms—Wireless sensor and actor network, event detection, cooperative algorithm, type matching, similarity measure

Cite: Haoyun Wang, Yuhan Fei, Jiaojiao Liu, Yingjun Xiong, Shougang Ren, and Huanliang Xu, “A Novel Event Detection Framework for Wireless Sensor and Actor Networks," Journal of Communications, vol. 11, no. 1, pp.1-14, 2016. Doi: 10.12720/jcm.11.1.1-14