Welcome to the InfoSci Platform
Reference Hub4
Black Hole Traffic Anomaly Detections in Wireless Sensor Network

Black Hole Traffic Anomaly Detections in Wireless Sensor Network

Tu-Liang Lin, Hong-Yi Chang
Copyright: © 2015 |Volume: 7 |Issue: 1 |Pages: 10
ISSN: 1938-0259|EISSN: 1938-0267|EISBN13: 9781466676657|DOI: 10.4018/ijghpc.2015010104
Cite Article Cite Article

MLA

Lin, Tu-Liang, and Hong-Yi Chang. "Black Hole Traffic Anomaly Detections in Wireless Sensor Network." IJGHPC vol.7, no.1 2015: pp.42-51. http://doi.org/10.4018/ijghpc.2015010104

APA

Lin, T. & Chang, H. (2015). Black Hole Traffic Anomaly Detections in Wireless Sensor Network. International Journal of Grid and High Performance Computing (IJGHPC), 7(1), 42-51. http://doi.org/10.4018/ijghpc.2015010104

Chicago

Lin, Tu-Liang, and Hong-Yi Chang. "Black Hole Traffic Anomaly Detections in Wireless Sensor Network," International Journal of Grid and High Performance Computing (IJGHPC) 7, no.1: 42-51. http://doi.org/10.4018/ijghpc.2015010104

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

With the flourish of Internet of Things, the security issues in wireless sensor network (WSN), especially traffic anomaly detections, have attracted researchers' attentions. As a distributed wireless network, WSN is vulnerable to many attacks. In this research, the authors investigate the traffic anomaly detections of a well-known attack, black hole attack, in WSNs. With limited computation capacity, sensor nodes are unable to perform sophisticated detection techniques. Therefore, the authors propose a profile based monitoring approach with a restricted feature set to supervise the network traffic. The proposed profile based monitoring approach contains two components, feature selection and anomaly detection. In order to complement the limited computing capacity of a sensor node, feature selection component will extract features with high contribution or high relevance for future monitoring. The anomaly detection component monitors the selected features and alarms the administrator when an anomaly is detected. Two types of combination are proposed, graphic and non-graphic based models. The graphic based approach seems to surpass the non-graphic based approach, but the graphic based approach takes much longer time to select the important features than non-graphic based approach.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.