UP	Paper 1054	US-T-I	DOWN
Distributed Detection in the Presence of Byzantine Attack in Large Wireless Sensor Networks
Marano,Stefano		University of Salerno
Tong,Lang		Cornell University
Matta,Vincenzo		University of Salerno
Wireless sensor networks are vulnerable to Byzantine attacks in which a fraction of sensors are tampered. The intruder can reprogram the compromised sensors, making them behave as if they are authentic nodes. We consider the problem of distributed detection in wireless sensor networks in the presence of Byzantine attacks where the compromised sensors collaboratively send fictious observations to the fusion center. Can such Byzantine attacks completely destroy the ability of fusion center to make reliable detection? What is the performance loss from such attacks? What is the optimal detection algorithm at the fusion center in the presence of Byzantine attacks? In this paper, we define the power of the attacker as the fraction of the sensors that have been compromised for the attack. We present the optimal attacking strategy for any given attacking power. We show that there is a critical power level above which the fusion center is completely "blinded" in the sense that the optimal detection performs no better than a coin flip independent of collected data. We also present a robust detector to counter the presence of Byzantine attacks.
Stefano Marano received the Laurea degree in Electronic Engineering (cum laude) and the PhD degree in Electronic Engineering and Computer Science both from the University of Naples, Italy, in 1993 and 1997, respectively. Since January 2005 he was appointed the permanent position of Associate Professor at the University of Salerno, where he was formerly Assistant Professor (since November 1999). His current research interests include: detection and estimation theory with emphasis on sequential and decentralized techniques; data fusion; sensor networks, stochastic modeling of electromagnetic propagation in random media. About these and related topics he has (co-)authored about 70 papers, including some invited, mainly on international journals/transactions and proceedings of international conferences. Prof. Marano serves as referee for the main international journals in the field of signal processing; he is also in the Technical Program Committee of several international symposia. Prof. Marano is co-recipient of the S. A. Schelkunoff Transactions Prize Paper Award of the IEEE Antennas and Propagation Society for the best paper published in the Transactions in 1999. Currently, Prof. Marano is in the Scientific Committee of the Remote Sensing Laboratory for Environmental Hazard Monitoring (ReSLEHM), University of Salerno, and in the Organizing Committee of the 9th International Conference on Information Fusion (FUSION 2006).