Title :
Afriat´s Test for Detecting Malicious Agents
Author :
Krishnamurthy, Vikram ; Hoiles, William
Author_Institution :
Dept. of Electr. & Comput. Eng., Univ. of British Columbia, Vancouver, BC, Canada
Abstract :
How can one detect if a set of agents is deliberately trying to avoid being detected? By assuming malicious agents are utility maximizers, we use a remarkable result developed by Afriat to construct a decision test that identifies such malicious agents with pre-specified Type-I error probability. Also a stochastic gradient algorithm is given to adapt the probe signals in real time to minimize the Type-II error probabilities of the decision test.
Keywords :
security of data; Afriat test; decision test; malicious agents detection; prespecifíed type-I error probability; probe signals; stochastic gradient algorithm; type-II error probabilities; Noise measurement; Optimization; Probes; Quality of service; Sensors; Signal processing algorithms; Vectors; Afriat´s Theorem; anomaly detection; networks; stochastic gradient algorithm; utility maximization;
Journal_Title :
Signal Processing Letters, IEEE
DOI :
10.1109/LSP.2012.2221711
