Detection of Cognitive Interference in Wireless Environments: An IQ Test in the Air

Cognitive interference, which can adapt its strategy to the states of the environment and legitimate system, is a serious threat to wireless communication systems or even more general systems such as smart grids. The interferer can also disguise itself behind the randomness of the environment, which makes the legitimate system difficult detect the interference. An intelligence detection approach is proposed in this paper, which claims a cognitive interference when the hint of intelligence in the environment against the legitimate system is captured. The mutual information between the legitimate system and the environment is used as the metric measuring the intelligence. Two approaches are proposed to compute the metric in an online manner, namely the k-nearest neighbor algorithm and Gaussian approximation. Numerical simulations in the scenario of cognitive interference against cognitive radio networks demonstrate that the proposed mechanism can effectively detect the cognitive interference.