Reliability-Intercept Gap Analysis of Underlay Cognitive Networks Under Artificial Noise and Primary Interference

Abstract

Underlay mode of cognitive radio networks (CRNs) permits secondary users (SUs) to simultaneously operate with primary users (PUs), inducing mutual interference between them. Nevertheless, primary interference (interference from PUs to SUs) is either neglected or treated as Gaussian noise in existing works. Moreover, artificial noise, which is deliberately introduced to harm signal reception of wire-tappers, can enhance information securing capability of CRNs. This paper analyzes a gap between reliability probability (successful decoding probability of legitimate receiver) and intercept probability (successful decoding probability of wire-tappers), which accounts for artificial noise, peak transmit power constraint, interference power constraint, and fading channels, and considers primary interference as non-Gaussian noise. Towards this end, exact closed-form expressions of reliability probability and intercept probability are first derived and then validated by computer simulations. Numerous results illustrate that both probabilities are saturated at either large peak interference power or large peak transmit power, the primary interference dramatically deteriorates them while the artificial noise creates a large gap between them, demonstrating its effectiveness in securing legitimate information.