Abstract
In this work, the impacts of imperfect channel estimations on the performance of cognitive amplify-and-forward (AF) relaying system with antenna selection and mutual primary-secondary interference are investigated. For the considered model, statistical knowledge of the feedback channel is assumed to be available at the secondary system and closed-form and asymptotic outage and ergodic capacity expressions are derived to get insights about important system parameters including primary-secondary interference and impact of channel estimation errors. Theoretical results which are validated by the simulations show that even though antenna selection improves system performance substantially, imperfect channel estimations have detrimental effects on the performance of cognitive relay systems.