A Robust Beamforming Based Interference Control Technique and its Performance for Cognitive Radios

Abstract

Cognitive radios have the ability to efficiently utilize the spectrum, by intelligently sensing absence of licensed users and transmitting signal through the frequency bands that are not occupied. Spatial diversity techniques at the cognitive base station could further help in controlling the interference temperature levels of the licensed users when signals are transmitted to unlicensed users. This paper analyses performance of such interference control techniques in terms of bit error rate of the licensed and unlicensed users and proposes a worst-case performance optimization based robust technique to mitigate effect of inaccurate channel state information (CSI) on the performance of both the licensed and unlicensed users. We use an ellipsoid based convex hull for the CSI errors and demonstrate robust technique has the ability to force the interference temperature of the licensed users below a target value all the time, while the non robust technique does not satisfy this target most of the time.