Secure Robust Ergodic Uplink Resource Allocation in Relay-Assisted Cognitive Radio Networks

We investigate the ergodic uplink resource allocation problem for secure communication in relay-assisted orthogonal frequency-division multiple access (OFDMA)-based cognitive radio networks (CRNs) in the presence of a set of passive eavesdroppers where relay nodes assist the legitimate users to transmit their messages. Previous works have commonly assumed the availability of channel state information (CSI) for this type of problems. However, due to the decentralized nature of CRNs and hidden activities of eavesdroppers, the assumption of availability of exact values of CSI is not realistic. In this paper, we consider uncertainty on the estimated values of CSI between different transmitters and receivers, e.g., CSI between each legitimate transmitter and its corresponding receiver and CSI of each legitimate user and each eavesdropper. We utilize the worst-case robust formulation to find power and sub-carrier allocations in such a way that under the worst condition of error, the regulatory constraints imposed to CRN are satisfied and the secrecy rate of each secondary legitimate user is stabilized. It is well known that the robust approaches impose a high computational complexity to the system and reduce the system performance as they conservatively consider the error in the maximum extent. We demonstrate how the robust formulation can be significantly simplified and tradeoff parameters can be introduced to moderate the effect of the worst-case approach. Simulation results are provided to demonstrate the performance of CRNs for different uncertain system parameters.