Secrecy Performance of RIS-enhanced Communication under Imperfect CSI

Abstract

In this work, we consider an reconfigurable intelligent surface (RIS) enhanced aerial communication system and reconfigurable intelligent surface (RIS) aided uplink noma communication system, in RIS-enhanced aerial communication system we have an un manned aerial vehicle (UAV) intends to deliver confidential information to the legiti mate user in the presence of an eavesdropper with the aid of an RIS. For an accurate performance analysis, we consider the practical constraints of imperfect channel state information (CSI) and discrete phase shifts for the RIS-enhanced system. Furthermore, in order to compensate for the secrecy loss due to the practical constraints, we formulate a secrecy rate maximization problem under the target secrecy rate constraint to optimize the number of RIS elements. Our extensive simulation results demonstrate the impact of various factors, namely the channel estimation error, RIS phase shift design, number of RIS elements, and transmit power, on the secrecy performance of the considered system. Moreover, the simulation results reveal that the secrecy performance degrades to a great extent in the case of imperfect channel estimation of the RIS to a legitimate user’s link. In case of RIS-aided uplink noma communication system, the UE’s wants to transmit confidential information to the base station (BS) in presence of an eavesdropper. We evaluate the impact of transmit power, distance of UE’s and eavesdropper from RIS and split factor on secrecy performance under Nakagami-m fading channels of the considered system. The simulation results reveal that the secrecy rate increases with increase in the trasnmit power of UE’s and the distance of eavesdropper from RIS, while secrecy rate degrades with increase in the distance of UE’s from the RIS. Moreover, there is an increase in the secrecy rate of UE1 and decrease in the secrecy rate of UE2 with an increase in the split ratio as the number of RIS elements for the far user imcreases with increase in the split ratio.