Interference Violation Probability Constrained Underlay Cognitive Massive MIMO Network Under Imperfect Channel Knowledge

We investigate the use of massive number of antennas at the cognitive base station (BS) in reducing interference caused to primary users (PUs) under imperfect channel knowledge without deteriorating the data rate provided to the cognitive user (CU). To this end, we develop a simple back-off factor based power adaptation policy for the cognitive BS which ensures that its transmissions do not violate the interference violation probability constraint at the PUs. We derive a new lower bound on the complement of the interference violation probability and also deduce a lower bound on the achievable rate of the CU when the cognitive BS has an imperfect estimate of its channels to the PUs and the CU. Through our analytical and numerical results, we quantify that the interference violation probability at the PUs can be reduced while providing a fixed data rate to the CU by deploying more number of cognitive BS antennas. Furthermore, if the number of PUs in the network increase, we show that the interference violation probability at the PUs and the data rate of the CU can be maintained at the same level by increasing the number of antennas at the cognitive BS.