Sum Rate Analysis of Cognitive Massive MIMO Systems with Underlay Spectrum Sharing

Abstract

The asymptotic sum rate performance of multi-cell/multi-user cognitive massive multiple-input multiple-output (MIMO) systems with underlay spectrum sharing is investigated. Specifically, each cell consists of a licensed (primary) multi-user massive MIMO system and a cognitive (secondary) multi-user MIMO system which is allowed to utilize the licensed frequency spectrum provided that the intra-cell interference inflicted at the primary base-station due to the concurrent transmissions of the secondary user nodes is maintained below a predefined interference temperature. For the uplink transmission of the aforementioned system set-up, the signal-to-interference-plus-noise ratio and achievable sum rate expressions are derived for three specific antenna configurations; (i) infinitely many primary and secondary base-station antennas, (ii) infinitely many primary base-station antennas and finitely many secondary base-station antennas, and (iii) finitely many primary and secondary base-station antennas. Our asymptotic analysis reveals that the achievable sum rate expressions become independent of the primary interference threshold whenever the number of primary base-station antennas grows unbounded. Consequently, the secondary network can be operated at its maximum average transmit power level without degrading the asymptotic performance of the primary network. Therefore, the primary and secondary networks can be operated independent of each other as both intra-cell and inter-cell interference can be asymptotically mitigated by exploiting the zero-forcing detectors employed at the massive MIMO enabled base-stations.