Interference-Cancelled Asymmetric Traffic Cellular Networks: Dynamic TDD Meets Massive MIMO

Abstract

Over the last decade, the ever-growing demand for asymmetric uplink (UL) and downlink (DL) traffic in cellular networks have made the necessity of dynamic time-division duplex (TDD). However, it is plagued by the cross-link interference of adjacent cells. To suppress it, we consider dynamic TDD with massive multiple-input multiple-output (MIMO). For a highly realistic system model, we consider the transceiver noise at the base station (BS), which would further aggravate the coupling between DL and UL. By using random matrix theory, we derive tight deterministic equivalents of the UL and DL achievable rates with a matched filtering receiver and maximum ratio transmission precoder. Exploiting these deterministic equivalents further, we analyze cross-link interference and transceiver noise, and then prove that their effects both scale down as O(M <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> ), thus their elimination is possible by massively increasing the number of antennas. Consequently, the UL-DL decoupling would be achieved in dynamic TDD systems by combining massive MIMO, which is a desirable feature in future cellular networks. Based on the analysis results, we also propose a power control scheme for maximizing the total sum rate, which is decoupled into two tractable geometric programming problems. The effectiveness of this power control is validated by extensive simulation results.