A Smart Pilot Assignment in Multi-Cell Massive MIMO Systems Using Virtual Modeling of Assigning Cost

Abstract

In this paper, a smart pilot sequence assignment method is proposed to minimize inter-cell interference generated in a massive multi-input multi-output (MIMO) system due to pilot contamination in uplink TDD (Time Division Duplex) mode. The proposed method employs a zero-one integer linear programming method as the assignment algorithm. The amount of intercell interference imposed on the target cell users is considered as assigning cost. The introduced assigning cost is composed of the steering vectors of the desired users in the target cell, and the sum of the channel correlation matrices of interference users in adjacent cells. By exploiting the virtual channel modeling technique, we develop a simple alternative approach to calculate the assigning cost, which has low complexity. We prove that the new assigning cost, called virtual assigning cost, is equal to the assigning cost defined based on the physical channel model. Moreover, we develop a token-based protocol to manage the coupling between cells in the whole multicellular network based on a distributed manner without the need for data exchange between the BSs. Simulation results demonstrate that the proposed low complex smart pilot sequence assignment method achieves a good performance, which is better than those of some other related works in multicellular design regarding normalized mean square error (NMSE) and achievable rate criteria.