Pilot Allocation and Power Optimization of Massive MIMO Cellular Networks With Underlaid D2D Communications

Abstract

Pilot pollution and limited power have become two important factors limiting the throughput of massive multi-input–multioutput (MIMO) systems. To improve system performance, we consider device-to-device (D2D) communication underlay massive MIMO (denoted “massive MIMO-D2D” for short) cellular networks and then perform pilot allocation and power optimization under this network. To begin with, the closed-form spectrum efficiency (SE) expressions for different types of users are derived in the massive MIMO-D2D cellular network. Then, we analyze the deficiencies of the existing pilot allocation schemes and propose a new pilot allocation problem, i.e., the SE product is maximized for enhancing the system SE and ensuring fairness among the users simultaneously. To solve the maximum SE product problem, we develop a pilot gray wolf prey (PGWO) algorithm by designing the fitness value used to measure pilot pollution and the global objective function used to evaluate the quality of pilot allocation. The PGWO algorithm can find the optimal SE accurately through a global search, and it is suitable for the pilot allocation of different models from single cell to multicell. Besides, we formulate the maximum–minimum fairness problem for power optimization and prove that the power objective function conforms to linear programming, and a bisection algorithm is provided to optimize the power. Simulation results show that the proposed SE product problem with the proposed PGWO algorithm promotes fairness for users while further enhancing SE compared to the existing pilot allocation schemes, and joint pilot allocation and power optimization achieves great sum SE over only pilot allocation.