Abstract: Compared with time-multiplexed pilot, superimposed pilot has been widely concerned for high bandwidth utilization and fast transmission rate. Therefore, the intelligent reflecting surface-aided massive multiple-input multiple-output ( MIMO) system in superimposed pilot mode is studied. Besides, the closed-form expression of sum achievable rate is derived when base stations use least square channel estimation and maximum ratio combining detection. In addition, an iterative optimization algorithm based on geometric programming and genetic algorithm is proposed to improve the mutual interference between the pilot signal and the data signal in superimposed pilot mode. The proposed algorithm can maximize the sum achievable rate of the system by cooperatively controlling the power control coefficient of the pilot signal, the power control coefficient of the data signal and the phase of the intelligent reflecting surface. The simulation results verify the accuracy of the closed-form expression of the achievable rate, and show that the proposed algorithm improves effectively the system rate performance.

Key words: intelligent reflecting surface, massive multiple-input multiple-output, superimposed pilot, geometric programming