Research On Adaptive Hybrid Beamforming Algorithm In Massive MIMO System

With the development of wireless communication,massive multiple input multiple output(Massive MIMO)technology has become one of the research hotspots.In conventional MIMO system,the number of antennas is usually small,where usually the full digital beamforming is applied.However,in Massive MIMO system,due to the huge number of antennas,the number of required RF chains will also be enormous.The hybrid beamforming scheme which can reduce the number of RF chains is an effective solution.In the hybrid beamforming systems,the previous research is mainly based on infinite precision phase shifters.However,the use of infinite precision phase shifters will bring about high hardware complexity,high overhead,and it is difficult to achieve infinite precision.Based on the above background,this paper designs the adaptive hybrid beamforming schemes in Massive MIMO system in the 5G millimeter wave scenario,which is to solve the problem of the compromise between system performance and energy consumption.The performance of the algorithm was simulated and evaluated.The main work of this paper is divided into the following points:Firstly,in order to reduce the hardware complexity and device implementation difficulty,this paper considers the finite precision phase shifter,that is,the constant mode discrete phase condition.Under this constraint,a hybrid beamforming scheme based on genetic algorithm is proposed.In this paper,the performance of the proposed algorithm under different discrete phase precisions is simulated.The simulation results show that the minimum transmit power under some certain discrete phase conditions can approach the transmit power under continuous phase conditions,but the accuracy of the phase shifter in the Massive MIMO system is reduced,thus reducing the overhead and complexity of the system.Secondly,further considering reducing the energy consumption and overhead,in this paper we adopt the switch structure instead of the phase shifter,and the adaptive connected method is used to improve the system performance of the switch structure.Then the hybrid beamforming algorithm is designed under the switch adaptive connected structure.In the single-user system,based on the switch structure,with the objective function of maximizing user data rates,we design a two-stage hybrid beamfonning scheme.In this paper,the performance of the proposed algorithm is simulated and evaluated.The simulation results show that the proposed hybrid beamforming algorithm has high energy efficiency,and it can reduce the energy consumption with less loss of user data rates,so that the performance and energy consumption of the system can achieve a compromise.Finally,this paper researches on the hybrid beamforming algorithm under the adaptive switch-based structure in multi-user Massive M3MO systems.The hybrid beamforming schemes based on different group intelligent optimization algorithm are given owing to the non-convex objective function.The simulation results show that in the multi-user system,compared with the phase shifter-based structure,the hybrid beamforming algorithm under the switch-based adaptive connected structure can achieve a better performance,which can ensure the system performance while reducing the system energy consumption to a certain extent.