Low Complexity Beam Selection Scheme Based On Group Iteration Algorithm

Millimeter wave(mmWave)communication technology can provide abundant spectrum resources and well support high-dimensional multiple-input multiple-output(MIMO)system.However,MIMO systems need to be equipped with a large number of radio-frequency(RF)chains in practical applications,which leads to unbearable hardware costs.Beam selection technology based on beamspace channel can effectively reduce the number of RF chains in communication system without significant performance loss.This paper studies the beam selection technology,and mainly completes the following two aspects:Firstly,aiming at the problem that the existing beam selection schemes are difficult to balance the system performance and computational complexity,this paper designs a beam selection scheme based on group iteration(GI)algorithm.According to whether users share the same maximum gain beam,all users are divided into fixed users and iterative users.For fixed users,the maximum gain fixed beam is allocated directly;For iterative users,the iterative algorithm designed in this paper is used to select the appropriate beam.The simulation results show that when the sum capacity reaches 250bits/S/Hz,the signal-to-noise ratio(SNR)of GI scheme can be reduced by about 10%compared with the traditional amplitude maximization beam selection scheme;When the number of users is 32,the energy efficiency can reach 58bps/Hz/W,which is nearly 50%higher than the traditional algorithm;At the same time,the computational complexity of GI scheme is O(K),which is at an ideal level compared with traditional algorithms.Secondly,considering that the demand of mobile users for real-time communication system is becoming higher and higher,this paper designs an adaptive neighbor iteration(ANI)algorithm based on user grouping in time-varying channel.Due to the discreteness and fixity of the beam transmitting angle in the beamspace channel,the change of the position of the user in the adjacent slot and the change of the beam number mapped to the transmitter is relatively small.Based on the above characteristics,we propose to construct the current slot beam selection matrix by using the previous slot beam result set.The simulation results show that in the time-varying channel environment with 32 users,the energy efficiency of ani scheme is about 20bps/Hz/W higher than that of amplitude maximization scheme,and the performance is nearly 50%;At the same time,the computational complexity of ani scheme is O(K2I),which has certain advantages over the traditional beam selection scheme.The above work is simulated and verified in MATLAB.