Abstract
Channel estimation and hybrid precoding are considered for multi-user millimeter wave massive multi-input multi-output system. A deep learning compressed sensing (DLCS) channel estimation scheme is proposed. The channel estimation neural network for the DLCS scheme is trained offline using simulated environments to predict the beamspace channel amplitude. Then the channel is reconstructed based on the obtained indices of dominant beamspace channel entries. A deep learning quantized phase (DLQP) hybrid precoder design method is developed after channel estimation. The training hybrid precoding neural network for the DLQP method is obtained offline considering the approximate phase quantization. Then the deployment hybrid precoding neural network (DHPNN) is obtained by replacing the approximate phase quantization with ideal phase quantization and the output of the DHPNN is the analog precoding vector. Finally, the analog precoding matrix is obtained by stacking the analog precoding vectors and the digital precoding matrix is calculated by zero-forcing. Simulation results demonstrate that the DLCS channel estimation scheme outperforms the existing schemes in terms of the normalized mean-squared error and the spectral efficiency, while the DLQP hybrid precoder design method has better spectral efficiency performance than other methods with low phase shifter resolution.
Highlights
D UE to the rich bandwidth resources of the millimeter wave, mmWave communication has attracted broad attention and become an important technology in future wireless communication systems [1], [2]
To remove the constraint that the phase shifter resolution was related to the number of antennas, a quantized angle linear search (QALS) precoding scheme was proposed [14], where the angular domain was quantized according to the limited resolution of phase shifters and a linear search method was used to obtain the optimal analog beamforming vectors aligning with the dominant channel paths
We compare the deep learning quantized phase (DLQP) method with the Exhaustion hybrid precoder design method, i.e., we generate the analog precoding matrix F R for 30, 000 times, where each entry of F R is randomly drawn from the set {ej2πn/Q, n = 1, 2, . . . , Q}, and the digital precoder is designed according to Algorithm 2
Summary
D UE to the rich bandwidth resources of the millimeter wave (mmWave), mmWave communication has attracted broad attention and become an important technology in future wireless communication systems [1], [2]. To remove the constraint that the phase shifter resolution was related to the number of antennas, a quantized angle linear search (QALS) precoding scheme was proposed [14], where the angular domain was quantized according to the limited resolution of phase shifters and a linear search method was used to obtain the optimal analog beamforming vectors aligning with the dominant channel paths. We investigate sparse channel estimation and hybrid precoding considering the limited resolution of phase shifters for multi-user mmWave massive MIMO systems. In the offline training stage, we obtain the training hybrid precoding neural network (THPNN) using the estimated channel vector and real channel vector of each user, where the approximate phase quantization is considered. The channel vector hu ∈ CNA for the BS and the uth user is represented as
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
