A Novel Scheme for Separate Training of Deep Learning-Based CSI Feedback Autoencoders

In this paper, we introduce a novel scheme for the separate training of deep learning-based autoencoders used for Channel State Information (CSI) feedback. Our distinct training approach caters to multiple users and base stations, enabling independent and individualized local training. This ensures the more secure processing of data and algorithms, different from the commonly adopted joint training method. To maintain comparable performance with joint training, we present two distinct training methods: separate training decoder and separate training encoder. It’s noteworthy that conducting separate training for the encoder can pose additional challenges, due to its responsibility in acquiring a compressed representation of underlying data features. This complexity makes accommodating multiple pre-trained decoders for just one encoder a demanding task. To overcome this, we design an adaptation layer architecture that effectively minimizes performance losses. Moreover, the flexible training strategy empowers users and base stations to seamlessly incorporate distinct encoder and decoder structures into the system, significantly amplifying the system’s scalability. .

A CSI Acquisition Approach for mm Wave Massive MIMO

Massive Multiple-Input Multiple-Output(MIMO)is considered as a key technology for 4G and 5G wireless communication systems to improve spectrum efficiency by supporting large number of concurrent users.In addition,for the target frequency band of 5G system,mmWave band,massive MIMO is pivotal in compensating the high pathloss.In this paper,we investigate the channel state information(CSI)acquisition problem for mmWave massive MIMO.With hybrid analog-digital antenna architecture,how to derive the analog beamforming and digital beamforming is studied.An iterative analog beam acquisition approach is proposed to save system overhead and reduce beam searching complexity.Regarding the digital beamforming,a grouping based codebook is proposed to facilitate CSI feedback.The codebook is then extended to incorporate also analog beam acquisition.Furthermore,channel reciprocity is exploited to save CSI reporting overhead and a two-stage approach is proposed to fully utilize the channel reciprocity at both mobile station and base station side and accelerate the CSI acquisition procedure.

Hybrid CSI-RS Transmission Mechanism-Based 3D Beamforming Scheme for FDD Massive MIMO System

Massive multiple-input multipleoutput(MIMO) refers to the idea that equipping base station(BS) with a large number of antenna elements,and features the ability of three dimensional(3D) beamforming technique to enable improvement in system performance.The prior works on massive MIMO generally rely on a common assumption of the whole channel characteristics are perfectly known at both user equipment(UE) and BS,however,this is quite difficult to realize in practical frequency division duplexing(FDD) system since the channel state information reference signal(CSIRS) overhead and CSI feedback overhead are proportional to the number of antenna elements.In this paper,two hybrid CSI-RS transmission mechanism-based 3D beamforming schemes are proposed in FDD massive MIMO system,in which non-beamformed CSI-RS configuration is utilized in conjunction with beamformed CSIRS configuration.In the proposed schemes,an angle quantization-based vertical codebook and a DFT-based horizontal codebook are designed,respectively,and an eigenvalue decompositionbased precoding matrix indicator(PMI)selection algorithm is also proposed for CSI acquisition.Simulation results show that the proposed 3D beamforming schemes achieve significant improvement in system capacity without incurring excessive CSI-RS overhead.