검색 상세

Data-Aided Codebook Design and Intelligent Beam Tracking Techniques for 6G Wireless Communications

6G 무선통신을 위한 데이터 기반 코드북 설계 및 지능형 빔 트래킹 기술 연구

  • 박상천 (SANGCHUN PARK, 서강대학교 일반대학원)

초록 (요약문)

In this dissertation, we propose data-aided codebook design and intelligent beam tracking techniques for 6G wireless communications. A data-aided codebook enhancement method is proposed to improve beamforming performance by efficiently utilizing reference signal resources. Unlike the 3rd Generation Partnership Project (3GPP) standard codebooks with pre-determined beamforming directions based on the discrete Fourier transformation (DFT) matrix, the proposed codebook enhancement method continuously updates the codebooks as channel conditions change. The proposed method utilizes kernel density estimation (KDE) and k-means++ clustering methods to determine beamforming directions based on precoding matrix indicator (PMI) feedback information. The generated codebook by the proposed method is shown to exhibit a negligible amount of performance degradation from the codebook constructed with full channel state information (CSI). We then propose a new method of transmitting accurate beams to highly mobile users with a substantially reduced amount of feedback overhead, by introducing a set of beam signatures that are composed of multiple beams along the trajectories of mobile users. Instead of forming a spot beam corresponding to the PMI reported by the user equipment (UE), the base station (BS) utilizes the history of previous reports to determine an appropriate beam signature and transmit beams to predicted UE positions. The proactive decision for the next beam position is made with the aid of machine learning (ML) using the train data obtained from typical mobile movements for given road conditions, thus providing the adaptability to the channel environment with progressively improving accuracy. The proposed beam signatures provide reliable beamforming over the UE trajectory, even when the CSI feedback interval is considerably longer than parameters supported by the current 5G new radio (NR) standard. We also propose an improved beam tracking method using the deep Q- network (DQN), eliminating the need for channel state information reference signal (CSI-RS) transmissions and PMI reports. The proposed DQN replaces the ϵ-greedy exploration method of conventional DQNs with a reward-dependent exploration method and dynamically optimizes the beam direction based on the environment and signal strength. Simulations show that the proposed DQN using a reward-dependent exploration method consistently outperforms the conventional DQN using ϵ-greedy exploration and the DFT codebook-based beam tracking methods with conventional CSI-RS transmissions and PMI reports.

more

목차

1. Introduction 1
1.1 Background and Related Works 1
1.2 Main Contributions 8
1.3 Organization of Dissertation 11
2. System Model for Vehicle Communications 12
2.1 Multi-user Transmission Signal Model 12
2.2 Channel Model 16
2.3 Beamformed CSI-RS 19
3. PMI-based Codebook Enhancement 23
3.1 Introduction 23
3.2 PMI-based Codebook Design 24
3.3 Performance Evaluation 30
3.4 Summary 36
4. Beam Signature Based Beam Tracking 37
4.1 Introduction 37
4.2 Proof of Concept for Beam Signatures 39
4.2.1 Location Information Based Generation of Beam Signatures 39
4.2.2 DNN-based Utilization of Beam Signatures using PMI 45
4.2.3 Performance Evaluation 54
4.2.4 Limitations 60
4.3 Practical Utilization Strategies for Beam Signatures 61
4.3.1 PMI-based Generation of Beam Signatures 61
4.3.2 Rule-based Utilization of Beam Signatures using PMI 66
4.3.3 Performance Evaluation 69
4.4 Summary 81
5. Deep Q-Network Based Beam Tracking 83
5.1 Introduction 83
5.2 Indoor Repeater Environment 84
5.3 Reward Dependent DQN Algorithm 87
5.4 Performance Evaluation 91
5.5 Summary 103
6. Conclusions 104
References 106

more
반출 Meta View 목록