Abstract:We consider a mission-driven multiple unmanned aerial vehicles (multi-UAVs) network with millimeter-wave (mmWave) transmissions, where a leader UAV (LUAV) communicates with a large number of follower UAVs (FUAVs) simultaneously via a uniform planar array with only a limited number of radio frequency chains. Since only a few orthogonal beams are available and mission-driven UAV networks are autonomous and delay-sensitive, non-orthogonal multiple access (NOMA) over these beams is considered for agility and efficiency. In particular, aiming to address these challenges of highly dynamic mobility, we propose a machine learning framework to enable agile analog beam management for mmWave-NOMA transmissions. The proposed beam management aims to achieve efficient beam tracking and NOMA-grouping-aware analog beamforming optimization to facilitate mmWave-NOMA transmissions by judiciously utilizing angular domain information (ADI). More specifically, a Gaussian process machine learning-based ADI prediction scheme is proposed to track the angular dynamics of FUAVs, which facilitates fast beam-tracking for mmWave-NOMA transmissions. Moreover, by exploiting the predicted ADI, an unsupervised-learning-based FUAV grouping scheme is proposed to facilitate mmWave-NOMA transmissions with high radio-frequency chain efficiency, while a deep learning-based NOMA-grouping-aware fast transmit beamforming optimization scheme is proposed to improve the coverage of mmWave-NOMA transmissions in highly dynamic multi-UAVs networks. Simulation results validate the performance advantages of our proposed beam management scheme against state-of-the-art schemes.

Position Prediction Based Fast Beam Tracking Scheme for Multi-User UAV-mmWave Communications.

Adaptive Prediction-Based Beam Tracking Scheme for UAV mmWave Communications.

Position-Attitude Prediction Based Beam Tracking for UAV Mmwave Communications.

Fast Beam Tracking for Base Station and UAV Communications

KF-LSTM Based Beam Tracking for UAV-Assisted Mmwave HSR Wireless Networks.

Sensing-assisted Accurate and Fast Beam Management for Cellular-connected mmWave UAV Network

Data-Driven Beam Management with Angular Domain Information for Mmwave UAV Networks

Trajectory Prediction and Channel Monitoring Aided Fast Beam Tracking Scheme at Unlicensed Mmwave Bands

Codebook-Based Beam Tracking for Conformal Array-Enabled UAV Mmwave Networks

Machine-Learning Beam Tracking and Weight Optimization for mmWave Multi-UAV Links

Codebook-Based Beam Tracking for Conformal ArrayEnabled UAV MmWave Networks

UAV Beam Alignment for Highly Mobile Millimeter Wave Communications

Machine Learning-Empowered Beam Management for mmWave-NOMA in Multi-UAVs Networks

Flight Sensor Data and Beamforming based Integrated UAV Tracking with Channel Estimation using Gaussian Process Regression

Adaptive Beam Design for UAV Network with Uniform Plane Array

Computer Vision Assisted Mmwave Beamforming for UAV-to-vehicle Links.

Learning-Enabled Radar-Assisted Predictive Beamforming for UAV-Aided Networks

Fast Multi-target Sensing and Tracking for Directional Array-Deployed UAV Systems

Beam Tracking for UAV Mounted SatCom on-the-Move with Massive Antenna Array

Location-aware Predictive Beamforming for UAV Communications: A Deep Learning Approach

Fast Beam Tracking Discontinuous Reception for D2D-Based UAV MmWave Communication.