Abstract:As an important part of the fifth generation (5G) mobile networks, unmanned aerial vehicles (UAVs) have been applied in various communication scenarios due to their high operability and low cost. In this paper, we investigate a multi-UAV communication system with moving users and consider the co-channel interference caused by the transmissions of all other UAVs. To ensure the fairness, we maximize the minimum average user rate during the observed time by jointly optimizing UAVs' trajectories, transmission power, and user association. Considering that UAVs can cover a large area for communications, UAVs do not need to move as soon as the users move. Therefore, a two-timescale structure is proposed for the considered scenario, where the UAVs' trajectories are optimized based on the channel state information (CSI) in a long timescale, while the transmission power and the user association are optimized based on the instantaneous CSI in a short timescale. To effectively tackle this challenging non-convex problem with both discrete and continuous variables, we propose a joint neural network (NN) design, where a deep reinforcement learning based Pointer Network named advantage pointer-critic (APC) is applied to optimize discrete variables and a deep-unfolding NN is used to optimize the continuous variables. Specifically, we first formulate a Markov decision process to model the user association, and then employ the APC network trained by the advantage actor-critic algorithm to address it. The APC network consists of a Pointer Network and a Multilayer Perceptron. As for the deep-unfolding NN, we first develop a block coordinate descent based algorithm to optimize the UAVs' trajectories and transmission power, and then unfold the algorithm into a layer-wise NN with introduced trainable parameters. These two networks are jointly trained in an unsupervised fashion. Simulation results validate that the proposed joint NN significantly outperforms the optimization algorithm with much lower complexity, and achieves good performances on scalability and generalization ability.

Predicting Spectrum Status Duration using Non-Linear Homotopy Estimation based HMM for UAV Communications

A Spectrum Status Duration Prediction Method Based on Homotopy Theory for UAV Communication

Predicting Spectrum Occupancies Using a Non-Stationary Hidden Markov Model

A non-stationary channel prediction method for UAV communication network with error compensation

Joint Resource Allocation and Trajectory Design for Multi-UAV Systems With Moving Users: Pointer Network and Unfolding

Time Allocation Optimization and Trajectory Design in UAV-Assisted Energy and Spectrum Harvesting Network

Machine Learning for Predictive Deployment of UAVs With Multiple Access

A 3-D Space-Time-Frequency Non-Stationary Model for Low-Altitude UAV Mmwave and Massive MIMO Aerial Fading Channels

Predicting the Health Status of an Unmanned Aerial Vehicles Data-Link System Based on a Bayesian Network

Collaborative Wideband Spectrum Sensing and Scheduling for Networked UAVs in UTM Systems

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

Learning-Aided Beam Prediction in mmWave MU-MIMO Systems for High-Speed Railway

Toward UAV-Enabled Stereoscopic UL Heavy NOMA: Joint Resource Allocation and 3D Trajectory Design

Frequency-Offset Information Aided Self Time Synchronization Scheme for High-Dynamic Multi-UAV Networks

Hybrid-Mode Multiple Access for UAV-BS Assisted Communications with UL-DL Rate Balancing

Proactive Blockage Prediction for UAV assisted Handover in Future Wireless Network

One-Step Bandwidth Assignemnt and Power Allocation for UAV-Enabled UL Heavy NOMA Systems

Unmanned Aerial Vehicle Flight Data Anomaly Detection Based on Multirate-Aware LSTM

UAV-Enabled Accompanying Coverage for Hybrid Satellite-Uav-Terrestrial Maritime Communications

A Hybrid Approach for UAV Flight Data Estimation and Prediction Based on Flight Mode Recognition.

Hybrid Offline-Online Design for UAV-Enabled Data Harvesting in Probabilistic LoS Channel