Abstract
Millimeter wave (mmWave) communications can potentially meet the high data-rate requirements of unmanned aerial vehicle (UAV) networks. However, as the prerequisite of mmWave communications, the narrow directional beam tracking is very challenging because of the three-dimensional (3D) mobility and attitude variation of UAVs. Aiming to address the beam tracking difficulties, we propose to integrate the conformal array (CA) with the surface of each UAV, which enables the full spatial coverage and the agile beam tracking in highly dynamic UAV mmWave networks. More specifically, the key contributions of our work are three-fold. 1) A new mmWave beam tracking framework is established for the CA-enabled UAV mmWave network. 2) A specialized hierarchical codebook is constructed to drive the directional radiating element (DRE)-covered cylindrical conformal array (CCA), which contains both the angular beam pattern and the subarray pattern to fully utilize the potential of the CA. 3) A codebook-based multiuser beam tracking scheme is proposed, where the Gaussian process machine learning enabled UAV position/attitude predication is developed to improve the beam tracking efficiency in conjunction with the tracking-error aware adaptive beamwidth control. Simulation results validate the effectiveness of the proposed codebook-based beam tracking scheme in the CA-enabled UAV mmWave network, and demonstrate the advantages of CA over the conventional planner array in terms of spectrum efficiency and outage probability in the highly dynamic scenarios.
Introduction
Unmanned Aerial Vehicles (UAVs) are increasingly deployed for high-bandwidth tasks such as 4K video streaming and emergency relaying, which necessitate the use of the Millimeter Wave (mmWave) spectrum. However, mmWave links rely on highly directional "pencil-thin" beams that are difficult to maintain during flight. UAVs do not just move linearly; they tilt, roll, and pitch, causing narrow beams to frequently lose alignment. Furthermore, conventional flat antennas (Uniform Planar Arrays) have limited fields of view, leading to link loss when a target moves behind the antenna's plane.
To address these issues, this research proposes the use of Cylindrical Conformal Arrays (CCA)—antennas wrapped around the aircraft's fuselage—to achieve 360-degree spatial coverage. By combining this hardware with a specialized hierarchical codebook and Gaussian Process (GP) machine learning for movement prediction, the system maintains high-speed links even during complex 3D maneuvers.
System Architecture and CCA Design
The proposed framework utilizes a CCA-enabled UAV network where multiple transmitter UAVs (t-UAVs) communicate with a central receiver UAV (r-UAV). Unlike flat panels, the CCA consists of Directional Radiating Elements (DREs) distributed across a cylindrical surface, allowing for full spatial coverage.
As illustrated in the system model, the "CCA view" shows how specific subarrays are activated to form beams toward different targets. This dynamic subarray partitioning is a core feature of the conformal design, allowing the system to physically "move" the active part of the antenna around the cylinder as the relative positions and attitudes of the UAVs change.
Methodology: Hierarchical Codebook and Predictive Tracking
The tracking solution relies on a specialized hierarchical codebook and a machine learning-based prediction engine to stabilize the high-speed links.
Cylindrical Conformal Array Geometry
The CCA geometry is defined by $M$ elements along the z-axis and $N$ elements around the xy-plane. This structure is critical for the codebook design, which must account for the curvature of the array and the directional nature of the individual elements.
Hierarchical Codebook and GP Prediction
A specialized hierarchical codebook was developed to jointly manage subarray activation and beamforming. Because the array is curved, only a subset of elements can effectively "see" a target at any given time. The codebook facilitates the rapid selection of the optimal subarray and phase weights.
To handle tracking latency, the system uses Gaussian Processes (GP) to predict the future position and attitude of the UAV. Crucially, the system calculates a Tracking Error (TE) bound based on the GP's variance. If the prediction is uncertain, the system adaptively widens the beam to ensure a higher "hit" probability; if the prediction is certain, it narrows the beam to maximize data rates.
Key Results and Insights
Simulations conducted at 60 GHz using a Smooth-Turn mobility model demonstrate the effectiveness of the proposed scheme:
- Spectral Efficiency: The CCA-enabled scheme significantly outperformed conventional Uniform Planar Arrays (UPA). While UPA performance drops sharply as UAVs move out of the flat array's effective range, the CCA maintains stable spectral efficiency across all angles.
- Outage Probability: The "TE-aware" beamwidth control ensured that the link remained active even during high-speed maneuvers, resulting in a much lower outage probability compared to non-adaptive methods.
- Latency: The total local processing time for the tracking algorithms was measured at less than 4ms on standard onboard hardware, meeting the requirements for real-time beam tracking in highly dynamic environments.
Conclusion
Integrating conformal arrays with predictive machine learning offers a robust solution for beam tracking in dynamic UAV mmWave networks. This approach overcomes the field-of-view limitations of traditional arrays and provides the agility needed for high-speed aerial communications. Future work will explore spherical conformal arrays and hybrid beamforming architectures to further enhance performance and reduce hardware complexity.
Summary updated