AI-Driven Low-Altitude Economy: Spectrum, Mobility, and Validation

TL;DR

This paper addresses the challenge of realizing AI-driven Low Altitude Economy (LAE) networks in densely populated, highly mobile airspace where traditional rule-based methods fail. It proposes a multi-faceted approach combining AI-enabled spectrum management (including FedSNR federated learning), interference mitigation, reinforcement-learning–driven resource allocation, and trajectory planning, all validated through realistic testbeds and digital-twin platforms. Key contributions include the FedSNR dynamic spectrum sensing framework with edge CNNs, joint RL-based optimization for spectrum and computing under dynamic conditions using a digital twin network, and practical insights from the AERPAW experimental platform that bridge simulation and deployment. The work lays out a roadmap for interoperable, scalable LAE ecosystems, highlighting open issues in standardization, security, explainability, and real-world validation to accelerate certification and deployment.

Abstract

The Low Altitude Economy (LAE) network, with its transformative capabilities, is a candidate to become one of the major technological developments of the next decade for air mobility. However, the expected unprecedented density, mobility, and heterogeneity pose challenges and require new approaches, as it renders traditional rule-based approaches inadequate. To address these challenges, this study introduces artificial intelligence (AI)-based approaches and validation frameworks for transitioning AI-enabled technologies from simulation-based studies to practical and deployable systems. This study discusses essential enablers for intelligent LAE networks. First, AI-based spectrum sensing and coexistence utilizing the distributed nature of LAE nodes is introduced. Then, joint resource allocation and trajectory optimization driven by reinforcement learning is discussed. Bridging the gap between simulation and deployment through experimental platforms such as Aerial Experiments and Research Platform for Advanced Wireless (AERPAW), which are critical for validating models under realistic and non-stationary airspace conditions, is also addressed. The study concludes by highlighting open issues and outlining a forward-looking roadmap for the development of efficient, interoperable, and scalable AI-driven LAE ecosystems.

Figures (4)

• Figure 1: Illustration of an AI-driven Low-Altitude Economy ecosystem integrating intelligent communication, sensing, and computing capabilities. Machine learning enhances coordination, autonomy, and efficiency across diverse aerial operations, supported by seamless air-ground collaboration and spectrum-aware infrastructure.
• Figure 2: Framework illustrating the interaction between LAE challenges, AI enablers, and experimental validation toward deployable intelligent aerial networks.
• Figure 3: FL-based spectrum sensing performance results.
• Figure 4: The AERPAW platform supports three operational modes: physical testbed, sandbox, and digital twin environments. In the CDF plot, solid and dashed curves correspond to measurements taken below and above 50 m altitude, respectively. The vertical black dotted line represents the threshold value of $-45$ dBm, which is used to determine the spectrum occupancy across frequency bands.