Agentic Low-Altitude Wireless Networks: Fundamentals, Design, and Future Directions

Abstract

The rapid expansion of the low-altitude economy (LAE), driven by the increasing deployment of large-scale drone systems, has significantly heightened the demand for flexible, low-latency, and highly reliable connectivity in rapidly changing environments. To this end, low-altitude wireless networks (LAWNs) have emerged as a unified aerial-terrestrial architecture that integrates communication, sensing, control, and computing to enable adaptive connectivity. However, LAWNs continue to rely on centralized scheduling and static configurations, which constrain their scalability, adaptability, and end-to-end closed-loop responsiveness. Recent advances in multi-agent systems (MASs) and agentic artificial intelligence (AI) have inspired the development of agentic LAWNs that support autonomous decision-making, cooperative behavior, and improved end-to-end responsiveness. In this paper, we present a comprehensive survey of agentic LAWNs covering architecture, functional design, enabling technologies, and applications. Specifically, we outline the foundational concepts of LAWNs, discuss the transition toward agentic architectures enabling self-organization and goal-directed coordination, review the design framework and enabling technologies, and explore the representative applications. Finally, we highlight open challenges in energy allocation, network management, cybersecurity, and ethical governance to inform future research toward intelligent, resilient, and mission-aware low-altitude ecosystems.