Research on Ground Wide-Area Network Communication Coverage Method Based on Multi-UAV Swarm

With the rapid development of 5G/6G, smart cities, and emergency communications, traditional ground communication infrastructure has exposed bottlenecks such as long deployment cycles and poor adaptability in rapid response, wide-area coverage, and dynamic scheduling. Due to their high mobility, flexible networking, and rapid deployment capabilities, multi-UAV swarms are regarded as a key technology to enhance ground wide-area network communication coverage. However, shortcomings remain in multi-dimensional environment perception, swarm adaptive control and self-healing, and edge intelligence collaboration and proactive resource scheduling capabilities. This paper proposes a comprehensive method integrating multi-source heterogeneous perception—edge computing—multi-agent reinforcement learning (MARL)—and user intention prediction, aiming to achieve wide-area communication coverage and resilience enhancement in complex dynamic environments. It also proposes a simulation–physical system integrated verification approach. The core ideas include: constructing a swarm self-organizing/self-healing model driven by multi-dimensional perception, designing edge intelligence collaboration and global consistency mechanisms, establishing a user-intention-driven proactive resource scheduling system, and building an evaluation framework based on NS-3/OMNeT++ and physical platforms. The proposed solution is oriented toward scenarios such as emergency communication, smart cities, and coverage in remote areas, demonstrating significant potential for engineering application.