Fusion of Deep Learning and Adaptive Networking for UAV Swarm Communication Performance Optimization

A distributed management system for multi-UAV swarm networks is designed based on mobile ad hoc networks. The system features a hierarchical architecture with dual-mode wireless communication and edge computing units. Through optimized protocol stacks and management algorithms, the system demonstrates stable performance in a 30-node swarm test. Results show successful dynamic topology reconstruction with end-to-end delay under 18.5ms and network throughput of 875Mbps, offering a reliable solution for UAV swarm management.