Dynamic Migration Scheme for Load Balancing and Average Delay Optimization in SDN-Based Multi-UAV Relay Network

Multi-UAV relay network can provide information relay and forwarding services for ground users who cannot communicate directly with the help of building reliable connectivity links. Software Defined Networking (SDN), a new network architecture, provides a promising solution for addressing problems such as load balancing, topology changes, and information sharing in multi-UAV relay network. Aiming at the load balancing problem of multi-UAV relay based on SDN, a dynamic migration algorithm based on load balancing and delay (DMA-LBD) is proposed by using the idea of dynamic migration. This ensures that users within the coverage area of UAVs with larger loads are migrated to other UAVs, thereby establishing new connection relationships that balance the load between all UAVs within the relay network. Experimental results demonstrated that, in comparison to other methods, the proposed algorithm makes a suitable compromise between the relay load balancing and the average delay of relay link. It can avoid the single optimization situation of an imbalanced UAV relay load or an excessive average delay, providing a more effective solution to the problem of relay load balancing of multi-UAV relay network.