On Genetic Algorithm-Particle Swarm Optimizationbased Cross-Layer Dynamic Networking for LEO Constellations

With the advancement of the 6 G vision, Low Earth Orbit (LEO) satellite networks, leveraging their advantages of low latency and extensive coverage, have become a crucial component of the future integrated air-space-ground network. As satellite networks rapidly evolve toward super-dense and multi-layered architectures, cross-layer networking faces significant challenges, including the increasing difficulty of selecting appropriate nodes for cross-layer link establishment and frequent link interruptions caused by harsh space environments. To address these challenges, this paper proposes a Cross-Layer Dynamic Networking (CLDN) algorithm based on the genetic particle swarm algorithm. The algorithm integrates key parameters such as signal quality, node load, and link duration to construct a link reliability utility function, enabling intelligent screening and optimization of candidate links to significantly enhance network performance. A primary-backup link redundancy mechanism is also introduced to improve network fault tolerance. Simulation results demonstrate that the CLDN algorithm outperforms the traditional Distance Vector Networking (DVN) algorithm in key metrics: the average end-to-end(E2E) hop count is reduced by $23 \%$, network latency decreases by $\mathbf{2 5 \%}$, and in simulated satellite failure scenarios, the fluctuation range of E2E latency is reduced by $35 \%$. These results highlight the algorithm’s stronger robustness and validate its effectiveness and reliability in practical applications.