FTTR Adaptive Routing Optimization Algorithm Based on Graph Convolutional Network

With the rapid expansion of Fiber To The Room (FTTR) networks, intelligent routing optimization has become crucial to maintaining high-speed, low-latency data transmission. Traditional routing methods struggle to adapt to dynamic network conditions, leading to congestion and suboptimal performance. To address this, we propose a Graph Convolutional Network (GCN)-based adaptive routing optimization algorithm that leverages graph representation learning to model the complex topological structure of FTTR networks. By embedding topological and traffic features into a latent space, the proposed GCN model predicts the transmission performance of different routing paths in real-time. These predictions enable an adaptive path selection strategy that dynamically optimizes data packet transmission, mitigating congestion and improving overall network efficiency. Experimental results on synthetic and real-world FTTR topologies demonstrate that our approach significantly outperforms traditional heuristic-based routing algorithms, achieving lower transmission latency, improved load balancing, and higher network throughput.