This is the authors' abstract. We don't add key points for this paper.
Elastic optical networks (EONs) have emerged as a flexible and spectrally efficient solution for future backbone infrastructures, where dynamic routing and spectrum assignment (RSA) is a key but complex resource allocation task. Existing RSA schemes are mainly based on heuristics, mathematical optimization, or machine learning and deep reinforcement learning, often trading off blocking performance, computational complexity, and interpretability. In this paper, we propose a graph signal correlation -based RSA (GSC-RSA) scheme that exploits graph signal processing (GSP) tools to perform dynamic and proactive optical resource allocation. The physical topology is first mapped to its line graph, and both the spectrum availability and the candidate paths are modeled as graph signals defined on the line graph nodes. A correlation-based metric is then defined between path signals and spectrum availability signals constructed from contiguous slot blocks. This metric jointly captures the feasibility of each path–block pair, the path length, and the load distribution over the network. For each incoming request, GSC-RSA evaluates this metric over all candidate path–block combinations and selects the one with the highest score in a single analytic decision step. We evaluate the proposed scheme under dynamic traffic on representative EON topologies and compare it against two classical K -shortest-path first-fit and exact-fit heuristics and two recent fragmentation-aware RSA algorithms. Simulation results indicate that GSC-RSA can significantly reduce request and bandwidth blocking probabilities while maintaining competitive spectrum utilization and per-request computational cost compatible with online operation. These results demonstrate that GSP provides an effective and scalable framework for dynamic RSA in EONs.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Zarei et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: