Fragmentation-Aware Spectrum Allocation for Dynamic Traffic in Elastic Optical Networks

Efficient spectrum allocation and routing in elastic optical networks (EONs) are essential for improving resource utilization, especially under dynamic traffic conditions where fragmentation is a major challenge. This paper proposes a fragmentation-aware spectrum allocation scheme using an integer linear program (ILP) to optimize routing and spectrum allocation. The ILP evaluates k-shortest paths and selects the one that minimizes spectrum usage and fragmentation for each incoming request. If a feasible solution is found, the request is accepted; otherwise, it is blocked. While the ILP effectively reduces blocking and spectrum waste, its computational complexity limits scalability in large networks. To address this, a heuristic approach is introduced. Numerical results demonstrate that the proposed scheme outperforms existing schemes in reducing blocking probability and enhancing resource utilization under dynamic traffic conditions.