Dynamic on-demand defragmentation in flexible bandwidth elastic optical networks.

Area of Science:

• Computer Science
• Optical Networking
• Telecommunications

Background:

• Flexible bandwidth elastic optical networking offers promising future network capabilities.
• Spectral fragmentation is a key challenge, increasing blocking probability and degrading performance.
• Efficiently managing spectrum is crucial for network efficiency.

Purpose of the Study:

• To address the spectral defragmentation problem in flexible bandwidth elastic optical networks.
• To propose and evaluate novel defragmentation algorithms and node architectures.
• To reduce blocking probability and minimize connection disruptions.

Main Methods:

• An auxiliary graph-based approach is used to model the defragmentation problem.
• The problem is transformed into finding the maximum independent set (MIS) in the auxiliary graph.
• Heuristic defragmentation algorithms are developed and simulated.

Main Results:

• The proposed min-cost defragmentation algorithms significantly reduce blocking probability.
• A substantial minimization in the number of disrupted connections is achieved.
• The effectiveness of defragmentation-capable node architectures is demonstrated.

Conclusions:

• The auxiliary graph-based approach effectively solves the spectral defragmentation problem.
• The proposed algorithms offer a viable solution for improving optical network performance.
• This work contributes to the advancement of spectrally efficient optical networks.