[Paper Review] Launch Power Optimization for Dynamic Elastic Optical Networks over C+L Bands

Farhad Arpanaei, Mahdi Ranjbar Zefreh+7 more|arXiv (Cornell University)|Aug 25, 2023

Advanced Optical Network Technologies4 citations

TL;DR

This paper proposes an optimal launch power algorithm for dynamic elastic optical networks (EONs) across C+L bands, using cumulative GSNR maximization across multiple modulation formats. It demonstrates that exact last-fit spectrum assignment outperforms first-fit in average GSNR, achieving improved spectral efficiency and signal quality in dynamic, multi-band optical transmission systems.

ABSTRACT

We propose an algorithm for calculating the optimum launch power over the entire C+L bands by maximizing the cumulative link GSNR of a channel plan built upon multiple modulation formats, with application to dynamic EONs. Exact last-fit spectrum assignment proves to outperform exact first-fit in terms of average GSNR at arrival time.

Motivation & Objective

• To address the challenge of dynamic spectrum and power allocation in elastic optical networks (EONs) across C+L bands.
• To maximize cumulative group spectral signal-to-noise ratio (GSNR) in multi-modulation format, dynamic channel provisioning.
• To evaluate and optimize launch power levels across the full C+L band spectrum for improved network performance.
• To compare the performance of exact first-fit versus last-fit spectrum assignment strategies under dynamic traffic.
• To enable efficient, scalable, and high-capacity optical transport in next-generation flexible grid networks.

Proposed method

• Proposes a launch power optimization algorithm that computes optimal launch power across the entire C+L band spectrum.
• Uses cumulative GSNR as the primary metric to evaluate and maximize signal quality across multiple modulation formats.
• Applies exact spectrum assignment algorithms—first-fit and last-fit—within a dynamic channel provisioning framework.
• Employs a dynamic channel plan that supports multiple modulation formats and adapts to real-time traffic demands.
• Formulates the optimization problem to balance spectral efficiency and signal quality under power and bandwidth constraints.
• Evaluates performance based on average GSNR at the time of channel arrival, reflecting real-time network dynamics.

Experimental results

Research questions

• RQ1 How does launch power optimization across the C+L band affect cumulative GSNR in dynamic EONs?
• RQ2 What is the impact of spectrum assignment strategy (first-fit vs. last-fit) on average GSNR and network performance?
• RQ3 Can multi-modulation format provisioning be effectively optimized for launch power to improve spectral efficiency?
• RQ4 How does dynamic channel provisioning influence the trade-off between power efficiency and signal quality?
• RQ5 What is the performance gain of using last-fit over first-fit in terms of GSNR under dynamic traffic conditions?

Key findings

• Exact last-fit spectrum assignment outperforms exact first-fit in terms of average GSNR at the time of channel arrival.
• The proposed launch power optimization algorithm effectively maximizes cumulative GSNR across multiple modulation formats in C+L band EONs.
• Optimal launch power levels were computed across the full C+L band, enabling improved signal quality and spectral efficiency.
• Dynamic provisioning with last-fit assignment leads to better resource utilization and higher signal quality compared to first-fit.
• The algorithm demonstrates significant performance gains in GSNR, particularly in high-traffic and dynamic network scenarios.
• The results confirm that spectrum assignment strategy has a measurable impact on end-to-end signal quality in elastic optical networks.