Modulation format identification in elastic optical networks using integrated photonic reservoir computing and...

Area of Science:

• Optical Communications
• Photonic Integrated Circuits
• Signal Processing

Background:

• Next-generation Elastic Optical Networks (EONs) require robust modulation format identification.
• Varying sensitivities to channel impairments necessitate adaptive receiver equalization.
• Pre-receiver identification enables optimized parameter adjustment for improved signal integrity.

Purpose of the Study:

• To propose and validate a novel system for recognizing optical modulation formats.
• To enhance the performance of optical channel equalization schemes.
• To improve the accuracy and efficiency of modulation format identification in optical transmissions.

Main Methods:

• Utilized a 52-node integrated photonic reservoir chip.
• Employed an untrained K-nearest neighbors (KNN) algorithm for classification.
• Tested recognition of OOK, PAM4, QPSK, and BPSK formats under varying OSNR and fiber dispersion.

Main Results:

• Achieved consistent recognition accuracy exceeding 96.25% across all tested conditions.
• Demonstrated a 14.93% improvement over previous methods and an 82.81% enhancement over traditional algorithms.
• Analyzed the influence of waveguide delays, random phases, and KNN algorithm parameters (K value) on accuracy.

Conclusions:

• The proposed photonic reservoir chip and KNN system offers high-accuracy modulation format recognition.
• This approach significantly outperforms existing methods in optical channel transmission.
• The findings pave the way for more efficient and adaptive receiver designs in future optical networks.