CN

AI Enables Optical Network, FiberHome Realizes a Major Breakthrough in OSNR Intelligent Monitoring Technology

Time:2018-12-27  

With the development of a new generation of artificial intelligence technologies such as reinforcement learning and migration learning, machines are becoming more and more intelligent. The introduction of artificial intelligence technology into the optical network to achieve optical network independent planning, independent configuration, independent optimization, and independent cure, and finally achieve the goal of optical network self-awareness, achieve closed-loop autonomy and self-evolution, is the future development trend of optical network.

Break through intelligent monitoring technology and consolidate the foundation of AI
The prospects of optical network autonomy is good, but currently still faces many challenges: the combination of artificial intelligence general technology and optical network specific knowledge, how to let AI empower optical network, and the deep learning and other intelligent technologies in the field of optical network. A series of problems need to be solved. Among these challenges and problems, how to realize optical signal-to-noise ratio (OSNR) intelligent monitoring based on artificial intelligence technology is the key. At present, FiberHome has successfully broken through the key technology of optical network OSNR intelligent monitoring. Based on the deep learning network, the OSNR value is calculated. The accuracy of the transceiver and intermediate nodes is within 0.5dB, which is in the leading position in the industry. The breakthrough of optical network OSNR intelligent monitoring technology provides valuable experience for artificial intelligence technology in the field of optical network, and lays a solid theoretical and practical foundation for the development of intelligent optical network.

Focus on optical network accurate assessment and the core calculation method
In the optical transport network, the optical signal-to-noise ratio OSNR can accurately reflect the quality of the transmitted signal, and is an important reference indicator for network optimization and upgrade. The traditional OSNR calculation and evaluation method requires many parameters that are difficult to obtain, and the accuracy is not high and the calculation amount is also very large. These methods are difficult to apply in practical optical communication networks. In response to the above problems, Fiberhome established a core technical research team to focus on the accurate evaluation of optical network performance, analyze the deficiencies of traditional methods for optical network OSNR which include nonlinearity and other related difficulties, and determine the optical network performance intelligent evaluation scheme based on deep learning. A series of core technical research work is carried out in a planned manner. In the process of tackling the project, the Fiberhome technology team adheres to the spirit of the optical network artisans, delves into the details of each intelligent monitoring technology, and achieves a comprehensive breakthrough in the basic methods, basic principles, core models and other technical points.

Fiberhome builds an optical network intelligent monitoring deep learning neural network, receives optical network parameters such as optical power and dispersion collected by optical network probes, optimizes and enhances based on reinforcement learning and migration learning, and comprehensively considers the cost impact of nonlinearity. Accurate evaluation and monitoring of the optical signal-to-noise ratio (OSNR) of the transceiver and intermediate nodes. Deep learning technology has achieved great success in the fields of image recognition, speech translation, and chess games. Applications in other fields are still in the process of exploration and research. Beacon communication realizes deep learning technology in the field of optical network, breaking through the core technologies of optical network data collection, knowledge representation, modeling reasoning, intelligent monitoring, etc., and finds a correct path for industrial network artificial intelligence research.

AI enhances optical network intelligence, 5G leads digital intelligence
The optical network has never stopped evolving, and 5G is gradually getting closer. Introducing artificial intelligence technology into optical networks, and letting optical networks have the wisdom of self-configuration, self-optimization, self-healing, and self-evolution, is the goal pursued by FiberHome. The major breakthrough of OSNR intelligent monitoring technology laid the foundation for the introduction of artificial intelligence technology in optical networks, and accumulated experience gained during the technological breakthrough process, further accelerating the extension of artificial intelligence technology to 5G cloud network, and building with intent identification. Deep perception, forward-looking security, and a new generation of communication networks under the control of the company, will drive the smart 5G networks to lead the development of digital society.

Relevant Information