The evolution of optical network technology continues to move toward ever greater infrastructure disaggregation, and for good reason. Open optical networks allow communications service providers (CSPs) and data center operators to mix and match components from different vendors, avoiding proprietary vendor lock-in to speed innovation, increase agility and reduce total cost of ownership (TCO). Yet, open optical line systems (OLS) can introduce a certain degree of complexity.
The challenge is that most currently available network planning and design tools were designed for proprietary ROADM and optical transponder specifications. That means these tools are unable to accurately estimate the performance of third-party transponders over an open line system, requiring manual field measurements and calculations from in-service multivendor architecture deployments — a task that is highly difficult, if not impossible.
This lack of modern design tools presents considerable operational challenges, complicating and hampering network planning, design and deployment. Furthermore, the complexity of multivendor planning often results in static optical network models and costly stranded capacity that goes unused, significantly impacting performance throughout the network lifecycle.
Nurture intelligent twins
As state-of-the-art artificial intelligence (AI) and machine learning (ML) algorithms advance, these technologies provide an ideal solution to address the inherent challenges of multivendor network planning and design. To harness this powerful intelligence, Fujitsu developed Virtuora® OLS-Designer, which generates ML-powered digital twin models from real-time optical network data to optimize multivendor network planning, design and testing.
With the ability to generate digital twin models, the OLS-Designer microapplication uses real-time ROADM topology, optical routes, transponder configurations and pre-forward error correction (pre-FEC) bit error rate (BER) to validate transponder reach and interoperability. This ability enables more accurate optical signal performance predictions with transponder reachability assessments, optimal modulation format and data rate recommendations.
The OLS-Designer intelligence allows for enhanced alien wavelength planning without the need for proprietary ROADM information, eliminating the need for costly field calculations and simulation studies. This capability allows the highest possible capacity modulation format when new transponders are deployed, ultimately maximizing network capacity per link and wavelength.
Learn from real-time insights
Leveraging AI/ ML intelligence, the Virtuora OLS-Designer also allows OLS network designs to be continuously trained for improved planning accuracy. With the power of supervised machine learning, this intelligent microapplication suggests optimum modulation format and data rates, generating valuable insights that enable peak network performance, scalability and the highest possible capacity.
As a result, CSPs can simplify and streamline OLS multivendor network planning and deployment, reducing operational costs and maximizing total network capacity while speeding service delivery.
