Machine learning helps extract hidden network insights
Machine learning refers to the ability for machines to learn without being explicitly programmed. It focuses on recognition of data patterns or trends, enabling software programs to take actions based on them. It also involves processes like data classification and clustering, dealing with the complexity of disparate data sources and myriad variables. Unlike traditional analysis, machine learning thrives on growing datasets. The more data fed into a machine learning system—assuming it’s ‘good’ data—the more it can learn and apply the results to higher quality insights. Machine learning applied to networking involves big data analytics and the following data-driven insights:
- Descriptive: Interprets historical data to determine “what has happened”
- Diagnostic: Determines why something has happened using techniques such as data mining, discovery, and correlation for a root-cause analysis
- Predictive: Finds results that predict “what will happen” in the future based on historical patterns that are sometimes combined with external data
- Prescriptive: Predicts multiple outcomes for a given scenario based upon actions that are taken. The idea is to show how a different set of actions will affect the situation and steering the user toward the best possible option
For network operators, machine learning and big data analytics can address a wide range of business and operational initiatives. For example, predictive analytics based on machine learning can enhance network assurance. Analytics can learn from historical performance metrics and trouble ticketing data and then combine that knowledge with real-time network telemetry to anticipate future network failures. Analytics can also be used to optimize network security, traffic management, and capacity planning.
Machine learning benefits from Blue Planet Analytics’ open design and micro-services-based architecture
Ciena’s Blue Planet Analytics (BPA) provides advanced analytics capabilities within the Blue Planet software suite. BPA provides an open framework with libraries for cluster computing, data processing, and machine learning. BPA leverages Blue Planet’s robust Resource Adapter (RA) framework to allow for quick and easy integration of multiple data sources, whether physical or virtual. Integration with cluster-based storage systems allows for more scalable datasets and more efficient machine learning. And, Blue Planet’s micro-services architecture provides modularity and scale to enable rapid adoption of machine learning and data-driven automation.