Big data analytics are an extremely important tool for web-scale networking, providing insight on network traffic, resources and health. Service providers and network operators are tasked with managing ever-growing networks comprised of cloud-based applications and services, mobile traffic and virtualized and SDN components. Automation is becoming essential for monitoring and managing traffic in real time, and programmable automation required solid data analytics along with AI.
There are a number of steps network operators can take to integrate analytics into their operating and management strategies. According to optical networking company Ciena, operators need to understand how analytics work in web-scale networking as well as develop a management strategy for use in order to glean the most value and facilitate network intelligence and automation.
One of the key ingredients for successful implementation of analytics relies on network operators working together to create visibility and interoperability among networks. That means agreement on a common set of information and data models as well as APIs, enabling network orchestrators in different domains to communicate effectively. This challenging but critical step is absolutely necessary for automated service provisioning, configuration and quality assurance from end to end.
The role of analytics also needs to be better defined, as they have two key uses. The first is for building closed control loops that autonomously configure, assure and optimize services by predicting problems. And the second use is to analyze huge amounts of network and customer data with the goal of offering personalized services and better meeting customer demands and expectations. Operators should also come up with a big data strategy, aggregating data from disparate sources into a centralized platform for analysis.
Other important components for successfully applying analytics include developing a common event data model. Ciena points to AT&T (News - Alert) as an early contributor with its OPNFV code for a virtualized event streaming framework. Network operators should also have a clear understanding of the business benefits of analytics, including saving on OpEx and CapEx as well as increasing revenues. They may also use analytics to experiment with machine learning, including proactively identifying network events that may impact performance. Machine learning algorithms may also be used to dynamically adjust capacity and network response to security threats, and can also make VNFs self healing and able to restart, move or rebuild automatically. And finally, network operators need to begin experimenting with 5G deployments to take advantage of their many benefits.
Ciena is aiming to help network operators better understand and take advantage of analytics and AI through its Blue Planet Analytics offering. The solution assimilates data from throughout the network to enable intelligent automation and operations while offering a framework for collecting and normalizing data. The solution supports data collection from any source, including multiple network vendors, and is built on an open design and micro-services-based architecture.
Edited by Maurice Nagle