The fundamentals of network security are being redefined. Driving this process are the changes inherent in the adoption of the cloud, open-source software, distributed low-latency reference architectures, and evolution of the enterprise service delivery model.

Looking beyond the hype around AI, operators see value in adopting embedded AI technologies including machine learning in customer care, network operations, policy and service assurance, etc. Goals include applying machine intelligence to drive predictive analytics, automate processes, cut costs, and improve customer experience. We look at the leading operator use cases of AI technologies and its potential to disrupt.

CSPs will ultimately face one stark reality: Automate or die. Rather than focus on virtualization and automation as solely a cost and efficiency strategy, operators need to understand that automated networks are an essential requirement to service the growing automation (software, robotics, drones, IoT, etc.) of industrial, enterprise, healthcare and government customers.

CSPs are moving to telco cloud networks for economies of scale and to run programmable networks on standard platforms in a data center. While virtualizing network functions was an important first step, few VNFs or its associated management software are cloud ready, let alone cloud-native. Without a plan to decompose into flexible micro services that can be orchestrated, scaled out and rolled back, it will be difficult for CSPs to get the full commercial benefits of cloud networks.

Intent is a new abstraction layer within the network management hierarchy. Interest in the intent paradigm has resurfaced as operators struggle to get to grips with the orchestration of NFV. But for intent to be useful to service providers it needs to be standardized so that every vendor’s network functions can be managed using the same intent abstractions.

Optical networks are in a perpetual state of transition as the ever-changing requirements of applications and services demand innovation in the underlying optical networks that support them. Conversely, optical networking advances enable next-gen applications and services. Looking forward, the optical industry has a three-fold mandate: support data center interconnection across the metro, long-haul, and submarine networks; build a fronthaul/midhaul/backhaul infrastructure for coming 5G; and adopt SDN-based automation to rapidly scale while dramatically lowering the costs of transport.

The delivery of services as mobile edge computing, enterprise cloud, IoT, big data and 5G depends on a highly reliable, distributed and software-defined data center infrastructure. This infrastructure includes software-defined connectivity and virtualized network functionality as well as new compute and storage architectures and a management and orchestration stack that can support the increasingly broad range of telco service requirements.

The transition to 4K UHD, the increase in streaming, and the growth of social video is driving an explosion in video volumes. And that’s not including the growth of Immersive Media Experiences (IME) such as virtual reality (VR), 360-degree video and Augmented Reality (AR). With four out of every five bits of network traffic expected to be video within five years, operators need to develop effective technological solutions to manage these experiences, and viable business strategies to monetize them.

SD-WANs are the perfect example of the benefits of SDN-based programmability and automation applied to the enterprise WAN. Industry adoption has been rapid as traditional telecom network operators, competitive service providers, managed service providers and, most recently, cable operators have all seized on this promising new managed services opportunity. With the initial steps taken, service providers must now tackle the challenge of wide-scale rollouts that boost top line and bottom line revenues.