Building a Network for the Future: How Service Providers Can Prepare Now

By Jatinder Khurana

While most sectors are grappling with declines in business currently due to COVID-19, according to a report by GSMA, LTE subscribers in India rose around 26 percent year-on-year to around 644 million by June 2020. Now, the increased data traffic on mobile networks resulting from COVID-19 combined with increased tariffs translated into growth in ARPU and revenues1. Clearly, the Indian telecom sector is slated to grow amidst the economic slowdown.

These figures clearly compel us to look towards the fast-approaching future, which is NOW. With staying home being the new norm, many businesses are considering making this a permanent model globally. In India, all schools and learning systems have gone online, and Indians have shifted to OTT content and social media as their primary sources of entertainment. For service providers, this trend clearly implies heavy traffic management and an increase in a number of subscribers in the current data bands and this will grow even further.

To maintain continuous Quality of Experience (QoE) for consumers, it is now time to adapt to the future of networking. Moving from high spend on OPEX and CAPEX, legacy systems which are difficult to scale in a heavy traffic period – factors such as adaptability, scalability, cost, automation, self-management play a key role in defining our future readiness to support new tech such as 5G, SD-WAN and use cases like online gaming competitions, smart homes, and connected vehicles.

Moving beyond legacy architecture and ways of network management, it is time to automate and scale telecom networks. Here is how we at Ciena believe service providers can maximize the exciting opportunities of networking:

Highly scalable and programmable infrastructure

Considering the ever-growing data demand from current and new users, and the speed of innovation at which next-generation technologies are accelerating, service providers need a scalable network without impacting the OPEX, TTM, and TTR2 at access, aggregation and metro networks.

Moreover, rather than waiting for the system to detect errors and the engineer to fix these errors during high traffic and new demands manually, we recommend service providers program their infrastructure (which can be in their own data centers or the cloud) remotely, and in advance. The convenience derived from remote management during unexpected demand and the ability to program multiple devices is the future. The ideal infrastructure will evolve towards an open, flexible, and scalable Adaptive Network.

Data migration to Edge Cloud

OTT content, online gaming, industrial IoT and automation, automotive apps, AR, and VR are just a few of the applications which have made real-time latency and bandwidth speeds impertinent. The storage and compute requirements of such use cases require physical data centers. This implies heavy infrastructure management and scaling these data centers – given the accelerating demand – poses many logistical challenges.

Given the requirements of new-age apps, the easiest, most cost-effective solution is the migration to Edge Cloud, that is, migrating data from physical centers to the cloud ecosystem encompassing commodity-off-the-shelf compute and storage components coupled to highly scalable and programmable networking components.

Fully automated systems for self-diagnosis and self-healing

Often, the underlying problem in the network architecture is recognized when there is an overcapacity or when a system alarm indicates an error, which requires manual checks in the current network ecosystem. Automation for problem recognition and awareness is the answer to these challenges. The need of the hour is AI and ML3 based capabilities which use the information provided by applications and telemetry, allowing the network to continually self-diagnose, self-optimise, and self-heal. Future-proofed network infrastructure is required to support the ongoing evolution of networks.

Disaggregated infrastructure

Instead of utilizing the costly, full-featured IP stacks on devices across the network, here we are talking about simplifying management and support by migrating the functionalities of these devices into an orchestration layer on the cloud. By doing this, we are disaggregating the physical and cloud functions while bringing them together seamlessly. The byproduct of this disaggregation will be - less processing power, less storage, reduction of overall network footprint, massive reduction in real estate, power, cooling, operational and maintenance costs.

Virtual management and control operations

New applications will require computing power to be located at the edge of the network, delivering scale and performance dynamically as needed while significantly increasing the number of deployed IP router nodes. This associated complexity makes intelligent automation a critical network requirement.

A multi-vendor, hybrid network environment can make it easy for any network provider to start their journey from a box-centric legacy IP approach to a simple, automated network design, efficiently supporting legacy services most while preparing for the next wave of applications requirements.

Wherever you are in your network journey, Ciena’s future network expertise can help you get started or help you create the Adaptive Network experience. Read more about Adaptive Network here: https://www.ciena.com/insights/white-papers/Introducing-the-Adaptive-Network-Vision.html

 Khurana is Director, Regional Sales, Ciena India