Operators know it all too well: streaming video, cloud computing, the Internet of Things and the evolution to 5G place massive pressure on today's networks, requiring capacity increases by orders of magnitude and the ability to respond to even greater unpredictability in traffic patterns. The optical network sits at the heart of communications, connecting people, data centers and an increasing number of devices across any distance, from next door to an ocean away.
With both the cloud and the IoT coming to dominate the enterprise data environment, the need to push connectivity across greater distances becomes paramount.
Recently, numerous platforms and service offerings have that aim to forge tighter links not only between remote data centers, but between individual server and storage components within those data centers. Ultimately, the aim is to produce a single federated ecosystem that spans local, co-located and cloud-based infrastructure, all defined on an abstract, virtual layer to achieve limitless flexibility and scalability.
This requirement for high-bandwidth services comes particularly from cloud service providers, who seek both higher bandwidth networking and software defined network-enabled connectivity solutions. Data center-based traffic is expected to nearly double by the end of the decade, growing to 20.6 zettabytes in 2021 from 11.6 zettabytes in 2017, with the cloud accounting for 95% of this traffic, according to Cisco's Global Cloud Index: Forecast and Methodology, 2016-2021.
The significant escalation in enterprise traffic is manifest in enterprises' migration of IT to public data centers and cloud services, driving demand for cloud connect services and increasing Ethernet service bandwidth by close to 30% annually, Coriant found.
Advances in Ethernet switches drives the data center interconnect (DCI) market, reports research firm Ovum. The 100Gbit/s equivalent DCI market will enjoy a compound annual growth rate of 56% between 2017 and 2021, driving the need for low cost 100Gbit/s over single lambda (wavelength) solutions, Ovum found. At the same time the industry will start adopting higher speed 400Gbit/s links also using single lambda technology. As 400G switches enter the market, optical transceivers must keep pace and rapidly transition to 400Gbit/s.
Platforms built specifically for high capacity interconnects address the need for incredible speed and massive density. These platforms deliver up to 400Gbit/s per wavelength "with a server-like operational model, giving operators a vital new tool for increasing capacity" says Kent Jordan, advisor of Technology Marketing at Ciena, in a company blog. "Enterprise, government and research and education (R&E) customers are applying these platforms and leveraging them in capacity-exhausted connections all across the network, increasing bandwidth across congested links in hours while saving tens of thousands of dollars."
Not surprisingly, 5G will be a huge driver for the next round of innovation in optical transport over the next two to four years.
5G will require 10- to 1000-times the bandwidth used for LTE today in a given service area so providers will no longer be able to dedicate and overprovision capacity, says an Ekinops spokesman. Because 5G networks will be more distributed and must support different applications in diverse network segments, optical networks must provide flexible, on-demand bandwidth that operators can quickly spin up and down to meet service level agreements for uptime and quality of service, he noted.
Ahead of NGON & DCI Europe 2018, we present a series of three reports diving deeper into the state of the next generation optical networking market as technologies and deployments ramp up from 100Gbit/s to 400Gbit/s and beyond.
— Adrian Pennington is a journalist and editor specializing in the creation, business and technology of moving image media. Published in The Guardian, The Financial Times and The Hollywood Reporter, he also copy writes a range of marketing materials for brands and marketing agencies. Follow Adrian on Twitter at @pennington1; on LinkedIn at Adrian Pennington.
Over the next two years, approximately 60% of service providers (both large and small) will adopt virtualization on a wide scale across their networks, according to the latest survey report from Ovum. Why are providers making these moves? Is there an easy way to start?
Learn how and why service providers are using virtualization to transform their networks. This webinar will look at how providers are leveraging virtualization to create more flexible and agile networks while also providing a better customer experience. Expert speakers from netElastic and Heavy Reading will address the industry drivers for network virtualization, the benefits that can be realized, the challenges to face and the results of virtualization being achieved by providers today.
Key topics will include:
Current network infrastructure and the move to virtualization
Benefits and challenges of network virtualization
How providers can get started
Service provider success stories: the decision to virtualize, the solution, and results