Why 5G Wireless Will Leverage High-Speed Optical Networks and Machine Learning
At this year’s Optical Fiber Conference (OFC) in San Diego, CA, we saw how advances in cognitive Software-Defined Networking (SDN) and fiber-optic electronics are enabling the rollout of 5G mobile wireless, which in turn will deliver new services such as augmented reality, drones, industrial IoT, self-driving vehicles and massive connectivity.
OFC is an annual conference sponsored by two IEEE societies: the IEEE Communications Society and the IEEE Photonics Society, as well as the Optical Society. This year’s event brought together 15000 attendees and nearly 700 exhibitors, who showcased the latest products in the growing optical-network infrastructure market, along with SDN solutions.
This year’s show highlighted the convergence of electronic, optical and wireless technologies that will support the growth in connectivity and network traffic over the coming decade.
New services like self-driving vehicles demand higher capacity in the fronthaul, backhaul and metropolitan optical networks that connect 5G wireless access points with data centers. Alex Jinsung Choi, a senior vice president at Deutsche Telekom, explained at OFC that “5G network operators must deliver 100Gb/s connectivity to the base station to support the exponential growth in global internet traffic, which is expected to reach 5 Zetabytes per year by 2022.”
5G is also driving the demand for 400Gb/s optical connectivity in hyperscale cloud data centers. Benny Mikkelsen, Chief Technology Officer at Acacia Communications, explained that “electro-photonic integration will be the key to achieving that target, while at the same time, reducing the size, power and cost of optical transceivers.”
Decreasing the power consumption in 5G networks is a key challenge, too. “By 2030, 20% of all global electricity is expected to be consumed by information technology, particularly in data centers and wired access,” said Mikkelsen during his keynote.
Cognitive SDNs are primed to deliver low-latency performance in data centers and 5G base stations by leveraging machine learning algorithms to predict and balance network traffic patterns. “We need more robots in the network to make it more efficient,” said Vinay Rathore, Senior Director of Marketing at Infinera.
“5G offers many monitoring points in the network,” explained IEEE Fellow George Rouskas, Professor of Computer Science at North Carolina State University, during an IEEE Future Networks session at OFC 2019. Together with 5G wireless access, new technologies such as IP-over-DWDM, which integrates IP routers and switches in the optical transfer network, would also help enable cognitive SDN.
Internet video providers such as Netflix are looking to take full advantage of advances in cognitive SDNs, 400Gb/s fiber-optic electronics and 5G. “Our goal is to support all customers on all devices, over 3G, 4G, 5G, regardless of infrastructure,” says David Temkin, Vice President of Networks at Netflix. “Advances in optical technology drive down cost and improve our customer reach.” Netflix operates one of the world’s largest content delivery networks, with over 11000 servers in production, and provides online video content to 150 million customers in over 100 countries.
China is also driving the optical networking market, with plans to roll out a 1Tb/s Data Center Interconnect (DCI) and Fiber-To-The-Home (FTTH) by 2025.
With 400Gb/s coherent (400-ZR) transceivers expected to hit the cloud data center market by 2020, we are likely to see highlights of successful 5G rollouts at OFC next year.
Originally published by Mario Milicevic at IEEE Transmitter
This article is from and published after authorization. It does not represent Gigalight Community's position. Before reproduced, please contact the original author.