Verizon and Cisco completed a lab trial of Information-Centric Networking (ICN) technology that uses “named data” rather than location identifiers like IP addresses to support streaming video content. The trial also looked at ways to increase usage of multi-access edge computing (MEC) architectures.

The companies demonstrated the ICN software over standard TCP/IP platforms in a Verizon lab located in Waltham, Massachusetts. The demo showed dynamic adaptive streaming and load balancing of video media. This can support different caching strategies to help manage traffic on the front end, and preserve backhaul and core capacity on the back end.

Giovanna Carofiglio, distinguished engineer at Cisco, explained the technology basically uses a content-specific naming method instead of an IP address. This allows for a more efficient handoff of streaming content as a user moves between either different cell sites or different network access methods like 5G or WiFi.

She said that the name designation is application agnostic and allows for seamless transport regardless of access layer. It also does not change the IP address to the data itself.

“This rethinks the network and transport layer, and allows for multipath capabilities across something like 5G and WiFi,” Carofiglio said. “The goal is decoupling the content-based usage on the Internet from the underlying network connectivity.”

Anil Guntupalli, executive director for technology architecture and planning at Verizon, said the trial was mostly focused on management of streaming video content. He said the carrier expects video traffic to account for up to 80 percent of network traffic.

“We’ve been looking for ways to optimize our core network that makes sense and doing caching in the network at points where it makes sense,” Guntupalli said.

He said that the industry in the past has relied on content delivery networks (CDNs) where everything was centralized at a location point. “But, now we are looking to see if there are better ways to optimize rather than using a purpose-built CDN and layering on top.”

Multicasting is typically done by streaming video content from one source out to many devices. This allows for a more efficient use of network resources compared with a unicast model where each device gets its own content stream.

An ICN deployment would allow for the streaming of content from a specific MEC location that would further diversify the network resources required and allow for a better user experience.

Guntupalli said the trial results showed that in some instances ICN was better than multicasting, though noted it was too early to pinpoint exact numbers. However, it did show enough promise that it could replace current multicast technologies at some point.

“There are some areas where we would need to keep multicasting, but this can be a benefit in a lot of areas,” Guntupalli said.

The companies also demonstrated the ability for the Hybrid-ICN variation to simplify the insertion of ICN technology in existing IP infrastructure and support coexistence with legacy IP traffic. Hybrid-ICN supports IPv4 or IPv6 compliant packet formats.

Carofiglio said that the hybrid example is likely to be the most widely used as it requires the least amount of modification for operators. However, she explained that the non-hybrid format is “not as heavy” and would be better for edge deployments.

“Our view is that you would deploy hybrid unless in some use cases where latency would be most critical and you would want to have the lighter ICN version,” Carofiglio said.

Cisco’s ICN work builds on technology it acquired from Xerox’s PARC division last February. PARC had been working on the technology for a decade under the content-centric network label.

As part of that acquisition, Cisco also created an open source project within the FD.io community of the Linux Foundation. Carofiglio said that work was continuing as a way to further bolster the community.

“Standardization activity is not required, but it’s something we are doing because it’s important with the work we do with our customers,” Carofiglio said.

For Verizon, Guntupalli said the carrier would like to see more work in terms of scaling the technology to a critical mass of more content providers.

“It’s very critical to get the naming portion right and we want to open this up so other content providers can easily provide their content to different providers,” Guntupalli said.