NFV Startup Could Challenge Incumbents

For Carsten Brinkschulte, one of the big problems with today's mobile networks is that they are simply not mobile enough.

Using the NFV-based technology developed by Core Network Dynamics GmbH (CND), the company Brinkschulte leads, an operator could run an entire core mobile network on a Raspberry Pi, a single-board computer about the size of a credit card. "You could put the core network into a vehicle, attach an antenna to it and then have a true mobile network," he says. "Wherever the vehicle goes, you have a mobile network, and if you have a fleet of vehicles connected using a mesh network, you've got resilience."

It is the kind of technology that could turn the telecom industry on its head, with major implications for traditional operators, equipment vendors and end users.

Take public safety networks, for example. A number of governments are now looking to replace their old-fashioned TETRA systems with 4G technology. Instead of investing in new cellular infrastructure costing billions of dollars, authorities could turn police vehicles into miniature mobile networks at a fraction of the cost. "Why not put an antenna into every police car and just a few core base stations in the country and you don't need to deploy 30,000 radio towers?" says Brinkschulte.

CND's technology could also be used to build mobile networks in remote locations, such as at an oil rig. The orthodox approach would be to invest in a lot of specialized hardware from the likes of Alcatel-Lucent (NYSE: ALU) or Ericsson AB (Nasdaq: ERIC), even though a relatively small number of people would be using that network. A much more cost-effective and more flexible alternative, says Brinkschulte, would be to use a standard Linux server running CND's OpenEPC-branded technology. "It does the same thing and will be a lot cheaper -- you won't pay a million for it," he says.

Perhaps the biggest opportunity for CND is the much-hyped Internet of Things (IoT), which will force operators to completely overhaul the way they design networks, according to Brinkschulte. "The network topology has to change for IoT," he says. "If 2 billion devices are connecting to the same centralized core network, it will become smoking hot." A decentralized infrastructure, based on the same principles as in the example of the police cars, would help to distribute the load.

These use cases are among several that CND is targeting as it looks to move its technology out of R&D environments and into live networks. Spun out of Germany's Fraunhofer Institute in 2013, CND currently has about 60 testbed customers, including major Tier 1 service providers like AT&T Inc. (NYSE: T), BT Group plc (NYSE: BT; London: BTA), Deutsche Telekom AG (NYSE: DT), NTT DoCoMo Inc. (NYSE: DCM), Orange (NYSE: FTE), Telefónica and Vodafone Group plc (NYSE: VOD). There is nothing, however, to stop any of those players from using OpenEPC in a commercial setting -- and not just in the scenarios that Brinkschulte presents but in more typical environments, too.

Next page: Operators urged to take NFV risks

Operators urged to take NFV risks
So why is that not already happening, given all the NFV promise? "Operators are traditional companies and relatively conservative," says Brinkschulte, who was appointed CEO of CND last month to lead its expansion drive. "Decisions are still being made and the technology, of course, wasn't there in the past."

Indeed, the immaturity of the standards underpinning NFV has been recognized as one of the barriers to commercial deployment, with concerns about open source protocols and interoperability continuing to hold up progress. During the recent SDN & Openflow World Congress in Dusseldorf, UK incumbent BT publicly complained that OpenStack -- an open source technology seen by many, including CND, as one of the building blocks of NFV networks -- was still not ready to be used with carrier services. (See BT Threatens to Ditch OpenStack.)

Brinkschulte disagrees that OpenStack is not "carrier grade" but acknowledges it is still unproven. And while he has a vested interest in promoting NFV, he issues a stark warning for operators taking an overly cautious approach to NFV investment. "If you want to wait until everything is perfect, you are not going to innovate," he says. "That old thinking is a source of the industry's problems and it is obviously not the way Silicon Valley works."

For more NFV-related coverage and insights, check out our dedicated NFV content channel here on Light Reading.

Government conservatism could also hinder CND's short-term opportunities in the public safety area. The company already appears to have missed the boat in the UK, where authorities recently awarded a contract to EE , the country's biggest mobile operator. "We are not involved at this stage, although we would like to be," says Brinkschulte. "But it's a radical proposal and I don't know if the government would entertain it."

Clearly, one of Brinkschulte's immediate priorities will be to present CND's vision to governments that have yet to make a firm decision about overhauling their TETRA networks. He estimates that about 300 such networks will need replacing over the next decade and that catering to the needs of the military -- besides the police -- could be another opportunity for CND. Encouragingly, however, at least one contract has already materialized: CND is currently building a public safety network based on 4G technology for a European government department whose identity remains confidential.

Even so, while the company has built a profitable business as a supplier of testbed technology, it will need to raise funding from external sources to fully realize its ambitions and grow at the pace Brinkschulte thinks possible. "This company has huge potential and existing technology representing 100 man years of investment, but to scale it ahead of revenues is a classical investment case," he says.

Ultimately, of course, the technology could have a much wider application than some of the use cases Brinkschulte describes, although CND will obviously bump up against the equipment-making giants as it looks to make its vision a reality. Brinkschulte resists suggestions that this type of organization could trigger a major upheaval in the vendor landscape, but he thinks CND's lack of legacy will give the 13-employee company a big advantage when it is competing against the incumbents.

"We don't have a history of building specialized kit and because of that we're where the others want to be in a year or two," he says. "Nokia and Ericsson are going to support NFV and SDN but as a small, fast-moving company we are already there."

Persuading operators to use a supplier that has not previously been on the radar will be another challenge for CND, but NFV will create opportunities for smaller companies that did not previously exist, Brinkschulte thinks. "An operator could run a segment of the network on OpenEPC and the rest using Ericsson kit," he says. "There is a new level playing field coming with NFV and SDN that wasn't there before."

— Iain Morris, , News Editor, Light Reading