Leveling 5G Battlefield Through Open Source and Open Spectrum
Every week we see a new announcement about some strategic move by the US government to out innovate China in 5G. Most of those either involve a partnership between the government and large tech or putting some new bans on Huawei. The geopolitical importance of controlling the connectivity stack is understandable: with everything from humans to thermostats integrated together through a cellular wireless network, he who controls 5G, controls the universe.
Yet protectionism isn’t known to spur innovation. Banning Huawei or throwing money at incumbent vendors is unlikely to do the trick. Innovation is driven by the free flow of ideas and free market competition — both somewhat absent in the cellular wireless space, where a handful of players, protected to the teeth with patents, dominate the market.
Almost a decade ago, Mark Collier and Jonathan Bryce approached me and pitched the idea of creating the OpenStack foundation, where Mirantis could become a gold sponsor. Less than 5 minutes into their pitch I told them we were onboard. I didn’t know what the right model to monetize open source infrastructure would be at the time, but I knew that it was going to be a thing — everyone was fatigued by the VMware virtualization monopoly. Shortly after, we authored our first ever Mirantis blog, making an argument for open source innovation in the infrastructure space. Today VMware has gone all in with open source and much of enterprise infrastructure is defined by open source innovation.
A year ago, I was approached by a team at Facebook Connectivity with a proposal: We are building an open source evolved packet core (EPC), would you be interested in helping? For those unfamiliar with EPC, you can think of it as an equivalent of SDN in enterprise networking. I saw the evolution of SDN in the enterprise firsthand and the proposal, while intriguing, gave me a pause. On the one hand, like with enterprise infrastructure ten years ago, cellular wireless industry is fatigued by the oligopoly of a few large players. On the other hand — we don’t see Ericsson going all-in on OpenRAN the way VMware has gone all in on Kubernetes. Furthermore, peaking under the hood of EPC architecture, I was discouraged over how utterly different cellular wireless networking is from anything that I’ve ever seen in the open source IT world. For somebody with enterprise networking background, LTE packet core is alien technology.
I soon learned that cellular wireless is an oligopoly black box primarily due to historic government regulation. A big part of that regulation is FCC’s grasp on wireless spectrum. To run an enterprise Wi-Fi access point or a router — you don’t need a spectrum license from the FCC. To run an LTE base station — you do; and those licenses cost millions. Consequently, the demand side for a cellular network stack has been limited to those who can pay millions for spectrum — mobile network careers (MNOs). And there are only a handful of those per country, which trickles down through all layers of the market and stifles the innovation dynamic.
If you are a startup innovating in enterprise SDN space, your market is thousands of enterprises who are tired of Cisco. You go directly to those customers and you sell a better solution. 90% of those won’t take you seriously right away, but you can get the early adopters and then cross that chasm over to the early majority. Conversely, if you are a startup innovating in cellular wireless — there are no early adopters. Your market is the few cellular wireless carriers who own the licensed spectrum and they are anything but early adopters. Good luck building a company. And without healthy bottoms up competition — good luck America winning in 5G and goodbye any prospects for a successful open source packet core …but something big happened recently…
While we were all busy reading anti-Huawei headlines, in what I believe to be the most under-appreciated regulatory shift to spur cellular wireless innovation, earlier this year the FCC approved commercialization of a large chunk of 5G-ready spectrum using a new, revolutionary spectrum-sharing model. If you haven’t heard about it — Google CBRS. The new model has many exciting things to geek over, but the most impactful ramification is that it destroys the biggest barrier to competition and innovation in the cellular wireless space — access to clean wireless spectrum. The price of an entry ticket to build a cellular network just changed from “millions of dollars and years of FCC paperwork” to “zero dollars and a single API call.”
With API-driven, on-demand licensing of clean, 5G-ready spectrum becoming available to anybody in the US, the floodgates for open source innovation in cellular wireless have finally opened. After years of waiting around, the market for open initiatives like OpenRAN and Open Network Core has expanded from 4 MNOs to thousands of enterprises. Starting 2020, any enterprise can build a Private LTE network at the economics of Wi-Fi, yet superior range and reliability of LTE. Redefining cellular wireless and making 5G happen with open source is the next big frontier. I want to be a part of that movement and, just maybe, 10 years from now we’ll see Ericsson and Huawei going all in on open source 5G.
