Software eats telco: The coming disruption – VentureBeat

The telco industry has a problem. In moving to 4G and soon to 5G, the cost and complexity of constructing and running networks has gotten out of hand. Building out the telco infrastructure with conventional equipment is slow and expensive work, and the amount of human labor required to manage those networks is just no longer practical.

As data is considered to be the new oil, the status quo is also an issue for governments, who are wary of their nations data infrastructures being controlled by foreign players. Just three equipment vendors Huawei, Nokia, and Ericsson hold 80% of market share between them. And their proprietary hardware, architecture, and software systems are closed black boxes. That makes a monolithic triopoly in telco equipment more than just a business liability it is also a national security problem.

However you cut it, telco is ripe for disruption.

So why has there been such a drought in new infrastructure players? Until recently, disrupting telco was a Herculean task. The capital requirements were too high and the road to ROI was too long for investors. But now thats all about to radically change with Open RAN (O-RAN) technology. And a cast of new telco players and fresh startups are poised to jump into the game.

If you think about the telco problem, it sounds quite like what happened with IT infrastructure over the past decade. The industry was dominated by a few monoliths with expensive, proprietary, solutions. But then that changed. We got open source, cloud computing, containerization, non-relational databases, and other tools that eliminated the need for a soup-to-nuts solution from a single, powerful vendor.

Today in IT infrastructure, there is a thriving ecosystem of players large and small as a result of this shift. And the hardware and labor costs of running IT data centers are a fraction of what they are for a telco. So why cant we do the same thing in telecommunications? Why cant we build 5G on an open source software model, following in the footsteps of the transformation of IT?

The answer is, we can, and we must. Currently, each telco antenna tower requires multiple hardware units (the black boxes) to connect to devices, access user and system information, and process data. With O-RAN, much of that functionality is virtualized and containerized by software at a core mobile communications facility. That eliminates the proprietary, vertically-integrated nature of the system and allows network operators to customize their own services using any number of vendors, instead of being beholden to the big three suppliers.

Although it isnt talked about that much yet, telco is already being radically reshaped in some global markets. Rakuten Mobile, a renegade telco entrant in Japan, is building 4,000 mobile edge nodes for its 5G network. The system will run on commodity X86 computers and standard data accelerators that anyone can buy.

Then there is Reliance Jio Infocomm. In just three years, Jio has become the largest mobile network operator in India, serving 600 million users. How did they do it? When Jio announced its intention to build a 4G network using more open source components, incumbent players dropped their equipment prices to stay in the game. The mere threat of virtualization was enough to create strong competitive cost advantages.

Its not just the benefits of reduced capital investment in hardware that could benefit new telco players, though. The shift from hardware to software will drastically reduce operating expenses as well. Cloud and edge technologies used by Google, Amazon, and Microsoft like OpenStack, Kubernetes, QEMU, and Akraino will provide the basic network management stack. Virtualization, automation, and containerization solutions from companies like VMWare, Red Hat and emerging startups like Portworx, Rancher Labs, and Robin.io will replace the black boxes with full network transparency. (Full disclosure: My firm is an investor in Robin.io.) AI will reduce operational costs as it does in IT. Consequently, operators will be able to run their networks at one tenth of the current labor cost.

The United States is now catching on. In January 2020, a $1 billion bi-partisan bill was introduced in the Senate, proposing to help US companies develop 5G alternatives to the big three. Of note, it offers financial incentives specifically to develop O-RAN technology. Between this bill and the recent announcement by the FCC Chairman to free up mid-band spectrum for 5G, the FCC seems to be doing its part right.

That makes a lot of sense for the US economy. Moving to an O-RAN paradigm would not only eliminate the worries of monolithic foreign vendors hoarding data under opaque circumstances, it would also stimulate growth. All those X86 boxes? Thats more revenue for Intel and AMD. The accelerators? Theyll come from NVidia, Xilinx, and others. Meanwhile, Telco suppliers like Altiostar, Parallel Wireless, Mavenir, and Affirmed Networks will have an opportunity to grow into major global players. (Full disclosure, I am an investor in Parallel Wireless.)

FCC chairman Ajit Pai seems to support this vision. At CES in January, he said There is no America based supplier of equipment as we currently conceive it. But people are innovating here in the United States and in other parts of the world to virtualize the radio access networks. Using the software layer to address not just the security elements, but also the cost element, I think is a win for everybody.

It is. Telcos should jump on the O-RAN opportunity now. If they dont, they might see themselves leapfrogged. Dish Network, for instance, flush with spectrum from the T-Mobile/Sprint merger agreement, could disrupt the space with a much more cost-effective network. (In fact, DISH has already committed to the mentality of virtualization espoused by O-RAN).

The drought in greenfield telco players is coming to an end, and with it, over-reliance on a foreign equipment vendor triopoly. Open source computing will soon transform telco the way that it did to IT. And for both greenfield and brownfield players who embrace the technology now, O-RAN will prove fertile ground in which to prosper.

Rajeev Madhavan is Founder and General Partner of Clear Ventures.

Continued here:
Software eats telco: The coming disruption - VentureBeat