The Wider View

Investing in Digital America

September 24, 2021

In the coming months, the U.S. federal government will likely pass an infrastructure bill. The total amount is in flux — $1, $2, $3.5 or $4.5 trillion. The following is a short list of what seems to be some of the items common across the various iterations of the legislation:

  • More than $110 billion to rebuild, repair, and modernize roads and bridges: This could include money for fiber, cell sites (small and maybe macro), C-V2X roadside units, IoT sensors, cameras, etc.
  • $66 billion to improve Amtrak and modernizing public transit: Again, more fiber along the tracks, cellular/wireless capabilities, sensors, cameras, etc.
  • More than $7 billion for electric vehicles and charging stations. Each of those could provide cellular service across multiple bands and would be good spots for new fiber. Also, IoT sensors could also be installed.
  • More than $25 billion to upgrade airports; some of that money could go to fiber and DAS, small cells, edge compute, V2X, IoT, etc.
  • $100 billion to upgrade and build public schools: The pandemic is showing how school district are using CBRS-based FWA to provide broadband service to their communities. Schools are ideal “hubs” for fiber, cellular, etc. And while they are at, why not introduce courses that take students through the sites, so they understand how the Internet is getting to their devices? Maybe some will even want to learn how to install and/or fix those sites – or get involved in related STEM programs.
  • $55 billion to modernize drinking water and waste processing. Again, fiber, sensors, cameras, etc.
  • $48 billion to develop the American workforce – match older workers with students. They can learn from each other about how to build, fix and otherwise digitally transform America.

Don’t hold me to the numbers cited; they may change, just as the items themselves might.

My point is simply that forthcoming infrastructure investments could include – should include – digital components. As an example, most of the cost of deploying fiber is in the physical process and labor of associated with trenching and/or boring and then installing the conduit and then pulling the fiber through. The cost of making that trench a little deeper and/or the conduit a little bigger and pulling more fiber through, while not insignificant, is minor compared to the expense associated with going back in a few years’ time. Dig once, benefit for years.

V2X – A new driver of digital infrastructure investment

June 8, 2021

The idea of connecting vehicles to each other (V2V) and to other things (V2X) has been around for a while – in fact, this was one of the first IOT use cases way-back-when.  Since then, little appears to have happened – while many cars have modems for 4G LTE/WiFi access and apps on the screen, etc, vehicles still do not connect to each other or to roadside furniture, signs, control signals etc.

But there are now signs that things are starting to change and that V2X could be the big driver of digital infrastructure investment – after all, if vehicles are going to communicate with assets at the side of the road, those assets are going to have to be connected.  That means fiber, edge compute, radios and towers to connect signs, stop lights, bridges, roads, etc to data centers and applications.  In short, V2X means a whole range of infrastructure at the side of the road, as well as new communications capabilities in the vehicle itself.

The term “connected vehicle” is used to describe vehicles that can “speak” to each other, along with roadside infrastructure and other devices (smartphones). There are – or were – two competing connected vehicle technologies. Dedicated Short Range Communications (DSRC) is the older of the two as it’s been around for more than 20 years. The relative newcomer is Cellular Vehicle to Everything (C-V2X) technology.

Both technologies enable basically the same use cases some of which include:

  • Basic safety, signal phase and timing, and information messages
  • Forward collision warning, pre-crash sensing, hard braking warning, emergency vehicle warnings
  • Traffic jam and route information.

The messages are sent or relayed between/among on-board units (OBUs) installed in vehicles and they can be augmented by interaction with roadside units (RSUs).

Note that smartphone technology and various applications already supply some of these use cases, while cameras, ultrasonics, LIDAR, etc., provide some of the vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) functionality without needing any type of wireless networking.

DSRC is also known as 802.11p. It is a standardized, wireless communications technology that allows vehicles to communicate with each other via OBUs RSUs. In the U.S., DSRC equipment operates in the 5.850 to 5.925 GHz band (5.9 GHz band). It is not a cellular technology.

C-V2X is a cellular technology that is defined by the 3GPP in Release 14 of its specifications. C-V2X also operates in the 5.9 GHz band. The 5G Automotive Association (5GAA) does a good job of summarizing what C-V2X is here (

Both C-V2X and DSRC require OBUs and RSUs and both can be made to interact with the cellular network. Note that those RSUs are basically small cell sites as they will have power, backhaul (fiber) and proximity to vehicles and therefore end users. They will also need edge compute.

DSRC has its own ecosystem of devices and while C-V2X does, as well, those devices and their components are part of the wider cellular ecosystem. And while C-V2X is LTE today, it will eventually migrate to 5G New Radio (NR) just like everything else.

In November 2020, the FCC released a report and order that did two major things:

  • Split the 5.9 GHz band into two pieces: the lower 45 MHz will be for unlicensed use (Wi-Fi) while the upper 30 MHz will be dedicated to C-V2X.
  • Current DSRC devices must migrate to the upper 30 MHz by year-end 2021 and that technology will itself be phased out.


The FCC’s “modernization” of the 5.9 GHz band in that November 2020 report and order remains contentious and could well be over-ruled by the current administration.

There appear to be two warring camps. One side says the entire band should be dedicated to CV technology while the other side says that the upper 30 MHz is sufficient.  Regardless of what happens to how the band is allocated, the C-V2X standard is not the issue – the stake the FCC pounded into DSRC’s heart may well remain untouched and the world will move forward with C-V2X, and all of the associated digital infrastructure needed to enable it.

If Prime: 5G Core by Thursday, RAN by Saturday

April 22, 2021

Remember the Rokr E1? It debuted about 16 years ago and resulted from a partnership between Motorola and Apple. Back then, Motorola was a big deal in phones – along with Palm, RIM (which is BlackBerry, btw) and Nokia. The Rokr E1 failed, largely due to the lack of letters.

I remember thinking it odd that Motorola and Apple partnered particularly since the Rokr E1 was so bad. Thanks to hindsight, I must also have thought “Maybe Apple’s learning what goes into making the cellular components of a phone.” Eighteen months later, Apple disrupted the cellular and computing industries.

Which brings me to this news. A month ago, Nokia and Amazon Web Services (AWS) announced a collaboration in which “engineering teams from both companies will research how the combination of Nokia’s RAN (Radio Access Network), Open RAN, Cloud RAN and edge solutions can operate seamlessly with AWS Outposts.”

Per the press release, the project will run Nokia’s 5G virtualized distributed unit (vDU) and 5G virtualized centralized unit (vCU) on AWS Outposts using Amazon Elastic Kubernetes Service (EKS) for far edge cloud or on-premises deployments. (Kubernetes is a container management platform.)

In 5G New Radio (NR), baseband processing is split between the DU and the CU. The goal there is to put those processing elements in physical locations where it makes the most sense (costs the least and/or provides the greatest functionality). This functionality is likely to be virtualized as well, but what the radios do is not changing that much.

The third part of the collaboration will build a “proof of concept for an end-to-end solution with Nokia’s 5G Cloud RAN and 5G standalone Core network running on AWS, where end enterprise users can leverage 5G for use cases such as an industrial application.”

Undeniably, Nokia has thirty plus years of cellular radio expertise which I would guess must transfer into its virtualization efforts. And it has a large existing market that is under threat thanks in no small part to the very mobile operators it sells to.

Ostensibly, a 5G RAN service (CU/DU) is another way for AWS to position itself as an enabler of 5G services. Already, they have outposts, wavelength zones, local zones, snow-family devices, a partnership with Athonet for 4G/5G core, and one with Federated Wireless for connectivity-as-a-service (which Federated does not directly have since they only run a SAS).

Will AWS launch 5G NR RAN-based service? Who knows. All I’m suggesting is that AWS with all the resources at its disposal, is more than capable of disrupting the cellular market. Yesterday’s news is a good example – Dish is using AWS to build its 5G network.