Spectrum policy: The missing voice of enterprise
Businesses have many wireless & spectrum needs, but a fragmented voice
I’ve been to a number of events about radio spectrum recently, including the Dynamic Spectrum Alliance (DSA) summit in Geneva, and the Spectrum Americas conference in Washington DC. These have generally had a huge focus on spectrum innovation, covering dynamic sharing, new unlicensed bands, as well as representatives from government and the mobile, satellite and Wi-Fi industries.
However, the business sector has been under-represented and under-appreciated – something that is also quite common in regulatory proposals from telecom-affiliated spectrum specialists, especially from the cellular industry.
Much of the debate at these events concerned the goals and objectives for 5G / 6G networks, as well as the need (or lack of it) for new exclusive spectrum bands for high-power use by public carrier networks. There has been plenty of back-and-forth over the requirements for mobile broadband, fixed wireless access, unlicensed use indoors – and their coexistence with the shifting needs of satellite connectivity and incumbents such as defence and audiovisual sectors, which also have their own evolution paths and ambitions.
But while there has been no shortage of speakers and panellists from telcos, the cellular and Wi-Fi vendor communities, regulators and industry associations – plus a presence from military and utility sectors – there is an important missing stakeholder group.
There is no effective “spectrum voice for enterprise” – either large businesses or SMEs. In theory, there should be some alliance of companies like Boeing, John Deere, Walmart, Tesco, Marriott, Coca Cola, Shell and Johnson & Johnson, that takes a collective stance on licensed, unlicensed and shared spectrum.
They are at the forefront of using different wireless connectivity, sensing and communications mechanisms. They use:
As a result, they collectively see the need for, and relative benefits of, different spectrum regimes, and complex landscapes of service providers and vendors. One size does not fit all. Optionality and fitness-for-purpose are the key. In some cases, they want to purchase services, and in others they prefer to own connectivity assets, with direct access to spectrum to meet the needs of their applications. They work with a variety of integrators, VARs and specialist service providers. They also have some unique views on the importance of network reliability, especially for business-critical and safety-critical uses – and the importance of long-term stability and predictability in regulation, in order to make investments.
Who speaks for enterprises as a group?
But today, they are only represented indirectly. Some participate in technology-specific forums such as OnGo Alliance for CBRS, Europe’s EUWENA for private wireless, Wi-Fi Alliance for Wi-Fi, 5G-ACIA for private 5G, and the Bluetooth SIG. They work with the traditional telecom community through GSMA and CTIA for public connectivity.
A few industries have representative bodies, such as UBBA and EUTC for utilities in the US and Europe. There is also some regional representation through groups like the US Enterprise Wireless Alliance (focused on narrowband private radio) and the UK’s Spectrum Policy Forum.
But there is no coordinated “Enterprise Spectrum Advocacy” group, either to give the collective voice of business users at conferences, or to respond to initiatives such as FCC, NTIA, Ofcom or EU consultations and national spectrum strategies.
Existing technology-specific industry associations are mostly unwilling to discuss multi-technology systems, although Spectrum for the Future (SFTF), DSA[1] and the Wireless Broadband Alliance at least cover aspects of both convergence and the need for various different spectrum models.
Case study: proposed changes to US CBRS
A good example of the need for a coordinated voice for enterprise spectrum users is the recent AT&T suggestion[2] in the US, to completely reorganize the 3GHz band. This essentially involves moving and clearing the whole CBRS band between 3.55-3.7GHz and relocating it to a new shared band in the lower-3GHz range.
In addition to the widely discussed difficulty of coordination with defence-sector incumbent users between 3.1-3.45GHz, such a plan would be hideously disruptive to businesses now adopting CBRS private networks at scale, as well as the existing SAS setup and the ecosystem of equipment and device suppliers. It would presumably also mean moving Navy radars to different bands or expanding coastal protection zones to prevent interference.
Consider a factory that is implementing a CBRS private 5G system for its robots, cameras and industrial automation systems, plus human workers and visitors with handhelds and smartphones. It will have gone through a lengthy and expensive process of planning and designing its network, selecting and integrating applications and devices, and then the disruption of installing the equipment and commissioning the system. It may have linked it to existing Wi-Fi in the factory and relied upon CBRS modules built-into a variety of other systems such as industrial robots, gateways and guided vehicles.
Ripping out all of this, sourcing new equipment and devices (where the industry would have to support a new and unusual band, at the same prices), and learning the complexities of accessing a new SAS type database would all be horribly expensive. Worse, any downtime for the factory’s production would be unlikely to be covered by any “incentive” compensation scheme related to the spectrum repackaging.
Recommended by LinkedIn
The fact that AT&T has seriously proposed such a disruptive concept points to either a complete lack of understanding of the importance of wireless systems to enterprises – or a deliberate attempt to undermine confidence in CBRS and shared spectrum, just as acceleration is occurring in adoption, innovation and iteration of the use of that band. Perhaps it is hoping that much-delayed and overhyped concepts such as 5G slicing will make its own network more competitive compared to dedicated private 5G for such businesses.
Other examples
Numerous other global spectrum-related policy initiatives and proposals would also benefit from a unified voice – or even just a discussion forum – for enterprises, in a similar way that GSMA speaks collectively for mobile operators, or GSOA for satellite.
Some of the domains that could benefit from this type of interest group could be:
There is also a need for such as group to "sanity check" and contextualise the assertions of their peers. When the inevitable forecasts of potential "GDP uplift" appear for #6G and assert the value for improved B2B connectivity as a basis for spectrum needs, there would be ample scope for enterprise spectrum advocates to chime in with "yes, but only if we get to license spectrum directly and build our own private networks".
Conclusion – a voice for enterprise spectrum users is needed
Overall, it seems to be hugely important for businesses to collectively stand up for the importance of a good mix of licensed, shared and unlicensed spectrum, both at national and international levels.
Otherwise, their voices are fragmented and lost in the noise of well-organized policy megaphones, such as that of the mobile industry. A collective messenger should be present at both spectrum-specific events, and as contributors to policy debates and public consultations and calls for comment.
Such a group would likely extol the need for:
Potentially, this type of advocacy could take place within other groups such as SFTF and DSA as an enterprise membership / supporter tier, but it would be important to assure its ability to be “neutral” (or at least “wanting a bit of everything”) in terms of licensed vs. unlicensed vs. shared spectrum options.
Or perhaps there needs to be an Enterprise Spectrum Users Association - ESUA - with its own lobbying platform, working groups and messaging?
Either way, corporate spectrum users deserve to take a seat at the table – and on conference panels, and in regulatory consultations.
#spectrum #5G #6G #wifi #private5G #privatewireless #spectrumpolicy #networks #wireless #mobile #enterprisenetworks #neutralhost
[1] Disclosure: both SFTF and DSA are advisory clients of Disruptive Analysis / Dean Bubley
I help wireless engineers master Private 5G & IoT 📶 & help enterprises adopt 📶🛜
3whappy to play a role in this and voice the needs of the enterprise on the different spectrum & wireless networks.
Senior Director Corporate Marketing @Celona
3wThis is such an important and well-reasoned perspective, as usual Dean. With CBRS in the US and spectrum dedicated for private use across the globe, Enterprises absolutely need a seat at the spectrum policy table. Many have now invested in these networks and are counting on them for their automation and connected worker projects. To suggest that we should tear down an ecosystem that took 5-10 years to develop is ludicrous, and would set back US economic development at a time of global supply chain uncertainty. Well said Dean Bubley.
Analyst and Founder at Recon Analytics LLC
1moThe Navy radars are moving. The US Navy's SPN-43 systems which was the reason for the potential interference with CBRS are being replaced by SPN-50 which are not using the 3 GHz band.
Software Networking and Automation Leader
1moSecond: (i promised two, didn't I), the possibly cynical reframing of the AT&T CBRS initiative is a great example of regulatory capture as a tool to maintain market control. Moving CBRS spectrum in the US with the knock-on effect of destabilizing the spectrum currently used for US Defense contracting (you mention the US Navy) would have profound positive business effects for the purveyors of owned spectrum (such as AT&T) and would have huge positive effects on any number of defence contractors building new gear to fit into new spectrum needs. The only people left out would be the enterprises that are trying to use the free CBRS spectrum to optimize their own work, forcing them into a position where subbing their networks out to the AT&Ts of the world would be the only sane choice in an increasingly risky world. Hopefully that is just being cynical. On the other hand, companies like AT&T and others invest hugely in influencing the regulatory environment and they likely wouldn't if it didn't have direct revenue and profit value.
Software Networking and Automation Leader
1motwo (completely different) thoughts about this article. First: this paragraph resonated with me in a big way: "As a result, they collectively see the need for, and relative benefits of, different spectrum regimes, and complex landscapes of service providers and vendors. One size does not fit all. Optionality and fitness-for-purpose are the key. In some cases, they want to purchase services, and in others they prefer to own connectivity assets, with direct access to spectrum to meet the needs of their applications. They work with a variety of integrators, VARs and specialist service providers. They also have some unique views on the importance of network reliability, especially for business-critical and safety-critical uses – and the importance of long-term stability and predictability in regulation, in order to make investments." With my background in networked applications and enterprise business goals, the need for integrated networking solutions seems completely obvious (and I see you and several other prominent Telecoms folk pressing for it as part of 6G). Whatever happens, applications will need to integrate multiple networking technologies together to function and pressing for more representation is critical.