Some say 5G slicing is a viable alternative to spectrum sharing or private networks — but it suffers from severe constraints
Some wireless industry commentators have suggested that “5G network slicing” is a viable alternative to either spectrum-sharing or the deployment of private (dedicated) networks by enterprises or other organizations. Network slicing, so the misconceived argument goes, should therefore allow for more high-power, wide-area spectrum to be cleared of its current users and made available for mobile carriers via auction.
In reality, 5G slicing has severe constraints related to its capabilities, maturity and usefulness.
It cannot replace shared-spectrum models such as the Citizens Broadband Radio Service (CBRS) which blends commercial wireless and defense/federal bands, nor enable the creation of substitutes for proper, independent private 4G/5G networks.
It also cannot be used to offer “frequency segments” suitable for military systems — a claim based on misunderstood international examples of wireless deployments and slicing.
In reality, network slicing is just a way to create “virtual” partitions of a 5G network, each (theoretically) with its own defined capabilities such as throughput speed, latency, security or priority. Much of the original vision centred on a dynamic model where each 5G customer or application could ask for its own custom-created slice, which was designed and priced to be “just right” for their needs.
There are three critical flaws in the slicing rhetoric:
- Contrary to assertions, network slicing cannot divide spectrum between cellular networks and other uses of spectrum, such as radar, aviation and maritime communications, satellite connectivity, Wi-Fi or scientific observation. It is designed to divide a single 5G network, not make it coexist with completely different radio systems with their own “waveforms” and control mechanisms.
- In the real world, network slicing is very hard to operationalise. The “theoretical” concept does not translate to live networks for a wide variety of reasons, especially variable coverage and capacity. The complexity of juggling hundreds or thousands of network users with unique requirements within a constrained resource is immense. Few carriers are able to offer slices today, and those are usually for isolated use-cases.
- Network slicing cannot generally provide localised, dedicated and customised networks for enterprises, with the same coverage, technical capabilities and degree of control that is available from genuine private wireless systems. In particular, it is not possible to slice a public network where there is no coverage to begin with — say, in remote areas, inside metal-walled factories, or down a mine.
Addressing the misconceptions
To be fair, network slicing sounds simple to describe. But it is, in reality, quite complicated to implement and operate, with lots of technical nuances. It is a capability being made available only on some of the most advanced carrier public 5G networks, or on 5G private networks. It can only work in situations where the network has excellent, proven coverage and adequate capacity, ideally including indoors and remote areas. It needs a complex new “standalone” core network (also called “5G SA”) for control, which few carriers have implemented across their whole footprints.
In other words, slicing is at early stages of commercialization. It potentially holds promise for certain applications in certain places, but it is not a general-purpose or fully-flexible approach. For instance, it is used in T-Mobile’s T-Priority service for public safety agencies wanting to use public networks, set up to rival to AT&T’s FirstNet.
To highlight the misconceptions, consider an April 2025 article by Roslyn Layton titled “Spectrum Supremacy: Reclaiming America’s Edge in a Contested Domain.” This painted an unrealistic scenario, suggesting that “slicing” might somehow be a technique that could allow existing military bands to be auctioned for high-power cellular use. It suggested that slicing could substitute for the advanced spectrum-sharing mechanisms currently being explored by military and commercial entities, via a misinterpretation of some other announcements around the world.
The article stated: “The militaries of China, Norway, Finland and Sweden all align spectrum to maximize commercial benefits, then allocate “slices” — dedicated frequency segments with tailored service levels — for military use”
This is wrong in several ways. These examples do involve partitioned access to the public 5G networks and services, where the military has been allocated a slice. But this is not something that enables coexistence for public 5G with radar, military aviation connectivity or missile defence systems. Nor is it being used for critical battlefield or training communications between soldiers.
It is actually for routine (but secure and prioritised) communications in peacetime, or for interacting with civil defense organisations or disaster preparedness. In essence, this is similar to the T-Mobile example for public safety. In an emergency, it may also be a way for defense forces to use a civilian network slice as a backup to a dedicated network. Examples cited in Finland and Sweden are of this type — although mainly still for trials rather than full deployment.
There is certainly a general move towards use of 5G in NATO and other defense forces around the world for a range of such applications, but it is largely as an alternative to private radio systems (such as “walkie-talkies”) rather than an across-the-board solution. Indeed, the US Department of Defense (DoD) is itself already following this path.
In October 2024 the DoD published a document called the Private 5G Deployment Strategy, which detailed its wide use of commercial 5G networks where feasible and appropriate. It cites routine mission needs, mobile users, office functions and mobile broadband for “quality of life” for warfighters, staff and others such as families and visitors on bases.
While this is a valid and interesting use of 5G, it certainly doesn’t address non-mobile technologies such as radar or command/control/communications platforms, such as those used in the U.S. lower 3GHz spectrum.
Another 5G slicing scenario has nothing to do with carrier networks at all: where an enterprise or military organisation deploys its own 5G private network in shared spectrum bands like CBRS or military-dedicated frequencies. For instance, various defense agencies are creating private 5G networks on bases, ships or forward-deployed locations. They may create separate slices for internal operational reasons, such as prioritising drone or vehicle connections on one slice, and handheld radio systems on another. It could look to segregate non-critical functions such as basic Internet access for housing on a base, from classified or operational systems running on the same network. One of the examples cited in Layton’s article — in Norway — is using this model.
In summary, 5G network slicing is an interesting technology, with various military and enterprise use-cases on both commercial and private networks — but it does not change the debate over spectrum allocation. In particular, you cannot use slicing to accommodate radar or other non-cellular wireless systems in the same frequencies; that needs more advanced forms of spectrum-sharing or coexistence mechanisms.
