Opinion: Was the industry wrong about mmWave?

Paul Wright, Chief Revenue Officer at Cambridge Broadband Network Group (CBNG), returns to Mobile Industry Review to revisit one of the industry's longest-running debates: was mmWave ever really a failure? I'm delighted to feature his post here.

Over to you Paul!

For more than two decades, mmWave spectrum has occupied an unusual place within the telecom industry. It has long been recognised for its enormous capacity potential, yet repeatedly questioned for its commercial viability.

From the ambitious Local Multipoint Distribution Service (LMDS) deployments of the late 1990s through to today’s dense 5G networks, the industry has consistently returned to mmWave as a potential solution for delivering high-capacity wireless connectivity. Yet despite impressive technical demonstrations and periodic waves of enthusiasm, many operators and industry veterans continue to associate mmWave with deployment complexity, limited range and difficult economics.

The conventional narrative is that mmWave never quite delivered on its promise.

But what if the industry has been asking the wrong question?

Perhaps the story of mmWave is not one of technological failure at all. Perhaps it is the story of a technology whose capabilities, economics and deployment models took longer to mature than the industry expected.

From early promise to industry scepticism

The original vision for mmWave was compelling.

In the late 1990s, LMDS promised high-speed broadband wireless access using large blocks of relatively uncongested spectrum. At a time when fibre deployments remained limited and broadband demand was growing rapidly, the prospect of delivering fibre-like speeds wirelessly was understandably attractive.

The technical proposition itself was sound. Large channel bandwidths offered the potential for significantly higher throughput than conventional wireless systems, while the abundance of available spectrum suggested a path toward scalable future capacity.

However, many early deployments struggled to meet commercial expectations.

In hindsight, part of the challenge was that the industry often approached mmWave using planning assumptions developed for lower-frequency networks.

Coverage models that worked effectively in traditional macro-cellular environments proved far less effective when applied to much higher frequency spectrum. Many deployments underestimated the impact of blockage, line-of-sight requirements, environmental attenuation and the level of densification required to maintain consistent performance.

As a result, some operators concluded that mmWave itself was the problem.

Yet a closer examination suggests a different interpretation.

The physics were never broken

One of the most persistent misconceptions surrounding mmWave is that its propagation characteristics somehow represent a flaw in the technology.

In reality, many of the characteristics later cited as evidence of mmWave’s shortcomings were never unknown limitations. They were well-understood aspects of mmWave radio propagation.

Unlike lower-frequency spectrum, mmWave signals exhibit minimal building penetration, limited diffraction around obstacles and a strong dependence on clear radio paths. Rain attenuation and environmental conditions also have a greater impact than they do at lower frequencies.

None of this was a surprise to radio engineers.

The challenge was that these realities often conflicted with decades of network design experience built around spectrum that could penetrate buildings, bend around corners and tolerate partial obstructions.

In many cases, planners were attempting to apply familiar deployment methodologies to a fundamentally different propagation environment.

Links were sometimes deployed beyond practical design limits. Expectations around non-line-of-sight performance proved unrealistic. Coverage assumptions developed for conventional cellular networks were applied to a spectrum band that was never intended to behave in the same way.

Viewed through this lens, many early mmWave deployments may be better understood as examples of planning assumptions conflicting with propagation realities rather than examples of technology failure.

The physics were never broken.

The industry simply had to learn how to design around them.

The technology ecosystem has finally matured

At the same time, it would be wrong to suggest that the technology itself has stood still.

The mmWave ecosystem has evolved dramatically since the LMDS era.

Advances in semiconductor technology, phased-array antennas and beamforming have transformed what is possible. Modern systems can direct energy with far greater precision, improving both reliability and spectral efficiency while reducing interference.

RFIC integration has significantly reduced equipment size, power consumption and manufacturing costs. Capabilities that were once available only in highly specialised systems are increasingly becoming commercially scalable.

Progress has also extended beyond the radio layer.

Cloud-native architectures, intelligent scheduling and AI-assisted network management now allow networks to dynamically optimise resources based on traffic patterns, environmental conditions and user demand.

These capabilities reduce operational complexity while helping operators maximise the value of available spectrum assets.

The result is a technology ecosystem that looks fundamentally different from the one that existed twenty years ago.

The industry problem has changed

Perhaps even more important than the technology evolution is the fact that the industry's problem has changed.

Historically, operators were primarily focused on extending coverage.

The dominant challenges of the 2G, 3G and early 4G eras revolved around increasing geographic reach, expanding subscriber footprints and ensuring broad service availability.

Today, many operators face a different challenge.

The bottleneck in modern networks is increasingly not geographic reach, but localised capacity concentration.

Dense urban environments, stadiums, campuses, transportation hubs, enterprise applications, fixed wireless access deployments and AI-driven workloads are creating demand patterns that are highly concentrated rather than evenly distributed.

In many locations, the challenge is no longer reaching users. It is serving them efficiently once they are connected.

This shift fundamentally changes the relevance of mmWave.

For years, the industry often evaluated mmWave against coverage-centric success metrics inherited from lower-frequency networks.

Today, operators increasingly recognise that future network performance may depend less on universal coverage and more on delivering extreme capacity exactly where it is needed most.

The rise of precision capacity

This shift is accelerating the emergence of what might be described as "precision capacity" architectures.

Rather than attempting to maximise coverage everywhere, operators are increasingly deploying targeted layers of high-performance infrastructure in locations where traffic demand and performance expectations are highest.

This approach aligns naturally with the strengths of mmWave.

The same propagation characteristics once viewed as limitations can become advantages when deployed within appropriately engineered architectures. Highly directional links enable exceptional spectral reuse. Large channel bandwidths support multi-gigabit performance. Dense deployments can deliver extraordinary capacity within relatively small geographic areas.

The future success of mmWave may depend less on overcoming its propagation characteristics and more on intelligently architecting networks around them.

In that sense, the industry is increasingly learning to work with the physics rather than against them.

The economics are changing too

Historically, another challenge facing mmWave was economic.

The enormous bandwidth available within mmWave spectrum has always been compelling. However, operators often struggled to justify the additional cost associated with specialised radio equipment, denser deployment models and more complex installation requirements.

For many years, the economic equation was difficult.

Today, that equation is changing.

Equipment costs continue to fall as semiconductor integration improves and production volumes increase. At the same time, the demand for capacity continues to rise, increasing the value of spectrum capable of supporting multi-gigabit services.

The economics of fibre deployment have also evolved. Rising labour costs, civil engineering expenses, permitting complexity and deployment timescales have increased interest in alternative approaches capable of delivering high-capacity broadband services more rapidly.

Public funding programmes are also reshaping the landscape.

Initiatives such as BEAD in the United States, alongside a variety of broadband investment and subsidy programmes across Europe and the United Kingdom, increasingly focus on delivering measurable broadband outcomes rather than prescribing specific access technologies. As a result, high-capacity wireless solutions are receiving greater consideration as practical tools for addressing connectivity challenges in both urban and rural environments.

Taken together, these factors are creating commercial conditions that are considerably more favourable than those faced by earlier generations of mmWave deployments.

Looking ahead

The story of mmWave is not one of technological failure.

It is a story of industry evolution.

Many of the challenges that shaped early perceptions were rooted in deployment strategies, economic realities and market expectations rather than the capabilities of the spectrum itself.

Today, the technology ecosystem is more mature. Network demand is changing. Economic conditions are evolving. Operators increasingly understand where mmWave delivers the greatest value and where it does not.

The question may no longer be whether mmWave can replicate traditional network architectures.

The more relevant question is whether future high-density networks can economically scale without it.

As operators build networks designed for AI-driven services, enterprise-grade connectivity and increasingly capacity-intensive digital experiences, success will depend on the intelligent deployment of precision capacity.

In that future, mmWave is positioned not as a replacement for existing spectrum layers, but as an increasingly essential component of next-generation network architecture.

Thank you very much Paul!

Connect with Paul on LinkedIn and read more about CBNG at www.cbng.co.uk.