Sustaining FWA Without Spare Capacity

Roberto Yanez and Greg Agami

Great While It Lasts: FWA Monetizes Unused Capacity

Wireless networks have always been built on a critical asset: spectrum. You can’t really see it, but it exists, it works, and it costs money. However, not all spectrum is created equal, and not all of it gets used equally. In practice, traffic across cellular networks often follows a Pareto principle pattern where roughly 20% of cell sites carry around 80% of the load, leaving many sites lightly utilized. Yet mobility requires mobile network operators (MNOs) to maintain broad coverage even in areas where demand is low, so a meaningful share of network capacity generates little to no revenue. This underused capacity is what we can think of as fallow spectrum.

The introduction of Fixed Wireless Access (FWA) was an incredible breakthrough, turning that fallow spectrum or spare network capacity into broadband service and a new revenue stream without major new build or capital investment. MNOs seized the opportunity, rapidly growing FWA subscriber bases. What could go wrong?

The fallow spectrum model works great as long as excess capacity exists. But when mobile and FWA demand increases, networks hit their dimensioning limits and overall performance suffers, something most MNOs will not tolerate as customers begin to experience degraded service including slower throughput and dropped connections. Left unaddressed, these impacts can lead to increased customer dissatisfaction and ultimately, customer churn. At that point, MNOs face a balancing act: delivering a broadband-grade FWA experience while still protecting the mobility performance their highest-value customers expect. As adoption grows and usage climbs, the real question is no longer whether FWA works, but how far it can scale before network investment decisions are needed.

FWA Traffic Growth Trends — Demand Will Close in on Available Capacity

Without continued innovation, excess capacity won’t last forever. FWA subscribers continue to grow and are already consuming broadband-scale data. T-Mobile ended Q4 2025 with about 8.5 million 5G broadband customers and now targets 15 million by 2030. Verizon reported more than 5.7 million FWA subscribers in Q4 2025 and remains on track towards its previously stated 8–9 million goal by 2028.

Total FWA demand is driven by two variables: subscriber growth and usage per subscriber. Per-household FWA consumption continues to climb as video quality improves, connected devices multiply, and cloud and AI applications become more common. Current FWA usage estimates settle in around ~580 GB per month for T-Mobile users and ~635 GB for Verizon. Projecting from today’s usage using a conservative ~10% annual growth rate suggests average FWA households will approach ~1.2 TB per month by 2030.

In wireless networks, growth is easy when capacity is available. The real challenge will come when capacity starts to run out and investing to support growth isn’t as profitable.

Not only is all spectrum not equal, neither is all traffic that rides on it. Yet the scale of FWA demand is too large to ignore. Analyst Craig Moffett estimates that FWA may drive about half of Verizon’s traffic but only ~3% of revenue, and for T-Mobile could represent roughly two-thirds of traffic but only ~6% of revenue. That mismatch forces MNOs to look beyond raw traffic growth and focus on FWA traffic value.

Revenue Reality: FWA vs. Mobile Value

Traffic volume tells only half the story. Historically, MNOs have solved congestion by adding capacity and investing in their networks as margins from mobile service supported it. But FWA economics suggest it won’t be that straightforward anymore.

Pricing per subscriber between mobile and FWA looks similar:

Usage, however, is not:

Translate that into approximate revenue per gigabyte, it becomes obvious that 

MobileFWA
Monthly revenue$60$55
Monthly usage25 GB600 GB
Revenue per GB$2.40$0.09

By 2030, average FWA subscribers using ~1.2 TB per month while pricing remains largely unchanged could push revenue per GB down to roughly ~$0.05, nearly half of today’s level.

So as FWA usage continues to grow, the profitability of each additional gigabyte falls. If FWA depends on monetizing fallow spectrum, then its long-term viability will depend on how efficiently that spectrum can be used. The upcoming phase of FWA will be defined not only by how much spectrum MNOs own, but also by how much performance they can extract from it and the economics of using it.

When Fallow Spectrum Starts To Run Out

As FWA adoption grows, it competes directly with mobile traffic for the same radio resources. That’s manageable when networks have headroom. However, when networks get loaded, FWA usage can eat into the capacity needed to keep mobile performance stable, especially during peak hours. And because mobile traffic carries higher value and stricter expectations, MNOs will always prioritize protecting mobility first.

Traditionally, when networks approach their limits, MNOs would invest either adding spectrum, densifying sites, or upgrading radios. That model worked because mobile margins justified it. But FWA changes the math. With significantly lower revenue per gigabyte, simply building more capacity becomes hard to justify economically. The solution can’t just be more spectrum or more radios. It requires a more strategic approach.

Features such as network slicing can help by reserving resources and keeping FWA demand in check. But slicing doesn’t create capacity. It only allocates what capacity already exists. Once networks approach their limits, even smart policy control can’t eliminate congestion. At that point, the question isn’t whether traffic will be managed. It’s how operators manage resources and network performance expectations without eroding profitability.

Capacity Approaches for Scalable FWA and Techniques to Solve Operator Pain

Licensed mid-band spectrum has been a boon for FWA. After spending over $100B in spectrum auctions on mid-band, US operators experienced massive growth in FWA subscribers, with more than 3.7 million FWA net adds the year after deployment compared to 1 million net adds the year before. While Upper C-band is planned for auction, data demand will eventually exhaust mid-band capacity. Millimeter wave spectrum offers ultra-high capacity but with limited coverage, so is most effective in targeted FWA deployments with high user density such as multi-dwelling units (MDUs).   

Lightly licensed and unlicensed spectrum can be a cost-effective alternative for FWA, although power restrictions and interference are significant challenges. CBRS is lightly licensed using desirable mid-band frequencies, and freely available using General Authorized Access (GAA). Priority Access Licenses (PAL) are also used in this spectrum for prioritized access, and all access is regulated by a Spectrum Access System (SAS) to mitigate interference through power adjustments. Unlicensed frequencies are available in 5GHz and 6GHz bands, but can suffer from Wi-Fi and other interference in congested spaces. Given the constraints of interference mitigation for lightly licensed and unlicensed spectrum, these frequencies are often best utilized in targeted deployments in rural high-clutter environments.

MNOs also need to decide whether to use standardized 3GPP technology or proprietary solutions. 3GPP enjoys a wide ecosystem scale and interoperability benefits across CPEs, chipsets, and infrastructure vendors including Ericsson, Nokia, and Samsung, with a transparent evolution path including feature roadmaps and a migration path towards 6G. 5G NR already employs advanced scheduling, beamforming, and MIMO capabilities to increase spectral efficiency along with carrier aggregation to best utilize fragmented spectral assets. That said, LTE and 5G NR were designed primarily for mobility use cases and traffic patterns, supporting FWA as an added capability.

Proprietary solutions can be optimized specifically for FWA, avoiding mobility-related complexity and potentially lowering costs in targeted deployments. Vendors such as Tarana (ngFWA) and Intracom (WiBAS) have deployed purpose-built point-to-multipoint systems with hundreds of service providers in lightly licensed, unlicensed, or mmWave spectrum, using advanced interference mitigation to improve spectral efficiency. When deployed on ancillary spectrum, these systems can add FWA capacity without affecting the mobile user experience and without requiring a traditional cellular core, instead feeding traffic into the operator’s network at an aggregation point. However, their narrower ecosystem introduces scaling risk, with upgrade paths tied to a single vendor, limited chipset diversity, and a smaller CPE portfolio lacking interoperability.

A hybrid deployment technique with 3GPP solutions deployed at scale with proprietary solutions deployed for FWA capacity offload in targeted areas can maximize the benefits of each approach.

The Broader Picture: FWA as a Sustainable and Continued Revenue Stream

While FWA has emerged as a successful use case for 5G using fallow capacity, it suffers cost and performance headwinds as a “bolt-on” feature for a system primarily designed for mobility.  The 3GPP migration path towards 6G offers an opportunity to re-position FWA as a primary 6G service and design target.  Towards this end, 3GPP TR 22.870 proposes new FWA requirements to provide a quantitative improvement in service when using 6G services and capabilities such as AI/ML, noting that FWA devices have a very different mobility profile, traffic usage pattern, and power capability with improved link performance compared to smartphone devices.  Specifically, 6G will provide optimized, secure, and spectrum-compliant support for FWA, including mechanisms for efficient bandwidth utilization, and awareness of service characteristics to enable real-time RAN and core network resource allocations for FWA.

We can expect 6G FWA CPEs to continue to evolve with more outdoor-friendly, self-install designs, higher-order MIMO, and stronger beamforming that improves performance in tougher RF conditions. Also, 6G networks are expected to be able to recognize FWA devices, understand where they are, and retain location context during sessions so optimization decisions can be far more precise. 

Open RAN, and specifically RIC automation with rApps, offers another opportunity starting in 5G to optimize the performance and business case for FWA while protecting the mobile user experience.  RF prediction and network traffic models can be trained using AI/ML techniques to employ capacity-aware traffic steering and energy-efficient scheduling to optimize the use of spectrum and dynamically manage fixed and mobile traffic.  With 6G, there would likely be additional network parameters available to optimize, and integrated sensing and positioning capabilities to optimize fixed access performance.

While FWA has often been seen as a competing solution to fiber, it can be positioned as a complementary technology for underserved fiber markets or those with challenging terrain.  Given the rapid deployment for a FWA system compared to lengthy fiber construction timelines, FWA can serve as a bridge to provide broadband service while fiber is built out, and then continue as a fail-over service once fiber is deployed.  This approach is especially applicable for capital-intensive broadband programs such as E-ACAM and BEAD, addressing rural high-cost areas where fiber construction risk is greatest. Beyond fiber, satellite technology presents a growing alternative for rural subscribers, pushing FWA to differentiate its offerings through superior performance and resiliency.

The FWA Road Ahead

FWA has been the 5G breakthrough use case, utilizing spare network capacity without requiring densification or significant infrastructure investment.  However, continued exponential data usage growth will exhaust existing capacity within the next few years, requiring optimal management of spectrum resources to sustain FWA growth and monetization.  Furthermore, hybrid approaches utilizing multiple spectrum bands and different access technologies will add complexity to managing network capacity, as streamlined proprietary solutions will continue to play an important role in driving down costs in the near-term.  

Looking forward, Open RAN rApps offer an opportunity to take advantage of AI/ML techniques to optimally manage capacity and maximize FWA revenues while protecting the experience of mobile users.  As FWA becomes a primary use case for 6G, these AI-powered applications will gain further importance to meet subscriber demand, enable continued growth and monetization for operators, and minimize network infrastructure expenses.  Operators that treat FWA as a spectrum efficiency problem, and use intelligent network automation to manage it, will be the ones that sustain its profitability into the 6G era.

References

origami networks

Copyright 2025 © Origami Networks LLC. All rights reserved.

Privacy Policy. Terms of Use.