What is 5G Fixed Wireless Access?

Fixed Wireless Access – usually shortened to FWA – is the use of a cellular network as a fixed broadband connection. Instead of running a fibre or copper cable to a building, you install a 5G router, point it at the nearest 5G mast, and get a broadband-grade internet connection over the air.

That is not a new idea. Mobile broadband routers have been available since 3G. What makes 5G FWA genuinely different is throughput, latency, and the introduction of 5G Standalone (SA) architecture – a mode that gives 5G its own core network rather than borrowing 4G’s.

Under 5G SA, properly configured connections can deliver consistent real-world download speeds of 200 Mbit/s to over 600 Mbit/s, with latency below 20ms on a well-loaded cell. That puts 5G FWA in direct competition with mid-tier fibre broadband for many premises – and ahead of it for temporary or remote deployments where running a physical cable is impractical or prohibitively expensive.

How 5G FWA works – the technical picture

A 5G FWA deployment has three components: the 5G radio access network (RAN) operated by the mobile network operator, the customer premises equipment (CPE) at your site, and the SIM card that authenticates your device onto the network.

The 5G radio layer

UK 5G currently operates on multiple spectrum bands. The most commonly deployed are n78 (3.4-3.8 GHz) – the primary 5G mid-band – and n1 (2.1 GHz), which provides deeper indoor penetration. mmWave (above 24 GHz) exists in limited trial deployments but is not relevant to FWA at this point. For Sub-6GHz 5G FWA, n78 is the band that delivers the best combination of speed and range.

From mast to router

The UF51 picks up the 5G signal through its internal or external antennas, establishes a radio bearer with the base station, authenticates via its SIM card, and receives an IP address from the mobile network’s packet core. From that point, it routes traffic just like any other gateway – NAT, firewall, DHCP, and LAN switching are all handled internally. Devices on your local network connect via Gigabit Ethernet or Wi-Fi and are entirely unaware that their internet connection is cellular rather than fibre.

Latency and throughput in practice

On a mid-loaded 5G SA cell with a strong n78 signal, you should realistically expect 100-400 Mbit/s download, 30-80 Mbit/s upload, and 10-25ms round-trip latency. That covers the vast majority of business and residential broadband use cases. Video conferencing, VoIP, cloud applications, remote desktop sessions, and large file transfers all run comfortably on those figures.

The Milesight UF51 – what it is and what it does

The Milesight UF51 is a compact Sub-6GHz 5G cellular router designed for both indoor and outdoor CPE deployment. Milesight pitch it at industrial and commercial applications, and the build quality reflects that – it is not a consumer hotspot in a white plastic shell. It is a properly engineered gateway with a feature set that goes well beyond getting devices online.

Specification	Detail
5G mode	SA and NSA dual-mode, Sub-6GHz NR
Fallback	LTE-A Cat 20, 3G, with optional Cat M/NB-IoT support
SIM slots	Dual nano-SIM with automatic failover
Wi-Fi	802.11 a/b/g/n/ac dual-band (2.4 GHz + 5 GHz)
Ethernet	Multiple Gigabit ports (WAN/LAN configurable)
Antenna	Internal with external SMA connector support
Edge computing	Milesight Development Platform (Node-RED based)
Remote management	DeviceHub cloud management, SNMP, SSH
Operating temperature	-40 to +70 degrees C
Power	12V DC with PoE option depending on variant

What distinguishes the UF51 from cheaper 5G routers is the SA mode support combined with the onboard programmability. Most consumer 5G routers are NSA-only and offer no edge logic layer at all. The UF51 gives you native 5G SA connectivity and the ability to run custom logic directly on the device – which opens up applications that a standard router simply cannot touch.

Why, Where and When – the deployment decision

Why choose 5G FWA over fibre?

Fibre broadband is generally the preferred choice where it is available, stable, and affordable. 5G FWA wins in specific situations:

• No fibre at the site. Rural locations, industrial estates, construction sites, and listed buildings where digging or cabling is impractical.
• Temporary or short-term deployments. Events, pop-up operations, site offices, or temporary retail where a 12-month fibre contract makes no financial sense.
• Fast deployment needed. A SIM-based connection can be live within hours of the hardware arriving on site. Fibre provisioning typically takes weeks.
• Fibre backup. Adding a 5G FWA connection alongside a fibre primary is a proven resilience strategy for sites that cannot afford downtime.
• Locations with good 5G coverage but poor fixed-line infrastructure. Some urban areas – particularly new developments or areas with ageing copper – have excellent 5G availability but poor FTTC/FTTP provision.

Where does 5G FWA work well?

• Sites within 1-2 km of a 5G mast with reasonable line of sight
• Business premises needing 50+ Mbit/s with low latency
• Temporary site offices, portakabins, and construction compounds
• Rural SME premises on slow ADSL or FTTC with no FTTP timeline
• Secondary offices needing a cost-effective WAN connection
• Event venues and seasonal operations

Where does 5G FWA struggle?

• Basement or sub-street-level locations with no outdoor antenna option
• Sites in deep rural areas beyond current 5G coverage
• Applications requiring SLA-backed guaranteed throughput (enterprise-grade MPLS or dedicated fibre remain the right choice there)
• Very high concurrent user density – a busy hotel or conference centre with hundreds of simultaneous users will stress any shared cellular connection

How to install the Milesight UF51 for 5G FWA

Installation is straightforward. The UF51 is designed to be configured via a local web interface and, once set up, managed remotely via Milesight DeviceHub or a third-party management platform.

The Milesight Development Platform – why it matters for FWA

Most 5G routers are dumb pipes. Traffic comes in one side, goes out the other, and that is the full extent of their capability. The UF51 is different because it includes the Milesight Development Platform (MDP) – a Node-RED based edge computing environment running directly on the device.

For a detailed technical breakdown of what MDP is and how it works, the team at routerstore.com have published a thorough explainer: Milesight Development Platform – complete guide.

In the context of 5G FWA specifically, MDP opens up capabilities that no ordinary broadband router can match:

Bandwidth monitoring and alerting

Build a Node-RED flow that tracks real-time throughput against your SIM’s data allowance. Configure automated alerts when you hit defined thresholds – 50%, 80%, 90% of monthly allowance – and trigger email or webhook notifications without needing any external monitoring infrastructure. This is genuinely useful when managing SIM data costs across multiple sites.

Custom failover logic

The UF51 has built-in dual-SIM failover, but MDP lets you build logic that goes further. You can define failover conditions based on latency, packet loss, or specific destination reachability rather than just raw link state. You can also build asymmetric routing rules – route latency-sensitive VoIP traffic via one path while bulk data takes another.

Local data processing

For IoT-heavy deployments where the UF51 is also providing connectivity for sensor devices, MDP can aggregate, filter, and pre-process sensor data at the edge before it leaves the building. This reduces cloud processing costs and keeps the cellular link free for high-value traffic.

API integration and webhooks

MDP can post connection status, throughput metrics, and device health data directly to any HTTP endpoint – your own dashboard, a Slack channel, a monitoring platform. This turns a connectivity device into a visible, manageable asset rather than something you only think about when it breaks.

Which SIM to use in the UF51 – and why this matters more than most people realise

SIM choice for 5G FWA is arguably the most important decision you make in the whole deployment, and it is consistently underestimated. Get it wrong and you will have a router that connects on 4G rather than 5G, or on 5G NSA rather than 5G SA, regardless of what your signal meter says.

The 5G SA problem with MVNO and roaming SIMs

Many businesses default to multi-network or MVNO SIM cards for IoT and connectivity deployments – and with good reason. They offer network flexibility, roaming across multiple MNOs, and simplified commercial management. But there is a critical limitation that catches people out:

The majority of MVNO and multi-network roaming SIM cards do not currently support 5G SA.

The reason is commercial and technical. 5G SA is a distinct core network architecture. Roaming onto another operator’s 5G SA core requires bilateral roaming agreements at the SA level – not just at the radio level. Those agreements are still being established across the industry. Most MVNO roaming arrangements were built on LTE and 5G NSA agreements. The result is that a roaming SIM inserted into a 5G SA capable device like the UF51 will often register on 5G NSA – or fall back to LTE entirely – even when you are standing next to a 5G SA base station.

Which UK MNO for 5G FWA?

All four UK MNOs – EE, Three, O2, and Vodafone – have 5G SA in deployment to varying degrees. EE and Three currently have the most extensive 5G SA coverage in UK urban and suburban areas. Three’s 5G SA rollout has been aggressive following the merger completion. Coverage varies significantly at a local level, so the right answer depends on your specific site.

The practical approach: test two or three MNO SIMs on site before committing. A brief signal walk with each SIM in a 5G SA capable handset, checking reported 5G mode (NSA vs SA) in engineering mode, will tell you more than any coverage map.

What happens if EE goes down?

This is a fair question and one worth addressing directly. Mobile networks – including EE – do experience outages. Planned maintenance windows, base station hardware failures, backhaul issues, and software updates all cause intermittent disruption. If your site is entirely dependent on a single MNO SIM for internet access, a network event at that MNO will take your connection down.

The honest answer is: if your site cannot tolerate any connectivity loss, a single SIM on a single MNO is not a complete solution. It is, however, a very good primary path – and the UF51 is designed to support resilience alongside it.

Resilience options with the UF51

How often do UK MNO networks actually go down?

Major outages affecting a significant portion of a network are relatively rare – a few times a year at most for any given MNO. Local cell-level issues – a single mast with a backhaul fault or software issue – are somewhat more common but typically affect a small geographic area. For most business deployments, a dual-SIM configuration covering two different MNOs provides sufficient resilience for all but the most critical applications.

For mission-critical connectivity – SCADA systems, payment processing, alarm monitoring – a dedicated secondary path on a different physical infrastructure is the correct design. In those cases, the UF51 is typically the primary router in an architecture that also includes a secondary LTE router or a fixed line as a completely independent failure domain.

Monitoring and alerting for proactive resilience

One of the most valuable uses of the Milesight Development Platform in an FWA context is building proactive monitoring into the device itself. A simple Node-RED flow can ping a known-good external address every 60 seconds, log response times, and send an alert if the primary link degrades or fails – before your users start complaining. That gives you and your users visibility of network events in real time rather than finding out retrospectively.

Should you wait for SGP.32? The honest answer.

SGP.32 is the GSMA’s IoT-specific eSIM specification. It builds on the M2M eSIM architecture (SGP.02) and the consumer eSIM standard (SGP.22) with a design optimised specifically for IoT devices – remote profile management, low-bandwidth provisioning, and support for resource-constrained hardware. In practical terms, SGP.32 will eventually allow you to switch MNO profiles on a device remotely, over the air, without physically swapping a SIM card.

That is genuinely significant for large-scale FWA and IoT deployments. The ability to move a device from EE to Three to O2 remotely – without a field visit – changes how you manage connectivity at scale. It also simplifies the resilience picture considerably: instead of managing two physical SIMs across a fleet of devices, you manage profile assignments from a central platform.

Is it worth waiting for now?

For most deployments in 2025, the honest answer is no. Here is why:

• Hardware availability. SGP.32-capable consumer and commercial CPE devices are still rare. The ecosystem is in early commercial deployment. Waiting means delaying your connectivity project by an uncertain amount of time.
• MNO readiness. SGP.32 requires MNO support for the IoT eSIM Manager (eIM) interface. UK MNOs are building this capability but full commercial availability at scale is still developing.
• Your immediate problem still exists. If you have a site with poor broadband and good 5G coverage, installing a UF51 with a native MNO SIM today solves that problem today. Waiting 12-18 months for SGP.32 hardware means 12-18 months of living with the problem.

Frequently asked questions

---