4G vs 5G Router: Which One Does Your Site Actually Need?

The 4G versus 5G question gets asked as though it has one answer. It does not. It depends on the coverage at your site, what the router is connecting, how long you expect it to stay installed and how much you care about the price of each unit. This guide works through those factors honestly, including the cases where 5G is a waste of money.

The headline differences

The numbers above are what you see in the field with decent antennas, not the theoretical peaks on a datasheet. A 4G modem quoted at 600 Mbps or a 5G modem quoted at several gigabits will almost never reach those figures, because they assume perfect signal, multiple wide carriers and an empty cell.

Understanding 4G categories

4G modems are graded by category, which tells you roughly what they can do. Cat 1 and Cat 4 cover most IoT and telemetry. Cat 6 adds carrier aggregation, which combines two frequency bands for more speed. Cat 12 and above add more aggregation and more MIMO. Cat 20 modems can, in theory, pass 2 Gbps, although nobody in the UK will see that.

For low-data work a Cat 1 router is ideal. It uses little power, costs less and is designed to run for years. Cat 1 bis takes this further with a single antenna, which makes very small devices possible. Neither is a sensible choice for video or replacing a broadband line.

5G modems also support 4G, and they usually support a lot of it. A modern 5G module will typically handle more 4G carrier aggregation combinations than an older 4G-only router. That means a 5G router on a 4G-only site is often still faster than the 4G router it replaced, just not by as much as the marketing implies.

Where 5G clearly wins

Replacing a fixed line. If a router is going to be the main connection for an office, a shop or a site cabin full of people, 5G gives you the headroom. One person on a video call, two people uploading files and a backup job running overnight will saturate a weak 4G connection quickly.

Video. CCTV is upload-heavy. Several high resolution cameras streaming to a cloud service or being viewed remotely need sustained upload capacity, and 5G has much more of it where n78 coverage exists.

Busy locations. In town centres, stations and event venues, 4G cells get congested at peak times. 5G has more capacity per cell, so performance holds up better when everyone is on their phones.

Latency-sensitive work. Voice, remote desktop and some control applications benefit from lower and more consistent latency, especially on standalone 5G.

Long service life. Hardware installed in infrastructure tends to stay there. If you expect a router to be in a cabinet until the mid 2030s, buying 5G now is cheaper than a truck roll later.

Where 4G is still the right answer

Low data telemetry. Smart meters, tank level sensors, environmental monitoring, alarm panels and most SCADA polling move kilobytes, not gigabytes. A 4G Cat 1 router will handle that for years and nobody will notice the difference.

Battery and solar sites. Power budget matters. A 5G router drawing 10 watts all day needs a much bigger solar panel and battery than a Cat 1 device drawing a couple of watts.

Large fleets on tight budgets. Multiply the price difference by five hundred units and the business case changes quickly.

Sites with no 5G and no prospect of it. If the nearest 5G is many miles away and the application is light, paying for 5G hardware is buying a feature you will not use.

The network lifecycle question

The UK has already switched off 3G. Equipment that relied on it had to be replaced, often in a hurry. 2G is due to go by 2033. 4G is not going anywhere soon, and will almost certainly outlast both, but networks will gradually move spectrum from 4G to 5G as more devices support it.

That does not mean 4G routers will stop working in a few years. It does mean the 4G experience may get thinner in some places over time as capacity shifts to 5G. For low-data devices that is unlikely to matter. For a router carrying an office’s traffic it might.

There is a middle path worth knowing about. 5G RedCap is a reduced capability version of 5G designed for exactly the kind of devices that would otherwise use mid-range 4G. It runs on standalone 5G and offers modest speeds with lower cost and power than full 5G. See our 5G RedCap router guide.

Antennas: the hidden cost of 5G

Most 4G routers have two antenna ports. Most 5G routers have four. If you are using external antennas, which you should be for anything installed in a cabinet or a building with thick walls, the 5G installation needs a four-port antenna or two dual-port antennas and four cable runs. That costs more and takes more care.

Cutting corners here wastes the 5G hardware. A 5G router with only two antennas connected loses a large part of its potential throughput on n78. See our 5G antenna guide for the detail.

Power, heat and enclosures

5G modems run warmer than 4G modems, especially when pushing high throughput. In a sealed outdoor enclosure in summer that matters. Industrial 5G routers are designed for it, with metal cases and wide temperature ranges, but cheaper units can throttle or reboot when they get hot. If the router is going in a cabinet that sits in the sun, check the operating temperature range and allow some ventilation or shade.

Power is the other consideration. Most industrial 5G routers accept a wide DC range and draw somewhere between 6 and 15 watts depending on load. Budget for the peak, not the average, when sizing a power supply or UPS.

Cost comparison over the life of the router

The purchase price gap between 4G and 5G hardware is real, but it is only part of the cost. Over five or more years the larger costs are usually data, management platform fees and site visits. A single unplanned truck roll can wipe out the saving from buying a cheaper router, so the question is really whether 4G will do the job for the whole life of the deployment.

Data tariffs for 5G are not automatically more expensive than 4G, but not every business SIM includes 5G access. Check that the SIM you plan to use allows 5G, and if you need it, standalone 5G.

What the router says versus what you get

Router status pages are optimistic by design. A 5G icon tells you the modem has a 5G carrier available. It does not tell you how much of your traffic is actually going over it. On non-standalone 5G the router can be anchored to 4G with a 5G carrier added only when there is enough data to justify it, so a router idling on a status page can look different from the same router mid-download.

The figures worth looking at are the signal measurements for each connection. For 4G that means RSRP, RSRQ and SINR. For 5G the equivalents are usually labelled SS-RSRP, SS-RSRQ and SS-SINR. As a rough guide, RSRP better than minus 90 dBm is healthy, minus 90 to minus 105 is workable, and below that you are fighting for every megabit. SINR above 13 dB is good, and below 3 dB things get unreliable whatever the RSRP says. A strong but noisy signal is often worse than a weaker clean one.

When you compare 4G and 5G at a site, compare like with like. Run speed tests at the same time of day, from the same position, with the same antennas, and repeat them. One test at 6am on a Sunday tells you almost nothing about Tuesday lunchtime.

Running a mixed fleet

Many organisations end up with both. A water company might use 4G Cat 1 for hundreds of small outstations and 5G for a few treatment works with CCTV and remote engineering access. A retailer might put 5G in flagship stores and 4G backup in small branches. That is perfectly sensible, as long as the two sets of hardware can be managed from the same platform and configured the same way.

The trap is ending up with three manufacturers, three management portals and three sets of firmware to keep patched. If you can, choose a manufacturer whose 4G and 5G ranges share the same operating system. Configurations can then be copied between models, engineers learn one interface, and the security patching job stays manageable.

It is also worth standardising on SIMs. A multi-network SIM that works in both the 4G and 5G hardware simplifies stock, provisioning and support. See SIM options for 5G routers.

A simple way to decide

Testing before you commit

The single best thing you can do is test at the site. A coverage map tells you what the network models predict. A test device tells you what actually happens. Look at RSRP and SINR, not just the bars on a screen, and test at the time of day the site will be busiest.

If you are planning a rollout across many locations and cannot test every one, choose hardware that will work acceptably on 4G and better on 5G, and pair it with a SIM that can use more than one network. That combination covers most of the uncertainty.

If you would like help deciding which way to go for a particular project, email sales@5grouter.co.uk with the site details and what you are connecting.