Nmschoolbuildings
Upgrading Industrial IoT Devices from 4G to 5G: Why the Antenna Change Is Not Straightforward
Industry Guide November 17, 2026

Upgrading Industrial IoT Devices from 4G to 5G: Why the Antenna Change Is Not Straightforward

Industrial IoT deployments built on 4G LTE are facing pressure to migrate toward 5G as network operators expand NR coverage and as new application requirements — lower latency, higher throughput, network slicing — push against what LTE can deliver. The hardware upgrade path seems straightforward on the surface: swap the modem module, update the firmware, replace the antenna. In practice, the antenna portion of that upgrade is where many deployments run into performance problems that weren’t anticipated during the planning phase.

The core issue is that 4G and 5G NR operate across a substantially different set of frequency bands, and the antenna characteristics that work well for one don’t necessarily transfer to the other. An antenna that was correctly specified and installed for a 4G application may underperform significantly in a 5G NR deployment at the same location — not because the antenna is broken, but because the new radio environment makes different demands on it.


How the frequency band landscape changed

4G LTE in most industrial deployments operates primarily in low-band and mid-band spectrum: 700 MHz, 850 MHz, 1700/2100 MHz (AWS), 1900 MHz, and 2100 MHz are the most common bands in North America and Europe. These frequencies have wavelengths in the range of roughly 15 cm to 43 cm, and the antennas designed for them are physically sized and tuned to these wavelengths.

5G NR adds a new set of bands on top of the existing LTE spectrum. Sub-6 GHz 5G NR uses many of the same low and mid bands as LTE, adding new mid-band allocations around 2.5 GHz, 3.5 GHz (n78, the most widely deployed 5G mid-band globally), and 3.7–3.98 GHz (n77). It also introduces high-band millimeter wave spectrum above 24 GHz in markets where operators have deployed it.

For industrial IoT applications, millimeter wave 5G is largely irrelevant — its range is too short and its penetration through building materials too poor for most industrial deployment scenarios. The practically relevant change is the addition of mid-band 5G NR in the 3.5 GHz range. A wavelength at 3.5 GHz is approximately 8.6 cm, compared to roughly 42 cm at 700 MHz. An antenna optimized for 700 MHz LTE is physically much larger than what 3.5 GHz demands — which means it has poor radiation efficiency at 3.5 GHz and cannot be reused for 5G NR mid-band operation.

What “replacing the antenna” actually requires

The assumption that a 4G antenna can be directly swapped with a 5G antenna, keeping the same mounting location and cable run, breaks down for several reasons.

First, 5G NR mid-band antennas need to cover a wider frequency range than typical 4G antennas. A single antenna that covers both legacy LTE bands (for fallback and voice) and 5G NR mid-band must operate across a frequency ratio of roughly 5:1 — from 700 MHz to 3.5 GHz. Achieving acceptable VSWR across this entire range from a single antenna element is a genuine engineering challenge, and the compromise designs that cover the full range tend to have lower gain at any individual frequency than a narrowband antenna optimized for a specific band.

Second, the installation location that worked for a 4G antenna may not be suitable for the 5G mid-band element. At 3.5 GHz, any object within a few wavelengths of the antenna — a metal enclosure wall, a cable tray, another antenna — has a stronger near-field interaction with the antenna than at 700 MHz. An antenna mounted on the outside of a metal industrial enclosure, 5 cm from the nearest edge of the enclosure, has adequate clearance for 700 MHz operation. At 3.5 GHz, that same 5 cm is less than two wavelengths — well within the near-field zone — and the enclosure edge will distort the radiation pattern and detune the antenna’s impedance match.

Third, cable loss increases with frequency. A 10-meter coaxial cable run that introduces 2 dB of insertion loss at 900 MHz introduces roughly 5–6 dB at 3.5 GHz for the same cable type. If the 4G installation used a long cable run to reach a good mounting location, that cable run may now be imposing a significant loss penalty on the 5G signal path that didn’t exist in the 4G deployment. In a marginal coverage area, 3–4 dB of additional cable loss is the difference between a usable 5G connection and an unusable one.

Reassessing mounting location for 5G NR

The practical consequence of these changes is that a proper 5G NR antenna upgrade requires evaluating the mounting location as if starting from scratch, not treating it as a fixed constraint from the previous installation.

For the 4G 5G Cellular Antennas used in industrial applications, the key questions when evaluating a mounting location are different for 5G mid-band than they were for 4G. At 3.5 GHz, the antenna needs clearance from metal structures in all directions — not just below and to the sides, but also from any structure above it that might block the elevation angles where the base station signal arrives. 5G NR mid-band base stations often use beamforming, which means the signal can arrive from a specific azimuth and elevation angle rather than uniformly from the direction of the tower. An antenna that is physically blocked in the direction of the beamformed signal will not benefit from the beam gain that the base station is directing toward the device.

Mounting height also matters more at mid-band than it did at low-band LTE. The Fresnel zone radius for a 5G NR link at 3.5 GHz is smaller than at 700 MHz — approximately 60% of the radius for the same link geometry. This means a lower antenna can clear the first Fresnel zone adequately at 3.5 GHz in some installations where it couldn’t at 700 MHz, but it also means that obstacles closer to the antenna have a larger relative impact on the link. A rooftop mounting that clears the parapet wall by 30 cm might have been fine for 4G; at 3.5 GHz, the same 30 cm clearance may put the antenna in the diffraction zone of the parapet edge for base stations in certain directions.

MIMO antenna placement for 5G NR

5G NR devices typically use 2×2 or 4×4 MIMO, which requires multiple antenna elements with sufficient spatial or polarization diversity to provide independent signal paths. The antenna separation requirements for 5G NR MIMO at mid-band are physically smaller than for 4G MIMO at low-band — a half-wavelength separation at 3.5 GHz is about 4.3 cm, compared to about 21 cm at 700 MHz. This makes it easier to fit a properly spaced 5G MIMO antenna pair into a compact installation.

However, the smaller required separation means that carelessly placed antennas — mounted very close together because space is tight — have higher mutual coupling at 3.5 GHz than the equivalent installation did at 700 MHz. At 1 cm separation, two 3.5 GHz antenna elements are coupled to each other in a way that degrades the MIMO correlation properties. The practical guidance is that the spacing should still be at least a half-wavelength, with a full wavelength preferred if the installation geometry permits, even though the absolute physical dimension is small.

Testing before committing to an installation

Given the number of ways that a 5G NR antenna installation can underperform without obvious cause, field testing before finalizing the mounting arrangement is worth doing. A cellular signal meter or the diagnostic mode of a 5G modem can report signal level and signal quality (RSRP, RSRQ, SINR) at the candidate mounting location before the cable is routed and the installation is buttoned up. Moving the antenna trial position by 10–20 cm in different directions and measuring the signal change identifies whether the current location is in a local optimum or whether there’s a significantly better position within reach.

This takes an extra hour at installation time and prevents the kind of underperforming 5G deployment that gets attributed to “5G coverage not being good here yet” when the actual cause is an antenna that’s mounted in a position that made sense for the previous 4G deployment but doesn’t work for the new frequency bands.

Related Articles