IoT Antenna Range: Gain, Cable Loss & Link Budget

An antenna does not have a fixed range. IoT range comes from the entire RF link: transmit power, both installed antenna systems, cable and connector losses, propagation, receiver sensitivity, interference and the margin reserved for real-world variation. An IoT antenna link budget puts those terms into one dB calculation. It will not promise a field distance, but it will show whether the design is plausible, how much uncertainty it can absorb and whether changing the antenna or fixing the cable is the better move.

The RF link-budget equation

For a one-way terrestrial link, received power can be written as:

Prx = Ptx + Gtx - Ltx - Lpath + Grx - Lrx

Then compare received power with the receiver threshold:

Raw link margin = Prx - receiver sensitivity

Design margin = raw link margin - fade and uncertainty reserve

Keep the signs explicit. Transmit power and received power are absolute levels in dBm. Antenna gains and losses are ratios in dBi or dB. Adding the transmit antenna gain and subtracting its feed loss gives EIRP, but the link still needs the path, receiving antenna, receiving feed and sensitivity.

This arithmetic is protocol-independent. The input values are not: a LoRa receiver’s sensitivity changes with bandwidth and spreading factor, while a cellular modem’s threshold changes with band, channel, coding and network conditions.

Three outputs are often confused. Received power predicts the signal level at the receiver port. Raw link budget is sometimes used for the radio’s maximum theoretical loss—roughly transmit power minus sensitivity plus both antenna gains—but may omit the installation. Allowable path loss is the useful design number because it subtracts feed losses and the required reserve:

Lpath,max = Ptx + Gtx - Ltx + Grx - Lrx - sensitivity - reserve

If the modeled path loss is below Lpath,max, the design closes by the difference. This rearranged form is also a fast way to write a requirement. Instead of asking a supplier for “the longest-range antenna,” specify the frequency, installation losses and minimum realized gain needed to preserve the target margin.

Convert antenna gain into realized gain

Datasheets often lead with peak gain, but the budget needs gain in the direction of the other radio after installation. Four effects commonly spend the advertised advantage:

• Pattern: a higher-gain omni narrows its vertical beam. That can help two fixed antennas at similar height and hurt a tilted device or a gateway above its nodes.
• Efficiency and match: power accepted at the port is not all radiated. Enclosures, batteries and nearby metal can detune the element.
• Polarization: two linearly polarized antennas lose signal when their orientations diverge; a moving node makes that loss variable.
• Feed system: every cable and connector subtracts margin, and loss rises with frequency and length.

The correct comparison is not “2 dBi versus 5.5 dBi.” It is the complete installed path. If moving to the higher-gain antenna adds 3.5 dB but requires a cable with 3 dB more loss, only 0.5 dB remains—and the narrower pattern may make the field result worse.

The IoT antenna buyer’s guide applies the same band-first logic across LoRa, cellular and Wi-Fi hardware. Use the link budget after selecting the correct band—not to justify an antenna tuned for the wrong one.

Path loss turns margin into a range estimate

Free-space path loss is the clean reference model. ITU-R P.525-5 gives the current formal treatment of free-space attenuation. With frequency in MHz and distance in kilometres, the familiar form is:

FSPL = 32.44 + 20 log10(fMHz) + 20 log10(dkm)

Frequency and distance both matter. Double distance and free-space loss rises by about 6 dB. At the same distance, a 2.4 GHz link has about 8 dB less free-space loss than a 6 GHz link.

Free space is not a factory, farm, city or basement. Add—or measure—the losses created by walls, foliage, terrain, Fresnel-zone obstruction, human bodies, vehicle motion and local interference. Keep a separate reserve for variation that cannot be predicted precisely. That separation is useful: deterministic cable loss can be fixed in the BOM, while fading is managed with margin, placement and field validation.

Worked example 1: 433 MHz terrestrial IoT

Consider an illustrative 5 km line-of-sight planning case. These are assumptions, not a product range claim:

The receive level is:

Prx = 14 + 2 - 0.5 - 99.2 + 5.5 - 1 = -79.2 dBm

Raw margin is -79.2 - (-123) = 43.8 dB. If the planning model reserves an illustrative 20 dB for fading, obstruction and implementation uncertainty, 23.8 dB remains.

That large number does not guarantee 5 km. It says the model can absorb another 23.8 dB before reaching its design threshold. A blocked Fresnel zone, a node at ground level, wet foliage, a tilted whip or local interference can consume it. The 433 MHz antenna guide explains the form-factor and placement choices behind those variables.

Regulation is a separate gate. 433 MHz power, duty cycle and permitted applications vary by jurisdiction. In the United States, for example, operation around 433.92 MHz is not a universal general-purpose ISM allowance; the applicable FCC Part 15 rule and device behavior must be checked. The US Government Publishing Office provides the current 47 CFR §15.231 text. A positive link budget never overrides local radio rules.

Worked example 2: private cellular IoT

Now take an illustrative private 900 MHz link over 2 km. Free-space loss is about 97.6 dB, and the planning model adds 15 dB for clutter and penetration, producing 112.6 dB total path loss.

Prx = 23 + 2 - 1.8 - 112.6 + 6 - 1 = -84.4 dBm

Raw margin is 15.6 dB. After an illustrative 10 dB reserve, only 5.6 dB remains. This is where procurement and installation choices become visible:

The example is deliberately near the edge. It shows why a 6 dBi label cannot rescue an unknown cable or a blocked installation. It also models only one direction. Cellular uplink and downlink use different transmit powers, antenna systems and receiver thresholds, so calculate both—or use the network operator’s planning data. Our 4G LTE external antenna guide and 5G IoT antenna guide cover the band and form-factor decisions that precede the math.

What to change when the link budget does not close

Start with the variables that are both controllable and measurable. Move the antenna into the clear, verify the mounting surface and replace a needlessly long or damaged feed assembly before buying more peak gain. Those changes can improve the link without narrowing the pattern or increasing wind load.

Next, revisit the radio configuration. A lower data rate or narrower bandwidth may improve receiver sensitivity, but it also changes airtime, latency, capacity and sometimes duty-cycle compliance. That is a system trade, not a free antenna improvement. For a managed cellular network, band selection or a different site may be more valuable than another few dB on the device.

Only then add antenna gain. Confirm that the new pattern covers the full movement and height range, and calculate both directions again. If the remaining deficit comes from a concrete wall or terrain obstruction, a directional antenna, higher site or additional gateway can solve a problem that an omnidirectional gain upgrade cannot.

How much fade margin should an IoT link keep?

There is no universal “good” margin. A fixed indoor sensor with easy service access does not need the same reserve as a remote safety alarm, moving vehicle or seasonal agricultural network. The 10 and 20 dB reserves above are transparent planning assumptions, not recommendations for every system.

Set the reserve from evidence:

• variation in measured RSSI/SNR across locations, orientations and weather;
• installation and production tolerances;
• expected interference and network loading;
• mobility and obstruction;
• the operational cost of a dropped link.

If the system must survive a measured 12 dB seasonal swing and 4 dB unit-to-unit/installation spread, a 5 dB paper margin is already inadequate. Document the worst credible case and preserve margin beyond it.

Use the right calculator

For a terrestrial IoT path, use the equation above, a transparent worksheet or a protocol-specific tool with the exact modem settings. Semtech’s LoRa Calculator is useful when a LoRa device and its bandwidth, spreading factor and coding settings are known.

GNSource’s link-budget calculator serves a different job: it models an active GNSS receiver chain with antenna/LNA gain, noise figure, cable stages and receiver input performance. It does not accept distance, terrestrial path loss, LoRa settings, cellular bands or fade margin. Use it when an IoT device also contains an active GNSS antenna—not as the engine for the two range examples above.

Field validation before freezing the BOM

1. Use the exact conducted power and sensitivity for the chosen band, bandwidth, modulation and error criterion.
2. Measure the full cable/connector assembly at the operating frequency.
3. Install each antenna on its production ground plane and inside or outside the real enclosure.
4. Record RSSI/RSRP, SNR and packet delivery at worst-case positions, not only at the average site.
5. Test uplink and downlink independently.
6. Compare predicted and measured loss, explain the gap and preserve a documented design margin.

If the measured link misses the model, do not immediately buy more dBi. First inspect antenna matching, cable damage, connector polarity, mounting and Fresnel clearance. Those failures are common, measurable and usually cheaper to fix.

For an antenna review, send GNSource Engineering the band and radio/module, conducted power, sensitivity setting, enclosure, mounting surface, cable and connector, geography, target distance and environment. That is enough to evaluate whether a magnetic-mount, whip or fixed outdoor antenna is the right starting point.

Frequently asked questions

How do you calculate an IoT antenna link budget?

Add transmit power and the realized Tx/Rx antenna gains, subtract feed losses and path loss, then compare predicted received power with receiver sensitivity. Subtract an explicit fade/uncertainty reserve to obtain design margin.

How many times farther does 6 dB reach?

In ideal free space, 6 dB supports roughly double the distance because path loss rises by about 6 dB when distance doubles. Real terrain, walls, antenna height and interference prevent that from being a universal field rule.

Does higher antenna gain always increase range?

No. Gain improves the budget only in the directions covered by the new pattern. A narrower beam, wrong polarization, detuning or extra cable loss can eliminate the apparent advantage.

What is a good link margin for IoT?

There is no single value. Base it on measured fading, installation spread, mobility, interference and the consequence of failure. A positive zero-fade calculation is not a robust design; document the worst credible losses and keep additional margin beyond them.

Treat range as a system result, not an antenna specification. A link budget tells you which dB you can control; field testing tells you whether the model described the real installation.