Choosing the right spreading factor decides how far a sensor reaches, and the LoRaWAN gateways at Kilo Electronics, an online authorised dealer, are only as useful as the link budget behind them. This 2026 benchmark page shows LoRaWAN range by spreading factor (SF7 to SF12) and how each step behaves across open, suburban and dense urban environments.
- LoRaWAN range by spreading factor: SF12 gains 14 dB of link budget over SF7, which is the whole range advantage.
- In a free-space model SF12 reaches about 5.0 times the SF7 distance; in a dense urban model about 2.2 times.
- SF12 cuts the data rate from 5,470 bps to 250 bps, so each message occupies the channel far longer.
- Fix a weak link with antenna height and placement first; use a higher SF only for the sensors that still miss.
The short answer: what the average gain from a higher spreading factor is
Moving from SF7 to SF12 adds 14 dB of receiver sensitivity at 125 kHz bandwidth. Whether that becomes 2 times or 5 times more distance depends almost entirely on the environment. Open terrain rewards a higher spreading factor most. Dense urban streets reward it least, because buildings absorb signal faster than extra sensitivity can recover it.
The verdict in one sentence: SF12 reaches roughly 2.2 to 5.0 times further than SF7 on the same hardware, at the cost of about 22 times less data rate. Plan for the lower end of that range in towns and the upper end only with clear line of sight.
Why this matters when you buy hardware
Spreading factor is the setting that trades speed for reach. A sensor on SF7 sends fast and short. The same sensor on SF12 sends slowly and long, so it can be heard by a gateway further away or behind more walls.
That trade-off changes what you need to buy. If a site needs SF12 for most sensors, you are probably short of gateways or have them in the wrong place. One extra gateway at a good height often lets every sensor drop to a faster setting, which saves battery and channel time. Kilo Electronics lists gateways, antennas and sensors from several manufacturers, so you can plan the radio side and the hardware together instead of discovering the gap after installation.
The figures below apply to the EU868 band with 125 kHz channels, the most common setup across the EU, UK and Switzerland. Other regional plans use different channels and limits.
LoRaWAN spreading factor table: data rate and sensitivity
These values come from the published Semtech SX127x receiver specifications and the LoRaWAN regional parameters for 125 kHz channels. They describe the radio, not a field test.
| Spreading factor | Data rate (125 kHz) | Receiver sensitivity | Demodulation SNR limit |
|---|---|---|---|
| SF7 | 5,470 bps | -123 dBm | -7.5 dB |
| SF8 | 3,125 bps | -126 dBm | -10 dB |
| SF9 | 1,760 bps | -129 dBm | -12.5 dB |
| SF10 | 980 bps | -132 dBm | -15 dB |
| SF11 | 440 bps | -134.5 dBm | -17.5 dB |
| SF12 | 250 bps | -137 dBm | -20 dB |
SF7 is the fastest and least sensitive. SF12 is the slowest and most sensitive, at 14 dB better than SF7. Each step up from SF7 to SF10 adds 3 dB. The last two steps add 2.5 dB each. Airtime roughly doubles with every step for the same payload.
The bottom line: sensitivity improves by a modest, fixed amount per step, while airtime grows steeply. That asymmetry drives every decision in the rest of this page.
LoRaWAN range by environment: relative distance for each spreading factor
Absolute kilometres depend on gateway height, antenna gain, transmit power, terrain and interference, so a single km figure per spreading factor would be invented. What you can state reliably is the relative range, using the standard log-distance path-loss model: distance scales by 10 to the power of (dB gain divided by 10 times the path-loss exponent).
The table uses three modelling assumptions: exponent 2 for open line of sight (free space), exponent 3 for suburban, and exponent 4 for dense urban or heavily obstructed sites. These are illustrative exponents from the standard propagation literature, not measurements of a specific site. All values are relative to SF7 with identical transmit power, antennas and 125 kHz bandwidth.
| Spreading factor | Gain vs SF7 | Open (exponent 2) | Suburban (exponent 3) | Dense urban (exponent 4) |
|---|---|---|---|---|
| SF7 | 0 dB | 1.0x | 1.0x | 1.0x |
| SF8 | 3 dB | 1.4x | 1.3x | 1.2x |
| SF9 | 6 dB | 2.0x | 1.6x | 1.4x |
| SF10 | 9 dB | 2.8x | 2.0x | 1.7x |
| SF11 | 11.5 dB | 3.8x | 2.4x | 1.9x |
| SF12 | 14 dB | 5.0x | 2.9x | 2.2x |
Who gains most, who gains least
Open terrain leads: SF12 stretches the SF7 distance by 5.0 times. Dense urban trails at 2.2 times. Suburban sits between at 2.9 times. This is the sentence worth remembering: the harder the environment, the less each extra spreading factor buys you.
A second effect matters in towns. A wall or floor slab removes a fixed number of dB. Once that loss exceeds your link margin, the sensor is simply out of reach at any spreading factor you can use. In that case the fix is a gateway closer to the sensor, not a higher setting.
Methodology and limits of these benchmarks
Sensitivity and data-rate figures come from published Semtech receiver specifications and LoRaWAN regional parameters for 125 kHz channels. The relative-range figures are calculated from those sensitivities with the log-distance model, so they are modelled values, not field measurements.
One named limitation: the model ignores fading, antenna patterns, interference and gateway noise floor. Real links at the edge of coverage fluctuate, so treat the multipliers as a planning guide and test a representative sensor on site before ordering in volume.
What a higher spreading factor costs you
Range is only half of the trade. The other half is channel time and battery.
- Airtime. Each step up roughly doubles the time a message occupies the channel. SF12 is therefore about 22 times slower than SF7 in raw data rate.
- Duty cycle. In the EU868 band most sub-bands limit a device to 1% transmit time. Long SF12 messages use that allowance quickly, which caps how often a sensor may report.
- Network fair use. The Things Network community policy asks for a maximum of 30 seconds of uplink airtime per device per day and 10 downlink messages per day. An SF12 message uses that budget far faster than SF7.
- Battery. Longer transmissions draw current for longer, so an SF12 sensor reports less often or needs a bigger battery for the same life.
- Collisions. Slow messages overlap more often with other devices, which lowers gateway capacity as the network grows.
The practical verdict: use the lowest spreading factor that still gives a reliable link. Adaptive data rate (ADR) does this automatically for fixed sensors, by moving each device to the fastest setting its signal margin allows.
How to use these LoRaWAN range benchmarks in 2026 and 2027
For projects planned through 2026 and 2027, work through the following sequence rather than guessing a setting.
- Classify each site. Open yard, suburban estate or dense urban block. Pick the matching column in the range table.
- Raise the gateway first. Gateway antenna height and a clear first Fresnel zone often add more than 14 dB of effective margin. A better outdoor antenna or a mast position can beat any spreading factor change.
- Leave ADR on for fixed sensors. Turn it off for moving trackers, where the signal changes between messages and the network cannot keep up.
- Reserve SF11 and SF12 for the edge. Plan for the majority of sensors on SF7 to SF9, with the slow settings only for a small group of difficult locations.
- Add a gateway when more than a small share of sensors need SF12. The cost of one more gateway is often lower than the lost battery life and duty-cycle headroom.
Two hypothetical examples show the logic. A farm with sensors across open fields, where the model gives SF10 about 2.8 times the SF7 distance, can sometimes cover the ground with one well-placed gateway. A block of flats in a dense city, where SF12 only gives about 2.2 times, usually needs a gateway on each side or indoor coverage.
Check the radio plan for the country where you deploy. The EU868 figures here do not apply to US915 or AS923 networks, which have different channels and rules.
Antennas, power and placement: the levers that beat spreading factor
Spreading factor changes the receiver. Everything else changes the link budget on both ends. Three levers are worth checking before you accept slower settings.
- Antenna gain. A higher-gain outdoor antenna on the gateway raises received power for every sensor at once. Gain narrows the vertical beam, so it suits flat terrain better than a tall building.
- Cable loss. Long coaxial runs between gateway and antenna lose signal. Placing the gateway close to the antenna, or powering it over Ethernet at the mast, avoids that loss.
- Sensor antenna. Small internal antennas in metal enclosures can lose several dB. A sensor with an external antenna or a better mounting position can recover it.
In short, spend effort on the physical layer first. A 3 dB improvement from placement is worth one full spreading factor step from SF7 to SF10, without the airtime cost.
FAQ
What is the LoRaWAN range by spreading factor?
SF12 reaches about 5.0 times the SF7 distance in a free-space model, about 2.9 times in a suburban model and about 2.2 times in a dense urban model. Absolute distance depends on antennas, gateway height and terrain.
Which spreading factor has the longest range?
SF12 has the longest range, with -137 dBm sensitivity at 125 kHz. It is also the slowest, at 250 bps, so each message uses much more airtime than at SF7.
Is SF7 or SF12 better for battery life?
SF7 is better for battery life. Its messages are far shorter, so the radio draws current for less time. Use SF12 only for sensors that cannot reach a gateway at a faster setting.
How much more sensitive is SF12 than SF7?
SF12 is 14 dB more sensitive than SF7: -137 dBm against -123 dBm at 125 kHz. That gain is the whole range advantage of the higher setting.
Does adaptive data rate (ADR) choose the spreading factor for me?
Yes, for fixed sensors. The network server moves each device to the fastest spreading factor its signal margin allows. Turn ADR off for moving devices, because their signal changes between messages.
Should I use SF12 for every sensor to get maximum range?
No. SF12 uses the most airtime, drains batteries faster and consumes duty-cycle allowance quickly. Use the lowest setting that gives a reliable link and add a gateway if many sensors need SF12.
Does a higher spreading factor help through walls?
It helps by 14 dB at most, but a wall or floor slab can remove a similar amount. In dense buildings a closer gateway usually works better than a higher spreading factor.
One last thing
The last two steps are not equal to the first four. SF7 to SF10 adds 3 dB per step, but SF10 to SF11 adds 2.5 dB and SF11 to SF12 adds another 2.5 dB while airtime still doubles each time. SF11 and SF12 are the most expensive decibels in LoRaWAN. If a sensor needs SF12 to be heard, treat that as a signal to move the gateway or the antenna before accepting the cost.
