Skip to main content

LoRa Technology Explained

LoRa Technology Explained

LoRa stands for Long Range. It is a wireless radio modulation technique invented by a French startup called Cycleo and acquired by Semtech in 2012. LoRa chips appear in millions of devices worldwide today, from smart utility meters to wildlife trackers to community mesh nodes. To understand why LoRa is so effective for mesh networking, you need to understand the core technique it uses: Chirp Spread Spectrum.

What Is Chirp Spread Spectrum?

Most radio systems transmit data on one fixed frequency. A chirp spread spectrum (CSS) system does something different: it sweeps the signal continuously across a range of frequencies, rising or falling in pitch over time (like a bird's chirp — hence the name). The receiver looks for this specific sweep pattern rather than listening at one spot.

Why does this matter? Several reasons:

  • Noise rejection. Random noise and interference tend to be concentrated at specific frequencies. Because a CSS signal is spread across a wide band, any noise at one frequency only affects a small fraction of the signal. The receiver can mathematically reconstruct the original data even when most of the band is noisy.
  • Multipath resistance. Radio signals bounce off buildings, hills, and trees. These echoes arrive at the receiver slightly delayed and can cancel out the original signal. CSS is much more resilient to this effect than narrowband modulation.
  • Low detectability. A spread-spectrum signal looks almost like background noise to anyone not specifically looking for it, making it robust in RF-noisy environments.

Processing Gain and Extraordinary Sensitivity

The key metric for a radio receiver is its sensitivity — the minimum signal power it can decode. A conventional FM radio might have a sensitivity of around –100 dBm (decibel-milliwatts). LoRa achieves sensitivities as low as –148 dBm with its highest settings. That is almost 50 dB better than FM, which means LoRa can receive signals that are more than 100,000 times weaker.

This extraordinary sensitivity comes from processing gain: the mathematical process of correlating the received sweep against an expected template. The more time and bandwidth the receiver spends correlating, the more gain it recovers. LoRa lets you tune this trade-off with a parameter called the Spreading Factor.

What Spreading Factor Actually Does

The Spreading Factor (SF) is a number from 6 to 12. It controls how many "chips" (individual frequency steps) are used to encode each bit of data.

  • Low SF (e.g., SF7): fewer chips per bit → faster data rate → shorter range → lower battery use per packet. Used when nodes are close together.
  • High SF (e.g., SF12): more chips per bit → very slow data rate → very long range → more battery use per packet. Used when you need maximum range.

Think of it like this: shouting a word once at normal speed (SF7) versus repeating every syllable ten times very slowly (SF12). The slowly repeated version can be understood even with a lot of background noise, but it takes much longer to say.

A common community mesh preset uses SF10 or SF11, which balances range and throughput for typical text messaging workloads. Meshtastic's default "LongFast" channel uses SF11 on 915 MHz.

The Range / Speed / Power Triangle

In radio, you almost never get something for nothing. LoRa is no exception. The three variables — range, data rate, and power consumption — are always in tension:

  • To go farther (more range), you must slow down (lower data rate) or transmit louder (more power).
  • To go faster (higher data rate), you must accept shorter range or higher power.
  • To save power, you must accept either shorter range or slower speed.

LoRa's key achievement is that it pushes this triangle to extremes: it can achieve very long range at very low power, but only by accepting a very slow data rate. A LoRa packet might carry 50–250 bytes of useful data. The raw over-the-air data rate at SF12 is about 250 bits per second. That is slower than a 1990s dial-up modem. But for text messages, GPS coordinates, and sensor readings, it is entirely sufficient — and nothing else achieves that range at that power level.

LoRa vs. FSK vs. GFSK

To appreciate LoRa, compare it with the modulation techniques used by competing low-power radios:

Modulation How it works Typical sensitivity Typical range (open field) Used in
FSK (Frequency Shift Keying) Switches between two fixed frequencies for 0 and 1 –112 dBm ~1–2 km Many 433/915 MHz modules, older APRS
GFSK (Gaussian FSK) FSK with a Gaussian filter to reduce bandwidth –105 to –115 dBm ~1–3 km Bluetooth, Zigbee, older Nordic radios
LoRa (CSS) Frequency sweep across entire channel bandwidth –137 to –148 dBm 5–15+ km Meshtastic, MeshCore, LoRaWAN

The ~30–40 dB sensitivity advantage translates to LoRa reaching 30–100× the distance of FSK at the same transmit power. This is why LoRa took over the low-power, long-range wireless market.

  Transmit power:        +30 dBm   (1 watt)
  Transmit antenna gain: + 2 dBi
  Path loss (10 km):     -120 dB   (approximate for 915 MHz open field)
  Receive antenna gain:  + 2 dBi
  -----------------------------------------------
  Received signal:       -86 dBm

  LoRa sensitivity (SF11): -137 dBm
  Link margin:           +51 dB    (plenty of headroom)

LoRa Is Not LoRaWAN

One important clarification: LoRa is just the physical-layer radio modulation. It says nothing about how multiple devices share the channel, how data is addressed, or how the network is organized. LoRaWAN is one specific network protocol built on top of LoRa — but it is not the only one. Meshtastic and MeshCore are entirely different protocols, also built on LoRa. The next page compares all four of these in detail.

No comments to display

Back to top