T ThinkEdge Learn

LoRa and LoRaWAN: The Physics of Long Range, Low Power Radio

LoRa trades speed for range and range for battery life, in numbers you can calculate. Learn the trade-off, then decide if it beats Wi-Fi for your project.

lora lorawan lora module lora vs wifi
Maker Hardware and IoT · Uses real link budgets and airtimes instead of marketing range claims

LoRa is the radio, LoRaWAN is the network

LoRa is a physical layer: a modulation called chirp spread spectrum. The transmitter sweeps its carrier across the band in a fixed sequence of steps, and the receiver un-sweeps it. Because the energy is spread over a wide band, the receiver can pick the signal out well below the background noise level, which is what buys the range.

The knob that matters is the spreading factor, from SF7 to SF12. Each doubling of the spreading factor makes the receiver integrate the signal for twice as long, gaining about 3 dB of sensitivity, but it also makes the packet take roughly twice as long to send. In the Indian 865 to 867 MHz band with the standard 125 kHz channels, the rates run from 5470 bit/s at SF7 down to 250 bit/s at SF12, and a 51-byte application payload at SF12 takes roughly 2.8 seconds on air, because the LoRaWAN header adds a further 13 bytes. The consequence is a hard ceiling on what you can send. A LoRa node is a telemetry device, not a video link. One reading every few minutes, or a short status message, fits comfortably. Anything continuous does not, and no amount of antenna work will change that, because the limit is time on air rather than signal strength.

Gateway, network server and the rules you have to obey

LoRaWAN adds the network on top. End devices send to a gateway, which is just a listener that forwards every packet it hears to a network server. The server decides what to do with it. End devices join the network with an over-the-air activation, which gives each device its own session keys, rather than a static one shared by many.

Devices are also classified by how they listen. A Class A device is asleep, sends an uplink, and opens two short receive windows afterwards, one second and two seconds later. A Class B device adds scheduled ping slots on a GPS-synchronised schedule. A Class C device keeps its receiver on and can be commanded at any time, at the cost of power. Battery nodes are almost always Class A. The rules differ by region and they matter. In the Indian 865 to 867 MHz band, the regional parameters specify three mandatory default channels at 865.0625, 865.4025 and 865.985 MHz, and the PHY layer there carries no dwell-time or duty-cycle limitation. In Europe the 868 MHz band runs on a 1 percent duty cycle with 10 mW ERP, which is a very different constraint when you are planning how often a node can report. In the United States the band is 902 to 928 MHz. Know which band you are on before you buy a module.

Step by step

  1. Calculate the link budget before buying anything

    Add the transmit power to the antenna gain, subtract the receiver sensitivity and the path loss, and see what margin you have. This is a page of arithmetic that saves a week of range testing.

  2. Match the antenna to the band

    A 433 MHz antenna on an 868 MHz module will not simply be wrong, it will be badly wrong. Check the band of every module and antenna you buy, and write it on the box.

  3. Transmit a fixed packet and measure the airtime

    Send a 20-byte payload and time it with the serial output. Seeing the airtime grow as you raise the spreading factor makes the trade-off concrete.

  4. Compare spreading factors in the real world

    Measure the range at SF7 and at SF12. The SF12 test will reach further, and the difference in how long the packets take is worth measuring too.

  5. Join a network and publish a reading

    Activate a device over the air, subscribe to its topic and watch a sensor reading arrive on a screen somewhere else.

  6. Work out the battery life from the numbers

    Take the transmit current, the transmit time and the sleep current, and calculate the average current. Then check that calculation against a real battery for a week.

What you'll need

ItemQtyWhy
SX1276 LoRa module, 868 MHz or 915 MHz version2The transceiver that turns serial bytes into LoRa packets. Buy the band you are actually allowed to use.
Spring antenna with SMA connector2A quarter-wave antenna for the band, with a matching cable, is worth more than any change to the code.
Arduino Nano or Uno R32Two nodes let you test a real link rather than a loopback to the same board.
DHT22 temperature and humidity sensor1One small reading per node is the right payload size for a long-range link.
Breadboard and jumper wires1 setSPI connections to the radio need a clean, short path and good common ground.
18650 cell and protected holder2Long-range testing takes many transmission attempts, and a flat battery ends the experiment early.

Practise it before you build it

Every step above has a matching circuit in the ThinkEdge Learn simulator. The simulator runs a real Arduino sketch against a modified-nodal-analysis solver, so digitalWrite actually drives the LED on screen and the readings change when you change the wiring — no hardware, no burnt components.

Common questions

How far can LoRa actually reach?

With a matched antenna and clear line of sight, a 14 dBm node at SF12 can be heard several kilometres away in a rural area and about a kilometre or two through a town. The same hardware at SF7 might manage 20 to 30 percent of that distance and send data six times faster. Walls, hills and badly matched antennas take more off than any datasheet figure suggests.

Is LoRa better than Wi-Fi for a classroom project?

For anything inside one building, Wi-Fi wins on bandwidth and simplicity. LoRa wins when the nodes are spread out, run on a battery for months, or only need to send a number every few minutes. The honest comparison is a Wi-Fi node that sends a reading every minute against a LoRa node that sends the same reading every fifteen minutes on the same battery.

How much power does a LoRa node use?

An SX1276 transmitting at 14 dBm draws roughly 45 mA for the length of the packet, which is about 2.8 seconds at SF12 for a 51-byte payload. Between transmissions it can sleep at a few microamps. A node that reports every fifteen minutes therefore spends a tiny fraction of its life awake, which is why the battery maths works out in years rather than days.

Related guides

ESP32 IoT Projects: From a Single Sensor to a Live DashboardFive ESP32 IoT projects that teach sensors, MQTT publishing, deep sleep and reconnection, each staged so the logic runs in a browser simulator first. Mesh Networking: How Nodes Relay Data for Each OtherA mesh network lets nodes relay data for each other. How flooding and hop limits work, what mesh costs in latency, and when a simple star is better. ESP32 Microcontroller: The One With the Radio OnboardThe ESP32 is a 3.3 V microcontroller with Wi-Fi and Bluetooth built in. What it really is, how it differs from an Arduino Uno, and where it fits in IoT builds. Home Energy MonitoringHow to monitor energy use in a home or classroom: clamp meters, energy monitoring modules, the difference between watts and watt-hours, and what the data shows.

Try the full simulator with your class

103 guided circuit projects, 21 lessons and an in-browser Arduino simulator. The first 15 projects are free, no account needed. NCERT Science mapping on 47 projects, with Arabic and English copy in the regional builds.