Antennas for LoRa networks: choose the pattern, not just the biggest number

September 29, 2026

Try the idea

Coverage depends on the pattern

A conceptual comparison, not a measured radiation plot. Gain redistributes coverage; orientation and the actual antenna still matter.

Teaching model only. No radio transmission, measured coverage or delivery guarantee.

An antenna is part of a complete radio installation. Its frequency range, radiation pattern, polarisation, connector, cable and mounting position all matter. Buying the antenna with the largest advertised gain can make a particular link worse if the energy is directed away from the people you need to reach. This guide helps you choose deliberately and test the result.

Begin with the frequency and the radio port

Use an antenna specified for the operating band of the radio. Similar-looking antennas can serve completely different frequencies. A LoRa board can also have separate connectors for LoRa, Wi-Fi or GPS; attaching the right-looking antenna to the wrong port does not help. Check the exact board documentation and antenna specification together.

Most relevant radio installations use a nominal 50-ohm RF system, but confirm the components rather than treating this as a universal rule. Antenna, feed line and radio should be designed to work together. Adapters solve mechanical connection problems; they do not convert an antenna to another frequency band. RAKwireless’s antenna overview explains impedance and polarisation.

Understand the common antenna types

Type	Useful application	Important limitation
Compact whip or rubber-covered antenna	Portable nodes and simple installations	Performance depends on its actual construction and surroundings
Monopole with a suitable ground plane	Vehicles or a deliberately designed base	The mounting surface or counterpoise is part of the antenna system
Dipole	A simple reference or fixed installation	Feed arrangement, orientation and nearby objects matter
Collinear omnidirectional antenna	Coverage around a fixed site	Greater gain can narrow the vertical pattern
Directional panel or Yagi	A known point-to-point direction	Needs alignment; gives less coverage elsewhere

“Omnidirectional” normally describes coverage around the antenna in a plane, not a perfect sphere. Real antennas have lobes and nulls. A tall antenna intended to cover the horizon may be a poor answer for users almost directly underneath it. Read a radiation-pattern plot when the manufacturer provides one.

Illustration: whip, dipole, directional antenna and coax. Choose for frequency, radiation pattern, connectors and placement.

Gain and placement solve different problems

Gain in dBi compares an antenna’s strongest direction with an ideal isotropic reference. It does not mean amplification supplied by a passive antenna. Picture squeezing a balloon: changing the shape sends more of the pattern in some directions and less in others. The illustration is conceptual; real patterns depend on the antenna.

For a moving user, a forgiving pattern may be more useful than maximum peak gain. For a fixed link between two known points, a directional antenna can be appropriate. For a hilly area, avoid assuming that a narrow vertical pattern covers every elevation. The RAK antenna-selection guide is a useful manufacturer reference.

Do the cable arithmetic before buying

Antenna cable loses power in both transmit and receive paths. Loss depends on cable type, length and frequency, and connectors add their own contribution. Use the manufacturer’s loss figure at your operating frequency. A thick cable’s advantage is irrelevant if the connector cannot be mounted safely on your small board.

As an illustrative transmit-side budget, 14 dBm at the radio minus 2 dB of feed-line loss plus 3 dBi antenna gain gives 15 dBm EIRP. This is arithmetic, not an approved transmit setting. A 3 dB loss corresponds approximately to half the power. Increasing antenna gain may change whether the installation meets the applicable radiated-power limit, so check the complete installation against local rules.

Avoid an unnecessarily long thin pigtail. Sometimes placing the radio near the antenna and extending an appropriate power/data connection is better than extending RF cable. That choice introduces enclosure, access and weather-protection requirements; evaluate the whole installation.

Check connector details and strain relief

SMA and RP-SMA are not interchangeable names. Check the centre contact as well as the threaded shell. A connector can appear to screw together while failing to make the intended RF contact. Small board connectors such as u.FL/IPEX-style connections are delicate and are not handles for supporting a heavy cable.

Use a suitable pigtail and mechanical strain relief. Follow the connector manufacturer’s mating instructions, avoid repeated unnecessary disconnection and protect the board from cable leverage. For outdoor installations, consider water ingress, UV exposure and safe mounting. Mast work, grounding and lightning protection require appropriate expertise; a plastic box alone does not address them.

Polarisation and the installation environment

Keep the intended polarisation compatible at both ends. Rotating one antenna can change the received signal, particularly on a clear direct path. In reflective environments the result may be more complicated, which is another reason to test at the real location.

Metal cabinets, batteries, your body and even a mounting bracket can alter the installed behaviour. Test the antenna in the enclosure and orientation you will actually use. If a handheld node works on a windowsill but fails in a pocket, that is useful installation evidence, not necessarily a firmware defect.

What a VNA can and cannot prove

A vector network analyser can help examine impedance matching over frequency when correctly calibrated at the measurement reference plane. Return loss or VSWR can reveal that an antenna is poorly matched in the intended band. The cable, adapter and how you hold the assembly can affect the measurement.

A good match is not proof of good radiation efficiency or field coverage. A dummy load can present a good match without being a useful antenna. Combine matching measurements with manufacturer data and controlled over-the-air tests. See Meshtastic’s antenna-testing resources.

A repeatable comparison

Keep the same two nodes, firmware, preset, power, message length and test locations.
Establish a reference with the existing antenna and record messages attempted and received.
Change only one antenna. Keep its mounting and orientation documented.
Repeat in both directions; compare delivery and delay alongside RSSI/SNR.
Repeat at difficult locations and after fitting the final enclosure.
Retest after changing cable or mounting. Those are changes to the RF system too.

The longest successful message is an interesting observation, not a reliable coverage boundary. Select the arrangement that meets your actual route and maintenance needs with useful margin.

Continue with LoRa and airtime, Meshtastic network design or Reticulum interfaces. When asking eBits for help, include the exact board, radio band, connector photographs, cable length and intended mounting. We can help narrow the choice without pretending that a product photograph predicts your coverage.

Reviewed 28 September 2026. Illustrations and calculations are educational, not measured antenna patterns or eBits field-test results.

The radio workshop

Understand the radio. Build the network.

LoRa, Meshtastic and Reticulum: learn what each layer does, choose suitable parts and test your own links. Start with a learning path, then explore the electronics underneath.