RSSI, SNR and Packet Loss Explained

A practical interpretation of received signal strength, signal quality and delivery reliability in Meshtastic.

RSSI is received power

RSSI is the radio’s estimate of the power present in the received LoRa signal, expressed in dBm. Values are negative; a value closer to zero is stronger. Therefore −80 dBm is stronger than −110 dBm. RSSI describes one packet at one receiving radio. It includes the consequences of transmit power, antenna gain, feed-line losses, distance, obstructions and receiver installation.

RSSI is not a universal pass/fail score. Receiver sensitivity depends on the LoRa modem parameters, and a strong unwanted signal or local noise can still make reception poor.

SNR is signal relative to noise

SNR compares the wanted signal with the local noise level, in decibels. Positive SNR means the signal is above the measured noise; negative SNR means it is below it. LoRa can decode packets at negative SNR, so a negative value is not automatically a failure. Higher SNR generally means more margin and a clearer link.

Two sites can report similar RSSI but very different SNR because one has a noisier environment. A rooftop near electronics, switching power supplies or other transmitters may show disappointing SNR despite apparently strong received power.

Packet loss measures service

Packet loss is the proportion of expected packets that were not successfully received. If 100 numbered packets are transmitted and 83 are recorded, the observed loss in that direction is 17 percent. This requires knowledge of what was actually sent. Counting only received packets cannot reveal the missing ones.

Loss includes more than weak RF. Collisions, duty-cycle restrictions, queueing, relay contention, busy-channel behavior, resets and power failures can all remove packets. End-to-end delivery through several hops also combines the reliability of every hop.

Read the three together

Observation	Likely interpretation	Next check
Strong RSSI, poor SNR	High local noise or interference	Compare noise at another site/time and inspect electronics
Weak RSSI, usable SNR	Quiet path near the link margin	Repeat in changing conditions
Good RSSI/SNR, high loss	Congestion, collisions, configuration or power issue	Inspect airtime, logs and sequence timing
Metrics only one way	Asymmetric path or test method	Compare antennas, power and receivers

Beware of relays and samples

The displayed RSSI and SNR normally describe reception of a particular packet at a particular receiver or hop. They are not necessarily the quality of the complete path back to the original sender. Confirm whether the packet was direct, and use traceroute or controlled topology where appropriate.

A single excellent packet does not establish a stable link. Record a distribution over time: median, weaker tail, delivery ratio and conditions. For infrastructure acceptance, margin and repeatability matter more than the best number seen.

Use metrics to guide action

If RSSI is consistently weak, improve line of sight, antenna placement, connectors or feed-line loss. If RSSI is reasonable but SNR poor, investigate noise and interference. If both look healthy but delivery is unreliable, inspect airtime, traffic, hop behavior, firmware stability and power. Change one factor at a time and repeat the same test.

Link margin is more useful than a magic number

Operators often ask for a single “good RSSI” threshold. That ignores modem sensitivity, noise, packet size and the stability required. Think in terms of margin: how far the received conditions are from failure, including the weaker observations, not only the average. A link that works at midday with little margin may disappear after rain, foliage growth, a vehicle moving or a new noise source switching on.

Compare measurements only when the preset and receiving system are known. Values from a different radio, antenna or modem preset do not form a controlled benchmark.

Measure each direction and each hop

Radio links are paired receiver observations. A reports the quality of packets received from B; B separately reports packets from A. These values need not match. On a relayed path, each reception has its own metrics. End-to-end success can be poor even if one displayed hop looks excellent, because another hop or the return path is weak.

Use sequence-based delivery tests alongside metrics. RSSI and SNR exist only for packets that were decoded; the missing packets have no reported measurement and may represent the worst conditions.

Related guides

Official sources

Firmware, applications and documentation change over time. Confirm current setting names and applicable regional rules before commissioning or changing live infrastructure.