What Is Cellular Signal (RSSI)?
When a remote gateway drops offline, the first thing a technician checks is signal. RSSI is the classic number for that - a measure of how strong the received cellular signal is. This guide explains RSSI, how it reads in dBm, and how it relates to the LTE metrics RSRP, RSRQ, and SINR that matter on modern networks.
Cellular Signal (RSSI) in one line: RSSI (Received Signal Strength Indicator) is a measure of the total power a cellular device receives, expressed in dBm; a less negative number (closer to zero) means a stronger signal, and it is used to gauge whether a remote site has usable connectivity.
How to Read RSSI
RSSI is reported in dBm and is always negative for cellular. As a rough guide, around -65 dBm or higher is excellent, -75 to -85 dBm is good, -85 to -95 dBm is fair, and below about -100 dBm is poor - marginal enough to cause dropped sessions and retries. The exact thresholds vary by network and hardware, but the direction is universal: closer to zero is better.
The catch is that RSSI measures total received power, including noise and interference from neighboring channels, so a high RSSI does not guarantee a clean, usable signal. That is why on LTE and 5G engineers lean on more specific metrics rather than RSSI alone.
RSSI vs RSRP, RSRQ, and SINR
On LTE, RSRP (Reference Signal Received Power) isolates the power of the network's reference signals, giving a cleaner strength figure than RSSI: roughly -80 dBm and up is strong, while below -100 dBm is weak. RSRQ (Reference Signal Received Quality) and SINR (Signal-to-Interference-plus-Noise Ratio) describe quality - how clean the signal is relative to noise. A site can show decent RSRP but poor SINR, and it will still perform badly.
For a remote SCADA site, the practical takeaway is to read both strength and quality. A gateway that reports strong RSRP but low SINR often benefits from a better antenna, repositioning, or locking to a cleaner band rather than simply amplifying the signal. Getting signal right up front prevents the intermittent, hard-to-diagnose dropouts that plague marginal sites.
Running a Survey That Predicts Real Performance
A signal survey is only useful if it measures what the installed gateway will actually experience. That means reading signal at the exact mounting position and antenna height - not at head height in the open, and not from a phone, whose radio, antenna, and carrier may all differ from the gateway's. Log the serving cell and band along with the numbers, because a modem can camp on different cells at different times, and a site that surveys well against one tower may end up running against another once traffic loads shift.
Take readings at more than one time of day if the visit allows it; congestion and even atmospheric conditions move the quality numbers. Record all four metrics - RSSI, RSRP, RSRQ, and SINR - not just the headline one, since the failure modes they reveal are different. The mechanics of doing this properly are covered in reading RSRP and RSRQ during a site survey.
Antennas and Cabling: Where Improvement Comes From
When the numbers are marginal, the instinct is to buy gain, but height and a clear path usually beat gain. Raising the antenna above the metal clutter of a site, or moving it to the side of a building facing the serving tower, often does more than any antenna swap. A directional antenna adds gain toward one tower at the cost of needing to be aimed and staying committed to that azimuth, while an omnidirectional antenna trades peak gain for indifference to which cell serves the site. LTE gateways generally want two antennas for MIMO and diversity; feeding only one port quietly gives up performance.
Cabling subtracts from whatever the antenna gains. Every meter of coax and every connector has loss, per the manufacturer's datasheet, and a long run of cheap cable can consume the entire benefit of a mast - the arithmetic is worth doing explicitly, as in budgeting antenna cable loss. Outdoor antennas also need a surge arrestor and proper grounding; a lightning path through the modem is a more expensive problem than weak signal.
Trending Signal Metrics as SCADA Tags
Most industrial gateways expose their modem diagnostics - RSSI, RSRP, RSRQ, SINR, serving cell, band - and it is worth bringing them in as ordinary tags alongside the process data. A survey is a snapshot; a trend catches what a snapshot cannot: seasonal foliage attenuation, degradation after nearby construction, an antenna connector slowly admitting water, or a modem that changed serving cell after a carrier reconfiguration.
Alarm on sustained degradation rather than instantaneous dips, with a generous deadband, because signal metrics are inherently jumpy. The real value comes at incident time: when a site drops, the trend shows whether the radio metrics deteriorated first - pointing at the RF path - or stayed healthy while the connection died, pointing at the SIM, the carrier, or the backhaul beyond the tower. That one distinction saves truck rolls.
When the Numbers Are Good and the Site Still Drops
Healthy signal metrics with an unhealthy connection move suspicion up the stack: a SIM or data plan in the wrong state, an APN misconfiguration, carrier-side session teardown, network address translation timing out idle sessions, tower congestion at particular hours, or gateway firmware trouble. Each leaves a different fingerprint - disconnects at regular intervals suggest a timer, time-of-day patterns suggest congestion, failures only under load suggest something else again. A structured walk through these causes is in diagnosing a gateway that keeps dropping.
Before escalating to the carrier, capture evidence: timestamps of the drops, the serving cell identifiers, and the metric trends around each event. A ticket that says the site drops daily at shift change on a named cell with healthy RSRP and collapsing SINR gets a categorically different response from one that says the site keeps going offline. If the root cause is congestion or a misprovisioned plan, only the carrier can fix it - the site evidence just determines how fast.
Frequently Asked Questions
What is a good RSSI for a cellular SCADA site?
As a rule of thumb, around -65 dBm or higher is excellent, -75 to -85 dBm is good, and below about -100 dBm is poor. Closer to zero is stronger. On LTE, also check RSRP and SINR, since RSSI alone does not reflect signal quality.
What is the difference between RSSI and RSRP?
RSSI is the total received power including noise and interference, while RSRP measures only the power of the LTE reference signals, giving a cleaner strength reading. On modern networks RSRP is the more reliable indicator of coverage.
Why does my site have strong signal but still drops out?
Strong RSSI or RSRP shows the signal is present but not that it is clean. Poor RSRQ or SINR - from interference or noise - can cause retries and dropped sessions despite a strong reading, often fixed with a better antenna or a cleaner band.
Do the signal bars on a gateway or phone mean anything for SCADA?
Very little. Bars are a device-specific rendering of one metric or another with arbitrary thresholds, and two devices can show different bars on the same signal. Read the actual numbers - RSSI, RSRP, RSRQ, SINR - from the modem diagnostics instead, and trend them over time.
Will a cellular booster fix a marginal SCADA site?
Sometimes, but fix the antenna system first - height, placement, cabling, and both MIMO ports connected. Boosters amplify noise along with signal, so they help least at interference-limited sites where SINR is the real problem. Whether one is appropriate is site-specific; follow the manufacturer's guidance and local rules on approved equipment.
Automation services
Need help turning this into a working system?
Merobix integrates SCADA, programs Allen-Bradley and Siemens PLCs, and designs and fabricates industrial control panels.
Meeting requests are reviewed before confirmation.