An RF link budget is the accounting that tells you, before you climb a single tower, whether a radio link will actually work. It sums every gain and subtracts every loss between the transmitter and the receiver, then compares the result against how weak a signal the receiver can still decode. If the arriving signal is comfortably above that floor, the link closes; if it is not, the link is a coin toss. This guide walks through the terms in the budget and how to leave enough headroom.
RF Link Budget in one line: An RF link budget is a calculation, done in decibels, that adds the transmitter power and antenna gains and subtracts the feedline losses and free-space path loss to find the signal strength arriving at the far receiver, then checks that this arriving level exceeds the receiver's sensitivity by a comfortable margin. A link only closes reliably when the received power beats the sensitivity threshold with a fade margin of roughly 20 to 30 dB left over to absorb weather, foliage, and multipath.
Everything in a link budget is expressed in decibels so the whole chain can be added and subtracted rather than multiplied. On the transmit side you start with the radio's output power in dBm, add the transmit antenna gain in dBi, and subtract the loss of the coaxial feedline and connectors between the radio and that antenna. The sum of these is the effective radiated power leaving the transmit antenna.
Between the two antennas sits free-space path loss, the single largest number in most budgets. It grows with both distance and frequency - double the distance and the loss rises, and a higher-frequency link loses more over the same span than a lower-frequency one. Real paths add further losses on top of the ideal free-space figure: partial obstruction of the Fresnel zone, foliage, and diffraction over terrain all subtract more decibels.
On the receive side you add the receive antenna gain and subtract that side's feedline and connector loss. What remains is the received signal level in dBm at the radio's input. The reference point you compare it against is the receiver sensitivity - the weakest signal, also in dBm and always a negative number, at which the radio can still decode data at an acceptable error rate.
The link is said to close when the received signal level is stronger than the receiver sensitivity. But equality is not enough. RF paths are not static: rain, humidity, moving foliage, temperature-driven ducting, and multipath reflections all cause the received level to fade up and down over time. If you design so the signal only just clears sensitivity on a good day, it will drop below the threshold and the link will fail on a bad one.
The cushion between the received level and the sensitivity is the fade margin, and it is the number that separates a link that works in a demo from one that works in a thunderstorm two years later. A common design target is roughly 20 to 30 dB of fade margin for a link expected to stay up through weather and seasonal change. Getting there usually means raising antennas for better clearance, choosing higher-gain antennas, using lower-loss coax, or, when the path is simply too long or too obstructed, inserting a repeater to split it into two shorter hops.
The budget is iterative. You compute it, find the margin is short, change one term - a taller mast, a Yagi instead of an omni, LMR-400 instead of thin coax - recompute, and repeat until the margin is healthy. Doing this on paper is vastly cheaper than discovering a marginal link after the crew has demobilized.
A SCADA network is only as trustworthy as its weakest link. When a remote site drops off the poll list, an operator sees a stale value or a comm-fail alarm and loses visibility of a well, a tank, or a compressor. Many of those dropouts trace back not to a broken radio but to a link that was designed with too little margin and now fades below threshold whenever conditions turn. A disciplined link budget at design time is the cheapest reliability investment in the whole telemetry chain.
The link budget also sets realistic expectations for what a site can support. A path with generous margin can tolerate a faster poll cycle and more remotes sharing the channel; a marginal path may need slower, more patient polling and aggressive retransmission just to stay connected. Understanding the budget tells the SCADA engineer which sites are solid and which are fragile before they ever go live.
A cloud SCADA platform such as Merobix surfaces the symptoms of a weak budget - rising retransmission counts, intermittent comm failures, and RSSI values creeping toward the noise floor - as data an operator can trend. Seeing a link's received signal drift down over a season is often the first sign that the fade margin was thin to begin with, and that a taller antenna or a repeater is overdue.
Work in decibels. Take the transmitter output power in dBm, add the transmit and receive antenna gains in dBi, and subtract the feedline and connector losses on both ends plus the free-space path loss and any obstruction losses. The result is the received signal level in dBm, which you then compare against the receiver's sensitivity to find your fade margin.
A common design target is roughly 20 to 30 dB of margin above the receiver's sensitivity for a link expected to stay reliable through weather and seasonal change. Less margin than that risks dropouts during rain, heavy foliage, or multipath fades. If you cannot reach a healthy margin, the usual fixes are higher antennas, higher-gain antennas, lower-loss coax, or a repeater.
Free-space path loss is almost always the largest single term. It increases with both distance and frequency, so longer links and higher bands lose more signal between the antennas. Feedline loss and partial Fresnel-zone obstruction add to it, but the raw spreading loss over the path usually dominates the budget.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.