Free Space Path-loss & RSSI Calculator

{{ formatDbm(rxPower, 1) }} dBm

{{ verdict.badgeLabel }} {{ formatFrequencyLabel(freqMHz) }} {{ formatDistanceFromKm(distanceKm, distUnit) }} FSPL {{ formatDb(fspl, 1) }} dB EIRP {{ formatDbm(eirp_dBm, 1) }} dBm {{ reserveBadge }}

Preset:

Choose Custom to keep manual values, or pick a LoRa, telemetry, Wi-Fi, or bridge baseline.

Frequency:

Use the actual channel center, for example 868 MHz, 915 MHz, 2.412 GHz, or 5.8 GHz.

Distance:

Enter the antenna-to-antenna path in meters or kilometers; choose m for short links such as 50 m.

TX power:

Enter transmitter power in dBm, for example 14 for LoRa or 20 for Wi-Fi.

dBm

TX antenna gain:

Use antenna gain in dBi from the data sheet; negative values can model inefficient antennas.

dBi

RX antenna gain:

Enter receive antenna gain in dBi; use 0 for a small omnidirectional whip.

dBi

Additional path loss:

Use dB penalties from surveys or estimates, for example 8 for indoor wall loss.

RX sensitivity:

Enter receiver sensitivity in dBm, usually negative, such as -123 or -72.

dBm

Target RSSI:

Optional dBm floor for deployment quality, for example -120 for LoRa or -67 for Wi-Fi.

dBm

Fade reserve:

Enter 0-60 dB of cushion; 10 dB is the default planning reserve.

Uses free-space attenuation by default. Put wall loss, foliage, rain fade, body loss, or other non-line-of-sight penalties into Additional path loss.

TX cable loss:

Enter total transmit-side insertion loss in dB; 0 means no modeled feed loss.

RX cable loss:

Enter receive-side insertion loss in dB between antenna and radio; 0 disables it.

Polarization mismatch:

Enter mismatch loss in dB; keep 0 for aligned antennas.

Noise figure:

Enter receiver NF in dB; 5-7 dB is typical for many radio front ends.

Noise bandwidth:

Enter channel bandwidth and unit, for example 125 kHz for LoRa or 20 MHz for Wi-Fi.

Section	Signal	Value	Detail
Planning snapshot	{{ tile.label }}	{{ tile.value }}	{{ tile.detail }}
Verdict	{{ verdict.title }}	Link status	{{ verdict.detail }}
What to do next	{{ item.title }}	{{ item.badge }}	{{ item.detail }}
Model assumption	Assumption {{ idx + 1 }}	Input model	{{ note }}

Stage	Delta	Running	Why it matters	Copy
{{ row.label }}	{{ row.deltaDisplay }}	{{ row.runningDisplay }}	{{ row.note }}

Predicted RSSI is swept over distance. If you set both sensitivity and target RSSI, the stricter floor is used for planning reserve.

Comparison keeps distance, antenna gains, and all entered losses fixed so you can see what the band choice alone is doing to receive power.

Reference floor: {{ referenceFloorLabel }}. Values below 0 dB mean reserve is available; values above 0 dB show how much extra system gain is still needed.

Export to PDF Fullscreen

Embed:

Customize

Include current inputs

Size

Advanced

Width

Height

Aspect ratio

Max height

Collapsible embed

Allow fullscreen

Referrer policy

Sandbox tokens

Free-space path loss is the baseline loss a radio signal suffers simply because its energy spreads as distance grows. For the same path length, higher frequencies lose more than lower ones, which is why an 868 MHz telemetry hop and a 5.8 GHz bridge can behave very differently before trees, walls, rain, or interference enter the picture.

A useful link budget turns that baseline into a planning answer. It combines transmit power, antenna gains, feed losses, extra path penalties, and receive thresholds, then asks whether the remaining signal still clears the floor that matters for the link. On this page that floor can come from a preferred receive target, bare sensitivity plus fade reserve, or whichever of those two is stricter.

Diagram showing transmit budget, path penalties, predicted receive level, and the planning floor used to decide reserve or shortfall.

That makes the result more useful than a bare loss number. The page estimates receive power, shows the transmit budget as EIRP, works out a Planning floor, and then reports whether the path has real reserve, only clears the bare threshold, or falls short. It also exposes Fresnel midpoint clearance, a thermal-noise estimate from bandwidth and noise figure, a planned free-space range, and charts that make the margin easier to see over distance or across common bands.

The strongest fit is early planning for line-of-sight or near-line-of-sight links such as LoRa field nodes, unlicensed telemetry paths, indoor Wi-Fi checks, and short outdoor bridges. It is much weaker when the path is dominated by clutter, diffraction, or interference that you cannot estimate, because the baseline remains optimistic until those penalties are added explicitly.

Technical Details:

Point-to-point radio planning usually starts with free-space basic transmission loss, the reference loss between isotropic antennas in free space. In practical units that loss rises with 20 log10(distance) and 20 log10(frequency), so doubling the path length costs about 6 dB and doubling the frequency costs about 6 dB. That is why lower bands often buy noticeable headroom when everything else in the link budget stays fixed.

Receive power is only one part of the decision. A link can clear bare sensitivity and still be too fragile for field use, which is why planners often work with a higher floor than the demodulation limit alone. Here the governing floor is the higher of the chosen target receive level and sensitivity plus fade reserve. That pushes the result away from the narrow question of simple decode toward the more practical question of whether enough signal is left after reasonable operating margin.

RF link budget symbols used in the formulas
Symbol	Page field	Meaning
`f`	`Frequency`	Carrier frequency, expressed in MHz for the main loss equation
`d`	`Distance`	Path length, expressed in km for the main loss equation
`Ptx`	`TX power`	Conducted transmitter output before antenna gain and feed loss
`Gtx`, `Grx`	`TX antenna gain`, `RX antenna gain`	Antenna gains referenced to isotropic
`Ltx`, `Lrx`	`TX cable loss`, `RX cable loss`	Feedline, connector, or inline hardware loss on each side
`Ladd`, `Lpol`	`Additional path loss`, `Polarization mismatch`	Non-free-space penalties such as walls, foliage, rain fade, or polarization loss
`Pf`	`Planning floor`	The stricter receive threshold used for reserve and range decisions

The main equations are short, but they carry the whole planning model. The first gives free-space loss, the second predicts receive power after gains and losses, the third defines the planning floor, and the fourth adds the thermal-noise estimate used by Noise / SNR.

L_{FS} = 32.44 + 20 \times \log_{10} (d_{km}) + 20 \times \log_{10} (f_{MHz})

P_{rx} = P_{tx} + G_{tx} - L_{tx} + G_{rx} - L_{rx} - L_{add} - L_{pol} - L_{FS}

P_{f} = \max (P_{target}, P_{sens} + R_{fade})

N = - 174 + 10 \times \log_{10} (B_{Hz}) + NF

The page also exposes two geometry checks that matter in real links. First Fresnel-zone radius grows with wavelength and path length, so lower bands and longer hops need a wider obstacle-free bulge around the direct line even when the antennas can still see each other. ITU-R guidance treats about 60% of the first Fresnel zone as the practical clearance target for free-space conditions, which is why Fresnel midpoint is worth checking on longer paths.

Range and verdict outputs are just different views of the same budget. Planned max range solves the main loss equation in reverse for the chosen floor. Available reserve is Predicted receive power - Planning floor. Receiver margin is still useful, but it answers a narrower question because it only compares receive power with bare sensitivity.

Verdict logic used by the path loss and RSSI calculator
Verdict outcome	Rule used here	What it means
`Model only`	No `Planning floor` is defined	The page can still predict receive power and loss, but it cannot judge reserve or planned range
`Robust`	`Available reserve >= 20 dB`	The path has substantial modeled headroom above the chosen floor
`Healthy`	`10 dB <= Available reserve < 20 dB`	The link looks workable with meaningful reserve, but not lavish reserve
`On plan`	`0 dB <= Available reserve < 10 dB`	The path clears the floor, but only by a slim cushion
`Short reserve`	`Available reserve < 0 dB` and `Receiver margin >= 0 dB`	The receiver may still decode, but the stricter planning target is not met
`Below floor`	`Receiver margin < 0 dB` once sensitivity is known	The current budget does not clear the chosen receive floor at all

Those equations do not cover every propagation effect. ITU-R line-of-sight guidance also calls out diffraction, multipath, rain attenuation, and gaseous loss as real contributors, especially at higher microwave bands and on long paths. This page is therefore best read as a disciplined free-space baseline plus any penalties you explicitly add under Additional path loss, not as a full propagation simulator.

Everyday Use & Decision Guide:

If you already know the rough link type, start with a Preset and read the first verdict before touching anything advanced. If the path is custom, enter Frequency, Distance, and TX power first, then add antenna gains only when you know them. That gives you the clean free-space baseline before extra assumptions start stacking up.

Most bad outcomes here come from one of three things: a unit slip, a missing threshold, or a free-space model that was never given the real penalties. A distance entered in metres instead of kilometres can move the answer by tens of decibels. Leaving both RX sensitivity and Target RSSI blank produces a receive estimate but no reserve judgment. Leaving Additional path loss at zero on a cluttered path usually makes the result look safer than it is.

Open Advanced for TX cable loss, RX cable loss, Additional path loss, and Polarization mismatch. Budget Ladder is the fastest place to see whether one bad assumption is hiding in the path.
Enter RX sensitivity as soon as you know the receiver limit. Add Target RSSI only when you need a stricter service target than bare decode, then set Fade reserve to match how conservative you want to be.
Use Band Comparison when the decision is mainly about 433 MHz, 868 MHz, 915 MHz, 2.4 GHz, 5 GHz, 5.8 GHz, or 6 GHz under the same geometry. Use Gain Gap Map when the real question is how many dB you still need.
If Noise / SNR says Not modeled, Noise bandwidth is still zero or invalid. A positive bandwidth plus a reasonable Noise figure is required before the thermal-noise view means anything.

This page is a good fit for open or mostly open paths where you can defend the gains, losses, and thresholds you enter. It is a weak fit for deep indoor coverage, blocked terrain, or heavy interference problems unless you can translate those penalties into realistic extra loss. Before trusting a positive verdict, check that Planning floor comes from the source you intended and that the added loss terms are not still sitting at their defaults simply because the path data was incomplete.

Step-by-Step Guide:

The safest way to use the page is to build the link budget in layers and watch the meaning of each result tile change as new thresholds appear.

Choose a Preset close to your scenario, or enter Frequency, Distance, and TX power manually. Once frequency and distance are above zero, the summary box and Link Verdict panel appear.
Add TX antenna gain and RX antenna gain, then open Advanced for cable loss, extra path loss, and polarization mismatch. Budget Ladder should now read like a believable signal path from the radio port to the receiver input.
Enter RX sensitivity if you know the minimum decode level. That fills Receiver margin and turns the page into a threshold check instead of a raw receive-power estimate.
Enter Target RSSI and set Fade reserve when you need more than bare decode. Planning floor, Available reserve, Planned max range, and Gain Gap Map all follow whichever floor is stricter.
Add Noise figure and a positive Noise bandwidth only when SNR matters. If Noise / SNR still reads Not modeled, the recovery step is to raise bandwidth above 0 and confirm the bandwidth unit.
Use Range Envelope to see how quickly receive power falls with distance, Band Comparison to isolate frequency choice, and Gain Gap Map to see whether you have reserve below zero or a positive dB shortfall above zero.
Finish with Budget Ladder or JSON. The result is ready to share when each gain, loss, and threshold still looks intentional outside the chart view.

Interpreting Results:

Once a Planning floor exists, Available reserve is the main planning number. Positive reserve means predicted receive power clears the stricter threshold. Negative reserve means the current budget is short by that many decibels. That is why a path can still show a positive Receiver margin yet receive a warning verdict: it may decode in clean conditions while still missing the safer operating target.

Receiver margin above zero means the link clears bare sensitivity, not that it is stable under fading, clutter, or congestion.
Planned max range assumes the same gains, losses, and thresholds you entered. If the real path has uncounted obstruction or weather loss, the range is optimistic.
Noise / SNR is a thermal-noise estimate from bandwidth and noise figure. It does not include co-channel interference or external noise sources.
Fresnel midpoint is a geometry warning. A strong power budget can still fade badly if obstacles intrude into the Fresnel bulge.

If the verdict looks better or worse than expected, verify Planning floor, Additional path loss, and the unit selectors before you start changing antennas, bands, or path length.

Worked Examples:

LoRa field node with comfortable headroom

Using the built-in 868 MHz field-node pattern, a 2 km path with 14 dBm transmit power, 1 dB cable loss on both ends, 2 dB of extra path loss, RX sensitivity -123 dBm, and Target RSSI -120 dBm produces FSPL 97.2 dB, Predicted receive power -87.2 dBm, Planning floor -113.0 dBm, and Available reserve 25.8 dB. Planned max range reaches about 38.9 km, so the current 2 km hop sits well inside the free-space budget. This is the kind of case the Robust verdict is meant to signal.

Indoor Wi-Fi pass with thin reserve

The 2.4 GHz Wi-Fi client preset uses a 50 m path, 18 dBm transmit power, 3 dBi antennas, 8 dB of extra path loss, RX sensitivity -72 dBm, and Target RSSI -67 dBm. The resulting outputs are Predicted receive power -60.1 dBm, Planning floor -66.0 dBm, Available reserve 5.9 dB, and Noise / SNR 33.9 dB. The link passes, but only with a small cushion. A few decibels of uncounted wall loss or poor antenna orientation would be enough to turn an On plan result into a shortfall.

Longer 5.8 GHz bridge that still decodes but misses the target

Take the outdoor 5.8 GHz bridge pattern, stretch it to 8 km, and add 3 dB of extra path loss. The page moves to FSPL 125.8 dB, Predicted receive power -71.8 dBm, Planning floor -65.0 dBm, Receiver margin 3.2 dB, and Available reserve -6.8 dB. That is the classic Short reserve case. The radio still clears bare sensitivity, but Gain Gap Map is telling you the budget is still about 6.8 dB short of the planning target.

Result looks incomplete because the floor was never defined

With a 5.8 GHz path at 2 km, 26 dBm transmit power, 2 dBi transmit gain, 14 dBi receive gain, 3 dB polarization mismatch, 20 MHz bandwidth, and no receive threshold, the page still shows Predicted receive power -76.7 dBm and Noise / SNR 20.3 dB. What stays empty are Planning floor, Available reserve, Receiver margin, and Planned max range. The calculation is not broken. The corrective path is to enter RX sensitivity, Target RSSI, or both so the page knows what the signal must clear.

FAQ:

Does free-space path loss include walls, trees, rain, or body blocking?

No. The free-space term only covers geometric spreading. Known non-free-space penalties belong in Additional path loss or Polarization mismatch. At higher microwave bands, rain and atmospheric loss can also become important even on otherwise clear paths.

Why is Available reserve or Planned max range missing?

Both depend on Planning floor. If RX sensitivity and Target RSSI are both blank, the page can still compute receive power, but it has no threshold to compare against.

Why can Receiver margin be positive while the verdict is still negative?

Receiver margin only compares receive power with bare sensitivity. The verdict uses Planning floor, which may be higher because of Target RSSI or Fade reserve. A link can decode and still miss the safer operating target.

Why does a lower band often look better in Band Comparison?

That chart keeps distance, gains, and losses fixed, then changes frequency alone. Under the free-space equation, path loss rises with frequency, so lower bands usually recover margin when the rest of the budget stays the same.

Why does Noise / SNR say Not modeled?

Thermal-noise math is disabled until Noise bandwidth is above 0. After that, the page uses the entered Noise figure to estimate noise floor and SNR.

Does the page send or store my link budget?

No server-side processing is present for this calculator. The calculations, charts, CSV output, DOCX export, and JSON export run in the browser from the values currently on the page.

Glossary:

Free-space path loss: The baseline spreading loss between isotropic antennas in free space, shown here as FSPL.
EIRP: Effective isotropic radiated power, equal here to transmitter power plus transmit antenna gain minus transmit-side feed loss.
Planning floor: The stricter receive threshold used for reserve and range decisions, taken from target RSSI, sensitivity plus reserve, or whichever is higher.
Available reserve: How far predicted receive power sits above the planning floor. Negative values show how much more system budget is still needed.
Fresnel zone: The clearance volume around the direct path that must stay mostly unobstructed to preserve near-free-space conditions.
Noise figure: Receiver-added noise above the thermal reference, used here with bandwidth to estimate noise floor and SNR.

References:

P.525: Calculation of free-space attenuation, International Telecommunication Union.
P.526: Propagation by diffraction, International Telecommunication Union.
P.530-19: Propagation data and prediction methods required for the design of terrestrial line-of-sight systems, International Telecommunication Union, September 2025.
The Essential Noise Figure Measurement Guide: Measuring Noise Figure with Signal Analyzers, Part 1, Keysight Technologies.