How to Verify a Radio Path with Fresnel-Zone Math

Back to Blog

Gather the Path Inputs You Need

Every Fresnel calculation needs four inputs, and getting them right matters more than the arithmetic. You need the great-circle path length between the two antennas, the operating frequency of the radios, a terrain profile along the path (from a topographic map, survey data, or an online path-profile tool), and the proposed antenna centerline heights above ground at each end. Keep the distance in kilometers and the frequency in gigahertz if you use the metric form of the radius equation, because the constant in the formula is tied to those units. Mixing miles and megahertz into a metric constant is the single most common reason a hand-calculated path is wrong.

You also need to know the tallest credible obstruction between the sites and how far it sits from each end, because the Fresnel zone is fattest at the midpoint and pinches to nothing at the antennas. A ridge two-thirds of the way down the path is evaluated at that ridge's own d1 and d2, not at the midpoint. List every candidate obstruction - ridgelines, tree canopy, buildings, tank batteries - with its ground elevation and its distance from the master site, so you can test the worst one rather than assuming the midpoint governs. Remember that trees and structures add height above the bare-earth terrain elevation.

This desk check is the front end of the same discipline covered in the RF link budget. The Fresnel math tells you whether the path is geometrically clear; the link budget tells you whether there is enough signal energy once it is. Both have to pass. A path that clears every Fresnel zone but has no fade margin still drops, and a path with plenty of margin that clips an obstruction still suffers diffraction loss the budget did not predict.

Confirm Optical Line of Sight First

Start with the simplest test: does a straight line between the two antenna centerlines clear the terrain at all? Draw or compute the path profile, mark the antenna heights at each end, and connect them with a straight sightline. If that line already intersects a ridge or the tree canopy, you do not have optical line of sight and no amount of Fresnel arithmetic will rescue the path at these heights. This is the go/no-go gate before you calculate anything, and it is explained in more depth in the reference on line of sight and the Fresnel zone.

If the straight line clears the bare terrain, you have optical line of sight but not necessarily radio line of sight, because the radio beam is not a pencil-thin ray. It is a fat ellipsoid, and objects that sit near the sightline but do not touch it still steal energy through diffraction. That is exactly what the Fresnel-zone radius quantifies, and it is why a path that looks clear on a straight-line drawing can still perform poorly. Treat optical line of sight as necessary but not sufficient.

Where the terrain is genuinely flat, such as a coastal marsh or a plowed field, the limiting obstacle is often the curvature of the earth itself rather than any feature on the ground. In that case skip ahead to the earth-bulge term, because the ground rises to meet your beam in the middle of the path even though every point of terrain is at the same elevation.

Compute the First Fresnel-Zone Radius

The first Fresnel-zone radius at any point along the path is r = 17.3 * sqrt( (d1 * d2) / (f * d) ), where d1 and d2 are the distances in kilometers from that point to each antenna, d is the total path length in kilometers, f is the frequency in gigahertz, and r comes out in meters. The 17.3 constant already bundles the speed of light and the unit conversions, so it only holds for kilometers-and-gigahertz. At the exact midpoint d1 equals d2 equals d/2, which simplifies the radius to r = 17.3 * sqrt( d / (4 * f) ), a handy form for a quick worst-case midpoint estimate.

Read the formula rather than just plugging into it, because it tells you the design levers. The radius grows with the square root of distance, so doubling the path length only makes the zone about 1.4 times fatter, not twice. The radius shrinks with the square root of frequency, so a higher-frequency link has a slimmer Fresnel zone and needs less clearance height, which is one quiet reason higher bands tolerate marginal terrain better. And because the zone pinches toward each antenna, an obstruction close to one end is far less demanding than the same obstruction at the middle.

Compute the radius at the specific location of your worst obstruction using that obstruction's own d1 and d2, not the midpoint values, unless the worst obstruction happens to be central. Then the clearance test is straightforward: the height of the sightline above the obstruction must exceed 60 percent of this radius. Sixty percent of the first Fresnel zone is the widely used rule of thumb below which diffraction loss becomes significant; full first-zone clearance is the conservative target where you can afford the antenna height.

Add the Earth-Curvature Bulge

On paths longer than a few kilometers the earth is not flat under your beam; it bulges up in the middle. The height that the earth's curvature adds at a point is h = (d1 * d2) / (12.75 * k), where d1 and d2 are in kilometers, h is in meters, and k is the effective-earth-radius factor. Under standard atmospheric refraction k is taken as 4/3, because the atmosphere bends the beam slightly downward and effectively flattens the earth; using k = 4/3 gives the everyday-conditions bulge, while k = 1 (a true-sphere, no-refraction case) is the pessimistic value some designers check against for robustness.

Add this curvature bulge to the physical height of the obstruction before you run the Fresnel clearance test. In effect the ground, the trees, and the earth-bulge all stack up under the beam, and your sightline has to clear the sum of them plus 60 percent of the Fresnel radius. On a long, otherwise flat path the earth bulge alone can be the deciding obstruction, which is why over-water and over-prairie links sometimes need surprisingly tall masts even with nothing visible in the way.

Because k varies with weather, the same path has more or less clearance on different days, and a path designed only for the k = 4/3 average will occasionally lose clearance when the atmosphere sub-refracts. Designing to a stricter k, or carrying extra height, is how you buy robustness against those atmospheric swings. This is the geometric cousin of the signal-level robustness you buy with fade margin.

Verifying the Result

Assemble the full clearance check at each candidate obstruction as a single inequality: sightline elevation at that point must be greater than obstruction elevation, plus any tree or structure height, plus the earth-curvature bulge, plus 0.6 times the first Fresnel-zone radius. Work it symbolically first with your terms in place, then substitute your site numbers, so you can see which term dominates. If the path fails, the equation shows you the cheapest fix: raise an antenna (increases sightline elevation), pick a higher frequency (shrinks the Fresnel term), or relocate a site (changes d1 and d2).

Do a sanity pass on units before you trust the answer. The Fresnel constant 17.3 and the curvature constant 12.75 are both tied to kilometers, gigahertz, and meters; if you fed the formula miles or megahertz the radius will be off by a large factor and the whole verification is invalid. Recompute the midpoint radius with the simplified r = 17.3 * sqrt( d / (4f) ) form as an independent cross-check against your obstruction-point calculation, because an order-of-magnitude disagreement usually means a unit slip.

Finally, treat the paper result as a prediction to be confirmed, not a guarantee. When the link is built, the commissioning step is to read received-signal level and compare it against the value the link budget predicted. A path that cleared on paper but reads far weaker than predicted usually means an obstruction was missed, a height was wrong, or the terrain data was coarse, and it sends you back to the profile with better inputs.

Common Mistakes

The most common error is evaluating the Fresnel radius only at the midpoint and then applying it to an off-center obstruction. The zone is genuinely narrower near the ends, so a ridge close to one antenna needs less clearance than the midpoint radius implies, and treating it as if it needed midpoint clearance can reject a workable path. Always use the obstruction's actual d1 and d2.

A second frequent mistake is forgetting to add tree canopy and structure height to the bare-earth terrain elevation. Topographic data and many path tools give ground elevation only, and a mature tree line or a nearby building can add substantial height that quietly eats the clearance you thought you had. Survey the actual vegetation and structures along the path, and remember that trees grow.

The third is ignoring earth curvature on long paths or, conversely, ignoring atmospheric variability by designing only for k = 4/3. Skipping the bulge on a long over-water path underestimates the obstruction; designing only for average refraction leaves the link short of clearance during sub-refractive conditions. Carrying height margin and checking a pessimistic k value are how experienced designers avoid both traps.

Frequently Asked Questions