C-DRONE GUIDE · 1 SEPTEMBER 2026

Antenna Radiation Measurement by Drone: Pattern, Tilt, Azimuth, Coverage

A transmitting antenna is an object whose function is invisible. The mast is sound, the fixings are tight, the radome is intact — and yet coverage has degraded on one side of the sector since the last crew visit. A visual check will say nothing : what changed is the radiation, not the structure. A mechanical tilt knocked two degrees out of true, an azimuth shifted after a storm, two jumpers swapped between sectors during an equipment upgrade, an antenna refitted upside down : none of these show up on a photograph, and all of them show up on a measured pattern. A drone carrying a receiver or a field probe makes that measurement possible in place, on the installed antenna, in its real environment, without dismounting it or shipping it back to an anechoic chamber. Here is what an airborne measurement adds compared with a ground-based one, how it is carried out, what it is worth legally under the French public-exposure framework — and above all what it does not prove.

Published on 1 September 2026, reviewed on 11 September 2026 — regulations in force as of September 2026.

Radio measurement or structural inspection: two missions not to be confused

A drone comes to a high site for two entirely distinct reasons, and confusing them wastes both parties' time. The first mission is structural and visual inspection : condition of the lattice or guyed mast, corrosion, lean, anchors, bolting, radome condition, abnormal hot spots on connectors and cabinets seen through a thermal camera. It produces photographs, a 3D model and a report of located defects. That is the subject of our guide to telecom tower and cell antenna inspection by drone, and it is not the subject of this one.

The second mission is a radio measurement. Here the drone is no longer a flying camera but an instrument carrier : it flies a measurement antenna and a receiver, or a field probe, along a trajectory known to the centimetre, recording the received level at every point. The deliverable is no longer an image of a defect but a curve : the radiation pattern in a given plane, the field level along a profile, the difference between theoretical and actual pointing. You are not looking for a visible fault, you are looking for a gap between the antenna as it is supposed to radiate and the antenna as it actually radiates.

The distinction is not academic. A mechanically flawless antenna can radiate wrongly — because the electrical tilt was never reprogrammed after a configuration change, because a jumper was plugged into the wrong port during an equipment swap, or because a bracket twisted under wind loading with nothing visible from the ground. Conversely, a degraded structure can keep radiating to specification right up to the day it fails. The two campaigns answer two different questions and are often scheduled together, on the same site visit, because access coordination is by far the heaviest item : the same pooling logic applies to an aerial fibre network drone audit, where structure and pole occupancy are surveyed in a single pass.

The closest parallel on this site is therefore not visual inspection but drone noise mapping with an acoustic camera : in both cases the drone carries a metrological sensor, the value of the deliverable rests on control of the measurement chain, and the question of the result's regulatory status arises immediately.

What a flight gives you that a ground measurement cannot

The first benefit comes from a geometric constraint. An antenna's radiation pattern only means something in the far field, beyond a distance at which the wavefront can be treated as planar ; the usual criterion is 2 D²/λ, where D is the antenna's largest dimension and λ the wavelength. Now, the panel antennas of a mobile site, like broadcast antennas, are mainly long in the vertical plane and radiate a narrow lobe in elevation. From the ground, that plane can only be sampled by moving away, which runs straight into terrain, buildings and vegetation. By climbing, the drone sweeps the elevation angle directly at constant range and makes accessible the part of the pattern that is hardest to obtain any other way.

The second benefit is checking actual pointing. A site file states an azimuth and a tilt ; the installed antenna has its own. An airborne survey lets you compare the two : finding the elevation angle where the received level peaks means finding the effective tilt, the sum of the bracket's mechanical tilt and the electrical tilt programmed into the radio unit. Likewise, a horizontal arc sweep at constant elevation gives the real direction of the main lobe. This is how you catch the errors that leave no visible trace : a sector whose azimuth was entered backwards, two adjacent sectors whose feeders were swapped, an antenna refitted upside down after a maintenance visit, a mechanical tilt shifted by a storm.

The third benefit is coverage in the literal sense. A coverage complaint — a shadow at the cell edge, a dead spot inside an industrial site, an unstable link on a private radio network — is classically handled by a ground measurement campaign, slow and constrained by roads. An airborne measurement along a vertical profile or a ring around the site gives a picture of the field that the ground cannot show, and helps separate two causes that look alike : an antenna not radiating as designed, and propagation degraded by terrain.

None of this is new, nor a sales extrapolation. A review by V. R. Kandregula, Z. D. Zaharis, Q. Z. Ahmed, F. A. Khan, T. H. Loh, J. Schreiber, A. J. R. Serres and P. I. Lazaridis published in 2024 in Sensors surveys some fifteen years of drone-based antenna and propagation measurement, from large radio-astronomy arrays to base-station and broadcast antennas : the authors report amplitude differences of the order of 0.5 to 1 dB against an anechoic-chamber reference, and identify ground reflections as the dominant error contribution, ahead of positional uncertainty (see the study, DOI 10.3390/s24227395).

The method: payload, trajectory, positioning, repeatability

The payload determines what can be measured. Two families coexist. The first carries a measurement antenna and a receiver — a compact spectrum analyser, a software-defined radio — recording the received level on a given frequency or band with a known polarisation : that is the pattern-survey configuration, where the antenna under test transmits and the drone listens. The second carries a broadband isotropic field probe, giving an overall level in V/m without separating sources : that is the exposure-mapping configuration. The two ends can also be swapped : the drone then carries a calibrated test source and the ground antenna becomes the receiver. That is the configuration adopted as early as 2014 by G. Virone, A. M. Lingua, M. Piras, A. Cina, F. Perini, J. Monari, F. Paonessa, O. A. Peverini, G. Addamo and R. Tascone in IEEE Antennas and Wireless Propagation Letters : a micro-drone carrying a source, flying autonomously, its absolute position tracked by a ground survey instrument, the pattern being reconstructed from the received power along the flight path — a setup validated on two standard wire antennas at 150 and 408 MHz with an estimated accuracy of 1 dB (see the study, DOI 10.1109/LAWP.2014.2298250).

The trajectory follows the question. For the vertical plane, you fly an arc at constant horizontal range, climbing in regular steps, which samples the elevation angle. For the horizontal plane, you fly a ring at constant altitude around the mast. For a full characterisation, the two are combined into a series of arcs forming a partial hemisphere. In every case, two parameters govern the quality of the result : the angular sampling step, which must be clearly finer than the width of the lobe being measured, and the measurement range, which must satisfy far-field conditions while keeping a usable signal-to-noise ratio.

Positioning is the keystone. A pattern survey is only usable if you know which angle each sample corresponds to : a position error translates directly into an angular error, hence into an error on the reported tilt. Standard GNSS, with its few metres of uncertainty, is not enough once the lobe is narrow ; centimetre-level RTK or PPK positioning becomes a condition of the measurement, not a refinement. That is exactly the point analysed by A. Y. Umeyama, J. L. Salazar-Cerreno and C. J. Fulton in a study published in 2020 in IEEE Access on far-field drone measurement of polarimetric weather radar patterns : they build a full error budget in which aircraft position and the orientation of the gimbal carrying the source bear directly on the accuracy of the reconstructed pattern (see the study, DOI 10.1109/ACCESS.2020.3027790).

Then comes repeatability, which is in practice where the real operational value lies. An operator rarely compares an antenna against an absolute : it compares the antenna with itself, before and after a maintenance visit, before and after a storm, between two sites meant to be identical. That means freezing the method and documenting it in the report : reference point coordinates, arc radius and altitudes, angular step, frequency and bandwidth, polarisation, onboard antenna type, transmission configuration during the flight. A survey whose method is not written down cannot be compared with the next one, and loses most of its point.

Use cases: site acceptance, coverage complaints, swaps, broadcasting, radars

Site acceptance is the clearest case. A site has just been built or upgraded ; the file describes three sectors, each with its azimuth, tilt and radio configuration. One airborne survey per sector verifies, before commercial service starts, that what was installed matches what was specified. The economics are simple : a pointing error found at acceptance is corrected while access equipment is still on site, whereas an error discovered three months later through degraded performance indicators forces a second full mobilisation, with its coordination and rework cost. The same reasoning applies to any technical acceptance : checking on delivery costs a fraction of what a return visit costs.

The coverage complaint is the second use case, and the most common in operations. A business customer reports an area where service is unstable ; prediction tools say the area is covered. Between prediction and reality, the airborne measurement settles it : if the measured pattern is compliant, the cause lies elsewhere — shadowing, interference, capacity — and crews stop being sent up the mast ; if it is distorted or offset, you know what to correct. Validation after an equipment swap follows the same logic : replacing radio units or antennas is a moment of high risk for cabling inversion, and that is precisely the error a measurement detects most easily.

Post-storm checking alone justifies having a repeatable method. After a severe wind event, the operator of a portfolio of high sites has to triage fast : which sites actually moved? A visual inspection sees a bent structure, not a bracket that swung three degrees. A quick azimuth and tilt survey on the most exposed sites, compared with the reference survey, directs resources to those that need them — a triage approach comparable to the mass loss assessments run after a weather event.

Finally, the field extends well beyond mobile telephony. In FM and DTT broadcasting, the vertical pattern of a panel or radiating array governs ground coverage and compliance with the frequency plan ; on a high-power transmitter, airborne measurement is sometimes the only way to reach the vertical plane without chartering a crewed aircraft. On radars and navigation aids — weather radars, radio beacons, landing systems — periodic pattern verification is a long-standing operational requirement, historically met by flight-check aircraft : this is one of the areas where the literature on drone-based measurement is richest. Private radio networks, finally — an industrial site's PMR, a local authority's safety network, an energy operator's telemetry — benefit most from the fall in access cost, because they never could afford a conventional airborne check campaign.

Public exposure to radio waves: the French framework, and where drone measurement does not fit

As soon as you talk about measuring an electromagnetic field in France, a precise regulatory framework applies, and it concerns not radio engineering but public protection. Decree no. 2002-775 of 3 May 2002, issued under point 12 of article L. 32 of the postal and telecommunications code, sets the public exposure limit values for fields emitted by telecommunication network equipment and radio installations ; it transposes Council recommendation 1999/519/EC of 12 July 1999. The reference levels usually quoted for mobile frequencies are around 41 V/m at 900 MHz, 58 V/m at 1,800 MHz and 61 V/m above 2,000 MHz (the 2,100, 2,600 and 3,500 MHz bands).

Verification runs through a single protocol. The order of 3 November 2003 defines the in-situ measurement protocol applicable to fixed transmitting stations ; it was amended by the order of 23 October 2015 and then by the order of 9 November 2017, which replaces the reference « ANFR/DR 15-3 » with « ANFR/DR 15-4 », the new protocol becoming mandatory one year after publication. The protocol describes a ground measurement : locating the point of maximum exposure with a broadband probe, taking the result as the average of three measurements at 1.1 m, 1.5 m and 1.7 m above ground, then a narrowband analysis service by service with an RMS value integrated over six minutes, across the 100 kHz - 6 GHz band. It distinguishes a broadband « case A », all sources combined, from a service-by-service « case B » ; according to ANFR, case B becomes mandatory when the case A result exceeds 6 V/m.

The decisive point for a buyer is this : that official measurement is entrusted to laboratories accredited by COFRAC, bound by the ANFR protocol and by independence criteria. The national monitoring scheme lets any individual or legal entity have their exposure measured, at home or in a place open to the public, free of charge to the requester : the form must be countersigned by an authorised body — a municipality or group of municipalities, a regional health agency, an approved environmental or health protection association —, the measurement is funded from a fund fed by a levy paid by the operators, and results are published on the Cartoradio map. A local authority facing residents' concerns therefore does not need to pay for a measurement : it needs to route the request into that scheme.

The third layer is local information. Law no. 2015-136 of 9 February 2015, known as the Abeille law, requires a town hall information file to be sent to the mayor before any new installation or substantial modification of a radio installation. Decree no. 2016-1211 of 9 September 2016, on local information regarding public exposure to electromagnetic fields and on ANFR's national dialogue committee, states that the file — and, where applicable, the requested exposure simulation — is made available to residents within ten days of the mayor receiving it. A drone engineering measurement can feed a local dialogue and document a situation, but it substitutes neither for the information file nor for the national scheme's measurement.

Worth noting : ANFR itself has taken an interest in drones as a radio measurement tool, targeting first hard-to-reach areas — transmitters on masts, rooftops or facades — and the location of interference sources in dense urban environments, where multiple reflections disturb ground-based measurement and direction finding. That confirms the technical relevance of the approach, without changing its legal status.

Flying around a live transmitter: height, interference, coordination

The first constraint is height. A vertical pattern survey means climbing at least to antenna level, often above it : on a broadcast mast, that frequently exceeds the 120 m ceiling of the Open category. Implementing regulation (EU) 2019/947 provides a useful exception here : within 50 m horizontally of an artificial obstacle taller than 105 m, the maximum height may be raised to 15 m above the top of the obstacle, at the request of the entity responsible for it. That suits a flight hugging the mast, far less an arc flown at constant range, which by construction leaves that 50 m cylinder. Beyond it, the flight requires a height application, covered in our guide to flying a drone above 120 m, and sometimes a switch to the specific category. This constraint shapes everything : it explains why a radio measurement campaign is prepared over weeks rather than days.

The second constraint is the electromagnetic environment itself. A drone approaching the main lobe of a high-power transmitter enters an intense field, liable to disturb its command and video links, its sensors, and above all its satellite reception. Degraded or lost GNSS near a mast is a scenario to plan for, with consequences both for position hold and for the validity of the measurement itself — since position is the quantity that gives each sample its meaning. Our guide to GNSS signal loss, jamming and spoofing covers the fallback modes and what they mean for a deliverable. In practice the risk is bounded by setting a minimum approach distance, preparing a degraded flight mode, and, where appropriate, negotiating a power reduction on the sector being overflown.

The third constraint is human and contractual. A high site is an operational facility, often shared between several operators and an infrastructure owner. You need the obstacle manager's agreement for the height exception, the agreement of every hosted operator whose antennas you approach, and coordination of the transmission configuration during the measurement window. On an industrial or energy site, the site's own access rules apply on top, as covered in our guide on hosting a drone mission on an industrial site : prevention plan, co-activity, access permits. Finally, many high sites sit near aerodromes, precisely because they occupy commanding positions : checking airspace beforehand on the Géoportail drone zone map is a prerequisite, and a location inside a control zone triggers the procedure described in our guide to a drone mission inside an aerodrome CTR.

The limits: what a drone measurement does not prove

The most important limit is legal, and it deserves to be written explicitly into any tender. A field measurement carried out by drone has no value as a regulatory check on public exposure unless it is performed within the accredited framework described above : ANFR protocol, COFRAC-accredited laboratory, independence criteria. An airborne measurement report is an engineering document, usable internally, enforceable against a supplier in a contractual acceptance, useful in a technical discussion — but it does not demonstrate compliance with the 2002 decree and does not replace a measurement from the national scheme. A contractor implying otherwise would be selling a deliverable that fails in the one discussion where it counts.

The second limit is metrological. An airborne survey happens in the site's real environment, with the ground, buildings and neighbouring structures ; it does not reproduce anechoic-chamber conditions. Reflections add to the direct signal contributions whose phase varies with position, and the literature identifies this as the dominant error source. The measured pattern is therefore the pattern of the antenna as installed on its site — precisely what operations needs to know — but must not be confused with the antenna's intrinsic pattern quoted by its manufacturer. It is also why side and back lobes, where the level is low and the signal-to-noise ratio poor, are measured far less faithfully than the main lobe.

The third limit is operational. The measurement describes the antenna's state during the flight window, with the transmission configuration in force at that moment. On modern beam-steering equipment, what you record depends directly on the transmission mode configured for the test : a survey only means something if that configuration is described in the report, otherwise two campaigns are not comparable. Nor is the result permanent : an electrical tilt reprogrammed remotely the next day makes the survey obsolete, which argues for event-driven measurement — acceptance, swap, storm, complaint — rather than blind periodicity.

The last limit concerns the chain of responsibility. As with any technical deliverable, the report's value depends on what the contract demands of it. A tender ordering « a drone radiation measurement » without specifying method, frequency, polarisation, angular step, transmission configuration and raw data format buys nothing enforceable. Authorities going through a public contract will find the drafting framework in our guide to buying drone services in public procurement, and the contractor's professional liability insurance remains, as for any work on an operating facility, something to verify before site access.

Method and price of a drone radio measurement campaign in 2026

A campaign runs in four stages. First preparation : collecting the characteristics of the antenna under test (type, dimensions, frequency, polarisation, specified azimuth and tilt), computing the far-field distance, defining the arcs and angular step, checking airspace, requesting agreement from the obstacle manager and from every hosted operator, negotiating the transmission configuration and the time window. This is the longest item : counted in weeks, not days. Then the flight, short by comparison : a few tens of minutes of actual measurement per plane and per sector, wrapped in half a day of setup, calibration and reference-point survey. Then processing : matching samples to the RTK trajectory, conversion into angles, correction for the known gains of the chain, plotting the patterns and comparing them with the theoretical pattern. Finally the report, which must contain the full method, time-stamped and georeferenced raw data, and the deviations found on azimuth and tilt.

Orders of magnitude observed in France in 2026, excluding VAT, for engineering work of this kind : a pointing check (actual azimuth and tilt against the site file) on a three-sector site, with a photogrammetric survey of antenna positions, runs from €1,200 to €2,200 ; a radiation pattern survey on one antenna, in one plane and one band, from €1,800 to €3,500 ; a full radio acceptance of a new site, three sectors, vertical and horizontal planes, with a report comparing results to the design file, from €3,500 to €6,500 ; a field mapping around a high-power transmitter (broadcast, radar, private network), over several profiles or a partial hemisphere, from €4,000 to €8,000 depending on extent. A campaign across a portfolio of high sites is quoted individually, with a markedly degressive per-site rate from about ten sites upwards, since pooling travel and preparation engineering weighs more than the flight itself.

These brackets sit above those of a structural mast inspection — around €400 to €700 for a simple visual check, €900 to €1,600 for visual, thermal and 3D model — for three reasons : the payload is a measuring instrument rather than a camera, processing requires radio expertise rather than photogrammetry software, and regulatory preparation is heavier because of the height involved. Three levers reduce the unit cost : combining the radio check and the structural inspection in a single mobilisation, handling several sites in the same geographic area on one trip, and supplying a complete technical file of the antennas up front — every missing item is paid for in engineering time.

Conversely, two items should never appear on this quote. The official public exposure measurement belongs to the national scheme described above : it is carried out by a COFRAC-accredited laboratory and is free to the requester, so there is no reason to bill it. And the exposure simulation attached to the town hall information file is the responsibility of the installation's operator, not of a flight contractor. To scope a radiation measurement, a site acceptance or a post-storm portfolio check, request a quote stating the antenna type, frequency, support height, number of sectors and the result expected : we check aeronautical feasibility and metrological relevance before pricing.

Frequently asked questions

Request a free quote

Also worth reading

• Olive orchards and drones: water stress and yield forecasting, price
• Prune d'Agen and drones: water stress and ripeness, price
• Asparagus and drones: spotting weeds on the ridge before emergence, price