Understanding Near Field and Far Field for Antenna Performance
Understanding Near Field and Far Field for Antenna Performance
A field-strength reading becomes evidence only when the region, probe, geometry, reference plane and measured quantity are defined. Walking around an antenna with a handheld VNA or probe is not an efficiency test.
I often see an antenna comparison reduced to this: connect a small probe to a handheld instrument, walk around the antenna, and call the largest reading “most efficient.” That experiment may reveal a local field or a coupling problem. It does not, by itself, measure radiated power, accepted power, gain or efficiency.
Near field does not mean “no radiation,” and far field is not the only place from which a valid result can be derived. The decisive question is whether the selected method measures enough calibrated information, at a declared reference plane, to recover the quantity being claimed.
The Field Does Not Change Regime at a Sharp Wall
An antenna's exact field is a superposition. For familiar elemental-radiator solutions, different terms vary with range as 1/r³, 1/r² and 1/r. Their relative size and phase change continuously. It is therefore misleading to assign one decay law to an entire spatial region or to treat a published boundary as a physical discontinuity.
In the reactive near field, stored electric or magnetic energy makes a strong contribution. Electric and magnetic fields need not have the far-field phase, magnitude ratio or polarization relationship. That does not make the outward time-average radiated-power component zero: a transmitting antenna is still sending power through space while reactive energy is exchanged locally.
In a radiating near field, often called the Fresnel region, radiation is prominent but the angular field distribution can still change with distance. This region is especially important for electrically large apertures. An electrically small radiator may not present a useful, separately identifiable Fresnel interval.
In the far field, the angular pattern is essentially independent of range for the stated accuracy and direction. In free space the radiative electric and magnetic fields are approximately transverse and fall in amplitude as 1/r; their power density therefore falls as 1/r². Ground, buildings, lossy media and multipath can prevent a real outdoor site from behaving like that ideal free-space range even when the geometric distance is large.
| Region | What dominates the measurement problem | What a local reading cannot establish alone |
|---|---|---|
| Reactive near field | Stored-field contribution, strong spatial gradients and probe loading | Radiated power, gain or efficiency |
| Radiating near field | Range-dependent amplitude, phase and angular distribution | A far-field pattern without a valid transformation |
| Far field | Range validation, polarization, reflections, alignment and dynamic range | Efficiency unless accepted power or an equivalent calibrated quantity is also known |
Use a Boundary That Fits the Antenna and the Accuracy
Let D be the maximum dimension of the radiating structure and λ the wavelength. For an electrically large aperture,
r ≥ 2D²/λ
is a widely used Fraunhofer-distance criterion. It is tied to an aperture phase-error approximation; it is not a universal pass/fail rule for every antenna, observation direction or uncertainty target. The test plan must also account for the antenna geometry, desired amplitude and phase accuracy, probe aperture, quiet-zone quality, polarization, reflections and whether the entire radiating structure fits inside the declared dimension.
Expressions involving 0.62√(D³/λ) are useful approximate boundaries in conventional treatments of electrically large antennas. A scale near λ/2π appears in elementary electrically small-radiator analysis. Neither expression turns the field on or off, and combining them mechanically does not validate a range.
There is no general “three wavelengths is always enough” rule. For a large aperture, 2D²/λ can be much greater. For a compact antenna, the greater limitation may instead be site reflections, direct coupling, probe response or the uncertainty required by the measurement. Use the applicable measurement standard and validate the actual range.
Nearby earth, walls, masts, feed lines and operators do not merely shorten a textbook boundary. They become part of the electromagnetic problem: they can change antenna current, add image and return paths, scatter fields and create multipath. Measure the intended installed system, or remove and quantify those influences.
Why the Handheld Walk-Around Is Not Efficiency
A VNA calibrated to the antenna feed plane can measure input reflection coefficient and impedance. With a second characterized port, it can also be an instrument inside a valid transmission, range, chamber or near-field-scanner method. The instrument is not the problem; the incomplete experiment is.
A person walking with a whip, loop or uncalibrated probe normally lacks several quantities needed for efficiency:
- the probe's vector response, polarization, orientation and cable correction;
- position accuracy and enough spatial samples to represent the field;
- control of body, cable and probe perturbation of the antenna under test;
- separation of direct radiation, stored-field coupling, scattering and multipath;
- the accepted RF power at a declared antenna reference plane; and
- a calibrated relationship between the sampled voltage and total radiated power.
An electric-field probe and a magnetic loop deliberately respond to different field components. A higher reading from one probe at one point can mean stronger local coupling, different polarization, a sharper spatial gradient or a changed antenna current—not higher radiation efficiency.
The observer is part of the setup. A hand, body, instrument enclosure and trailing cable can change the antenna impedance and current distribution. Repeat a position with the operator, cable route or probe orientation changed. If the reading moves materially, that variation belongs in the uncertainty budget rather than in an efficiency claim.
Define the Power Ratio Before Measuring It
Radiation efficiency is the ratio of radiated power to power accepted by the antenna at a declared feed reference plane:
ηrad = Prad/Pacc
Pacc = Pinc − Prefl
For a one-port antenna, with incident and reflected powers evaluated at the same plane, the mismatch factor is 1 − |Γ|². A total or realized efficiency referenced to incident power at that plane can then be written as ηrad(1 − |Γ|²). Move the reference plane through a feed line, tuner, balun or matching network and their losses enter the result unless they are de-embedded. Always state which plane and which efficiency definition the number uses.
Input match alone does not provide Prad. A lossy structure can accept power and convert much of it to heat. Conversely, a mismatched antenna can have high radiation efficiency for the power it does accept. SWR, accepted power, radiation efficiency, total efficiency, directivity and gain are related but distinct quantities.
Methods That Can Produce Defensible Results
Validated far-field range
A gain-transfer, comparison or three-antenna measurement can derive gain from calibrated transmission data when the range meets its far-field and quiet-zone requirements. The result needs known cable and fixture losses, polarization, alignment, distance, reference antennas, mismatch treatment, reflections and uncertainty. Gain plus directivity can support an efficiency result when both quantities describe the same antenna state and reference conditions. A VNA may measure S21 here, but a minimum distance by itself does not validate the range.
Calibrated near-field scanning
Planar, cylindrical or spherical systems sample complex field—amplitude and phase—on a controlled surface. A defensible near-field-to-far-field result needs adequate spatial sampling and scan extent, accurate probe positions, a characterized probe, probe compensation, cable and drift control, coordinate alignment and a validated transformation. NIST's near-field scanning chapter explicitly treats the mathematical transformation and measuring-probe compensation; NIST also documents the principal uncertainty sources, including position error.
Near-field measurement is therefore not a contradiction. It is a complete vector measurement and inverse problem, not a few handheld spot readings.
Reverberation chamber
A characterized, mode-stirred reverberation chamber can determine antenna efficiency or radiated power statistically without free-space far-field separation. Chamber calibration, usable frequency, loading, stirring independence, mismatch, reference antennas or the chosen one-/two-/three-antenna method, and uncertainty still matter. IEC 61000-4-21 includes reverberation-chamber procedures and antenna-efficiency guidance; NIST's three-antenna method separates total and radiation-efficiency estimation without assuming a known antenna efficiency.
Wheeler cap for a bounded small-antenna problem
A Wheeler-cap method can estimate radiation efficiency for an electrically small antenna when the conductive enclosure suppresses radiation while preserving the loss behavior closely enough. Cap resonances, antenna-to-cap coupling, fixture loss, balance, matching and processing method can dominate the result. It is a specialized method, not a general recipe for arbitrary HF antennas or arrays. IEEE's current recommended practice for antenna measurements covers efficiency measurement and Wheeler-cap technique boundaries.
A Practical Test Plan
- Name the measurand: impedance, local field, pattern, gain, radiated power, radiation efficiency or total efficiency.
- Draw the reference plane: include or de-embed every cable, balun, tuner, fixture and adapter deliberately.
- Define the antenna state: frequency, power, geometry, mounting, return path, feed-line route, ground and surrounding objects.
- Select a complete method: validated far-field range, calibrated near-field scan, characterized reverberation chamber or a properly bounded small-antenna method.
- Calibrate the entire path: instrument ports alone are not enough when probes, cables, fixtures and range reflections remain.
- Quantify perturbation and uncertainty: repeat positions, orientations and cable routes; report noise floor, drift and environmental variation.
- Keep claims inside the evidence: a local-field map is a local-field map until a validated method transforms or integrates it into a radiated quantity.
Primary standards and measurement sources
- IEEE Std 145-2025: current antenna terminology.
- IEEE Std 149-2021: current recommended practice for antenna measurements, including range and efficiency methods.
- IEEE P1720: current project for recommended near-field antenna measurement practice, including scanning geometry, probe compensation, transformation and uncertainty.
- ITU-T K.91 (01/2024): field-region behavior and the limitations of distance criteria.
-
ETSI I-ETS 300 457: aperture phase-error basis for
2D²/λand validated near-field-to-far-field measurement. - ITU-R SA.1345: near-field measurement and modelling when conventional far-field range length is impractical for large antennas.
Mini-FAQ
- Does the reactive near field contain no radiated power? No. Reactive energy strongly affects the local fields, but a transmitting antenna still has an outward time-average radiated-power component. “Reactive” does not mean “radiation is zero.”
-
Is three wavelengths always enough for a far-field antenna test? No. The required range depends on antenna dimensions, geometry, direction, accuracy, probe aperture and site quality. For a large aperture, the common
2D²/λcriterion can require much more distance. - Can a handheld VNA measure antenna efficiency? Not from a walk-around probe reading. A VNA can measure impedance at a calibrated feed plane and can serve inside a complete range, chamber or scanning method, but efficiency also requires radiated and accepted-power information.
- Can gain or efficiency be obtained from near-field data? Yes, when calibrated complex near-field samples are sufficiently dense and accurate, the probe is characterized and corrected, and a validated near-field-to-far-field or radiated-power method with uncertainty is used.
- Does good SWR prove good radiation efficiency? No. SWR describes reflection at a reference plane. Radiation efficiency compares radiated power with accepted power; loss can produce a good match without useful radiation.
- What should I record before comparing two antennas? Record the measurand, feed reference plane, accepted power, geometry, return path, feed-line route, environment, polarization, probe and calibration, sampling plan, repeatability and uncertainty.