TEST & EVALUATION · METHOD

Hardware-in-the-loop radar testing

Hardware-in-the-loop testing puts a real radar or sensor through a synthetic world. Instead of flying real targets past the system, a test bench generates the radar returns the system would see and feeds them in at radio frequency, so the system under test behaves as if it is looking at aircraft, drones, clutter, and jamming that are not actually there. It is how a radar gets proven before it ever points at the sky. The phrase people use is HWiL. The hardware in the loop is the actual production radar or seeker, wired into a simulation that closes the loop around it in real time.
Why it matters

Why anyone bothers

Live-sky testing is expensive, slow, weather-dependent, and hard to repeat. You cannot summon a specific swarm geometry, a specific jamming technique, and a specific clutter background on demand, twice, identically. Hardware-in-the-loop can. The value is in three things a range cannot easily give you.

Repeatability. The same scenario runs again and again with one variable changed, which is what you need to actually characterize a system rather than collect anecdotes about it.

Controllability. You can set a target's radar cross section, its Doppler, its range, and the jamming environment to exact values, including values that would be unsafe, unavailable, or restricted to produce in the open air.

Coverage of the hard cases. The threats that matter most are often the hardest to fly: saturating raids, sophisticated electronic attack, low-observable targets. A bench can produce them on demand.

The physics

What the bench has to reproduce

A radar return is not one thing. To present a credible target to a real radar, the bench has to get several physical quantities right at once.

Range, by delaying the return. A radar measures distance from the time a pulse takes to come back, so the bench delays the synthetic return by the corresponding amount.

Velocity, by shifting the frequency. A moving target Doppler-shifts the return, and the bench applies that shift so the radar reads the intended closing speed.

Size and character, by setting amplitude and signature. A target's radar cross section sets how strong the return is, and finer structure in the return can carry signatures a radar may exploit.

Environment, by adding clutter and multipath. Real returns arrive against a background of ground, sea, weather, and reflections, and a credible bench includes that background rather than a single clean target.

Electronic attack, by injecting jamming. Range-gate and velocity-gate deception, noise, and false targets belong in the test, because a system that only works in a clean environment is not the system being bought.

Methods

Two ways to make the returns

There is a real engineering split in how a bench generates the synthetic scene, and it is worth understanding because it shapes what a given bench can and cannot do.

Digital methods capture the radar's own transmitted pulse, hold it, modify it in memory, and play it back. This is the family that includes digital radio frequency memory, or DRFM, techniques. It is flexible and precise.

Physical or scatterer-based methods construct the scene from controllable scattering elements, building the returns the radar expects from a structured field rather than from pure digital playback.

The neutral point: both families are legitimate, and which one fits depends on the radar, the test objective, and the budget. Rogue River Tech indexes the methods. It does not rank them.

Where it sits

From the bench to the range

Hardware-in-the-loop is one stage in a longer chain, not the whole of test and evaluation. A common progression runs from modeling and simulation in pure software, to hardware-in-the-loop that adds the real hardware into that simulated world, to integrated ground testing, to captive-carry or installed-system testing, and finally to live events on a range.

Each stage is more realistic and more expensive than the last, so hardware-in-the-loop earns its place by retiring risk cheaply before the costly stages begin. In the United States, the Department of Defense framework for operational and live-fire test and evaluation is set by the office of the Director, Operational Test and Evaluation.

The rules

Standards and spectrum

Two governance threads touch hardware-in-the-loop testing: the standards that define how systems are tested, and the spectrum authority that governs any test that transmits.

Test standards and methods
ASTM Committee F38, Unmanned Aircraft Systems ↗
Design, performance, and acceptance-test standards for unmanned aircraft systems.
NIST standard test methods for small UAS ↗
Quantitative methods for system capability and operator proficiency, standardized through ASTM E54.09.
DOT&E, Director, Operational Test and Evaluation ↗
Department of Defense operational and live-fire test and evaluation policy.
A single canonical counter-UAS sensor test standard does not yet exist. Rogue River Tech links the bodies that govern adjacent ground and indexes the field as it develops.
Spectrum authority

Even a bench radiates, and any open-air or installed test that transmits needs the right authorization. In the United States the relevant path is the FCC Experimental Radio Service under 47 CFR Part 5, administered by the FCC Office of Engineering and Technology. Experimental authority is granted on a non-interference basis, and the FCC coordinates shared federal bands with the National Telecommunications and Information Administration.

There is one notable exception worth knowing. Under 47 CFR 5.7(g), an experimental license is not required when a radiofrequency device operates fully contained within an anechoic chamber or a Faraday cage. A shielded bench that does not radiate into open air sits outside the licensing requirement, while a test that puts energy into the open air or into an installed system does not.

FCC Part 5, Experimental Radio Service ↗ 47 CFR Part 5 (eCFR) ↗

For the bands, incumbents, and open proceedings that shape what a test may transmit and where, see the spectrum reference, and the SIGINT and RF policy tracker.

Live companion
See the scatterer method in motion
The E-RES scatterer console is an interactive teaching visual for how a physical, scatterer-based scene builds the returns a radar expects to see. A conceptual model, with its limits stated on the page.
Open the E-RES console →
FAQ

Common questions

What does HWiL stand for?
HWiL stands for hardware-in-the-loop. The hardware is the real radar, seeker, or sensor under test, and it is wired into a real-time simulation that generates the signals it would encounter, closing the loop around the actual hardware rather than a software model of it.
How is hardware-in-the-loop different from pure simulation?
Pure modeling and simulation tests a software model of the system. Hardware-in-the-loop tests the real production hardware by feeding it synthetic radar returns at radio frequency. It catches behavior that a model misses, because the actual receiver, processor, and firmware are in the loop.
Do you need an FCC license to run a radar test bench?
It depends on whether the bench radiates into open air. Under 47 CFR 5.7(g), a device operating fully within an anechoic chamber or Faraday cage does not require an experimental license. An open-air or installed test that transmits generally needs authorization under the FCC Experimental Radio Service, 47 CFR Part 5, coordinated with NTIA for shared federal bands.
What are the two main ways a bench generates synthetic targets?
Broadly, digital methods such as digital radio frequency memory capture and replay the radar's own pulse with controlled modifications, while physical or scatterer-based methods build the scene from controllable scattering elements. Both are legitimate; the right choice depends on the radar, the objective, and the budget.
Rogue River Tech curates, aggregates, and summarizes public information on RF and test and evaluation. This explainer is educational and is not legal, engineering, or licensing advice. Standards and spectrum rules change, confirm the current text against the linked primary sources before relying on it. Last reviewed 2026-06-25.