Airborne radar measures distance by timing a radio pulse out and back, measures velocity by reading the frequency shift of the return, and — in modern AESA arrays — does both for hundreds of targets at once while painting ground maps and jamming hostile sensors in the same second.
See the F-22 Raptor and its AN/APG-77 AESAA radar transmits a short pulse of radio energy. The pulse travels outward at the speed of light, strikes a target, and a small fraction of the energy reflects back to the receiver. Range is simply distance = c × t / 2, where c is the speed of light (about 300,000,000 m/s) and t is the round-trip time. A target at 100 km returns the pulse in roughly 667 microseconds.
The radar then waits, listens, and fires another pulse. The rate at which pulses go out is the pulse repetition frequency (PRF). Long-range search radars use a low PRF — perhaps 1,000 pulses per second — so each pulse has time to return from a distant target before the next one goes out. High-PRF modes used for closing fighters can run 200,000 pulses per second, trading unambiguous range for a clean velocity measurement.
When the target is moving, the returning pulse comes back at a slightly different frequency than it left. A target closing on the radar compresses the wavefronts and shifts the return upward in frequency; one opening away stretches them and shifts it downward. This is the same effect that changes the pitch of a passing ambulance siren.
The shift is small — at X-band (10 GHz), a target closing at 1,000 mph (about 447 m/s) produces a Doppler shift of roughly 30 kHz. Modern signal processors measure this shift to about 1 m/s precision and use it to discriminate fast-moving aircraft from stationary clutter.
An airborne radar looking down sees the ground reflecting energy from every direction. A fighter cruising at 30,000 ft sees ground returns that are typically 40-70 dB stronger than a small target hiding in the clutter. Early radars such as the AN/APQ-72 on the F-4 Phantom could not look down — the ground washed out anything below the radar's horizon.
Pulse-Doppler radar fixes this by exploiting velocity. Ground returns directly below the aircraft have zero Doppler relative to the world. Returns slightly off to the side have predictable Doppler that varies with the radar's own forward speed. A target moving differently — a closing fighter, a low-flying cruise missile — appears as a Doppler return that does not fit the ground clutter band. The radar filters out the predictable returns and keeps what is left.
This is what gave the F-15 Eagle its first true look-down/shoot-down ability in 1976 with the AN/APG-63, and what lets the E-3 Sentry AWACS pick low-flying aircraft out of ground returns at 250 nautical miles.
Where pulse-Doppler is the airborne fighter's solution, surface and battlefield radars use moving target indication (MTI). The radar fires two consecutive pulses at the same patch of sky, subtracts one return from the other, and keeps only the difference. Static returns cancel; anything that has moved between pulses survives the subtraction.
MTI is simpler than pulse-Doppler and works well at fixed sites where the radar itself is not moving. Most ground-based air-defence radars — the S-400's 91N6E acquisition radar, for example — use MTI variants. Ground moving-target indication (GMTI) is the airborne version used by E-8 Joint STARS and the AN/APY-9 on the E-2D Hawkeye to track vehicles on the ground.
For decades, fighter radars used a mechanically-scanned dish — a parabolic antenna driven by a hydraulic gimbal that physically swept the beam left and right. The AN/APG-63 on the F-15 worked this way, as did the AN/APG-70 on the F-15E. Scan rates topped out near 60 degrees per second, and the dish could only point one direction at a time.
A passive electronically scanned array (PESA) replaces the dish with hundreds or thousands of fixed antenna elements driven by a single transmitter. Phase shifters at each element steer the beam electronically, with no moving parts. The Russian N011M Bars on the Su-30MKI is a PESA, as is the Su-35's Irbis-E, which puts 20 kW of peak power through a passive array steerable mechanically across ±120 degrees in azimuth.
An active electronically scanned array (AESA) replaces the single transmitter with one tiny transmit/receive module behind every element of the array. The AN/APG-77 on the F-22 Raptor has roughly 1,956 such modules; the AN/APG-81 on the F-35 has about 1,676. Each module radiates a few watts; the array sums coherently to produce kilowatts on target.
The benefits of AESA over a mechanically-scanned dish or a PESA are concrete:
The F/A-18E/F Super Hornet received the AN/APG-79 AESA from 2007, replacing the AN/APG-73 mechanical radar. The Su-35's successor, the Su-57, mounts the N036 Byelka — a distributed AESA with one main X-band array in the nose plus side-looking arrays in the cheeks and L-band arrays in the wing leading edges. The L-band elements double as IFF and as a counter-stealth sensor.
Pulse timing gives range. Doppler gives velocity. Neither tells you exactly which direction the target is in. Monopulse processing extracts target angle from a single pulse by comparing the return strength in four overlapping receive beams arranged in a clover-leaf pattern. A target dead-on-axis returns equally in all four; one slightly off produces an imbalance whose direction and magnitude give the angle error to a fraction of a beamwidth.
Monopulse is what lets a fire-control radar guide a missile to within metres of a target rather than within the beamwidth (typically 2-4 degrees, or roughly 1 km of error at 30 km range). Every modern fire-control radar uses some form of monopulse.
Radar bands trade resolution for range and atmospheric loss:
Conventional radar is limited by the horizon — about 250 nm for a target at 30,000 ft against a sea-level radar. Over-the-horizon (OTH) radar uses HF frequencies (3-30 MHz) that bounce off the ionosphere, giving detection ranges of 1,000-3,000 nautical miles. Australia's Jindalee Operational Radar Network (JORN) covers the entire northern approaches to the continent from three sites. Russia's Container 29B6 system covers Europe out to 3,000 km.
The trade-off: OTH radar resolution is poor — tens of kilometres typically — and ionospheric conditions vary by hour and season. OTH detects that a raid is coming; it cannot tell you which aircraft are in it.
The combination of pulse-Doppler clutter rejection, AESA beam agility, and monopulse angle precision is what makes a modern beyond-visual-range engagement possible. The F-22 can detect a 5 m² target at roughly 240 km with the AN/APG-77 in long-range search mode, track it through clutter, hand it off to an AIM-120 AMRAAM via two-pulse mid-course datalink, then drop the track from its own emissions while the missile's seeker takes over inside 15 km. None of this is possible without every layer of radar processing covered above working together.
Content adapted from publicly available aeronautical engineering and defence references. Vehicle data sourced from the Who That Plane?! gallery.