Cross-Range Resolution
How Radar Separates Targets Across the Beam
A radar locates targets in two orthogonal dimensions. Down-range separation comes from timing the echo and is governed by waveform bandwidth, but lateral, or cross-range, separation comes from where the energy points in angle. In a real-beam system the antenna cannot distinguish two scatterers that fall inside the same illuminated angular cell, so the cross-range cell size is simply the slant range times the 3 dB beamwidth. Because beamwidth is fixed by the aperture and wavelength while the range grows, the lateral cell stretches with distance. A surveillance radar that cleanly separates two aircraft at 5 km may merge them into a single blob at 80 km, even though its range resolution is unchanged.
The only way to shrink the real-beam cell is a larger aperture or a shorter wavelength, both of which have practical ceilings. A 3 m X-band antenna already pushes the limits of an airborne nose radome, and moving to W-band buys angular sharpness at the cost of atmospheric loss and shorter range. This is why high-resolution imaging radars abandon the real beam entirely and form a synthetic aperture, trading antenna size for coherent processing of many pulses collected as the platform translates past the scene.
From Real Beam to Synthetic Aperture
Synthetic aperture radar treats the sequence of pulses gathered along the flight path as elements of one enormous array. Each target's echo carries a Doppler frequency that sweeps as the geometry changes, and matched filtering that phase history focuses the energy into a cross-range cell far smaller than the real beam would allow. In stripmap mode the achievable cell equals half the physical antenna length, independent of range, so a deliberately small antenna with a wide beam keeps each target illuminated longer and yields finer resolution. Doppler beam sharpening is the unfocused cousin of this technique, improving cross-range by a modest factor without full SAR processing.
Cross-Range Resolution Equations
δcr ≈ R × θ3dB ≈ R × (k × λ / D)
Stripmap SAR (focused):
δcr = λ × R / (2 × Lsar) ≈ D / 2
Doppler-limited cross-range:
δcr ≈ λ × R / (2 × v × Tdwell)
Where R = slant range, θ3dB = azimuth beamwidth (rad), k ≈ 0.886 (uniform aperture), λ = wavelength, D = physical antenna length, Lsar = synthetic aperture length, v = platform velocity, Tdwell = coherent integration time. Example: X-band λ = 3 cm, R = 10 km, D = 1.5 m → θ3dB ≈ 1° and δcr ≈ 175 m real-beam, versus 0.75 m focused stripmap SAR.
Cross-Range Resolution by Radar Mode
| Mode | Cross-range driver | Range dependence | Typical δcr | Best application |
|---|---|---|---|---|
| Real-beam (surveillance) | Antenna 3 dB beamwidth | Linear with R | 50 to 300 m at 10 km | Search, air traffic control |
| Doppler beam sharpening | Unfocused Doppler gradient | Weakly with R | 10 to 30 m at 10 km | Forward-looking ground mapping |
| Stripmap SAR | Synthetic aperture (D/2 floor) | Independent of R | 0.5 to 3 m | Wide-area terrain imaging |
| Spotlight SAR | Extended steered dwell | Independent of R | 0.1 to 0.3 m | High-detail scene imaging |
| ISAR | Target rotational Doppler | Depends on aspect rate | 0.3 to 1 m | Ship and aircraft classification |
Frequently Asked Questions
How does cross-range resolution differ from range resolution?
Range resolution separates targets along the line of sight and equals c / 2B, depending only on transmitted bandwidth, so 150 MHz gives 1 m at any distance. Cross-range resolution separates targets across the beam; in a real-beam radar it equals R × θ3dB, so a 1° beam yields 175 m at 10 km but 1.75 m at 100 m. The two axes are independent. SAR removes the range dependence by synthesizing a long aperture, reaching a D/2 floor set by the physical antenna length.
Why is SAR cross-range resolution equal to half the antenna length?
In stripmap SAR a point target stays illuminated only while inside the real beam, so the longest usable synthetic aperture is Lsar = Rλ/D, the beam footprint. Focused resolution is λR / (2Lsar), and substituting gives δcr = D/2. A smaller physical antenna has a wider beam, illuminates each target longer, and therefore builds a longer synthetic aperture and finer resolution. Spotlight SAR steers the beam to dwell longer and beats the stripmap floor at the cost of coverage area.
What cross-range resolution does an automotive radar achieve at 77 GHz?
A 77 GHz radar with a roughly 47 mm receive aperture has about a 5° azimuth beamwidth, giving 4.4 m cross-range at 50 m, too coarse to split two cars in adjacent lanes. Imaging radars expand the MIMO virtual array to 100+ channels, narrowing the beam below 1.5° for about 1.3 m cross-range at 50 m, enough to separate a pedestrian from a guardrail. Resolution still degrades linearly with range because these are real-beam systems, not SAR.