Technical Discussion: Limitations of War Thunder’s Fixed-Altitude Multipath Model

Machinery of War Discussion Aircraft
,
1

In War Thunder, low-altitude flight—including terrain masking and sea-skimming—is widely used to evade radar-guided missiles. The June 19, 2024 “Seek & Destroy” update changelog [1] stated that the maximum altitude of the multipath effect for airborne radars and missile radar seekers had been reduced from 100 m to 60 m. The same changelog explained that this effect had originally been introduced for gameplay reasons.

That source alone does not reveal how the magnitude of the effect varies with altitude, whether system-specific coefficients are used, or how the mechanic is implemented in the current build. It does, however, show that the officially described model at that time applied a common “maximum altitude of the multipath effect” to airborne radars and missile radar seekers.

Based on publicly available radar-engineering theory, this post argues that real-world multipath tracking errors depend not only on target height, but also on engagement geometry, surface conditions, waveform and signal processing, and seeker characteristics. A game model based on a common maximum altitude is easy to understand from a gameplay perspective, but it substantially simplifies the underlying physics. The purpose of this post is therefore to provide a basis for discussing possible improvements to the model.

1. Engagement Geometry and Variation in Path-Length Difference

An important quantity when assessing multipath effects is the path-length difference between the direct signal from the target and the signal that reaches the seeker after reflecting from the surface. In a simplified model with a locally flat reflecting surface, the one-way geometric path-length difference (ΔL) can be approximated as follows:

ΔL = √[R² + (H+h)²] − √[R² + (H−h)²]

The variables are defined relative to the same local reflecting plane, rather than to mean sea level:

  • H: seeker height above the local planar reflecting surface
  • h: target height above the same reflecting surface
  • R: horizontal separation between the seeker and the target

When the target height h is sufficiently small relative to the seeker height H and horizontal separation R, a first-order approximation can be written using the reference angle α₀:

ΔL ≈ 2h sin α₀

α₀ = arctan(H/R)

Here, α₀ is a reference angle used for the path-difference approximation. It is not necessarily identical to either the exact grazing angle at the specular reflection point or the seeker’s look-down angle.

For a conventional monostatic radar, the nominal range resolution ΔR can be written in terms of the effective bandwidth B and the speed of light c:

ΔR ≈ c / (2B)

This equation alone, however, is not sufficient to determine whether an actual missile seeker can separate multipath components. In an active radar homing (ARH) system, for example, four round-trip propagation paths may exist [2]:

  • Direct transmit path / direct receive path
  • Direct transmit path / reflected receive path
  • Reflected transmit path / direct receive path
  • Reflected transmit path / reflected receive path

Whether these components can be separated depends on factors including their round-trip path differences, the waveform ambiguity function, matched-filter sidelobes, range-gate design, signal-to-noise ratio (SNR), the amplitudes and phases of the reflected components, and the seeker’s signal-processing methods.

In a semi-active radar homing (SARH) system, the illuminating radar—typically on the launching aircraft—and the missile seeker are spatially separated, forming a bistatic geometry. Direct and reflected paths may therefore exist on both the illuminator-to-target leg and the target-to-seeker leg. Consequently, conclusions based on the range resolution or path geometry of a monostatic ARH system cannot simply be applied to an SARH missile such as the AIM-7.

2. Surface Conditions and the Rayleigh Roughness Criterion

The electromagnetic roughness of the surface is an important factor in determining whether a strong coherent specular-reflection component is produced. Assuming a Gaussian surface-height distribution, the tangent-plane approximation, a sufficiently large surface radius of curvature, and a gentle root-mean-square (RMS) slope, the Rayleigh roughness parameter R_a and the coherent scattered-intensity attenuation factor A_coh can be expressed as follows [3]:

R_a = (2πσ_h / λ) sin γ

A_coh = exp[−4(R_a)²]

where:

  • γ: the actual grazing angle of the reflected path, measured from the reflecting surface
  • σ_h: the RMS surface height
  • λ: the radar wavelength

For fixed σ_h and λ, geometries in which the seeker is higher than the target and the horizontal separation is comparatively short tend to produce a larger grazing angle γ. This increases R_a and reduces A_coh. A reduction in the coherent specular component does not mean that diffuse scattering, clutter, or localized specular reflections disappear.

Furthermore, A_coh represents only the attenuation of the coherent component caused by surface roughness; it is not the total reflected-signal intensity. The actual reflected intensity also depends on the smooth-surface reflection coefficient, the medium’s dielectric properties, polarization, incidence angle, and other factors.

3. Seeker Types and the Limits of Public Information

David K. Barton’s general theory [4] discusses low-angle and monopulse tracking errors when the direct and reflected components fall within the radar beam. The actual tracking limits of a specific missile, however, also depend on the characteristics of its seeker.

According to a Raytheon “AIM/RIM-7 Sparrow” document included in a U.S. Navy source [5], the AIM/RIM-7M was developed around a digital monopulse seeker. For the AIM-120, a modern ARH missile, a U.S. Air Force fact sheet [6] provides only the general statement that it offers improved capability against low-altitude targets compared with the Sparrow.

That general capability statement is not sufficient to infer the AIM-120’s specific multipath-mitigation methods, bandwidth, minimum tracking altitude, range resolution, or other internal parameters. Research examples using wideband radar and super-resolution processing do exist [2], but they do not demonstrate that the same methods are implemented in any particular operational missile.

4. Proposal for Improving the Game Mechanic

Real-world missile avoidance at low altitude depends on a combination of factors, not multipath alone. These include terrain masking, surface and sea clutter, and the missile’s guidance, control, and kinematic performance.

Doppler notching is another important factor. Here, Doppler notching means maneuvering so that the line-of-sight component of relative velocity—and therefore the corresponding Doppler frequency as observed by the radar or seeker—approaches the clutter-rejection region or another unfavorable region of the velocity-processing gate. In an SARH engagement, the transmitter and receiver are spatially separated, so the exact condition depends on bistatic Doppler geometry.

A model based on a common maximum altitude, as described in the 2024 official changelog, simplifies many of these factors. Rather than removing multipath entirely, I propose moving toward a more continuous model that considers the following:

  1. Independent evaluation of engagement geometry: Evaluate the seeker-to-target look-down angle and the grazing angle at the reflection point separately, using the horizontal separation and the target and seeker heights above the local surface.

  2. Consideration of surface type: Apply roughness-related attenuation according to whether the reflecting surface is open water, flat ground, or another surface type.

  3. Gameplay-level categorization of seeker characteristics: Where public information is available, treat guidance architecture (SARH/ARH), angular-tracking method, and signal-processing generation as separate parameters. Where public information is insufficient, any balancing coefficients should be explicitly presented as gameplay abstractions rather than as a real-world performance ranking.

  4. Combination with Doppler conditions: Evaluate the combined influence of line-of-sight Doppler conditions and multipath instead of treating either factor in isolation.

References

2

wow

3

Multipath shouldn’t even be a thing on modern ARHs, period, at least not on the very flat surfaces where it is easy to abuse like the sea in Nuke thunder

Also can we please avoid AI