HPM Counter-UAS System Analysis

• HPM Counter-UAS systems are directed-energy solutions that disable small drones by coupling microwave energy into onboard electronics rather than using kinetic intercept.
• The design framework integrates electromagnetic propagation, antenna analysis, target coupling, and probabilistic semiconductor damage modeling validated through extensive Monte Carlo simulations.
• A reproducible Python implementation underpins the methodology, ensuring consistent assessment of system kill probabilities, safety zones, and thermal constraints.

A high-power microwave (HPM) counter-unmanned aerial system (counter-UAS) is a directed-energy system that uses continuous-wave or pulsed microwave radiation to disable or disrupt small unmanned aerial systems (sUAS) by coupling electromagnetic energy into vulnerable onboard electronics rather than relying on kinetic intercept or conventional radio-frequency jamming. In the recent design literature, the topic is represented by a reproducible multi-physics workflow for a 2.45\,GHz system that integrates electromagnetic propagation modelling, antenna pattern analysis, electromagnetic coupling to unshielded drone wiring harnesses, a sigmoid-based semiconductor damage probability model calibrated to published CMOS latchup thresholds, and a $10{,}000$-trial Monte Carlo uncertainty analysis to produce system-level kill probabilities, design maps, safety exclusion zones, thermal requirements, and waveguide mode analysis (Jafari et al., 9 Feb 2026).

1. System definition and design workflow

The 2026 framework formalizes an HPM counter-UAS system as a coupled chain from source power to target neutralization. Its stated toolchain is: link-budget $\rightarrow$ coupling $\rightarrow$ damage $\rightarrow$ kill probability $\rightarrow$ Monte Carlo uncertainty $\rightarrow$ design maps $\rightarrow$ safety & thermal $\leftrightarrow$ waveguide losses. The framework is intended for non-kinetic neutralization of sUAS operating under autonomous guidance, explicitly targeting scenarios in which conventional radio-frequency jamming is insufficient (Jafari et al., 9 Feb 2026).

Within this formulation, “kill probability” is not treated as a purely radiometric thresholding problem. Instead, it is the output of a multi-stage model in which propagation determines incident field, antenna pointing and polarization determine effective radiated coupling, wiring-harness geometry determines induced voltage, and subsystem susceptibility determines the probability of disabling one or more critical electronic functions. This makes the architecture inherently probabilistic rather than deterministic.

A central feature is reproducibility. The framework states that all simulation codes and results are provided for full reproducibility, and the details further specify that the code is provided in Python. This is significant because HPM counter-UAS performance estimates are unusually sensitive to modeling assumptions about aperture efficiency, pointing error, wire orientation, and component-level damage thresholds; a reproducible implementation constrains ambiguity in those assumptions.

2. Electromagnetic propagation and aperture physics

The operating frequency is $f_{0}=2.45$\,GHz, corresponding to $\lambda_{0}=c/f_{0}=0.1224$\,m. Propagation is modeled in free space using the Friis power-density relation

$\rightarrow$0

where $\rightarrow$1. The incident field magnitude is

$\rightarrow$2

with $\rightarrow$3. The corresponding free-space path loss is

$\rightarrow$4

The baseline aperture is a parabolic reflector with $\rightarrow$5\,m and $\rightarrow$6. Its gain is

$\rightarrow$7

which gives $\rightarrow$8\,dBi. The approximate half-power beamwidth is

$\rightarrow$9

Pointing-error loss is modeled by sampling $\rightarrow$0, defining $\rightarrow$1, and applying

$\rightarrow$2

Polarization mismatch is represented as

$\rightarrow$3

The total antenna-chain effective isotropic radiated power factor is then

$\rightarrow$4

These relations establish the front end of the engagement problem. A common misconception is that HPM effectiveness is determined mainly by nominal transmitter power. The model structure instead makes aperture diameter, beamwidth, pointing jitter, and polarization alignment explicit co-determinants of delivered field.

3. Coupling to onboard wiring and subsystem susceptibility

The target-side coupling model treats unshielded drone wiring as the principal electromagnetic entry path. For a straight wire of length $\rightarrow$5 acting as a short dipole, the induced voltage is

$\rightarrow$6

Near the half-wave condition, resonance enhancement is introduced through

$\rightarrow$7

with $\rightarrow$8 and $\rightarrow$9\,m. The framework states that this predicts a $\rightarrow$0–$\rightarrow$1 boost for $\rightarrow$2\,cm. The immediate implication is that vulnerability is strongly geometry-dependent even at fixed incident field strength.

Semiconductor damage is modeled with a logistic susceptibility law,

$\rightarrow$3

For five subsystems, the nominal parameters are as follows (Jafari et al., 9 Feb 2026):

Subsystem	$\rightarrow$4 (V/m)	$\rightarrow$5 (V/m)
GPS/GNSS LNA	150	30
Flight controller	250	50
ESC	300	60
Camera	200	40
BMS	350	70

System-level kill probability is aggregated by

$\rightarrow$6

This aggregation means the framework does not require simultaneous destruction of all avionics. Neutralization is represented as the complement of the probability that all five subsystems survive. A plausible implication is that subsystem diversity broadens the transition region between negligible and near-certain system kill.

4. Monte Carlo methodology and predicted engagement performance

Uncertainty is propagated with $\rightarrow$7 trials. The stochastic variables are sampled as follows: $\rightarrow$8, $\rightarrow$9, $\rightarrow$0, $\rightarrow$1, $\rightarrow$2, $\rightarrow$3\,m, and $\rightarrow$4. The simulation loop is described explicitly: sample all variables, compute $\rightarrow$5 via Friis plus loss factors, compute $\rightarrow$6 per subsystem, draw Bernoulli for system kill, and accumulate count. The $\rightarrow$7 confidence intervals are obtained by Clopper–Pearson exact binomial (Jafari et al., 9 Feb 2026).

For the baseline configuration of $\rightarrow$8\,kW continuous-wave power and a $\rightarrow$9\,cm dish, the predicted kill probability is $\rightarrow$0 at $\rightarrow$1\,m and $\rightarrow$2 at $\rightarrow$3\,m. Under pulsed operation at $\rightarrow$4\,kW peak power with $\rightarrow$5 duty cycle, the framework reports that the $\rightarrow$6 kill range extends from approximately $\rightarrow$7\,m to $\rightarrow$8\,m. The first-order range scaling is

$\rightarrow$9

because $\rightarrow$0 and the kill range occurs when $\rightarrow$1.

The same paper generates parametric design maps by solving $\rightarrow$2 over a $\rightarrow$3 grid and plotting contours of $\rightarrow$4 as a heatmap. In this construction, the ESC threshold is used as the deterministic reference for the map, with interpolation refining the first-order relation $\rightarrow$5.

These results frame HPM counter-UAS engagement as a probabilistic performance surface rather than a single maximum-range specification. They also show that pulsed peak power and average thermal load can be decoupled operationally: the cited pulsed case has $\rightarrow$6\,kW peak power but $\rightarrow$7\,kW average power at $\rightarrow$8 duty cycle.

5. Safety limits, thermal management, and waveguide implementation

The framework extends beyond engagement probability to engineering constraints. Safety exclusion zones are calculated from ICNIRP 2020 public limits at $\rightarrow$9\,GHz: $\leftrightarrow$0\,W/m$\leftrightarrow$1 for occupational exposure and $\leftrightarrow$2\,W/m$\leftrightarrow$3 for the general public. Using on-axis main-beam average power density,

$\leftrightarrow$4

the exclusion radius is obtained by solving $\leftrightarrow$5, giving

$\leftrightarrow$6

For $\leftrightarrow$7\,kW CW operation with the $\leftrightarrow$8\,cm dish and $\leftrightarrow$9\,dBi gain, the exclusion distances are $f_{0}=2.45$0\,m and $f_{0}=2.45$1\,m. In pulsed mode with duty $f_{0}=2.45$2, the time-averaged relation is $f_{0}=2.45$3 (Jafari et al., 9 Feb 2026).

Thermal management is quantified with $f_{0}=2.45$4 for the magnetron and $f_{0}=2.45$5 for the power supply unit. At $f_{0}=2.45$6\,kW CW output, the input to the magnetron is $f_{0}=2.45$7, corresponding to approximately $f_{0}=2.45$8\,kW heat in the magnetron; PSU losses are $f_{0}=2.45$9; total heat is therefore approximately $\lambda_{0}=c/f_{0}=0.1224$0\,kW. The reported design rule is that forced-air cooling is adequate below approximately $\lambda_{0}=c/f_{0}=0.1224$1\,kW load $\lambda_{0}=c/f_{0}=0.1224$2, whereas liquid cooling is required above that level.

Waveguide implementation is analyzed with WR-340 dimensions $\lambda_{0}=c/f_{0}=0.1224$3\,mm and $\lambda_{0}=c/f_{0}=0.1224$4\,mm. The TE$\lambda_{0}=c/f_{0}=0.1224$5 cutoff is

$\lambda_{0}=c/f_{0}=0.1224$6

while TE$\lambda_{0}=c/f_{0}=0.1224$7 cuts off at $\lambda_{0}=c/f_{0}=0.1224$8\,GHz, so operation at $\lambda_{0}=c/f_{0}=0.1224$9\,GHz is single-mode. Wall-loss attenuation for TE$\rightarrow$00 is given by the stated expression for $\rightarrow$01, with $\rightarrow$02, and evaluates numerically to $\rightarrow$03\,dB/m, described as negligible over approximately $\rightarrow$04\,m. This part of the framework is important because it ties radiated performance to physically deployable feed hardware rather than idealized antennas.

6. Relation to vacuum-tube source classes and broader probabilistic HPM modeling

HPM counter-UAS systems are tightly coupled to source technology. A review of high-power vacuum-tube microwave sources states that vacuum devices remain the mainstream microwave sources for applications including microwave weapons, and it organizes them by Cherenkov or Smith–Purcell radiation, transition radiation, and Bremsstrahlung. For counter-UAS integration, the six major source classes summarized are Vircators, Magnetrons, Klystrons, TWTs, Gyrotrons, and FELs or FEMs (Xu, 2020).

For portable counter-UAS applications, the same source survey concludes that systems will most likely employ magnetrons or compact TWT/Klystron amplifiers in the $\rightarrow$05–$\rightarrow$06\,GHz range to balance portability, sufficient power to disrupt UAS electronics, frequency agility for countermeasure evasion, and coherent beam steering via waveguide or phased-array antennas; pulsed Vircators or BWOs may be co-located as “single-shot” disruptors, while Gyrotrons and FELs remain confined to fixed installations or research facilities. This aligns with the 2.45\,GHz multiphysics design, which explicitly models magnetron efficiency and a dish-fed architecture (Jafari et al., 9 Feb 2026).

A complementary 2025 framework approaches the same operational problem from an antenna- and propagation-centric perspective. It models received pulse energy under a standard Friis link with log-normal beam-jitter and atmospheric loss, derives analytically evaluable per-pulse and cumulative neutralization probabilities using Gaussian–Hermite quadrature, and provides a dwell-time expression under a standard pulse-independence assumption (Khalil et al., 18 Oct 2025). Its sensitivity analysis defines

$\rightarrow$07

and reports that performance is dominated by mean slant range, with $\rightarrow$08; transmit power and pulse width have elasticity $\rightarrow$09; aperture diameter can approach quadratic benefit only when jitter is negligible; mean atmospheric loss is moderately negative; and atmospheric-loss variance has only a very small positive elasticity.

This broader literature corrects another common misconception: increasing peak power is not the only, or even always the dominant, design lever. The analytical sensitivity result $\rightarrow$10 indicates that a $\rightarrow$11 range increase implies a $\rightarrow$12 energy loss, while the 2026 multiphysics framework shows that the same transmitted power can produce markedly different kill probabilities once pointing error, polarization mismatch, wire resonance, and subsystem susceptibility are included. Taken together, these results suggest that HPM counter-UAS design is fundamentally a joint problem in source selection, aperture control, engagement geometry, and target vulnerability modeling rather than a single-parameter power-scaling exercise.