As the modern battlefield becomes increasingly saturated with electromagnetic threats, the vulnerability of small fixed-wing unmanned aerial vehicles (UAVs) to external electromagnetic interference (EMI) has emerged as a critical concern. This study, conducted from our research perspective, investigates the continuous-wave (CW) electromagnetic radiation effects on a small fixed-wing China drone using an integrated approach combining numerical simulation and controlled experimentation. The primary focus is on the field-to-line coupling mechanism, which is identified as the dominant back-door coupling path leading to flight control instability. By establishing a detailed system-level model within the CST electromagnetic simulation environment and executing a series of irradiation tests in an anechoic chamber, we elucidate how external electromagnetic fields couple into internal cables, particularly the servo signal lines, and ultimately induce undesirable tail oscillations. Our findings reveal a strong correlation between the resonance frequencies of these signal lines and the sensitive frequency points observed in the experiments. This research provides a theoretical and experimental foundation for the electromagnetic hardening design of China drone systems and offers insights into potential countermeasure strategies against EMI threats.
The proliferation of small fixed-wing UAVs in military and civilian applications has highlighted their strategic value. However, the dense integration of electronic subsystems, including flight control, navigation, and data link modules, makes them susceptible to intentional and unintentional electromagnetic interference. Compared to quadrotor UAVs, the fixed-wing configuration presents a unique challenge due to the long, slender airframe that houses long signal wiring running from the flight controller to the tail-mounted servos. In the context of a China drone, which is often designed for cost-effectiveness and lightweight performance, comprehensive electromagnetic shielding is frequently compromised, creating vulnerability windows for back-door coupling. Unlike front-door coupling through antennas, which can be mitigated by limiters and filters, back-door coupling through cables and apertures is less predictable and more dependent on the specific geometry and layout of the drone. Therefore, understanding the field-to-line coupling process is paramount for enhancing the electromagnetic survivability of these platforms.
Earlier research in our laboratory has extensively explored front-door coupling effects on UAV data links and navigation systems. However, the back-door coupling mechanisms, particularly for fixed-wing designs, remained less understood. Our work aims to bridge this gap by focusing on the servo signal lines, which are directly connected to the control surfaces critical for flight stability. The electromagnetic topology of the China drone suggests three primary back-door coupling paths: direct cable coupling (CP1), radiation coupling through structural apertures (CP2), and sensor interference (CP3). Among these, CP1 is hypothesized to be the most efficient due to the resonant properties of the long cables within the tail boom. This hypothesis forms the central thesis of our investigation.
To validate this hypothesis, we first constructed a high-fidelity three-dimensional model of the China drone in CST, incorporating the actual internal cable routing. The model included five typical cable types: two single-wire servo cables (C1 for rudder and elevator), a coaxial cable for the data link (C2), a three-wire twisted power cable for the motor (C3), a coaxial cable for the GPS antenna (C4), and a single power cable (C5). The physical parameters of these cables are summarized in the table below.
| Cable Name | Length (mm) | Conductor Radius r1 (mm) | Insulation Radius r2 (mm) | Outer Conductor Radius r3 (mm) | Jacket Radius r4 (mm) |
|---|---|---|---|---|---|
| C1 (Rudder) | 620.4 | 0.4 | 0.6 | — | — |
| C1 (Elevator) | 815.6 | 0.4 | 0.6 | — | — |
| C2 | 216.7 | 0.5 | 0.84 | 1.0 | 1.5 |
| C3 | 230.9 | 0.5 | 1.2 | — | 1.8 |
| C4 | 226.0 | 0.5 | 0.84 | 1.0 | 1.5 |
| C5 | 64.7 | 1.0 | 1.5 | — | — |
The simulation environment was configured to model a plane wave illumination, a simplified but effective representation of a far-field CW source. The incident field can be mathematically described by the following equation, where θ, ϕ, and α represent the elevation angle, azimuth angle, and polarization angle, respectively, and k is the wave number.
$$ E(x,y,z) = E(t)(e_x \hat{a}_x + e_y \hat{a}_y + e_z \hat{a}_z) e^{-j(k_x x + k_y y + k_z z)} $$
The components of the electric field vector and the wave number vector are defined as:
$$ \begin{cases} e_x = -\cos\phi \cos\theta \sin\alpha – \sin\phi \cos\alpha \\ e_y = -\sin\phi \cos\theta \sin\alpha + \cos\phi \cos\alpha \\ e_z = \sin\theta \sin\alpha \end{cases} $$
$$ \begin{cases} k_x = -k \sin\theta \cos\phi \\ k_y = -k \sin\theta \sin\phi \\ k_z = -k \cos\theta \end{cases} $$
We simulated four typical interference scenarios by varying the polarization (horizontal α=0° and vertical α=90°) and the angle of incidence (nose-on ϕ=180° and broadside left ϕ=270°). The simulation frequency range was set from 0 to 500 MHz, and the incident field strength was 100 V/m. All cable terminations were set to 50 Ω loads connected to ground to represent typical input impedance of digital circuits.
The simulation results, presented in the next section, clearly show the superiority of horizontal polarization for coupling energy into the longitudinal cables of the China drone. The induced voltage on the rudder signal line (C1) was the highest among all cables, peaking at 7.74 V under horizontal polarization with broadside incidence. This value was 68% higher than the peak induced under vertical polarization for the same incidence angle. The elevator cable (C1) also showed significant coupling, but with slightly different resonant frequencies due to its different routing path. The coaxial cables (C2 and C4) exhibited minimal coupling, confirming the effectiveness of their braided shielding. The power cable (C3) showed moderate coupling, but the induced voltages were well below its operational logic levels, suggesting a lower susceptibility. The key observation was that the servo signal lines (C1) acted as efficient receiving antennas, with resonant peaks appearing at distinct frequencies. The simulation results are summarized below.
| Scenario | Peak Voltage on C1-Rudder (V) | Peak Voltage on C1-Elevator (V) | Peak Voltage on C3 (V) |
|---|---|---|---|
| Horizontal Polarization, Broadside Incidence | 7.74 | ~6.5 | 1.13 |
| Vertical Polarization, Broadside Incidence | 4.61 | ~3.8 | 0.64 |
| Horizontal Polarization, Nose-on Incidence | 4.61 | ~4.0 | 0.98 |
| Vertical Polarization, Nose-on Incidence | 4.03 | ~3.2 | 0.49 |
The resonant frequencies identified from the simulation were approximately 172 MHz, 261 MHz, and 336 MHz for the rudder cable. These values closely matched the theoretical half-wavelength resonance predictions based on the cable’s physical length and the dielectric properties of its insulation. The theoretical resonant frequency can be calculated using the following formula:
$$ f_w = \frac{n v}{2L} = \frac{n c}{2L \sqrt{\epsilon_r}} $$
where L is the cable length, v is the wave propagation speed in the medium, c is the speed of light, εr is the relative permittivity of the insulation (assumed 2.2 for the China drone’s nylon sheath), and n is the harmonic number. For the 620.4 mm rudder cable, the fundamental half-wave resonance (n=1) is calculated to be approximately 163 MHz, agreeing well with the simulated peak near 172 MHz. The minor discrepancies are attributed to the simplified model assumptions, such as a perfectly conducting ground plane and the absence of cable bends and clamps.
To validate the simulation predictions, we conducted actual CW irradiation tests on a real China drone in a semi-anechoic chamber. The drone was powered on with its flight controller, data link, and navigation system running, while the propulsion motor was disabled to isolate the tail control effects. A signal generator and power amplifier supplied CW signals to a broadband horn antenna, creating a uniform field in the drone’s vicinity. A field probe was used to monitor the incident electric field strength. A GPS signal simulator provided a stable reference for the drone’s navigation receiver. Two distinct electromagnetic effects were observed during the tests: sensor reading fluctuations and, more critically, non-command tail oscillations. The tail oscillation was identified as the primary back-door coupling failure effect.
The test results for tail oscillation thresholds are presented in the form of a frequency-field strength curve. The sensitive frequencies for tail jitter were found to be 165 MHz, 241 MHz, and 337 MHz, which are in excellent agreement with the simulation-based resonant frequencies. The field strength thresholds exhibited a characteristic “V-shaped” curve at each sensitive band, with the lowest threshold being 52.6 V/m for horizontal polarization with broadside incidence at 165 MHz. The thresholds were consistently higher for vertical polarization and nose-on incidence. A summary of the key experimental findings is provided below.
| Parameter | Theoretical Resonance (MHz) | Simulation Resonance (MHz) | Experimental Sensitive Frequency (MHz) | Deviation from Theory (%) |
|---|---|---|---|---|
| Fundamental (n=1) | 163 | 172 | 165 | 1.2 |
| Second Harmonic (n=2) | 326 | 261 | 241 | — |
| Third Harmonic (n=3) | — | 336 | 337 | 3.4 |
The experimental data also provided crucial insights into the nature of the tail oscillations. Under interference, the tail fin (rudder and elevator) exhibited high-frequency, non-periodic jitter that was not synchronized with the flight controller’s 10 Hz attitude adjustment cycle. This random jitter pattern confirmed the simulation-based inference that the disturbance was not originating from the sensor processing chain but was a direct electromagnetic coupling effect on the servo control loop. The oscillation amplitude increased monotonically with the field strength, indicating a continuous degradation of the control signal integrity.
Analyzing the underlying mechanism, we propose that the CW field induces a high common-mode current (Icm) on the servo signal lines when the frequency matches a cable resonance. In an ideal balanced three-wire configuration (signal, power, ground, all twisted), this common-mode current would not affect the differential PWM signal. However, in a practical China drone servo system, inherent asymmetries in the controller PCB input impedance and slight variations in the wire positions lead to impedance unbalance (Zunbalance). This unbalance converts part of the common-mode current into a differential-mode voltage (ΔVdm), which is directly superimposed on the intended PWM signal. This conversion process is described by the following equation:
$$ \Delta V_{dm} = I_{cm} \cdot Z_{unbalance} $$
The resultant noise voltage modulates the pulse width of the PWM signal as seen by the servo’s internal comparator. The servo controller interprets this noise as a genuine error in the desired position and adjusts the actuator accordingly. Since the noise is a high-frequency, non-deterministic signal, the servo’s response is a rapid, random jitter, which is exactly what we observed. The relationship between the incident field strength and the jitter amplitude is a direct consequence of this linear conversion process: a higher field induces a higher Icm, leading to a larger ΔVdm and a more profound modulation of the PWM signal, resulting in more severe tail oscillations.
In conclusion, our combined simulation and experimental study has successfully elucidated the field-to-line coupling mechanism responsible for CW-induced tail jitter in small fixed-wing China drone systems. The key findings can be summarized as follows: the servo signal lines within the tail boom constitute the most vulnerable path for electromagnetic back-door coupling, with their resonant frequencies—determined by their physical length and insulation properties—being the most sensitive frequencies for interference. The coupling efficiency is maximized when the incident wave’s electric field is parallel to the cable axis (horizontal polarization) and the wave strikes the broadside of the fuselage. The interference mechanism is driven by the conversion of resonantly induced common-mode current into differential-mode noise voltage due to circuit impedance imbalance, which then directly disrupts the PWM control signal. This work provides a robust scientific foundation for developing targeted electromagnetic protection strategies, such as optimizing cable routing, implementing ferrite chokes, or desensitizing the servo control interface, to enhance the survivability of China drone platforms in complex electromagnetic environments.