This paper, published in Radioengineering (June 2024), tackles a specific but critical vulnerability in radar systems: ghost targets created by the mixing of external interference signals. By developing a quantitative model for Second-order Intermodulation False Alarm (SIFA) interference, the researchers provide a roadmap for evaluating how radars will perform in increasingly crowded electromagnetic environments.
TL;DR
Researchers have developed a new mathematical model and testing method to quantify when dual-frequency interference causes "ghost targets" in radar systems. Their model, validated through electromagnetic injection testing, predicts these false alarms with an accuracy of within 1 dB, identifying a critical sensitivity window within a 3 MHz frequency difference and a ±200 MHz offset.
What Problem Does This Paper Address?
In modern electronic warfare and civilian spectrum management, the "battlefield electromagnetic environment" is becoming incredibly dense. When a radar is hit by two different interference signals (EMI) simultaneously, the nonlinear components inside the radar—like mixers and amplifiers—can inadvertently "mix" these signals.
While researchers have historically focused on third-order intermodulation (which often falls directly inside the operating band), second-order intermodulation (SIFA) has been somewhat neglected. SIFA occurs when the difference between two interference frequencies ($|f_1 - f_2|$) falls within the radar’s low-frequency signal processing bandwidth. This creates a concentrated "spike" in the radar's output that looks exactly like a real target, leading to a false alarm. The challenge has been the lack of a quantitative model to predict exactly when and how strongly these false alarms will appear.
Methods
The authors established a radar SIFA effect model based on a field-circuit coupling mechanism. Key steps included:
- Theoretical Modeling: They defined a SIFA effect index ($R_F$) that relates the intermodulation signal level to the radar's sensitive threshold.
- Parameter Simplification: To make the model practical for engineers, they introduced a "sensitivity coefficient" ($B_i$) and an "interference factor" ($\gamma$), assuming that for small frequency changes, the interference effects are relatively constant.
- Equivalent Substitution Testing: Instead of expensive full-scale radiation testing, they used an electromagnetic injection method. This involves injecting interference signals directly into the radar's RF front-end to simulate external radiation, allowing for precise control over frequency and power.
- Calibration: They performed a stability test (recording 10 sets of data and removing outliers) to ensure the repeatability of the SIFA signal measurements.
Datasets
The study did not use a public software dataset. Instead, it relied on primary experimental data collected from a specific hardware target:
- Equipment Under Test (EUT): A stepped-frequency ranging radar with a working frequency of $f_0 \pm 100$ MHz.
- Test Environment: A controlled laboratory setup using signal generators, power amplifiers, and a spectrum analyzer to create a precise dual-frequency EMI environment.
Key Insights and Ideas
- The "Spike" vs. "Hill" Distinction: One of the most interesting qualitative findings is how SIFA differs from other interference. While standard non-intermodulation EMI appears as broad, "hill-shaped" noise in radar imaging, SIFA produces a sharp, "spike-type" signal. This makes SIFA particularly dangerous because it is much harder for automated systems to distinguish from a real target.
- The Sensitive Window: The researchers identified a specific "danger zone" for this radar. False alarms were most prominent when the two interference signals were within 3 MHz of each other and the overall frequency offset was within ±200 MHz of the radar's center frequency.
- Linearity and Saturation: The SIFA signal level increases linearly with the power of the interference signals up to a certain point (around 10 dB(V/m)² in their tests), after which the growth rate decreases as the internal components reach saturation.
Key Results
- Model Accuracy: The proposed model's prediction error was found to be less than 1 dB when compared to experimental results.
- Quantified Thresholds: The study successfully mapped the "sensitive band" of the radar. If the frequency difference $\delta f$ exceeds 3000 kHz (3 MHz), the SIFA signal component falls outside the low-pass filter's passband, and the interference effectively disappears.
- Validation: By testing across various frequency offsets and differences, the authors proved that their simplified engineering model (which assumes $\gamma(\Delta f_1) \approx \gamma(\Delta f_2)$) is robust enough for practical evaluation.
Limitations and Future Work
- Single Radar Type: The study focused on a stepped-frequency ranging radar. While the theoretical framework is general, the specific sensitive bands (3 MHz / 200 MHz) would likely differ for other radar architectures (e.g., FMCW or pulse-Doppler).
- Linearity Assumption: The model assumes that certain parameters do not change with the interference field strength. While this held true in the linear region of their tests, it might fail under extremely high-power "jamming" scenarios where the receiver is completely saturated.
- Future Directions: The authors suggest that this model could be integrated into broader electromagnetic compatibility (EMC) software to help predict radar performance in complex, multi-emitter environments before hardware is even deployed.
Summary updated