Passive Radar Drone Detection with RTL-SDRs

ISEF Category: Embedded Systems

Subcategory: Signal Processing  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

Your school already sits inside a radar network. FM stations and TV towers are throwing out radio waves all day. If you can catch the echoes, you can detect motion without sending out a single pulse. That makes this project feel like secret listening, but with real signal processing.

What Is It?

Passive radar is a way to detect moving objects without transmitting your own signal. Instead, you listen to a strong broadcast source, like FM radio or digital TV, and compare the direct signal with the reflected signal from a target. The delay and shift between those two copies can reveal motion, distance, and speed.

Think of it like hearing an echo in a canyon, except the canyon walls are drones, cars, or other moving objects, and the shout comes from a nearby broadcast tower. The core math often uses a cross-ambiguity function, which is a search that checks many delay and Doppler shift pairs to find where an echo best matches a target. Your job is to build a system that can separate real motion from noise, clutter, and strong direct-path signals.

Why This Is a Good Topic

This topic is strong because you can test one clear signal-processing chain and measure how each design choice changes detection quality. It connects to real problems in surveillance, airspace monitoring, and low-power sensing, but you can study the basics with a small-scale setup. You can learn antenna placement, synchronization, filtering, clutter rejection, and statistical performance. That gives you real engineering depth, not just a demo.

Research Questions

• How does antenna spacing affect the signal-to-clutter ratio in a passive radar setup?
• What is the effect of using FM versus DVB-T as the illuminator on target detectability?
• Does a cross-ambiguity-function search on an MCU detect moving objects more reliably than a host-only implementation?
• To what extent does direct-path interference reduce detection of slow-moving targets?
• Which preprocessing method improves peak detection in the range-Doppler map the most?
• How does the target's speed change the strength and position of the Doppler peak?

Basic Materials

• Two RTL-SDR receivers.
• Two antennas matched to the chosen broadcast band.
• A laptop or desktop computer with USB ports.
• An MCU coprocessor board with enough processing power for signal preprocessing.
• Coaxial cables and adapters for the SDRs and antennas.
• A tripod or mounting hardware for stable antenna placement.
• A notebook for recording geometry, timing, and environmental conditions.
• A test target such as a car, bicycle, or drone, where permitted by school and local rules.

Advanced Materials

• Two phase-stable SDR receivers or a synchronized SDR platform.
• Calibrated antennas with known gain patterns.
• An MCU or embedded processor with DMA support and timing control.
• A spectrum analyzer or RF power meter for setup checks.
• An external reference clock if the receiver pair supports one.
• A GPS-disciplined oscillator if timing stability is a project variable.
• A controlled moving target platform, such as a vehicle or mobile reflector.
• A field laptop for logging IQ data and metadata.

Software & Tools

• Python: Processes IQ samples, computes the cross-ambiguity function, and plots range-Doppler maps.
• GNU Radio: Helps build and test the receiver chain and preprocessing blocks.
• NumPy: Handles array math for correlation, filtering, and peak finding.
• Matplotlib: Plots detection maps, spectra, and comparison charts.
• ImageJ: Can help inspect static screenshots of heatmaps if you document results as images.

Experiment Steps

1. Define the target you want to detect, the broadcast source you will use, and the motion conditions you can measure safely.
2. Design the receiver geometry, including reference and surveillance antenna placement, so you can separate direct-path energy from echoes.
3. Choose the signal-processing pipeline, then decide what the MCU will do locally and what the host computer will finish.
4. Build a calibration plan that turns raw correlation peaks into repeatable detection metrics.
5. Plan controls that test clutter, timing drift, and false alarms before you collect target data.
6. Pre-register the analysis method you will use to compare setups, so you can judge performance consistently.

Common Pitfalls

• Using receivers that are not frequency or time aligned, which smears the cross-ambiguity output and hides weak targets.
• Placing the antennas too close together, which lets the direct broadcast signal swamp the reflected echo.
• Testing near large metal surfaces, which creates clutter peaks that look like real motion.
• Changing the receiver gain during runs, which makes one trial impossible to compare with the next.
• Treating every bright spot in the range-Doppler map as a target, which inflates false detections.

What Makes This Competitive

A competitive version of this project goes past a working demo. You would compare multiple illuminators, multiple receiver layouts, or multiple preprocessing methods, then back the claims with clear statistics. Strong entries also measure false alarm rate, detection threshold, and sensitivity to motion speed. If you can explain why one design wins, not just that it works, your project gets much stronger.

Project Variations

• Test whether FM or DVB-T gives better passive detection performance in your area.
• Compare a schoolyard setup with an open field setup to quantify clutter effects.
• Use different moving targets, such as cars, bicycles, and drones, to compare Doppler signatures.

Learn More

• MIT OpenCourseWare Signals and Systems: Search MIT OpenCourseWare for lectures on correlation, convolution, and frequency-domain thinking.
• NOAA National Weather Service Radar Basics: A free overview of radar concepts and signal interpretation, found on NOAA sites by searching radar basics.
• NASA Earthdata: Search for remote sensing and radar articles that explain how reflected signals reveal motion and structure.
• PubMed: Search for review articles on passive radar signal processing and cross-ambiguity function methods.
• IEEE Xplore open access papers: Search for passive radar and illuminators of opportunity to find recent peer-reviewed examples.