Second-Order Characterization of Micro Doppler Radar Signatures of Drone Swarms | IEEE Journals & Magazine | IEEE Xplore

Monostatic radar at range R from the centre C of a drone swarm situated in a region of diameter D much smaller than R. Each drone in the swarm has Nr rotors with Nb blades.

Scientists Crack the Code: How Radar "Sees" Drone Swarms

New mathematical breakthrough could help defense systems automatically spot groups of attacking drones

Imagine a swarm of 50 drones approaching a military base in the dead of night. Each drone is no bigger than a pizza box, flying silently through the darkness. Traditional radar might spot them as tiny blips, but how do you tell if those blips are dangerous drones or just a flock of migrating birds?

Danish scientists at Aalborg University think they've cracked this puzzle with a mathematical formula that could revolutionize how we detect drone swarms. Their research, published this month, provides the first complete "fingerprint" of how groups of drones appear on radar screens.

The Spinning Signature

Here's the fascinating part: every drone leaves a unique radar signature, like a fingerprint, created by its spinning propeller blades. When radar waves bounce off these rapidly rotating blades, they create what scientists call "micro-Doppler signatures" – tiny frequency shifts that are completely different from the wingbeat patterns of birds.

"Think of it like hearing the difference between a helicopter and a hummingbird," explains lead researcher Anders Malthe Westerkam. "Even if you can't see them, the sound patterns are completely different. Radar works the same way with these micro-Doppler signatures."

But here's where it gets tricky: what happens when you have dozens of drones flying together? Until now, scientists could only study drone swarms by either collecting massive amounts of real-world data or running computer simulations that took forever to complete.

A Mathematical Breakthrough

The Aalborg team took a completely different approach. Instead of collecting data, they built mathematical equations that predict exactly how a drone swarm should appear on radar. It's like having a recipe that tells you what the cake will taste like before you even bake it.

Their equations consider everything that matters: how long the propeller blades are, how fast they spin, how many blades each rotor has, and how many drones are in the swarm. The math even accounts for the fact that drones don't all fly at exactly the same speed – some might be fighting headwinds while others cruise in calm air.

The real breakthrough? Their mathematical expressions come in the form of infinite series with coefficients that drop to zero at predictable limits, meaning that in practical applications, the series can be truncated without losing accuracy. In plain English: their equations are so elegant that you can stop calculating once you have enough precision, making them perfect for real-time defense systems.

Why This Matters Now

The timing couldn't be more critical. The global market for drone detection systems is exploding, projected to grow from $659 million in 2024 to over $2.3 billion by 2029. This isn't just about military threats – unauthorized drones are increasingly being used for smuggling, surveillance, and even terrorism.

According to defense contractor Lockheed Martin, drone swarms present a unique challenge: "You might only have seconds between each critical decision" when facing 10, 20, or even 100 small drones approaching simultaneously.

Current detection systems struggle with what experts call the "bird problem." Radar systems can detect objects as small as 6 inches, but they often can't tell the difference between a drone and a bird of similar size. This leads to false alarms that can shut down airports or put military bases on high alert for no reason.

The Arms Race of the Skies

The drone threat is evolving rapidly. The drone defense market is growing at an astounding 62% annually, driven by "escalated incidents of unlicensed drone operations near sensitive facilities" and "heightened awareness about national security and terrorist threats".

Modern anti-drone systems now use artificial intelligence combined with multiple detection methods – radar, radio frequency scanners, and cameras – to identify threats. But AI systems need accurate training data to work properly. That's where the new mathematical model comes in.

From Lab to Battlefield

The Danish research team tested their equations against real-world data and found remarkable agreement. They simulated a commercial drone similar to a DJI Mavic with realistic parameters: 21-centimeter blades, two blades per rotor, four rotors total, spinning at about 500 rotations per second.

The model successfully predicted both the autocorrelation function and power spectral density of the radar returns, creating what amounts to a mathematical "recipe" for how any given drone configuration should appear on radar.

But the researchers are honest about their model's limitations. To keep the math tractable, they ignored complications like background clutter, electronic noise, and the tilt of drones as they maneuver. In the real world, these factors matter, but the basic mathematical framework provides a solid foundation that can be built upon.

The Future of Airspace Security

Defense experts envision integrated systems that can "track dozens of small, low-flying drones" with "layered effectors capable of progressively thinning the swarm" guided by "intelligent battle management systems". The new mathematical model could provide the theoretical backbone for such systems.

Recent contracts worth tens of millions of dollars have been awarded for advanced counter-drone systems, including a $60 million deal with Elbit Systems and a $48 million contract with Saab for radar systems supporting US forces in Europe.

What's Next?

The Aalborg team believes their work could lead to "cognitive radar systems" that automatically adjust their detection parameters based on what they're seeing. Imagine a radar that learns – if it detects the signature of a four-rotor drone, it automatically fine-tunes itself to spot similar threats more effectively.

Companies like Robin Radar Systems are already developing AI-powered detection systems that can track drones moving up to 60 mph while distinguishing them from birds and other aircraft. The mathematical foundation provided by the Danish research could make such systems far more accurate and reliable.

As drone technology becomes cheaper and more accessible, the race between offensive capabilities and defensive countermeasures continues to accelerate. But with breakthrough research like this, defenders may finally be getting the mathematical tools they need to stay ahead of the threat.

The next time you hear about a mysterious drone sighting near an airport or military base, remember: there might be sophisticated mathematics working behind the scenes, using the telltale signatures of spinning blades to separate friend from foe in our increasingly crowded skies.

---

The complete research was published in IEEE Signal Processing Letters and represents a collaboration between multiple researchers at Aalborg University in Denmark, funded in part by the Thomas B Thriges Foundation.

I'll search for recent news and research about drone swarm detection and radar micro-Doppler signatures to provide current context for this IEEE paper.# Revolutionary Radar Model Provides New Framework for Detecting Drone Swarms

Danish researchers develop mathematical model to characterize micro-Doppler signatures from multiple drones, potentially transforming automated detection systems

A team of researchers at Aalborg University in Denmark has developed a groundbreaking analytical model that could revolutionize how radar systems detect and classify drone swarms. The research, published in IEEE Signal Processing Letters, presents the first comprehensive second-order characterization of micro-Doppler radar signatures specifically designed for swarms of rotor drones.

The study, led by Anders Malthe Westerkam and colleagues, introduces mathematical expressions for autocorrelation functions (ACF) and power spectral density (PSD) that directly reveal how key drone parameters affect radar signatures. The model considers swarms of identical drones, each with multiple rotors comprised of rotating blades, treating rotor orientation and speed as stochastic variables.

Critical Need for Advanced Detection

The research addresses an urgent security challenge as drone technology becomes more accessible and potentially threatening. The global drone detection market is projected to reach $2.33 billion by 2029, growing at a compound annual growth rate of 28.7%, driven by increasing unauthorized drone activities including surveillance and smuggling.

According to Lockheed Martin's Counter-UAS Director Tyler Griffin, defending against drone swarms requires "diverse, integrated sensors that can track dozens of small, low-flying drones, layered effectors capable of progressively thinning the swarm, and an intelligent battle management system". The new mathematical framework could provide the theoretical foundation for such intelligent systems.

Technical Breakthrough

Unlike previous approaches that relied on measurement-driven characterization or computer-intensive simulations, the Aalborg model provides closed-form mathematical expressions. The researchers derive expressions for both ACF and PSD in the form of infinite series with coefficients that drop to zero at predictable limits, allowing practical applications to truncate the series.

The model accounts for critical parameters including blade length, rotor speed, number of blades per rotor, and total number of drones. For the special case of deterministic rotor speed, the ACF can be expressed in closed form, significantly simplifying computational requirements for real-time applications.

Industry Applications and Market Impact

The research comes as the anti-drone market experiences explosive growth. The drone defense system market was valued at $33.04 billion in 2024 and is projected to reach $1.61 trillion by 2032, with a CAGR of 62.54%. Recent market analysis suggests the sector will grow from $13.17 billion in 2024 to $19.82 billion in 2025, representing a 50.5% growth rate.

Current anti-drone systems increasingly rely on AI-powered detection that combines artificial intelligence with machine learning and deep neural networks to identify UAVs and distinguish them from other airborne objects. The new mathematical model could enhance these AI systems by providing more accurate theoretical foundations for pattern recognition.

Commercial and Defense Applications

The timing of this research aligns with major industry developments. In January 2025, Elbit Systems secured a $60 million contract to provide multi-layered counter-UAS solutions to a NATO European country, while Saab received a $48 million contract for radar systems supporting US Air Forces in Europe.

Modern drone detection radars now incorporate micro-doppler classification and deep neural networks to distinguish rotating parts instantly, with systems capable of tracking at speeds up to 100 km/h. The Aalborg model could significantly improve the accuracy of such classification systems.

Research Context and Future Development

The research builds on extensive previous work in micro-Doppler signature analysis. Recent studies have demonstrated that drones and birds both produce distinctive micro-Doppler signatures due to propeller rotation and wingbeats respectively, which can be used for differentiation. Research has shown that micro-Doppler signals are particularly useful for radar applications, with detection performance varying significantly based on radar dwell time and other parameters.

Aalborg University's research group, led by Troels Pedersen, continues to advance radar signal processing techniques, with recent work also including distributed algorithms for cooperative tracking using multiple-input multiple-output radars.

Looking Ahead

The researchers acknowledge that their model makes simplifying assumptions, ignoring effects like macro-Doppler, clutter, noise, and hardware limitations to maintain mathematical tractability. However, they note that these effects could be incorporated in future extensions for more comprehensive modeling.

The model could potentially be leveraged for automatic drone swarm detection, interference mitigation, or target classification in real-time radar systems, representing a significant step toward more effective autonomous defense systems.

As drone threats continue to evolve, mathematical frameworks like this one provide essential tools for developing next-generation detection and defense systems capable of protecting critical infrastructure, military assets, and civilian airspace from increasingly sophisticated aerial threats.

---

Sources

1. Westerkam, A.M., Damkjær, A.S., Villadsen, R.E., Poulsen, M.Ø.B., & Pedersen, T. (2025). Second-Order Characterization of Micro Doppler Radar Signatures of Drone Swarms. IEEE Signal Processing Letters, 32, 3206-3209. DOI: 10.1109/LSP.2025.3596514

2. Using Drone Swarms as a Countermeasure of Radar Detection. Journal of Aerospace Information Systems. https://arc.aiaa.org/doi/10.2514/1.I011131

3. Dedrone. (2025). Layered anti-drone solutions integrate drone detection radars. https://www.dedrone.com/products/drone-detection/extensions/radar

4. Lockheed Martin. (2025). C-UAS Challenge: Closing the Gap in Drone Swarm Defense. https://www.lockheedmartin.com/en-us/news/features/2025/the-counter-uas-challenge-closing-the-gap-in-drone-swarm-defense.html

5. Defence Industries. (February 15, 2025). Next-Gen Anti-Drone Warfare: Top 5 Advanced Technologies in 2025. https://www.defence-industries.com/articles/next-gen-anti-drone-warfare

6. Robin Radar Systems. (2025). Drone Detection Radar. https://www.robinradar.com/solutions/drone-detection-radar

7. Unmanned Systems Technology. (December 16, 2024). Drone Detection Radar | C-UAS Radar Systems. https://www.unmannedsystemstechnology.com/expo/drone-detection-radar/

8. MarketsandMarkets. (2025). Drone Detection Market Size, Share and Industry, 2025 To 2030. https://www.marketsandmarkets.com/Market-Reports/drone-detection-market-199519485.html

9. AIAA SciTech Forum. (2022). Using Drone Swarms as Countermeasure of Radar Detection. https://arc.aiaa.org/doi/10.2514/6.2022-0855

10. Datategy. (July 21, 2025). AI-powered Management for Defense Drone Swarms. https://www.datategy.net/2025/07/21/ai-powered-management-for-defense-drone-swarms/

11. Airsight. (2025). Radar Drone Detection | Drones Detected Using Radar. https://www.airsight.com/en-us/knowledge-hub/drone-detection/radar

12. IEEE Xplore. Advances in applications of radar micro-Doppler signatures. https://ieeexplore.ieee.org/document/7003362/

13. Pramanik, S.K., Hossain, M.S., Islam, S.M.M. (2018). Radar micro-Doppler signatures of drones and birds at K-band and W-band. Scientific Reports. https://www.nature.com/articles/s41598-018-35880-9

14. EURASIP Journal on Advances in Signal Processing. (March 12, 2013). Developments in target micro-Doppler signatures analysis. https://asp-eurasipjournals.springeropen.com/articles/10.1186/1687-6180-2013-47

15. IET Digital Library. Radar Micro-Doppler Signatures: Processing and Applications. https://digital-library.theiet.org/doi/book/10.1049/pbra034e

16. Wang, et al. (March 13, 2024). Radar Micro‐Doppler Signature Generation Based on Time‐Domain Digital Coding Metasurface. Advanced Science. https://advanced.onlinelibrary.wiley.com/doi/full/10.1002/advs.202306850

17. Science.gov. radar micro-doppler signatures: Topics. https://www.science.gov/topicpages/r/radar+micro-doppler+signatures

18. ResearchGate. (October 29, 2015). Review of micro-Doppler signatures. https://www.researchgate.net/publication/283661270_Review_of_micro-Doppler_signatures

19. IEEE Xplore. Target classification and recognition based on micro-Doppler radar signatures. https://ieeexplore.ieee.org/document/8293404

20. IEEE Xplore. Radar-Based Analysis of Pedestrian Micro-Doppler Signatures Using Motion Capture Sensors. https://ieeexplore.ieee.org/document/8500656/

21. MDPI. (September 19, 2022). Detection of Micro-Doppler Signals of Drones Using Radar Systems with Different Radar Dwell Times. https://www.mdpi.com/2504-446X/6/9/262

22. Verified Market Research. (May 13, 2025). Drone Defense System Market Size, Share, Trends & Forecast. https://www.verifiedmarketresearch.com/product/drone-defense-system-market/

23. Grand View Research. Anti-drone Market Size And Share | Industry Report, 2030. https://www.grandviewresearch.com/industry-analysis/anti-drone-market

24. OpenPR. (1 day ago). Top Market Shifts Transforming the Drone Defense System Market Landscape. https://www.openpr.com/news/4153611/top-market-shifts-transforming-the-drone-defense-system-market

25. Grand View Research. Drone Detection Market Size & Share | Industry Report, 2030. https://www.grandviewresearch.com/industry-analysis/drone-detection-market-report

26. MarketsandMarkets. Drone Detection Market Size, Share and Industry, 2025 To 2030. https://www.marketsandmarkets.com/Market-Reports/drone-detection-market-199519485.html

27. Market Research Future. Drone Defense System Market Overview, Size, Share, Industry, Report -2030. https://www.marketresearchfuture.com/reports/drone-defense-system-market-10331

28. SkyQuest Technology. Drone Defense System Market Size - Industry Forecast 2025-2032. https://www.skyquestt.com/report/drone-defense-system-market

29. MarketsandMarkets. Anti-Drone Market Size, Share and Industry Report, 2025 To 2030. https://www.marketsandmarkets.com/Market-Reports/anti-drone-market-177013645.html

30. Business Wire. (February 21, 2025). Anti-drone Market Global Outlook & Forecast 2025-2030. https://www.businesswire.com/news/home/20250221089864/en/Anti-drone-Market-Global-Outlook-Forecast-2025-2030-Rising-Drone-Illegal-Activities-Technological-Advancements-Integration-with-AI-and-ML-Fueling-Growth---ResearchAndMarkets.com

31. Allied Market Research. Drone Defense System Market Size, Share & Trends, Forecast. https://www.alliedmarketresearch.com/drone-defense-system-market-A12507

32. Aalborg University. Drone Research Lab - Aalborg University's Research Portal. https://vbn.aau.dk/en/organisations/drone-research-lab

33. Aalborg University. Drone Research Lab — Aalborg University's Research Portal. https://vbn.aau.dk/en/equipments/drone-research-lab

34. ResearchGate. Troels Pedersen's research works | Aalborg University and other places. https://www.researchgate.net/scientific-contributions/Troels-Pedersen-2272864723

35. Aalborg University. Drone and Robotics Laboratory. https://www.energy.aau.dk/laboratories/drone-and-robotics-laboratory

36. Aalborg University. (July 22, 2025). Students build new hybrid drone — watch it fly in the air and then seamlessly dive underwater. https://vbn.aau.dk/en/clippings/students-build-new-hybrid-drone-watch-it-fly-in-the-air-and-then-

37. Aalborg University. Drone Research Lab - Publications. https://vbn.aau.dk/en/equipments/drone-research-lab/publications/

38. Aalborg University. Drone-Obtained Electromagnetic Signatures. https://vbn.aau.dk/en/projects/drone-obtained-electromagnetic-signatures

39. 3D Printing Industry. (3 weeks ago). Aalborg students build a drone that doesn't stop at water's edge. https://3dprintingindustry.com/news/aalborg-students-build-a-drone-that-doesnt-stop-at-waters-edge-242767/

40. Aalborg University. Drone Research Lab - Projects. https://vbn.aau.dk/en/equipments/drone-research-lab/projects/?status=FINISHED

41. Aalborg University. Drones in society. https://www.cs.aau.dk/education/education-list/drones-in-society

 

Second-Order Characterization of Micro Doppler Radar Signatures of Drone Swarms | IEEE Journals & Magazine | IEEE Xplore