A Four-Branch Radar Target Simulator with Fractional Delay on a PlutoSDR

summary

The gist

A low-cost, open-source radar target simulator based on an ADALM-Pluto without hardware modification supports four independently configurable targets and analog bandwidths up to 56 MHz…

In short

A low-cost, open-source radar target simulator based on an ADALM-Pluto supports four configurable targets and up to 56 MHz bandwidth. It uses fractional-delay filters to generate echoes with precise delays and Doppler shifts, allowing for the development of integrated sensing and communications (ISAC) hardware and algorithms.

Key concepts

Fractional-Delay Filters
These are digital filters used in the simulator to adjust echo timing beyond integer sample periods. They allow for sub-sample adjustments to accurately model target motion, which is crucial for realistic radar simulations.
Moving Target Modeling
The system models targets moving in monostatic radar geometry by linking range and velocity to the round-trip delay. This requires reproducing both the changing propagation delay and the carrier phase shift caused by Doppler shifts for motion consistency.
Self-Interference Analysis
When multiple targets are simulated, echoes appear near their sum of ranges with combined Doppler shifts. This highlights that antenna isolation is important; reducing this interference improves the accuracy of over-the-air setups.
ADALM-Pluto
This is the hardware platform used as the base for the simulator. It is an FPGA system that allows for custom digital signal processing, enabling the creation of a flexible and low-cost radar target simulation tool.

Terminology used across episodes

This episode discusses

The paper

A Four-Branch Radar Target Simulator with Fractional Delay on a PlutoSDR · Read on arXiv

Florian Euchner, Stephan ten Brink

Institute of Telecommunications, University of Stuttgart

Transcript

Introduction to the show: ident: Electrical Engineering. The day in electrical engineering: signals, systems and the hardware that carries them.

Rosa: We're looking at the paper: "A Four-Branch Radar Target Simulator with Fractional Delay on a PlutoSDR".

Dev: Here's the gist of the paper. A low-cost, open-source radar target simulator based on an ADALM-Pluto without hardware modification supports four independently configurable targets and analog bandwidths up to 56 MHz, allowing for the development and evaluation of integrated sensing and communications (ISAC) hardware and algorithms Simulator Capabilities The simulator is based on an ADALM-Pluto without hardware modification, supporting four independently configurable targets and analog bandwidths up to 56 MHz The FPGA generates echoes with independent delays, Doppler shifts and gains, using fractional-delay filters for sub-sample adjustment Over-the-air measurements demonstrate simultaneous echoes, sub-sample delay adjustment, and smooth target motion between sample positions Moving Target Modeling The system models moving targets in monostatic radar geometry where the round-trip delay is given by a formula involving range and velocity To generate motion-consistent echoes, the simulator must reproduce both the changing propagation delay and the carrier phase in (4) This is achieved by varying the delay of the baseband samples and applying a phase rotation corresponding to the carrier Doppler shift fD,k The target gain ak is assumed to be constant and frequency independent over the signal band Signal Processing Implementation The received waveform is represented by discrete samples x

n: at sample rate fs, where n is the sample index A delay line implemented in memory delays the waveform by an integer number of sample periods, and FIR interpolation is used to implement the fractional part of the delay The total delay at sample n is therefore τk (n/ fs) = τ0 + dk

n: / fs The carrier phase must satisfy φk

n: = -2π fcτk (n/ fs), up to a constant initial phase Performance and Accuracy The fractional-delay filters are evaluated numerically against an ideal delay with unit gain, resulting in the largest difference over all evaluated frequencies and settings being −60.6 dB in standard mode and −61.8 dB in overclocked mode The largest absolute median errors are 2.6 mm in overclocked mode and 1.6 mm in standard mode Continuous motion of multiple targets demonstrates that the two targets follow their configured trajectories independently, with consistent range and Doppler throughout the recording Self-Interference Analysis The observed self-interference echoes highlight the importance of antenna isolation in over-the-air setups For target indices i and j (including i = j), this produces an additional echo near range Ri + Rj with Doppler shift fD,i + fD,j, plus the short antenna-to-antenna propagation delay Improved isolation between the simulator’s antennas or lower transmit gain reduces these echoes The results show that a low-cost PlutoSDR can generate repeatable virtual echoes for sub-6 GHz ISAC experiments with bandwidths up to 56 MHz The open-source FPGA design, firmware and browser interface allow the simulator to be adapted to different experiments A useful next step is coordinated operation of several simulators placed at different angles, enabling repeatable tests of angular separation together with range and Doppler estimation Acknowledgment This work is supported by the German Federal Ministry of Research, Technology and Space (BMFTR) within the SENSATION project (grant no.

Paper summary: Rosa: So this paper, "A Four-Branch Radar Target Simulator with Fractional Delay on a PlutoSDR," it’s basically showing how you can make a low-cost simulator for integrated sensing and communications hardware. It’s based on an ADALM-Pluto, which is pretty cheap and open source, right?

Dev: Yeah, the core idea is that instead of buying expensive commercial radar simulators that cost a lot of money, you can use something much simpler for testing those sub-six GHz experiments <ref:2611.15674#pg1>. The thesis here is about making something low-cost but capable enough to test algorithms and hardware development.

Taro: So what makes it matter for the real world? It claims it supports four different targets and analog bandwidths up to fifty-six MHz, which is pretty solid for testing things in that spectrum <ref:2611.15674#pg1,targets and analog bandwidths up to 56 MHz>. But the real kicker seems to be how they handle motion—they claim you can get smooth target motion echoes even when you adjust the delay and carrier phase together, which is a big deal for simulating real radar signals.

Rosa: Exactly. They’re using fractional-delay filters to do this sub-sample adjustment, meaning they can tweak the timing more finely than just adding whole sample delays. That lets you reproduce things like changing range and frequency dependent Doppler shifts consistently over time, which is crucial when you're testing motion models.

Dev: From an engineering standpoint, the implementation details are interesting because they have to deal with the PlutoSDR’s limited on-chip memory and hardware multipliers. They mention supporting four target paths with a maximum programmable delay of twelve thousand two hundred seventy-six radio samples and interpolation filters of length six or eight, which shows they had to make some trade-offs in the FPGA design <ref:2611.15674#pg3,with a maximum programmable delay of 12276 radio samples and interpolation filters>.

Taro: And those trade-offs are where you see the real failure modes, right? When you push the limits on what the hardware can handle while keeping things motion-consistent, that’s what determines how robust your system is when it hits a real test setup.

Paper summary: Rosa: They also talked about self interference echoes that happen in over-the-air setups, and they pointed out that antenna isolation is super important there. It mentioned that if you improve the isolation between the simulator's antennas or lower the transmit gain, you can reduce those unwanted echoes near range Ri plus Rj with Doppler shifts of fD i plus fD j.

Dev: That makes sense from a practical standpoint because in a real setup, those kinds of self-interference artifacts can totally mess up your measurements if you don't manage the hardware layout carefully. So, they’re giving you some concrete advice on how to set up the physical test bench for this kind of simulator.

Taro: I wonder what happens when the world misbehaves in a real test? If you try to run these four targets at high bandwidths like fifty-six MHz, does the system start breaking down quickly, or is it surprisingly stable across those different configurations <ref:2611.15674#pg1>?

Rosa: They did some testing on continuous motion of multiple targets and found that the two targets follow their configured trajectories independently with consistent range and Doppler throughout the recording. That suggests a good level of stability for sustained testing scenarios.

Dev: That consistency across multiple targets is important for validating complex integrated sensing algorithms, so having that kind of predictable behavior is what makes this simulator useful beyond just generating simple echoes. It’s about simulating a more realistic environment for testing those algorithms under dynamic conditions.

Taro: So, the paper’s title suggests it's a simulator with fractional delay on the PlutoSDR. What does that actually mean for someone who is trying to build a system? Are they saying you can get very precise timing control just by using fractional delay filters on this specific hardware?

Rosa: They are showing that you can get repeatable virtual echoes for sub-six GHz ISAC experiments with bandwidths up to fifty-six MHz using this low-cost PlutoSDR setup <ref:2611.15674#pg1,sub-6 GHz ISAC experiments>. The open source FPGA design and firmware, plus the browser interface, means anyone can adapt it to different experiments they are doing.

Dev: The implications here seem to be that you don't need a massive budget for testing those integrated sensing and communications algorithms anymore if you can use this approach. It opens up a way for people to develop and evaluate hardware and algorithms in a much more accessible way.

Paper summary: Taro: I think the real value is in the open source part, because it lets researchers look under the hood at how that fractional delay works, which is important for understanding exactly where those motion-consistent echoes come from. It’s not just a black box; you can probe the mechanism.

Rosa: They also mentioned that even when evaluated numerically against an ideal delay with unit gain, the largest difference they found over all frequencies and settings was negative sixty point six decibels in standard mode and negative sixty one point eight decibels in overclocked mode. That’s a pretty small error margin for this kind of system.

Dev: A minus sixty point six decibel error is a good metric for how accurately the fractional delay filter is performing compared to an ideal scenario, which speaks to the accuracy of their implementation details. But we have to remember that these numerical tests don't tell us if it will perform as well in a noisy, real-world environment.

Taro: Right, so they show excellent theoretical performance in isolation, but the next step has to be testing how it handles real noise or unexpected system variations when you run these four targets at full speed.

Rosa: That’s the next logical step for any hardware simulator—moving from perfect numerical simulation to real-world validation where things get messy. The paper points toward coordinated operation of several simulators at different angles as a useful next step, which could be used for repeatable tests of angular separation along with range and Doppler estimation.

Dev: That sounds like a way to test how well you can separate targets in space without the simulator itself having to be perfectly aligned or placed in one spot. It’s about testing the sensing capability across different geometries.

Taro: So, this paper provides a solid foundation for building those more complex testing environments, moving beyond just single target motion simulations to multi-target scenarios where you test angular separation too. The open source nature makes it easy to build on top of that work.

Conclusion: Rosa: So we’re looking at "A Four-Branch Radar Target Simulator with Fractional Delay on a PlutoSDR." Basically, this paper shows how you can build a low-cost radar simulator using an ADALM PlutoSDR without needing any special hardware changes.

Dev: Yeah, the authors are focused on making this accessible for integrated sensing and communications hardware development. They’re talking about testing things in the sub-six GHz range up to fifty six MHz bandwidth.

Taro: What’s really interesting is how they manage the motion of targets in this monostatic radar geometry. They have to reproduce both the changing delay and the carrier phase smoothly while adjusting sample delays and applying a phase rotation for Doppler shift.

Rosa: That fractional-delay filter part is key here. They used it to get sub-sample adjustment, which lets them simulate smooth target motion between sample positions over time.

Dev: The performance numbers are pretty tight too. They tested those fractional filters against an ideal delay with unit gain and the largest difference they found was around negative sixty point six decibels in standard mode. That’s a small error margin for this kind of system.

Taro: And when it comes to accuracy on the targets themselves, they saw median errors of about two point six millimeters in overclocked mode and one point six millimeters in standard mode. That’s decent for a simulator, but we gotta remember that’s just the numerical test.

Rosa: Right, so what does this mean practically? It means you can generate repeatable virtual echoes for ISAC experiments without needing a massive budget for expensive commercial simulators.

Dev: It opens up a path where researchers can develop and evaluate hardware and algorithms in a much more affordable way using open source designs, firmware, and an interface they can use.

Taro: The authors also pointed out that self-interference echoes happen in over-the-air setups, which shows the importance of antenna isolation when you build these things physically.

Rosa: They suggested that improving the isolation between the simulator’s antennas or lowering the transmit gain helps reduce those unwanted echoes near range Ri plus Rj with Doppler shifts fD i plus fD j.

Dev: So they’re giving advice on how to set up a physical test bench for these simulations, moving beyond just the software side.

Taro: The conclusion is that this low-cost PlutoSDR setup can generate repeatable virtual echoes for sub-six GHz ISAC experiments with bandwidths up to fifty six MHz.

Rosa: It’s an open source FPGA design, firmware, and browser interface that lets you adapt it to different experiments you are doing.

Dev: A useful next step they suggest is coordinated operation of several simulators placed at different angles. That would allow for repeatable tests of angular separation along with range and Doppler estimation.

Taro: So, this work provides a solid foundation for building more complex testing environments where you can test how well you can separate targets in space without the simulator itself having to be perfectly aligned or placed in one spot.

Rosa: The open source nature makes it easy to build on top of that work.

Dev: That means we’re moving from just simulating single target motion to multi-target scenarios where you test angular separation too.

More episodes

← Home