Cyclic-Prefix OFDM Probing for Spatial-ISI-Free Distributed Acoustic Sensing via Frequency-Domain Channel Reconstruction

summary

The gist

In this work, we propose a cyclic-prefix orthogonal frequency-division multiplexing (CP-OFDM) distributed acoustic sensing (DAS) system for phase-sensitive optical timedomain reflectometry (ϕ-OTDR)…

In short

The paper proposes a Cyclic-Prefix OFDM (CP-OFDM) system for distributed acoustic sensing using optical time-domain reflectometry (OTDR). It uses a data-bearing CPOFDM waveform as the sensing probe to reconstruct the true distributed Rayleigh backscattering channel. This method eliminates waveform-induced spatial interference, significantly improving performance in dense event environments and enabling shared sensing and communication (ISAC).

Key concepts

Distributed Rayleigh Backscattering
This refers to how light signals scatter randomly off various points along an optical fiber. The paper models this as a finite-memory sensing multipath channel, meaning the received signal is a combination of the true backscattered signal and delayed copies of itself, which are characteristic of distributed fiber response.
Spatial-ISI-Free Reconstruction
This is the key achievement where the CP-OFDM receiver reconstructs only the physical propagation delay information from Rayleigh backscattering. It avoids 'waveform autocorrelation' blurring that plagues traditional methods, ensuring that strong acoustic events do not corrupt the measurement of weaker, nearby events.
Shared Waveform ISAC
Integrated Sensing and Communication (ISAC) means using the same signal waveform for both sensing (detecting acoustic waves) and communication (sending data). This system uses one CPOFDM waveform to perform both tasks simultaneously, achieving zero bit error rate in communication while sensing.
Finite-Memory Channel Model
The paper models the fiber's response as a finite-memory channel. This mathematical framework allows the authors to use cyclic prefixes to transform complex time-domain convolutions into simpler circular convolutions over OFDM blocks, making it easier to analyze and process the distributed sensing data.

Terminology used across episodes

This episode discusses

The paper

Cyclic-Prefix OFDM Probing for Spatial-ISI-Free Distributed Acoustic Sensing via Frequency-Domain Channel Reconstruction · Read on arXiv

Huan Huang, Zhiyang Xue, Ziang Chen, Zhongxing Tian, Dongdong Zou, Gangxiang Shen, Yi Cai

Optica

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: "Cyclic-Prefix OFDM Probing for Spatial-ISI-Free Distributed Acoustic Sensing via Frequency-Domain Channel Reconstruction".

Dev: Here's the gist of the paper. In this work, we propose a cyclic-prefix orthogonal frequency-division multiplexing (CP-OFDM) distributed acoustic sensing (DAS) system for phase-sensitive optical timedomain reflectometry (ϕ-OTDR) that uses a data-bearing CPOFDM waveform directly as the sensing probe, providing an initial demonstration of shared-waveform integrated sensing and communication (ISAC) The single most informative sentence about the paper's subject and finding is: "To the best of our knowledge, this is the first time that distributed Rayleigh backscattering has been formulated as a finite-memory sensing multipath channel." Modeling Distributed Rayleigh Backscattering The paper establishes a unified discrete-time model for pulse-compression ϕ-OTDR DAS and identifies the mechanism of waveform-induced spatial ISI The distributed fiber response is represented as a finite-memory sensing multipath channel This formulation uses the cyclic prefix to convert the finite-memory linear convolution into circular convolution over the useful OFDM block The received signal model is given by yp

n: = M X−1 m=0 hp,msCP

n − m: + wp

n: This equation shows that the distributed Rayleigh-backscatter response can be viewed as a finite-memory multipath channel along the fiber Proving Spatial-ISI-Free Reconstruction The authors rigorously establish the spatial-ISI-free property of the proposed CP-OFDM probing and clarify its spatial resolution Theorem 1 formalizes this property, stating that under sufficient conditions, the ZF channel reconstruction yields hˆOFDM p,m = hp,m + ˜wp,m This reconstruction contains no deterministic sidelobe leakage, i.e., P q̸=m hp,qcs

m − q: , and is therefore spatial-ISI-free The proof demonstrates that the periodic Kronecker delta reduces to the ordinary Kronecker delta over the physical range-bin support, yielding hˆOFDM p,m = hp,m Performance Gains in Simulations and Experiments Simulations and experiments verify the analysis under dense-event and practical link conditions In a 5.2-km fiber simulation, the proposed CPOFDM-aided frequency-domain receiver suppressed the off-event spatial floor and improved the phase-trace meansquare error by 1.94–29.55 dB for all ten events In a heterodyne coherent experiment, CP-OFDM improves the vibration-to-background ratio (VBR) by approximately 3.2 and 6.2 dB, respectively, compared to matched-filter-based LFM pulse compression Shared Waveform Integrated Sensing and Communication (ISAC) The same data-bearing probe also recovers an image with zero measured bit-error rate and a median error vector magnitude of-23.14 dB, providing an initial shared-waveform ISAC validation This experiment validates the CP-OFDMaided frequency-domain channel reconstruction for spatial-ISIfree DAS and demonstrates its potential for shared-waveform optical-fiber ISAC The same probe also recovers an image with zero measured bit-error rate and a median error vector magnitude of-23.14 dB Spatial Resolution Analysis The paper analyzes the spatial resolution of the proposed CP-OFDM probing, including its bandwidth-limited resolving interval For an aperture of Nact ≤ N contiguous active subcarriers with spacing ∆f = 1/Tu, the occupied bandwidth is Bocc = Nact∆f The bandwidth-limited spatial resolving interval is vg/(2Bocc) This resolves the spatial domain channel by reconstructing the range bins through frequency-domain equalization followed by an IDFT The adjacent reconstructionsample spacing is ∆τOFDM = 1/N∆f = 1/Fs This spatial sampling interval is vg/(2Fs) Experimental Verification of Performance The CP-OFDM heterodyne coherent DAS experimental setup is shown in Fig.

Rosa: 8 The CP-OFDM waveform is generated at a complex-baseband sampling rate of Fs = 128 MSa/s, digitally shifted to a 160-MHz intermediate frequency, and exported by the AWG at 512 MSa/s The cyclic prefix contains 7167 samples, corresponding to 55.992 µs This duration covers the approximately 52 µs round-trip delay of the 5.

Title and authors: Rosa: So we’re looking at this paper today called "Cyclic-Prefix OFDM Probing for Spatial-ISI-Free Distributed Acoustic Sensing via Frequency-Domain Channel Reconstruction". It sounds really technical with all those terms, but basically, they’ve built a new way to do distributed acoustic sensing using something called cyclic prefix orthogonal frequency division multiplexing.

Dev: Yeah, it’s about using this CPOFDM waveform not just for sensing the fiber—the distributed acoustic sensing—but also for sending data at the same time. It seems like they’re trying to combine two things that are usually separate: high-quality imaging and communication into one single probe.

Taro: I'm curious about the core finding they highlight on page one, which is that this is the first time anyone has formulated distributed Rayleigh backscattering as a finite memory sensing multipath channel. That’s a big theoretical step for how we model what happens when sound bounces around in fiber optic cables.

Rosa: Exactly. So before, they were dealing with spatial inter-symbol interference or ISI caused by compression sidelobes in matched filter-based systems, which can mess up the reconstructed traces and give you false signals. This paper says their CP-OFDM approach avoids that deterministic ISI by using a data-bearing waveform directly as the sensing probe.

Dev: That leads into what they call proving spatial inter-ISI free reconstruction on page two. They show that under certain conditions, this frequency domain channel reconstruction yields a result with no deterministic sidelobe leakage, which means it’s spatially ISI free.

Taro: So if we're talking about the practical implications for real-world sensing, how does that translate into what we actually see on the ground? I mean, is this something that works well outside of a clean lab setting?

Rosa: Well, simulations and experiments show it handles dense event and practical link conditions. In a five point two kilometer fiber simulation with ten simultaneous strong and weak events spread out, the proposed CPOFDM-aided frequency domain receiver suppresses the off-event spatial floor <ref:2606.19724#pg3,the proposed CPOFDM-aided frequency domain receiver>.

Dev: And it improved the phase trace meansquare error by a significant amount, specifically between one point nine four and twenty-nine point fifty-five decibels across all ten events. That’s a big jump in how well you can track the actual physical delay information from the vibration signal.

Taro: That improvement over matched filter based pulse compression is what really matters for detecting those weak acoustic events you mentioned earlier, Rosa. It means better sensitivity when there are lots of strong noises happening around them.

Rosa: Right, and in a heterodyne coherent experiment over a five point two kilometer fiber link, they recovered five hundred hertz vibrations near eighty meters using only one or five volt piezoelectric transducer drives with correlations of zero point nine nine seven five and zero point nine nine nine five to fitted sinusoids <ref:2606.19724#pg3,a heterodyne coherent experiment over a 5>.

Dev: That’s impressive correlation quality for something that is still capturing the vibration signal itself, not just the noise floor. But let's talk about what this means for shared waveform integrated sensing and communication or ISAC.

Title and authors: Taro: That part seems key because they use the same data-bearing probe to recover both the acoustic image and some forward communication data, achieving zero measured bit error rate. So it’s not just a sensing tool anymore; it’s a dual-purpose system for that specific waveform.

Rosa: It validates this shared waveform integrated sensing and communication idea by showing that you can simultaneously sense distributed acoustic fields and recover data without errors, which is an initial demonstration of ISAC.

Dev: And we also see improvements in how well the vibration signal stands out against background noise, what they call the vibration-to-background ratio or VBR. They improved this VBR by about three point two and six point two decibels compared to matched filter based linear frequency modulated LFM pulse compression at different drive levels.

Taro: That improvement in VBR is significant because it means the system can better isolate the target vibration signal from ambient noise, which leads to a more robust characterization of that vibration in real-world scenarios.

Rosa: So, to wrap up on these improvements: we have a spatial-ISI free reconstruction that eliminates those deterministic waveform artifacts, we see much better performance under dense event conditions with error reduction up to twenty-nine point fifty-five decibels, and we validate shared waveform ISAC by getting zero bit error rate in the recovery.

Dev: And the VBR enhancement is solid too, giving us those gains of three point two and six point two decibels over matched filter based LFM pulse compression. It shows that this CP-OFDM probing method really helps isolate those target signals from the noise floor.

Taro: What they do not explicitly state as a win is what happens when the world misbehaves in a way that breaks their assumptions—like if you push the system to its absolute limits with extreme environmental changes, or how gracefully it recovers if the fiber itself degrades over time.

Rosa: That’s a fair question. They do mention that subsequent developments have addressed limitations like phase noise compensation in pulse compression DAS and time-expanded OTDR for higher spatial resolution with reduced detection bandwidth.

Dev: So what's the final word on this paper, the "Cyclic-Prefix OFDM Probing for Spatial-ISI-Free Distributed Acoustic Sensing via Frequency-Domain Channel Reconstruction"? It establishes a new modeling framework for distributed Rayleigh backscattering as a finite memory channel and shows concrete performance gains in both sensing and communication tasks.

Taro: For me, it means we can start designing systems where the probe itself is an intelligent communication signal, not just a simple acoustic pulse. That opens up possibilities for much more complex integrated sensing applications down the road.

Rosa: It definitely points toward a future where we move beyond just sensing and into these kinds of shared waveform integrated systems. We’ll keep an eye on how this modeling concept evolves in real-world trials outside the lab.

The paper's summary: Rosa: So, this paper boils down to using this special Cyclic-Prefix OFDM probing method for distributed acoustic sensing that doesn't suffer from spatial intersymbol interference anymore because it reconstructs the actual channel itself rather than a blurred version of the waveform.

Dev: Right, so it tackles that deterministic ISI caused by matched filters, which is a major headache in these types of systems because it messes up your physical delay measurements.

Taro: The core finding they push is that under enough conditions, this frequency domain reconstruction doesn't leak any sidelobes in space, meaning the recovered information is actually free of that waveform-induced artifact.

Rosa: And they show this works in simulations and experiments—in a five point two kilometer fiber link, for instance—where the receiver suppresses those off-event spatial floors and improves the phase trace meansquare error by up to twenty-nine point five five decibels.

Dev: twenty-nine point five five decibels is a big bump over what you get from matched filter based pulse compression, and it happens even when you have ten events happening at once in that simulation.

Taro: That improvement in the error metric means the AI can reliably find and quantify weak acoustic events even when they're buried under many strong ones. That’s a significant gain for detection sensitivity.

Rosa: And they also validated shared waveform integrated sensing and communication, or ISAC, by using that same data-bearing probe to recover both the acoustic image and some forward communication data with zero measured bit error rate.

Dev: Zero measured bit error rate in a real-world link? That shows they’ve proven this is a practical framework for having one waveform do two things at once without losing information for either task.

Taro: It means the probe isn't just some sensor anymore; it's an intelligent signal carrier that can simultaneously sense the acoustic field and transmit data reliably.

Rosa: But what they don't really talk about is how this holds up when the world gets messy—like if we push the system to extreme environmental changes or if we have to worry about fiber degradation over a long time.

Dev: That’s fair, I mean, they did flag things like phase noise compensation in pulse compression DAS as something that would need future work to handle properly.

Taro: So the question becomes how gracefully this specific method recovers if the assumptions they made about the channel aren't perfectly met by real-world physics or fiber conditions.

Rosa: It sets a new modeling foundation for distributed Rayleigh backscattering, showing it can be treated as a finite memory sensing multipath channel, and that’s what opens up the door for building more complex integrated sensing applications.

The paper's improvements: Rosa: So, we’re looking at how they’ve improved the sensing system using this cyclic prefix method to eliminate that spatial intersymbol interference we talked about earlier and make it much cleaner.

Dev: They aren't just doing some small tweak; they are fundamentally changing how the fiber response is modeled, treating it as a finite memory channel instead of something simpler.

Taro: This modeling change allows them to prove that if you reconstruct the frequency domain channel correctly, you get a result with no deterministic sidelobe leakage at all.

Rosa: That means when they reconstruct the image, there’s no artifact left over from the waveform itself messing up the spatial resolution of where things are happening along the fiber.

Dev: So, what does that mean for a control engineer who has to manage latency and loop rates? It implies a much more predictable channel response in their real-time loops because you aren't fighting these deterministic sidelobes.

Taro: For autonomy research, it means the system can rely on the reconstructed phase trace to tell the difference between actual physical vibration and just noise from the transmitted signal.

Rosa: And when you look at the experimental numbers, they showed that this new approach suppresses that off-event spatial floor significantly and improves error metrics by up to twenty-nine point five five decibels in simulations.

Dev: That number is what matters for safety cases; a twenty-nine point five five decibel improvement in error means the system can detect weaker acoustic events much more reliably under dense event conditions.

Taro: It’s about making sure that when the environment misbehaves—when there are lots of strong signals—the AI still recovers those subtle, weak signals without getting overwhelmed.

Rosa: And they also showed it helps isolate the target vibration signal from background noise by improving the vibration to background ratio by three point two and six point two decibels in coherent experiments.

Dev: That VBR improvement is solid because it means the system can better separate what’s actually moving from ambient noise, which is a huge win for characterization.

Taro: It confirms that this isn't just theoretical math; they've shown it works in a heterodyne coherent setup, which means we’re seeing tangible results outside of just a clean lab bench.

Rosa: So the implication is that by using the data-bearing waveform as the sensing probe, we get better spatial resolution and cleaner signal quality without sacrificing our ability to do communication at the same time.

Dev: It shifts the whole design philosophy toward using a single, integrated waveform for both tasks, which is what they call shared waveform integrated sensing and communication.

Taro: That moves us closer to systems where sensing isn't just a separate step before communication; it’s inherently tied to how we send the data.

Conclusion: Rosa: So we’ve covered how this paper uses cyclic prefix OFDM probing to get spatial intersymbol interference free by reconstructing the channel itself and showing real performance gains in simulations and experiments for distributed acoustic sensing.

Dev: Right, it proves that using a data-bearing waveform as the probe is a viable path toward shared waveform integrated sensing and communication.

Taro: The big picture is that this modeling framework for distributed Rayleigh backscattering as a finite memory channel gives us much better tools to design next-generation optical fiber sensors.

Rosa: It means we can move beyond just seeing what’s happening in one spot and start designing systems where the probe itself acts like an intelligent communication signal.

Dev: I think the most important implication for real-time systems is that it gives us a much cleaner, more predictable channel model to work with when building those high-speed control loops.

Taro: For autonomy, it means we can trust the AI’s ability to resolve spatial information even in noisy environments because the reconstruction isn't biased by waveform artifacts anymore.

Rosa: What they didn't really detail is how robust this is against long-term fiber degradation or extreme environmental noise that might break their assumptions about the channel structure.

Dev: That’s a fair point; we still need to see what happens when the physical reality doesn't perfectly match the model they built for the fiber response.

Taro: We should definitely look at how this methodology scales if you try to push it into systems with much higher spatial resolutions or longer detection bandwidths.

Rosa: Anyway, this paper, "Cyclic-Prefix OFDM Probing for Spatial-ISI-Free Distributed Acoustic Sensing via Frequency-Domain Channel Reconstruction," sets a solid foundation for future distributed sensing.

Dev: It’s a very interesting piece of work that shows how frequency domain processing can solve spatial issues that used to plague time domain methods.

Taro: I'm looking forward to seeing what they do next with the ISAC aspect, because combining sensing and communication like this is where the real potential is.

More episodes

← Home