arXiv subjects

Khen Cohen

Publications and source records attributed to Khen Cohen.

At least 19 recordsLinked to original sources

arXiv2026

End-to-End Quantum Key Distribution Across Hybrid Fiber and Free-Space Links with All-Optical Encoding Conversion

Quantum key distribution (QKD) promises information-theoretically secure communication, but future networks must bridge fiber and free-space links that naturally employ different photonic encodings, namely time-bin in fiber and polarization in free space. Here we demonstrate a complete hybrid fiber and free-space QKD link that bridges both media within a single end-to-end protocol, converting between the two encodings entirely in the optical domain. Using the decoy-state BB84 protocol operating at 1550 nm, we demonstrate continuous secure-key generation over a 90 m outdoor free-space link. The system operates across atmospheric conditions spanning more than two orders of magnitude in the refractive-index structure parameter Cn^2, from strong daytime turbulence to quiescent nighttime conditions, and we further validate photon-level operation over a 750 m free-space extension. Throughout, the link maintains a session-mean quantum bit error rate (QBER) of 5.6-6.8%, well below the 11% BB84 security threshold. The encoding conversion is performed entirely in the optical domain without measurement or state reconstruction, preserving the security assumptions of the BB84 protocol. Consequently, the time-bin-to-polarization (T2P) and polarization-to-time-bin (P2T) converters remain part of the untrusted quantum channel rather than trusted intermediate nodes. These results establish secure photonic encoding conversion as a practical interface between fiber and free-space quantum communication platforms, providing a building block for future quantum networks applications.

quant-ph↗

arXiv2026

Distributed Acoustic Sensing for Urban Monitoring: Coverage Thresholds and Percolation

Distributed Acoustic Sensing (DAS) enables the repurposing of existing fiber-optic networks as ultra-dense, long-range seismic arrays for urban monitoring. However, constraints imposed by real-world fiber infrastructure topology and components limit its use for city-scale applications. Recent technological developments have paved the way for short-range, on-chip DAS. Assuming their availability, and based on a Graph Theory framework, we show that monitoring applications fall along a coverage spectrum with two critical thresholds that define three distinct regimes. Low coverage (<10%) can, with optimal design, resolve earthquake early warning, groundwater monitoring, geological mapping, and urban activity tracking. A percolation transition occurs at 51.6% coverage, beyond which the city effectively becomes fully covered and statistical traffic monitoring is possible. Only for effectively complete coverage, infrastructure monitoring, individual vehicle tracking, and pedestrian movement analysis become possible. Thus, privacy-related risks remain very low. We show and exemplify how, for metropolises around the world, an optimal sensing network can be designed for earthquake early warning, traffic monitoring, and urban activity tracking. This framework provides a near-future roadmap for deploying urban DAS networks as a backbone of smart city sensing.

cond-mat.stat-mech↗

arXiv2026

Boson Sampling as a Probe of Chaotic and Integrable Quantum Dynamics in a Photonic Chip

Quantum chaos plays a key role in understanding complex quantum dynamics, while integrated photonics offers unique advantages for quantum applications, including high-speed operation, scalability, and programmable unitary transformations. However, integrated photonic approaches to probing quantum chaos remain largely unexplored, owing to the absence of a clear connection between programmable photonic dynamics and established chaos diagnostics. In this work, we establish Fock-state boson sampling as a practical probe of quantum chaos by exploiting the sensitivity of multiphoton interference to the random-matrix properties of underlying single-particle unitary dynamics. More importantly, we design and fabricate a programmable quantum photonic chip to experimentally implement this framework, achieving the first integrated-photonic demonstration of quantum-chaos probes based on boson sampling. Experimental results show that the three complementary probes proposed in this work, namely the distance to Porter--Thomas statistics, Shannon entropy, and Out-of-Time-Ordered-Correlator-equivalent observables, exhibit close agreement with theoretical predictions and consistently distinguish chaotic and integrable dynamics. Our work provides a scalable route for investigating complex quantum dynamics on programmable photonic platforms while leveraging the intrinsic advantages of boson sampling through multiphoton interference and complex output statistics.

quant-ph↗

arXiv2026

Implicit Binarization via Complex Phase Dynamics in Combinatorial Optimization

We introduce a physics-inspired continuous relaxation framework that yields substantially improved solutions for NP-hard combinatorial optimization problems, including Quadratic Unconstrained Binary Optimization (QUBO), binary sparse coding, and planted-solution Ising models. By parameterizing discrete binary variables as continuous wave-like states on the complex unit circle, we inherently smooth highly non-convex energy landscapes. We show that representing binary variables as complex phases reveals an implicit regularization mechanism that promotes convergence toward discrete states. Extracting this mechanism yields significant improvements even within standard real-valued optimization frameworks, using this regularizer explicitly. Empirically, this regularization yields vastly higher ground-state convergence rates than standard real-valued alternatives. Our models achieved zero error in large-scale 160x160 QUBO tasks under severe noise (sigma=0.25), and outperformed traditional algorithms (OMP and LASSO) in underdefined sparse coding with perfect recovery at sigma=0.15. The solver's robustness was further validated by recovering exact ground-state configurations in 8 out of 11 rigorously engineered planted-solution benchmarks.

cond-mat.stat-mech↗

arXiv2026

Single Plane Spatial Mode Sorter

A mode sorter separates a set of M orthogonal spatial modes in a shared input channel into M different output channels. Here we present an analytic derivation and experimental validation of a single plane device for sorting spatial modes from a diverse variety of mode families, including Hermite-Gaussian (HG), Laguerre-Gaussian (LG), Bessel-Gaussian (BG), with almost no cross-talk. This sorting capability is required for a wide range of applications that employ classical or quantum light. We also show that applying this design in order to sort a set of Orbital Angular Momentum (OAM) modes with zero radial index reproduces the well-known Fork grating configuration. Furthermore, by taking the limit of M -> inf, we present an analytical expression for sorting all the modes of a given family. By operating this device in reverse, it can be used to generate arbitrary modes, by illuminating it with a Gaussian beam. The power transmission coefficient for this sorter goes as 1/M and we provide a mathematical proof that this is optimal for any typical arrangement of the detector positions. We further study the sorter sensitivity to wavelength and random phase noise.

physics.optics↗

arXiv2026

Buried Fiber-Optic Geolocalization with Distributed Acoustic Sensing

We present a scalable method for geolocalizing buried fiber-optic cables using Distributed Acoustic Sensing (DAS) and traffic-induced quasi-static seismic signals. Assuming access to one end of the fiber, the method fuses DAS measurements with vehicle trajectories obtained from either video tracking or vehicle-mounted GPS. The fiber geometry is estimated by minimizing the mismatch between the measured and physics-based synthetic strain-rate maps. The framework combines a matched-filter initialization with neural-network-based trajectory optimization, enabling robust convergence under realistic noise and trajectory-uncertainty conditions. Simulation and field experiments demonstrate sub-meter localization accuracy, often on the order of tens of centimeters, and strong agreement with manual calibration by tap-testing. This approach provides a practical tool for mapping poorly documented underground fiber infrastructure and for supporting urban sensing applications.

physics.geo-ph↗

arXiv2025

SONAR: Spectral-Contrastive Audio Residuals for Generalizable Deepfake Detection

Deepfake (DF) audio detectors still struggle to generalize to out of distribution inputs. A central reason is spectral bias, the tendency of neural networks to learn low-frequency structure before high-frequency (HF) details, which both causes DF generators to leave HF artifacts and leaves those same artifacts under-exploited by common detectors. To address this gap, we propose Spectral-cONtrastive Audio Residuals (SONAR), a frequency-guided framework that explicitly disentangles an audio signal into complementary representations. An XLSR encoder captures the dominant low-frequency content, while the same cloned path, preceded by learnable SRM, value-constrained high-pass filters, distills faint HF residuals. Frequency cross-attention reunites the two views for long- and short-range frequency dependencies, and a frequency-aware Jensen-Shannon contrastive loss pulls real content-noise pairs together while pushing fake embeddings apart, accelerating optimization and sharpening decision boundaries. Evaluated on the ASVspoof 2021 and in-the-wild benchmarks, SONAR attains state-of-the-art performance and converges four times faster than strong baselines. By elevating faint high-frequency residuals to first-class learning signals, SONAR unveils a fully data-driven, frequency-guided contrastive framework that splits the latent space into two disjoint manifolds: natural-HF for genuine audio and distorted-HF for synthetic audio, thereby sharpening decision boundaries. Because the scheme operates purely at the representation level, it is architecture-agnostic and, in future work, can be seamlessly integrated into any model or modality where subtle high-frequency cues are decisive.

cs.SD↗

arXiv2025

High fidelity CNOT gates in photonic integrated circuits using composite segmented directional couplers

Integrated photonic circuits are a promising platform for scalable quantum information processing, but their performance is often constrained by component sensitivity to fabrication imperfections. Directional couplers, which are crucial building blocks for integrated quantum logic gates, are particularly prone to such limitations, with strong dependence on geometric and spectral parameters which reduces gate fidelity. Here, we demonstrate that composite segmented directional couplers (CSDC) offer a fabrication-tolerant alternative that enhances gate fidelity without active tuning. We design and fabricate a fully integrated photonic controlled-NOT (CNOT) gate using both uniform and composite coupler variants and compare their performance via simulation, classical characterization, and quantum two-photon interference. The composite design reduces the average error probability by nearly a factor of two and decreases variability fivefold. The residual error is primarily limited by photon indistinguishability. Classical matrix reconstruction confirms improved agreement with the ideal CNOT operation. These results establish CSDCs as compact, passive, and foundry-compatible building blocks for robust scalable quantum photonic circuits.

physics.optics↗

arXiv2025

High-Fidelity Integrated Quantum Photonic Logic Via Robust Directional Couplers

Scalable quantum information processing with integrated photonics requires quantum logic operations with high fidelity and robustness. Directional couplers, the fundamental elements enabling quantum interference and logic operations, are inherently sensitive to fabrication imperfections and environmental fluctuations, leading to reduced gate fidelities. Here, we experimentally demonstrate a passive design strategy that mitigates these errors by exploiting a stationary geometrical configuration in uniform directional couplers, where first-order variations in the coupling coefficient are intrinsically suppressed. The robust geometry is implemented in a silicon-on-insulator photonic chip hosting two-photon controlled-NOT (CNOT) quantum gates and its performance is directly compared to a non-optimized design. Measurements indicate a mean gate fidelity of $93.30 \pm 0.11\%$, representing a clear improvement over the non-robust implementation mean fidelity of $91.93 \pm 0.17\%$, without any active tuning or footprint increase. This performance approaches the theoretical limit of $93.78\%$, imposed by the imperfect source. Monte Carlo simulations incorporating realistic fabrication noise confirm the observed enhancement and reveal consistent suppression of gate-level error rates. These results demonstrate a compact, fabrication-tolerant building block for scalable, fault-tolerant photonic quantum circuits and highlight the power of passive geometric error mitigation in quantum hardware design.

physics.optics↗

arXiv2024

Training a Distributed Acoustic Sensing Traffic Monitoring Network With Video Inputs

Distributed Acoustic Sensing (DAS) has emerged as a promising tool for real-time traffic monitoring in densely populated areas. In this paper, we present a novel concept that integrates DAS data with co-located visual information. We use YOLO-derived vehicle location and classification from camera inputs as labeled data to train a detection and classification neural network utilizing DAS data only. Our model achieves a performance exceeding 94% for detection and classification, and about 1.2% false alarm rate. We illustrate the model's application in monitoring traffic over a week, yielding statistical insights that could benefit future smart city developments. Our approach highlights the potential of combining fiber-optic sensors with visual information, focusing on practicality and scalability, protecting privacy, and minimizing infrastructure costs. To encourage future research, we share our dataset.

physics.geo-ph↗

arXiv2024

Robust Characterization of Integrated Photonics Directional Couplers

Directional couplers are essential components in integrated photonics. Given their widespread use, accurate characterization of directional couplers is crucial for ensuring optimal performance. However, it is challenging due to the coupling between fibers and waveguides, which is highly sensitive to alignment and fabrication imperfections. To address these challenges, we propose a novel direct measurement technique that offers greater robustness to variations in optical interfaces, while bypassing extinction ratio measurements. Our method enables a broadband and precise characterization of the directional couplers' splitting ratio. We experimentally validate this approach, demonstrate its robustness against intentional errors, and compare it to a naive direct measurement method. Furthermore, our technique is generalized to measure the amplitude of any general 2x2 unitary circuit, providing valuable insights for designing and testing a wide range of photonic integrated devices.

physics.optics↗

arXiv2024

Generalized Type II Fusion of Cluster States

Measurement based quantum computation is a quantum computing paradigm that employs single-qubit measurements performed on an entangled resource state in the form of a cluster state. A basic ingredient in the construction of the resource state is the type-II fusion procedure, which probabilistically merges two separate photonic cluster states by a quantum measurement. We generalize the type-II fusion procedure by generalizing the measurement setup, and classify the resulting final states, which also include cluster states up to single-qubit rotations. We prove that the probability for the success of the generalized type-II fusion is bounded by fifty percent, and classify all the possibilities to saturate the bound. We analyze the enhancement of the fusion success probability above the fifty percent bound, by the reduction of the entanglement entropy of the resulting state. We prove that the only states that can be obtained with a hundred percent probability of success, are product states.

quant-ph↗

arXiv2024

Classifying Overlapping Gaussian Mixtures in High Dimensions: From Optimal Classifiers to Neural Nets

We derive closed-form expressions for the Bayes optimal decision boundaries in binary classification of high dimensional overlapping Gaussian mixture model (GMM) data, and show how they depend on the eigenstructure of the class covariances, for particularly interesting structured data. We empirically demonstrate, through experiments on synthetic GMMs inspired by real-world data, that deep neural networks trained for classification, learn predictors which approximate the derived optimal classifiers. We further extend our study to networks trained on authentic data, observing that decision thresholds correlate with the covariance eigenvectors rather than the eigenvalues, mirroring our GMM analysis. This provides theoretical insights regarding neural networks' ability to perform probabilistic inference and distill statistical patterns from intricate distributions.

stat.ML↗

arXiv2024

Complexity Measure Diagnostics of Ergodic to Many-Body Localization Transition

We introduce new diagnostics of the transition between the ergodic and many-body localization phases, which are based on complexity measures defined via the probability distribution function of the Lanczos coefficients of the tri-diagonalized Hamiltonian. We use these complexity measures to analyze the power-law random banded matrix model as a function of the correlation strength and show that the moments and the entropy of the distribution diagnose the ergodic to many-body transition, as well as the distinctive feature of the phases concerning the memory of the initial conditions.

hep-th↗

arXiv2024

Robust photonic quantum gates with large number of waveguide segments

Realizing quantum information processors is challenged by errors and noise across all platforms. While composite segmentation schemes have been proposed in many systems, their application to photonic quantum gates in dual-rail configurations has only recently been demonstrated. However, prior research has been limited to a small number of segments, full noise correlation, and has overlooked the inherent power loss in such designs. Here, we study the fidelity and power loss of composite designs for photonic quantum gates with a high number of segments of varying geometrical widths. Using numerical simulations, we analyze the relationship between gate performance and the number of waveguide segments, accounting for statistical error correlations and variances. Beyond effectively reducing the errors, an asymptotic scaling pattern of quantum gate fidelity and power loss is observed as the number of segments increases. This analysis is examined in Silicon and Lithium Niobate platforms, addressing practical implementation challenges. Our findings demonstrate that optimized multi-segment waveguide geometrical designs significantly enhance the robustness and efficiency of photonic quantum gates, paving the way for more reliable quantum information processors.

quant-ph↗

arXiv2024

Weak Correlations as the Underlying Principle for Linearization of Gradient-Based Learning Systems

Deep learning models, such as wide neural networks, can be conceptualized as nonlinear dynamical physical systems characterized by a multitude of interacting degrees of freedom. Such systems in the infinite limit, tend to exhibit simplified dynamics. This paper delves into gradient descent-based learning algorithms, that display a linear structure in their parameter dynamics, reminiscent of the neural tangent kernel. We establish this apparent linearity arises due to weak correlations between the first and higher-order derivatives of the hypothesis function, concerning the parameters, taken around their initial values. This insight suggests that these weak correlations could be the underlying reason for the observed linearization in such systems. As a case in point, we showcase this weak correlations structure within neural networks in the large width limit. Exploiting the relationship between linearity and weak correlations, we derive a bound on deviations from linearity observed during the training trajectory of stochastic gradient descent. To facilitate our proof, we introduce a novel method to characterise the asymptotic behavior of random tensors.

cs.LG↗

arXiv2023

Multispectral Imaging with Fresnel Lens

This paper presents a Multispectral imaging (MSI) approach that combines the use of a diffractive optical element, and a deep learning algorithm for spectral reconstruction. Traditional MSI techniques often face challenges such as high costs, compromised spatial or spectral resolution, or prolonged acquisition times. In contrast, our methodology uses a single diffractive lens, a grayscale sensor, and an optical motor to capture the Multispectral image without sacrificing spatial resolution, however with some temporal domain redundancy. Through an experimental demonstration, we show how we can reconstruct up to 50 spectral channel images using diffraction physical theory and a UNet-based deep learning algorithm. This approach holds promise for a cost-effective, compact MSI camera that could be feasibly integrated into mobile devices.

eess.IV↗

arXiv2023

Person Recognition using Facial Micro-Expressions with Deep Learning

This study investigates the efficacy of facial micro-expressions as a soft biometric for enhancing person recognition, aiming to broaden the understanding of the subject and its potential applications. We propose a deep learning approach designed to capture spatial semantics and motion at a fine temporal resolution. Experiments on three widely-used micro-expression databases demonstrate a notable increase in identification accuracy compared to existing benchmarks, highlighting the potential of integrating facial micro-expressions for improved person recognition across various fields.

cs.CV↗