SEARCH · Search NASA

Results for “parallelism”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 307 records · Page 17

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Core inductive electric field during sawtooth crashes on DIII-D

Sawtooth crashes on tokamak plasmas exhibit relaxation much faster than resistive time scales via a mechanism not fully understood. Using core magnetic measurements from the Radial Interferometer-Polarimeter (RIP) diagnostic on the DIII-D tokamak, Grad–Shafranov equilibria constrained by internal magnetic measurements that have high time resolution (<1μs) can be computed, allowing analysis of how equilibrium parameters such as safety factor q, current density J, and parallel electric field E ∥ , particularly on-axis, evolve. At the sawtooth crash, on-axis safety factor q 0 is observed to rise by 5% but remain below 1 throughout the cycle, and on-axis current density J 0 is observed to drop by 5%. On-axis parallel electric field E ∥ (0) is found to be balanced by ηJ 0 (resistivity times on-axis current density) except during the 200 µs crash period, where E ∥ (0) reaches 22 V m –1 , exceeding ηJ 0 by a factor of more than 2000. These first measurements in tokamak plasmas verify that generalized Ohm's law is not balanced during the crash by resistive effects alone; this is a finding expected due to the relaxation being much faster than resistive timescales. As a result, measurement of the electric field during the tokamak sawtooth serves to illuminate the physical mechanisms at work.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

Determining the tilt of the Raman laser beam using an optical method for atom gravimeters

The tilt of a Raman laser beam is a major systematic error in precision gravity measurement using atom interferometry. The conventional approach to evaluating this tilt error involves modulating the direction of the Raman laser beam and conducting time-consuming gravity measurements to identify the error minimum. In this work, we demonstrate a method to expediently determine the tilt of the Raman laser beam by transforming the tilt angle measurement into characterization of parallelism, which integrates the optical method of aligning the laser direction, commonly used in freely falling corner-cube gravimeters, into an atom gravimeter. A position-sensing detector (PSD) is utilized to quantitatively characterize the parallelism between the test beam and the reference beam, thus measuring the tilt precisely and rapidly. After carefully positioning the PSD and calibrating the relationship between the distance measured by the PSD and the tilt angle measured by the tiltmeter, we achieved a statistical uncertainty of less than 30 µrad in the tilt measurement. Furthermore, we compared the results obtained through this optical method with those from the conventional tilt modulation method for gravity measurement. The comparison validates that our optical method can achieve tilt determination with an accuracy level of better than 200 µrad, corresponding to a systematic error of 20 µGal in g measurement. This work has practical implications for real-world applications of atom gravimeters.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2024

Investigation of pedestal parameters and divertor heat fluxes in small ELM regimes in DIII-D

Abstract Divertor heat flux and its correlation with pedestal parameters within various small edge localized mode (ELM) regimes, including high beta poloidal, type-II and ELMs with negative triangularity H-modes were investigated in DIII-D. The parallel energy fluences of type-II and high beta poloidal small ELM regimes fall below the linear scaling with pedestal electron pressure for type-I ELMs put forward in Eich et al 2017 ( Nucl. Mater. Energy 12 84–90). The negative triangularity of H-mode ELMs follow the Eich scaling for type-I ELMs. The parallel heat flux and total heat loads to the divertor were determined using high-time resolution infrared thermography, while pedestal parameters were obtained through self-consistent kinetic equilibrium reconstructions. Linear regressions for the type-II and high beta poloidal regimes demonstrate that an equivalent 7.5 MA small ELM scenario in ITER would fall below the ~5 MJ m − 2 leading edge melting limit for tungsten (Gunn et al 2017 Nucl. Fusion 57 046025). Utilizing fast thermography, the scrape-off layer power fall-off length for both inter-ELM and intra-ELM was determined and compared to the Eich scaling with poloidal magnetic field in Eich et al (ASDEX Upgrade Team and JET EFDA Contributors 2013 Nucl. Fusion 53 093031). Except for the high beta poloidal scenario, all the small ELM regimes during both inter- and intra-ELM periods had power fall-off lengths ( λ q ) larger then would be expected from the B pol , MP − 1 scaling associated with type-I ELMs, signifying their potential in managing heat loads and offering a solution for core–edge integration.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

First experimental confirmation of island SOL geometry effects in a high radiation regime on W7-X

Abstract This work characterizes the detachment behavior and radiation characteristics of the lowiota configuration in the Wendelstein 7-X (W7-X) stellarator. The island scrape-off layer (SOL) of thelow iota has a poloidal mode number of 6 islands surrounding the last closed flux surface (LCFS).The island geometry of the low iota configuration is significantly different to that of the standardmagnetic field configuration, whose detachment characteristics have already been described in previouswork[2, 3, 4]. Experimental results show that the radiation pattern in the low iota configurationis starkly different to that of the standard magnetic field configuration, with radiation concentratedat the island SOL O-points, rather than the X-points. Additionally, this O-point localized radiationpattern is associated with unstable detachment, with both radiation oscillations in experiments and thelack of a self-consistent plasma solution at high radiated power fraction in EMC3-Eirene simulations.EMC3-Eirene simulations are used to understand the radiation distribution. It was found that the O-point localized radiation arises first from local impurity accumulation near the parallel flow stagnation,which is located close to the geometrical center of the island (”O-point”). The local cooling in thisregion leads to plasma condensation in the islands in closest magnetic connection to the divertor targetplates. The heat source to this region of the island, which is thermally isolated from the upstream heatsource in terms of parallel transport, must arise via perpendicular heat transport. This heat source isexpected to be large for the low iota configuration due to its very small internal island field line pitch.This work highlights the importance (complementary to previous work, e. g. [5, 6]) of the internalisland field line pitch not only on the radiation pattern, but also the detachment performance of theisland divertor.

Winters, Victoria (ORCID:0000000181087774)↗

DOE OSTI2024

Evaluation of SPARC divertor conditions in H-mode operation using SOLPS-ITER

The predicted divertor conditions for the SPARC tokamak are calculated using SOLPS-ITER for a range of scrape-off-layer (SOL) heat flux widths λ q , input powers, and particle fueling locations. Under H-mode scenario conditions with an upstream separatrix density of 1 x 10 20 m -3 , the most conservative range of λ q extrapolations ( 0.15 mm) results in extremely high unmitigated particle and energy fluxes to the divertor, both under full field (12.2 T) and power (P SOL = 29 MW) conditions, and 2/3 field with P SOL = 10 MW. Increasing the cross-field SOL diffusivities by 2–10× reduces the magnitude of the mitigation challenge, however strategies such as impurity seeding or strike-point-sweeping will likely still be required. A combination of steady-state and time-dependent SOLPS-ITER simulations are used to map out phase space diagrams of upstream and divertor conditions. The simulations include parallel currents but neglect cross-field drifts. At low upstream density the inner and outer divertor conditions are highly asymmetric, with a large temperature difference and significant heat fluxes driven by parallel currents. The solution has sharp bifurcations with a region of hysteresis, depending on whether the initial state is at a low or high density. This behavior is observed even when the fueling location, cross-field diffusivity, and impurity level is changed, although the density window with asymmetry is reduced with increasing diffusivity. The addition of neon impurity seeding reduces the divertor heat fluxes, but also causes a drop in the upstream electron density with fixed particle throughput. This drop can be counteracted by increased main ion throughput, however too much neon results in a back transition into the asymmetric divertor regimes suggesting a need for control of both main ion and impurity seeding levels to achieve a desired divertor state.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2025

Measurements of the polarization of several instabilities in the DIII-D tokamak

Recently, a method to infer the polarization of modes with frequencies much less than the ion cyclotron frequency was published [X.D. Du et al., Phys. Rev. Lett. 132 (2024) 215101]. The method uses measurements of electron temperature and density fluctuations δTₑ and δnₑ at the same spatial position to infer the local ratio of “acoustic polarization,” |δϕ ∥ |/(|δϕ ∥ |+|δψ|), where δϕ ∥ is the effective parallel potential and δψ is related to the parallel magnetic vector potential A ∥ . This paper summarizes key formulas, with emphasis on their range of validity, and elaborates on the workflow required to infer the acoustic polarization from experimental data. The drift-acoustic polarization of ellipticity-induced, toroidicity-induced, and reversed shear Alfvén eigenmodes is nearly zero, as expected for modes with predominately shear-Alfvénic polarization. The polarization of beta-induced Alfvén eigenmodes contains an acoustic component that increases with poloidal wave number. “Low frequency modes,” (instabilities that appear transiently when the minimum of the safety factor qₘᵢₙ passes through rational values) have large and highly variable acoustic polarization. In both experiment and simulation, fishbones have non-zero acoustic polarization that increases as the mode chirps down in frequency.

Alfven eigenmode↗

DOE OSTI2026

Spherical tokamak physics research in preparation for the operation of NSTX-U

The National Spherical Torus Experiment Upgrade (NSTX-U) is preparing to resume operation, representing a crucial step toward realizing compact, cost-effective fusion pilot plants. In advance of this, extensive modeling and data analysis have been conducted to advance the physics basis for low-aspect-ratio, high-performance plasma regimes, focusing on three core objectives: confinement and stability, power and particle handling, and steady-state operation. Significant progress has been made in understanding the electron temperature flattening in high-β plasmas, which is shown to be driven by a complex interplay of magnetohydrodynamic instabilities (e.g. non-resonant infernal modes), fast-ion-driven Alfvén eigenmodes, and electron and ion-scale micro-instabilities, particularly Kinetic Ballooning Modes (KBMs), whose destabilization is strongly dependent on parallel magnetic field fluctuations (δB ∥ ). Furthermore, a new gyrokinetic critical pedestal model was developed, accurately predicting pedestal structure by identifying KBMs as the primary stability limit, offering a critical constraint for future high-confinement scenarios. To address the challenge of high heat flux, novel liquid lithium plasma-facing components were modeled. The analysis confirmed that lithium vapor shielding is a self-regulating mechanism for heat mitigation, while also emphasizing that strong main ion parallel flow is essential to minimize core lithium contamination. Finally, progress toward steady-state operation was anchored by developing the required physics basis and control tools. This includes predictive modeling for reversed magnetic shear sustainment, demonstrating that magnetic island-induced bootstrap current reduction is negligible in STs, and advancing real-time control and disruption avoidance capabilities. The development of high-speed surrogate models (e.g. MMMNet) provides computationally efficient tools vital for non-inductive scenario optimization and integrated, low-disruptivity operations planned for NSTX-U.

NSTX-U↗

DOE OSTI2024

Relative molecular orientation can impact the onset of plasticity in molecular crystals

Abstract Creating or moving dislocations is the first step to dissipating mechanical energy via plastic deformation under contact loading. In molecular crystals there is both a lattice that defines crystal orientation and a relative orientation of the basis of the molecules. We define a normalization parameter which relates strain at yield, the hardness of the bulk crystal, and a distance parameter analogous to a Burgers vector that nominally predicts the relative ease of initiating plasticity in this broad class of materials. Analyzing the yield behavior of 10 different molecular crystals of varying space groups shows the inter-molecular orientation predicts the experimentally observed applied stress needed to nucleate dislocations. When molecules are oriented ‘parallel’ relative to one another the normalized maximum shear stress at the onset of plasticity is on the order of 3–5 times lower than when molecules within the crystal are ‘anti-parallel’, and molecules with a more equiaxed shape fall in between these bounds. This provides an initial indication of a structural feature which predicts the relative ease of initiating plasticity during contact loading in molecular crystals.

36 MATERIALS SCIENCE↗

DOE OSTI2026

Linear complexity

We present factorization and solution phases for a new linear complexity direct solver designed for concurrent batch operations on fine-grained parallel architectures, for matrices amenable to hierarchical representation. We focus on the strong-admissibility-based $\mathscr{H}^{2}$ format, where strong recursive skeletonization factorization compresses remote interactions. We build upon previous implementations of $\mathscr{H}^{2}$ matrix construction for efficient factorization and solution algorithm design, which are illustrated graphically in stepwise detail. The algorithms are ‘blackbox’ in the sense that the only inputs are the matrix and right-hand side, without analytical or geometrical information about the origin of the system. We demonstrate linear complexity scaling in both time and memory on four representative families of dense matrices up to one million in size. Parallel scaling up to 16 threads is enabled by a multi-level matrix graph coloring and avoidance of dynamic memory allocations thanks to prefix-sum memory management. An experimental backward error analysis is included. We break down the timings of different phases, identify phases that are memory-bandwidth limited, and discuss alternatives for phases that may be sensitive to the trend to employ lower precisions for performance.

Boukaram, Wajih↗

DOE OSTI2025

Measuring and unbiasing the BAO shift in the Ly α forest with AbacusSummit

ABSTRACT The Dark Energy Spectroscopic Instrument (DESI) places sub- per cent constraints on measurements of the Baryon Acoustic Oscillation (BAO) scaling parameters from the Ly $\alpha$ forest. However, no systematic error budget stemming from non-linearities in the three-dimensional clustering of the Ly $\alpha$ forest is included in the DESI-Ly $\alpha$ analysis. In this work, we measure the size of the shift of the BAO peak using large Ly $\alpha$ forest mocks produced on the N-body simulation suite AbacusSummit, which adopt the Fluctuating–Gunn–Peterson Approximation (FGPA). Specifically, we measure the Ly $\alpha$ autocorrelation and the Ly $\alpha$-quasar cross-correlation functions. To mitigate the noise, we adopt a linear control variates technique, reducing the error bars by a factor of up to $\sim \sqrt{50}$ on large scales. From the autocorrelation, we detect a small positive shift in radial direction of $\Delta \alpha _{\parallel }= 0.35~{{\ \rm per\ cent}}$ at the 3$\sigma$ level and virtually no shift in the transverse direction, $\alpha _\perp$. From the cross-correlation, we see a similar shift to $\Delta \alpha _\parallel$, albeit with larger error bars, and a small negative shift, $\Delta \alpha _{\perp }=\sim$0.25 per cent, at the 2$\sigma$ level. We also make a connection with the Ly $\alpha$ forest effective field theory (EFT) framework and find that the one-loop EFT power spectrum yields unbiased measurements of the BAO shift parameters in radial and transverse direction for Ly $\alpha$ auto- and the Ly $\alpha$-quasar cross-correlation measurements. When using the one-loop EFT framework, we find that we can recover the BAO parameters without a shift, which has important implications for future Ly $\alpha$ forest analyses based on EFT. This work paves the way for novel full-shape analyses of the currently observing DESI and future surveys such as the PFS, WEAVE-QSO, and 4MOST.

Hadzhiyska, Boryana↗

DOE OSTI2025

Towards modelling AR Sco: calibration – reproducing high-energy pulsar emission and testing convergence to Aristotelian electrodynamics

In recent years, kinetic simulations have been crucial to further our understanding of pulsar electrodynamics. Yet, due to the large-scale separation between the gyro-period and the stellar rotation period, resolving the particle gyration has been computationally unfeasible for realistic pulsar parameters. The main aim of this work is comparing our gyro-phase-resolved model with a gyro-centric pulsar model, where our model solves the general equations of motion with included radiation reaction using a higher order numerical solver with adaptive time-steps. Specifically, we aim to (i) reproduce a pulsar’s high-energy emission maps, namely one with 10 per cent of the surface B-field strength of Vela, and the spectra produced by an independent gyro-centric pulsar emission model; and (ii) test convergence of these results to the radiation-reaction limit of Aristotelian electrodynamics. (iii) Additionally, we identify the effect that a large $E_{\parallel }$-field has on the trajectories and radiation calculations. We find that we can reproduce the curvature radiation emission maps and spectra well, using 10 per cent field strengths of the Vela pulsar and injecting our particles at a higher altitude in the magnetosphere. Using sufficiently large $E_{\parallel }$-fields, our numeric results converge to the analytic radiation-reaction limit trajectories. Additionally, we illustrate the importance of accounting for the $\mathbf {E}\times \mathbf {B}$-drift in the particle trajectories and radiation calculations, validating the Harding and collaborators’ model approach. Lastly, we found that our model deals very well with the high-radiation-reaction and high-field regimes present in pulsars.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2025

New scaling and nuclear structure aspects in heavy-ion fusion reactions

Three new behaviors have been found in comparisons of fusion cross sections for different collision systems. root (1) Replacing the energy E with a scaling one, E scal = (E-V g )/($\sqrt{2}$W g ), is successful for washing out the Coulomb interaction in the spectra of fusion cross sections, where V g and W g are barrier height and width of the single-Gaussian barrier distribution model. (2) In a representation of σE vs the scaling energy, E scal , all data sets display in parallel. Here, the ratio for sigma E from any two fusion systems over the whole range is a constant value. That behavior is also studied in another representation, in which the data sets display as parallel horizontal lines for any heavy-ion fusion system. (3) The constant ratio value is the ratio of parameter products, $R^2_gW_g$, of the two systems; where R g is the barrier radius obtained in the single-Gaussian barrier distribution model. Moreover, when comparing neighboring collision systems at the same E scal , the ratio of sigma is near a constant value within a few percent over the whole range. Thus a quantitative comparison for the fusion enhancement for neighboring systems is developed. The present finding could be beneficial for predicting unmeasured fusion cross sections.

Jiang, C. L. [Argonne National Laboratory (ANL), A↗

DOE OSTI2024

Geometric origin of the intrinsic transverse spin transport in a canted-antiferromagnet/heavy-metal heterostructure

We theoretically study the conditions under which an intrinsic spin Nernst effect–a transverse spin current induced by an applied temperature gradient–can occur in a canted-antiferromagnet insulator, such as LaFeO 3 and other materials of the same family. The spin Nernst effect may provide a microscopic mechanism for an experimentally observed anomalous thermovoltage in LaFeO 3 /Pt heterostructures, where spin is transferred across the insulator/metal interface when a temperature gradient is applied to LaFeO 3 parallel to the interface. We find that LaFeO 3 exhibits an intrinsic spin Nernst effect when inversion symmetry is broken on the axes parallel to both the applied temperature gradient and the direction of spin transport, which can result in a spin injection across the insulator/metal interface. Furthermore, our paper provides a general derivation of a symmetry-breaking-induced spin Nernst effect, which may open a path to engineering a finite spin Nernst effect in systems where it would otherwise not arise.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

DOE OSTI2026

Zonal magnetic fields regulate nonlinear edge-localized-mode dynamics via self-consistent force balance

Edge-localized modes (ELMs) eject intense bursts of heat and particles that threaten plasma-facing components in fusion reactors. Nonlinear full-torus BOUT++ simulations show that turbulence-driven zonal magnetic fields (ZMFs) play an essential role in nonlinear ELM evolution by maintaining self-consistent force balance. Zonal flows mitigate the initial crash through shear but do not prevent continued radial transport. When ZMFs are self-consistently included, turbulence-driven zonal currents modify the parallel current distribution and magnetic tension and are associated with a reduction of the axisymmetric (𝑛 = 0) perturbed radial force imbalance. This coincides with a transition from convective, bursty propagation to more localized, diffusive transport. Similar behavior is observed across the regimes considered, including both resistive-ballooning and peeling-ballooning cases. Finally, associated signatures, including radial electric field shear and parallel current redistribution, provide experimentally accessible diagnostics for present devices and ITER-relevant conditions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2026

Real-Time GPU-Accelerated OFDR With an Integrated Auxiliary Interferometer

A GPU-accelerated optical frequency domain reflectometry (OFDR) system with an improved integrated auxiliary interferometer is proposed. Unlike conventional approaches that require separate auxiliary interferometers and multiple detection channels, the proposed OFDR system embeds this functionality directly into the signal via an intentional beat component. This enables self-calibration of laser nonlinearity while maintaining a cost-effective hardware configuration. Building on this simplified configuration, the system leverages GPU acceleration with an NVIDIA RTX 4070 Ti to achieve real-time performance, delivering high-throughput signal processing for continuous OFDR interrogation. The signal processing pipeline comprises signal capture, resampling for nonlinearity compensation, and frequency shift computation, all optimized for parallel execution. Hardware benchmarking demonstrates substantial acceleration over CPU implementations, achieving up to a 45× speedup for resampling and frequency shift computations and enabling processing latencies below 30 ms. Thermal response validation is conducted under two complementary scenarios: localized heating using a water bath and cryogenic-temperature conditions using liquid nitrogen. Under localized heating, the system achieves an accuracy of 0.249 °C with a thermal sensitivity of 5.971 GHz/°C, while cryogenic-temperature validation demonstrates a frequency shift response with a sensitivity of 2.383 GHz/°C and an accuracy of 2.04 °C. The high acceleration of the proposed GPU-accelerated OFDR system and its accuracy are achieved by exploiting CUDA-based stride indexing, enabling efficient parallel segmentation and processing of large datasets without additional memory copies. The benchmarking results confirm the robustness, accuracy, and deployability of the proposed OFDR system across a wide temperature range, establishing it as a practical platform for real-time distributed fiber sensing in structurally dynamic environments.

Harb, Salah [Lawrence Berkeley National Laboratory↗

DOE OSTI2024

Design of High-Power Polyphase PCB Coil Systems for Wireless Power Transfer

Printed circuit board (PCB) coils have been proposed prior for implementation as inductive wireless charging coils to minimize size and cost. Utilization of PCBs can allow for a reduced cost, improved manufacturability, and a wide range of geometric customization options. To circumvent material limitations on insulation and thermal performance, parallel paths can be implemented to divide the current per path accordingly. Within this paper, an unconventional high-power PCB coil is designed employing all possible techniques for wiring with axial and radial parallel paths with equivalent transposition to minimize circulating currents. Design studies are simulated in 3D finite element analysis (FEA) to evaluate imbalance between phases with and without transposition. Two experimental prototype coils were fabricated with measurements for self-inductance and mutual inductance between phases. These measurements were validated to be sufficiently consistent with FEA results. Additionally, a method is proposed for a two-step optimization of coupling coefficient and coil losses.

Lewis, Donovin D.↗

DOE OSTI2025

Modified Andronov-Hopf Oscillator-Based Grid-Forming Converter with Emulated Virtual Cable for Enhanced Power Sharing Performance

Nonlinear oscillator-based grid-forming converters offer superior dynamic and steady-state performance, making them an attractive solution for interconnecting renewable resources. This paper proposes a novel modified Andronov-Hopf oscillator to enhance the operating spectrum and facilitate the integration of renewable energy sources. An inner loop controller based on the Lyapunov energy function is implemented to achieve robust stability and performance, while a virtual cable emulation strategy enables seamless parallel operation. Comprehensive modeling and simulation studies validate the effectiveness of the proposed system, demonstrating its capabilities in addressing diverse operating scenarios, including grid faults, renewable energy fluctuations, and parallel operation. The proposed solution exhibits fast transient response, robust stability, and flexible operation, making it a valuable contribution to the field of renewable energy integration. The results of this study can be used to inform the design and implementation of next-generation grid-forming converters, enabling a more sustainable and reliable energy future. Additionally, the proposed system's ability to operate in both grid-connected and islanded modes makes it an ideal candidate for remote and off-grid renewable energy applications. The proposed solution's scalability and modularity also make it suitable for large-scale renewable energy integration. The proposed system is verified through MATLAB/Simulink and PLECS simulations, demonstrating its effectiveness in ensuring robust and efficient operation.

Andronov-Hopf Oscillator (AHO)↗

DOE OSTI2026

Interpretable Models for Workflow Differentiation in High-Performance Scientific Networks

Scientific workflows in high-performance networks spawn hundreds of interdependent flows that must be managed collectively—yet existing network classifiers treat each flow in isolation, leading to fragmented QoS decisions and missed interflow patterns. We present a novel traffic classification solution that operates at the workflow level, distinguishing entire filetransfer operations from streaming analytics by capturing how concurrent flows interact and burst together. We introduce a workflow identification window (WIW) that ingests raw packet headers from parallel flows into unified tensors, preserving the spatial-temporal patterns that differentiate scientific workflows. This approach achieves 98.7% accuracy using CNN, LSTM, and hybrid architectures, while maintaining 84% accuracy on production traffic collected a week later—demonstrating robustness to temporal drift. By integrating SHAP and GradCAM explainability, we reveal that early-packet timing patterns and cross-flow correlations drive classification decisions, providing operators with interpretable insights. Our system enables coherent workflow-level QoS enforcement and dynamic bandwidth allocation in scientific networks, eliminating manual per-flow configuration while maintaining classification latency at millisecond level.

Giannakou, Anna [LBL, Berkeley]↗