Search NASA⌕ Search

SEARCH · Search NASA

Results for “model parameterization”

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.

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

At least 487 records · Page 27
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

A Lie algebraic theory of barren plateaus for deep parameterized quantum circuits

Variational quantum computing schemes train a loss function by sending an initial state through a parametrized quantum circuit, and measuring the expectation value of some operator. Despite their promise, the trainability of these algorithms is hindered by barren plateaus (BPs) induced by the expressiveness of the circuit, the entanglement of the input data, the locality of the observable, or the presence of noise. Up to this point, these sources of BPs have been regarded as independent. In this work, we present a general Lie algebraic theory that provides an exact expression for the variance of the loss function of sufficiently deep parametrized quantum circuits, even in the presence of certain noise models. Our results allow us to understand under one framework all aforementioned sources of BPs. This theoretical leap resolves a standing conjecture about a connection between loss concentration and the dimension of the Lie algebra of the circuit’s generators.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Synthetic method of analogues for emerging infectious disease forecasting

The Method of Analogues (MOA) has gained popularity in the past decade for infectious disease forecasting due to its non-parametric nature. In MOA, the local behavior observed in a time series is matched to the local behaviors of several historical time series. The known values that directly follow the historical time series that best match the observed time series are used to calculate a forecast. This non-parametric approach leverages historical trends to produce forecasts without extensive parameterization, making it highly adaptable. However, MOA is limited in scenarios where historical data is sparse. This limitation was particularly evident during the early stages of the COVID-19 pandemic, where the emerging global epidemic had little-to-no historical data. In this work, we propose a new method inspired by MOA, called the Synthetic Method of Analogues (sMOA). sMOA replaces historical disease data with a library of synthetic data that describe a broad range of possible disease trends. This model circumvents the need to estimate explicit parameter values by instead matching segments of ongoing time series data to a comprehensive library of synthetically generated segments of time series data. We demonstrate that sMOA has competitive performance with state-of-the-art infectious disease forecasting models, out-performing 78% of models from the COVID-19 Forecasting Hub in terms of averaged Mean Absolute Error and 76% of models from the COVID-19 Forecasting Hub in terms of averaged Weighted Interval Score. Additionally, we introduce a novel uncertainty quantification methodology designed for the onset of emerging epidemics. Developing versatile approaches that do not rely on historical data and can maintain high accuracy in the face of novel pandemics is critical for enhancing public health decision-making and strengthening preparedness for future outbreaks.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

A Novel Noise-Aware Classical Optimizer for Variational Quantum Algorithms

A key component of variational quantum algorithms (VQAs) is the choice of classical optimizer employed to update the parameterization of an ansatz. It is well recognized that quantum algorithms will, for the foreseeable future, necessarily be run on noisy devices with limited fidelities. Thus, the evaluation of an objective function (e.g., the guiding function in the quantum approximate optimization algorithm (QAOA) or the expectation of the electronic Hamiltonian in variational quantum eigensolver (VQE)) required by a classical optimizer is subject not only to stochastic error from estimating an expected value but also to error resulting from intermittent hardware noise. Model-based derivative-free optimization methods have emerged as popular choices of a classical optimizer in the noisy VQA setting, based on empirical studies. However, these optimization methods were not explicitly designed with the consideration of noise. In this work we adapt recent developments from the “noise-aware numerical optimization” literature to these commonly used derivative-free model-based methods. We introduce the key defining characteristics of these novel noise-aware derivative-free model-based methods that separate them from standard model-based methods. In conclusion, we study an implementation of such noise-aware derivative-free model-based methods and compare its performance on demonstrative VQA simulations to classical solvers packaged in scikit-quant.

classical optimizers↗
DOE OSTI2024

The vertical-velocity skewness in the atmospheric boundary layer without buoyancy and Coriolis effects

One of the main features of near-neutral atmospheric boundary layer (ABL) turbulence is the positive vertical velocity skewness $Sk_w$ above the roughness sublayer or the buffer region in smooth-walls. The $Sk_w$ variations are receiving renewed interest in many climate-related parameterizations of the ABL given their significance to cloud formation and to testing sub-grid schemes for Large Eddy Simulations (LES). The vertical variations of $Sk_w$ are explored here using wind tunnel and flume experiments collected above smooth, rough, and permeable-walls in the absence of buoyancy and Coriolis effects. These laboratory experiments form a necessary starting point to probe the canonical structure of $Sk_w$ as they deal with a key limiting case (i.e., near-neutral conditions). Diagnostic models based on cumulant expansions, realizability constraints, and constant mass flux approach routinely employed in the convective boundary layer as well as prognostic models based on third-order budgets are used to explain variations in $Sk_w$ for the idealized laboratory conditions. The failure of flux-gradient relations to model $Sk_w$ from the gradients of the vertical velocity variance σ$_w^2$ are explained and corrections based on models of energy transport offered. Novel links between the diagnostic and prognostic models are also featured, especially for the inertial term in the third-order budget of the vertical velocity fluctuation. The co-spectral properties of w′/σ w vs w′ 2 /σ$_w^2$ are also presented for the first time to assess the dominant scales governing $Sk_w$ in the inner and outer layers, where w′ is the fluctuating vertical velocity and σ w is the vertical velocity standard deviation.>

Boundary layer flow↗
DOE OSTI2024

Quantifying scalar field dynamics with DESI 2024 Y1 BAO measurements

Quintessence scalar fields are a natural candidate for evolving dark energy. Unlike the phenomenological w 0 ⁢w a parameterization of the dark energy equation of state, they cannot accommodate the phantom regime of dark energy w⁡(z) <–1, or crossings into the phantom regime. Recent baryon acoustic oscillation (BAO) measurements by the Dark Energy Spectroscopic Instrument (DESI) indicate a preference for evolving dark energy over a cosmological constant, ranging from 2.6⁢σ –3.9⁢σ when fitting to w 0⁢ w a , and combining the DESI BAO measurements with other cosmological probes. In this work, we directly fit three simple scalar field models to the DESI BAO data, combined with cosmic microwave background anisotropy measurements and supernova datasets. We find the best fit model to include a 2–4% kinetic scalar field energy Ω scf,k , for a canonical scalar field with a quadratic or linear potential. However, only the DESY-Y5 supernova dataset combination shows a preference for quintessence over Λ cold dark matter (CDM) at the 95% confidence level. Fitting to the supernova datasets Pantheon, Pantheon+, DES-Y5, and Union3, we show that the mild tension (n σ < 3.4) under Λ ⁢CDM emerges from a BAO preference for smaller values of fractional mass-energy density Ω m < 0.29, while all supernova datasets, except for Pantheon, prefer larger values, Ω m > 0.3. The tension under Λ ⁢CDM remains noticeable (n σ < 2.8), when replacing two of the DESI BAO redshift bins with effective redshifts z eff = 0.51, and z eff = 0.706 with comparable BOSS DR 12 BAO measurements at z eff = 0.51, and z eff = 0.61. Furthermore, canonical scalar fields as dark energy are successful in mitigating that tension.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2026

TRANSP-TGLF core predictive modeling of the JET DT baseline scenario

In recent years, an intense modeling activity has been focused on preparing and analyzing the second JET Deuterium–Tritium (D–T) experimental campaign DTE2. Among the numerous scientific outcomes of this campaign was the unique opportunity to test and validate the state of the art of modeling tools with fusion-relevant DT plasmas using the full metallic ITER-like wall in different scenarios. This work reports on the core predictive modeling of plasma density, electron and ion temperatures (n e , T e , T i ) performed using TRANSP (Pankin et al 2025 Comput. Phys. Commun. 312 109611) coupled with Trapped Gyro-Landau Fluid (TGLF)-SAT2 (Kinsey et al 2008 Phys. Plasmas 15 055908), (Staebler et al 2021 Nucl. Fusion 61 116007) for the JET D–T baseline scenario (I p = 3.5 MA, q 95 = 3, β N < 2, with pellet pacing) (Garzotti et al 2025 Plasma Phys. Control. Fusion 67 075011). The sensitivity to different input parameters, as the $\vec{E}$ x $\vec{B}$ shear parameterization and the values of the kinetic quantities at the boundary of the prediction domain (ρ = 0.85) has been assessed, identifying the confidence interval of the prediction results. In particular, the dependence of the electron density profile on the particle source parameters has been studied, identifying the ionization source as the main cause for the density gradient under-prediction obtained by TRANSP-TGLF, reported in (Hyun-Tae et al 2023 Nucl. Fusion 63 112004).

JET↗
DOE OSTI2025

SympGNNs: Symplectic Graph Neural Networks for identifying high-dimensional Hamiltonian systems and node classification

Existing neural network models to learn Hamiltonian systems, such as SympNets, although accurate in low-dimensions, struggle to learn the correct dynamics for high-dimensional many-body systems. Herein, we introduce Symplectic Graph Neural Networks (SympGNNs) that can effectively handle system identification in high-dimensional Hamiltonian systems, as well as node classification. SympGNNs combine symplectic maps with permutation equivariance, a property of graph neural networks. Specifically, we propose two variants of SympGNNs: (i) G-SympGNN and (ii) LA-SympGNN, arising from different parameterizations of the kinetic and potential energy. We demonstrate the capabilities of SympGNN on two physical examples: a 40-particle coupled Harmonic oscillator, and a 2000-particle molecular dynamics simulation in a two-dimensional Lennard-Jones potential. Furthermore, we demonstrate the performance of SympGNN in the node classification task, achieving accuracy comparable to the state-of-the-art. Finally, we also empirically show that SympGNN can overcome the oversmoothing and heterophily problems, two key challenges in the field of graph neural networks.

Deep learning↗
DOE OSTI2025

A Machine Learning Framework for Predicting Microphysical Properties of Ice Crystals From Cloud Particle Imagery

The microphysical properties of ice crystals are important because they significantly alter the radiative properties and spatiotemporal distributions of clouds, which in turn strongly affect Earth's climate. However, it is challenging to measure key properties of ice crystals, such as mass or morphological features. Here, we present a proof-of-concept framework for predicting three-dimensional (3D) microphysical properties of ice crystals from in situ two-dimensional (2D) imagery. First, we computationally generated synthetic ice crystals using 3D modeling software along with geometric parameters estimated from the 2021 Ice Cryo-Encapsulation Balloon (ICEBall) field campaign. Then, we used synthetic crystals to train machine learning (ML) models to predict effective density ($ρ_e$), effective surface area ($A_e$), and number of bullets ($N_b$) from synthetic rosette imagery. On unseen synthetic images, our ML models accurately predicted ice crystal properties. ResNet-18 performed best, achieving $R^2$ values of 0.99 and 0.98 for $ρ_e$ and $A_e$, respectively, and MAE of 0.10 for mathematical equation in single view tasks. Stereo view ResNet-18 further reduced RMSE by 40% for $ρ_e$ and $A_e$ and reduced MAE by 0.08 for $N_b$. This work provides a novel ML-driven framework for estimating ice microphysical properties from in situ imagery, which will allow for downstream constraints on microphysical parameterizations, such as the mass-size relationship.

Ko, J. [Columbia Univ., New York, NY (United State↗
DOE OSTI2025

Field intercomparison of ice nucleation measurements: the Fifth International Workshop on Ice Nucleation Phase 3 (FIN-03)

Abstract. The third phase of the Fifth International Ice Nucleation Workshop (FIN-03) was conducted at the Storm Peak Laboratory in Steamboat Springs, Colorado, in September 2015 to facilitate the intercomparison of instruments measuring ice-nucleating particles (INPs) in the field. Instruments included two online and four offline measurement systems for INPs, which are a subset of those utilized in the laboratory study that comprised the second phase of FIN (FIN-02). The composition of the total aerosols was characterized using the Particle Analysis by Laser Mass Spectrometry (PALMS) and Wideband Integrated Bioaerosol Sensor (WIBS) instruments, and aerosol size distributions were measured by a laser aerosol spectrometer (LAS). The dominant total particle compositions present during FIN-03 were composed of sulfates, organic compounds, and nitrates, as well as particles derived from biomass burning. Mineral-dust-containing particles were ubiquitous throughout and represented 67 % of supermicron particles. Total WIBS fluorescing particle concentrations for particles with diameters of > 0.5 µm were 0.04 ± 0.02 cm−3 (0.1 cm−3 highest; 0.02 cm−3 lowest), typical of the warm season in this region and representing ≈ 9 % of all particles in this size range as a campaign average. The primary focus of FIN-03 was the measurement of INP concentrations via immersion freezing at temperatures > −33 °C. Additionally, some measurements were made in the deposition nucleation regime at these same temperatures, representing one of the first efforts to include both mechanisms within a field campaign. INP concentrations via immersion freezing agreed within factors ranging from nearly 1 to 5 times on average between matched (time and temperature) measurements, and disagreements only rarely exceeded 1 order of magnitude for sampling times coordinated to within 3 h. Comparisons were restricted to temperatures lower than −15 °C due to the limits of detection related to sample volumes and very low INP concentrations. Outliers of up to 2 orders of magnitude occurred between −25 and −18 °C; a better agreement was seen at higher and lower temperatures. Although the 5–10 factor agreement of INP measurements found in FIN-03 aligned with the results of the FIN-02 laboratory comparison phase, giving confidence in progress of this measurement field, this level of agreement still equates to temperature uncertainties of 3.5 to 5 °C that may not be sufficient for numerical cloud modeling applications that utilize INP information. INP activity in the immersion-freezing mode was generally found to be an order of magnitude or more, making it more efficient than in the deposition regime at 95 %–99 % water relative humidity, although this limited data set should be augmented in future efforts. To contextualize the study results, an assessment was made of the composition of INPs during the late-summer to early-fall period of this study inferred through comparison to existing ice nucleation parameterizations and through measurement of the influence of thermal and organic carbon digestion treatments on immersion-freezing ice nucleation activity. Consistent with other studies in continental regions, biological INPs dominated at temperatures of > −20 °C and sometimes colder, while arable dust-like or other organic-influenced INPs were inferred to dominate below −20 °C.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Flow dynamics and heat transfer in simplified battery energy storage systems with heated battery modules

Large-scale energy storage systems (ESSs) composed of batteries show promise in addressing current energy challenges, but dissipation of generated heat is important. Here, this paper focuses on buoyant convective flows in simplified ESS battery racks. Natural convection is not generally the primary cooling strategy but can be important in abnormal scenarios where there is module overheat or potentially thermal runaway. We use computational fluid dynamics to investigate the flow dynamics and heat transfer mechanisms in a simplified parameterized rack design. Despite its simplicity, this configuration produces many of the relevant features expected in real ESSs without details of module geometry or hardware, allowing broad conclusions independent of manufacture-specific designs. We start by providing visualizations of the flowfield and measurements of entrainment, heat flux, and pressure. To characterize the dependence on the system parameters, we develop an integral-scale analysis of the average temperature equation to highlight the dominant source terms. We use results from this analysis to derive a steady network model composed of simple algebraic expressions to provide first-order predictions of entrainment through the rack. The network model leads to a linear scaling of the Reynolds number based on convective mass flux with respect to the Grashof number based on the heat source. We deduce empirical relationships that relate the heat exchanged between modules using a surface-averaged Nusselt number as a function of the local Reynolds and Rayleigh numbers. Lastly, we investigate how space between the modules and rack in the spanwise direction creates flow bypass, resulting in different flow pathways.

Battery thermal management↗
DOE OSTI2024

Deep Learning-based Parameterization of Complex 3D CO2 Saturation Data in Large-scale Geological Carbon Storage

In deep learning (DL), dimension reduction plays a pivotal role in improving training efficiency and minimizing overfitting, especially when working with complex datasets like three-dimensional (3D) saturation data. In the context of geological carbon storage (GCS), 3D saturation data introduces unique challenges due to its sparse nature and sharp transitions at plume boundaries, known as shock fronts. To tackle these challenges, we developed a novel DL framework that combines dimension reduction with advanced 3D reconstruction techniques. Our approach utilizes latent variables derived from 2D average saturation fields to efficiently capture the essential features of high-dimensional data while reducing the number of variables. This enhances both the robustness and accuracy of DL models, making the framework more practical for real-world applications. By offering a tailored solution for modeling complex 3D saturation dynamics, this framework holds significant potential for environmental monitoring, energy storage, and other geological applications.

Wang, Hongsheng [University of Texas at Austin]↗
DOE OSTI2026

Differences in cluster and internal wake effects from mesoscale and large-eddy simulations off the US East Coast

Mesoscale simulations are increasingly used to estimate wake effects within and between large wind farms, despite limited validation for large-scale wake effects. This study evaluates the capabilities and limitations of mesoscale simulations in capturing wake-induced impacts on wind turbine power production through a direct comparison with large-domain large-eddy simulations (LESs) for three planned offshore wind farms under realistic atmospheric conditions and a range of atmospheric stabilities. We assess mesoscale performance in replicating wake characteristics behind single and multiple turbine clusters and quantify the resulting variability in mean turbine power. Results show that mesoscale Weather Research and Forecasting simulations with the Fitch wind farm parameterization capture key features of the velocity deficit downstream of both single and multiple wind farms, with mean root-mean-square errors near 5 % and good agreement with stability-driven wake behavior. However, in these simulations, the mesoscale Fitch parameterization underestimates power losses from internal wake effects, particularly when turbines align with the prevailing wind direction or under stable stratification. In these conditions, individual wakes persist and dominate downstream power deficits. The coarse resolution of the mesoscale simulations limits their ability to resolve individual wind turbine wakes that drive power fluctuations within wind farms. Nonetheless, mesoscale simulations can yield accurate estimates of combined wake losses from internal and cluster effects across some wind direction sectors, where errors in wake representation may cancel each other out. These findings underscore the strengths of mesoscale simulations for capturing broader wake patterns while highlighting their limitations for modeling turbine-level power losses. Future work should explore hybrid modeling approaches to capture both long-range cluster wake propagation and localized internal wake dynamics.

17 WIND ENERGY↗
DOE OSTI2024

Extreme Risk Mitigation in Reinforcement Learning using Extreme Value Theory

Risk-sensitive reinforcement learning (RL) has garnered significant attention in recent years due to the growing interest in deploying RL agents in real-world scenarios. A critical aspect of risk awareness involves modelling highly rare risk events (rewards) that could potentially lead to catastrophic outcomes. These infrequent occurrences present a formidable challenge for data-driven methods aiming to capture such risky events accurately. While risk-aware RL techniques do exist, they suffer from high variance estimation due to the inherent data scarcity. Our work proposes to enhance the resilience of RL agents when faced with very rare and risky events by focusing on refining the predictions of the extreme values predicted by the state-action value distribution. To achieve this, we formulate the extreme values of the state-action value function distribution as parameterized distributions, drawing inspiration from the principles of extreme value theory (EVT). We propose an extreme value theory based actor-critic approach, namely, Extreme Valued Actor-Critic (EVAC) which effectively addresses the issue of infrequent occurrence by leveraging EVT-based parameterization. Importantly, we theoretically demonstrate the advantages of employing these parameterized distributions in contrast to other risk-averse algorithms. Our evaluations show that the proposed method outperforms other risk averse RL algorithms on a diverse range of benchmark tasks, each encompassing distinct risk scenarios.

Wang, Yu↗
DOE OSTI2024

Simulating Mixed‐Phase Open Cellular Clouds Observed During COMBLE: Evaluation of Parameterized Turbulence Closure

Marine cold-air outbreaks, or CAOs, are airmass transformations whereby relatively cold boundary layer (BL) air is transported over relatively warm water. To more deeply understand BL and mixed-phase cloud properties during CAO conditions, the Cold-Air Outbreaks in the Marine Boundary Layer Experiment (COMBLE) took place from late 2019 into early 2020. During COMBLE, the U.S. Department of Energy's first Atmospheric Radiation Measurement Mobile Facility (AMF1) was deployed to Andenes, Norway, far downstream (~1,000 km) from the Arctic pack ice. This study examines the two most intense CAOs sampled at the AMF1 site. The observed BL structures are open cellular with high (~3–5 km) and cold (–30 to –50 °C) cloud tops, and they often have pockets of high liquid water paths (LWPs; up to ~1,000 g m –2 ) associated with strong updrafts and enhanced turbulence. We use a high-resolution mesoscale model to explore how well four turbulence closure methods represent open cellular clouds. After applying a radar simulator to model outputs for direct evaluation, cloud top properties agree well with AMF1 observations (within ~10%), but radar reflectivity and LWP agreement is more variable. Results suggest that the turbulent Prandtl number may play an important role for the simulated BL and cloud properties. All simulations produce enhanced precipitation rates that are well-correlated with a cloud transition. Finally, the eddy-diffusivity/mass-flux approach produces the deepest cloud layer and therefore the largest and most coherent cellular structures. Furthermore, we recommend the use of a non-local turbulence closure approach to better capture turbulent processes in intense CAOs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Direct Numerical Simulation Database of High-Speed Flow over Parameterized Curved Walls

This study presents a direct numerical simulation (DNS) database of high-speed turbulent boundary layers (TBLs) subject to pressure gradients due to parametrically varied backward-facing and forward-facing wall curvatures, with an inflow Mach number of 4.9 and a friction Reynolds number of [Formula: see text] immediately before the onset of wall curvature. The Mach and Reynolds numbers are significantly higher than those reported in the literature for the DNS of pressure-gradient TBLs. The flow conditions and baseline wall geometries are representative of experiments in the high-speed blowdown wind tunnel at the National Aerothermochemistry Laboratory at Texas A&M University. The wall steepness of the baseline geometry for both the backward-facing and forward-facing walls was systematically varied to cause attached, incipiently separated, and fully separated flows. Precomputed flow statistics, including turbulent kinetic energy budgets, are available on the website of the Turbulence Modeling Resource of the NASA Langley Research Center, allowing other investigators to query any property of interest.

Engineering↗
DOE OSTI2025

A Large-Eddy Simulation Study in WRF on Wind over Broadband Waves of Different Directions and Spreading Widths

We use the Weather Research and Forecasting (WRF) Model coupled with moving waves to conduct large-eddy simulations (LESs) of wind over broadband waves with different propagating directions and spreading widths. Our results show that wind-opposing waves can double the form drag, and the wave propagating direction affects the mean wind velocity, velocity variances, and pressure stress. On the other hand, waves with wider spread tend to reduce the form drag in the streamwise direction. Results further indicate that the wave direction can impact the bulk drag coefficient by as much as 25%, while the wave directional spreading width can change it by 5%. Based on our wave-phase-resolved simulations, we demonstrate that the parameterization of sea surface roughness is significantly influenced by the direction of wave propagation relative to the wind, a factor that the commonly used Charnock relation does not account for.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Exposing Process‐Level Biases in a Global Cloud Permitting Model With ARM Observations

The emergence of global convective‐permitting models (GCPMs) represents a significant advancement in climate modeling, offering improved representation of deep convection and complex precipitation patterns. In this study, we evaluate the performance of the Simple Cloud‐Resolving E3SM Atmosphere Model (SCREAM) using its doubly periodic configuration (DP‐SCREAM) against large eddy simulations and modern observational data sets from the Atmospheric Radiation Measurement program. We introduce several new transitional cloud regime cases, such as the transition from shallow to deep convection and from stratocumulus to cumulus, as well as cold‐air outbreak scenarios. The results reveal both strengths and limitations of SCREAM, particularly in the accurate simulation of cloud transitions and midlevel convection, with varying degrees of sensitivity to horizontal and vertical resolution. Despite improvements at higher resolutions, key biases remain, including the abrupt transition from shallow to deep convection and the lack of congestus clouds. These findings underscore the need for further refinement in turbulence parameterizations and vertical grid resolution in GCPMs.

Bogenschutz, Peter A. [Lawrence Livermore National↗
DOE OSTI2025

TRIM: AI Guided Random Number Generation for Resource-Constrained IoT Systems

Random numbers often serve as the backbone for many security solutions in diverse domains such as cryptography, side channel leakage prevention, and moving target defense. However, generating true random numbers requires a physical source of entropy (e.g. hardware, quantum, environmental phenomenon) making it difficult to realize at a large scale and at a low cost. On the flip side, pseudorandom number generators (easy to implement) following a specific distribution (e.g. Gaussian) can be easily compromised given a sufficient amount of traces. In this work, we have developed a machine learning-guided generative approach that can be used to create portable, resource-efficient, and cost-effective random number generators with high throughput and true randomness characteristics. We implement the proposed approach as a highly parameterized framework and perform extensive evaluation for different settings. The framework was able to learn from true random sources such as irrational numbers and environmental audio noise and imitate those sources towards generating new good quality random numbers on demand. We have generated more than 1 billion bits and observed robust performance in terms of true randomness metrics obtained from NIST SP 800-22 and FIPS 140-1 randomness test suites achieving a throughput of up to 142.85 Mbps. Compared to the state-of-the-art (SOTA) technique, the iso-cost setup of our framework can achieve more than 500 Mbps in a distributed setting. We have evaluated the efficacy of running the true randomness imitation AI models on target edge devices such as Raspberry Pi 4 (Model B), Nvidia Jetson Nano, Nvidia Jetson Orin Nano and Nvidia Jetson Xavier. We have also looked at the security of the TRIM framework itself against different adversarial threat models.

Cybersecurity↗