Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “COMPUTING NETWORK”

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 595 records · Page 33

Deep Learning for Unsupervised Insider Threat Detection in Structured Cybersecurity Data Streams

Analysis of an organization’s computer network activity is a key component of early detection and mitigation of insider threat, a growing concern for many organizations. Raw system logs are a prototypical example of streaming data that can quickly scale beyond the cognitive power of a human analyst. As a prospective filter for the human analyst, we present an online unsupervised deep learning approach to detect anomalous network activity from system logs in real time. Our models decompose anomaly scores into the contributions of individual user behavior features for increased interpretability to aid analysts reviewing potential cases of insider threat. Using the CERT Insider Threat Dataset v6.2 and threat detection recall as our performance metric, our novel deep and recurrent neural network models outperform Principal Component Analysis, Support Vector Machine and Isolation Forest based anomaly detection baselines. For our best model, the events labeled as insider threat activity in our dataset had an average anomaly score in the 95.53 percentile, demonstrating our approach’s potential to greatly reduce analyst workloads.

Tuor, Aaron R.↗

Decommissioning the physics laboratory, building 777-10A, at the Savannah River Site (SRS)

SRS recently completed a four year mission to decommission {approx}250 excess facilities. As part of that effort, SRS decommissioned a 48,000 ft{sup 2} laboratory that housed four low-power test reactors, formerly used by SRS to determine reactor physics. This paper describes and reviews the decommissioning, with a focus on component segmentation and handling (i.e. hazardous material removal, demolition, and waste handling). The paper is intended to be a resource for engineers, planners, and project managers, who face similar decommissioning challenges. Building 777-10A, located at the south end of SRS's A/M-Area, was built in 1953 and had a gross area of {approx}48,000 ft{sup 2}. Building 777-10A had two main areas: a west wing, which housed four experimental reactors and associated equipment; and an east wing, which housed laboratories, and shops, offices. The reactors were located in two separate areas: one area housed the Process Development Pile (PDP) reactor and the Lattice Test Reactor (LTR), while the second area housed the Standard Pile (SP) and the Sub-critical Experiment (SE) reactors. The west wing had five levels: three below and three above grade (floor elevations of -37', -28', -15', 0', +13'/+16' and +27' (roof elevation of +62')), while the east wing had two levels: one below and one above grade (floor elevations of -15' and 0' (roof elevation of +16')). Below-grade exterior walls were constructed of reinforced concrete, {approx}1' thick. In general, above-grade exterior walls were steel frames covered by insulation and corrugated, asbestos-cement board. The two interior walls around the PDP/LTR were reinforced concrete {approx}5' thick and {approx}30' high, while the SP/SE reactors resided in a reinforced, concrete cell with 3.5'-6' thick walls/roof. All other interior walls were constructed of metal studs covered with either asbestos-cement or gypsum board. In general, the floors were constructed of reinforced concrete on cast-in-place concrete beams below-grade and concrete on metal beams above-grade. The roofs were flat concrete slabs on metal beams. Building 777-10A was an important SRS research and development location. The reactors helped determine safe operational limits and loading patterns for fuel used in the SRS production reactors, and supported various low power reactor physics studies. All four reactors were shut down and de-inventoried in the 1970's. The building was DD and R 2007, Chattanooga, Tennessee, September 16-19, 2007 169 subsequently used by various SRS organizations for office space, audio/visual studio, and computer network hub. SRS successfully decommissioned Building 777-10A over a thirty month period at a cost of {approx}$14 M ({approx}$290/ft{sup 2}). The decommissioning was a complex and difficult effort due to the building's radiological contamination, height, extensive basement, and thick concrete walls. Extensive planning and extensive hazard analysis (e.g. of structural loads/modifications leading to unplanned collapse) ensured the decommissioning was completed safely and without incident. The decommissioning met contract standards for residual contamination and physical/chemical hazards, and was the last in a series of decommissioning projects that prepared the lower A/M-Area for SRS's environmental restoration program.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Best practices mobile application for D and D KM-IT - 15712

The Deactivation and Decommissioning Knowledge Management Information Tool (D and D KM-IT) is the repository of useful information built specifically for D and D community. The knowledge provided by the D and D KM-IT is the experience and expertise gained by people within the field and fellow employees of the Department of Energy (DOE). This information has been the result of gathering, analyzing, storing, and sharing knowledge to help reduce time spent on discovering previously found data. In order to help in organizing all the data from all the professional contributors, the D and D KM-IT has been divided into 12 different modules. Each individual module pertains to a different area of information for users, such as vendor and technology information, collaboration tools, and even a hotline that contains problems posted by the users. Purpose: - Design and develop the module for the Best Practices on D and D KM-IT mobile platform deployed at https://m.dndkm.org; - The Best Practices mobile module should integrate with existing modules seamlessly; - Use the D and D KM-IT existing database to consume information for the mobile system. jQuery Mobile framework was used to develop Best Practices module for mobile devices. jQuery Mobile framework provides support for responsive web design (RWD). RWD allows adaptation of the layout and user interface of web sites and web apps to different device resolutions, screen densities, and views. UI design was implemented with three fundamental elements: CSS media queries, An adaptable grid, Flexible images and media. Advantages/benefits: - A responsive web design adapts to any-sized screen / resolution and orientation view; - ASP.NET compatible which allows the D and D KM-IT database to support all platforms accessing the system; - Navigating pages is more bandwidth efficient by refreshing data in small section with Ajax instead of reloading the entire page; - Multiple parts of a page can perform different tasks resulting in improved user's experience. Results: - Mobile system is accessible on compatible devices running Android, Apple, Blackberry, Palm, and Windows devices; - Best Practices searches through Characterization, Decontamination, Dismantlement, Material Disposition, and Worker Safety; - Keyword search is enabled for the best practices that are available on both the desktop and mobile systems - Documents are formatted into a PDF file format to ensure compatibility on any device. Conclusion: - Fully integrated with the desktop Best Practices module and displays documents that are uploaded on web platform; - Fast-loading with D and D KM-IT mobile system and downloadable documents; - Responsive design which gives the same user experience when either viewed on smart phone or tablet.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Vidi: a venti to go

Vidi is a Venti proxy that allows certain clients to work when there is no connection to the Venti server. Vidi can be used on computers, such as laptops, to create archives of the file system even when disconnected, and later to transfer the archives to the Venti server. This paper describes an archival configuration used by the author as well the design and implementation of a proxy that allows it to work in a disconnected state.

97 MATHEMATICS AND COMPUTING↗

The Application of Formal Software Engineering Methods to the Unattended and Remote Monitoring Software Suite at Los Alamos National Laboratory

The Unattended and Remote Monitoring (UNARM) system is a collection of specialized hardware and software used by the International Atomic Energy Agency (IAEA) to institute nuclear safeguards at many nuclear facilities around the world. The hardware consists of detectors, instruments, and networked computers for acquiring various forms of data, including but not limited to radiation data, global position coordinates, camera images, isotopic data, and operator declarations. The software provides two primary functions: the secure and reliable collection of this data from the instruments and the ability to perform an integrated review and analysis of the disparate data sources. Several years ago, the team responsible for maintaining the software portion of the UNARM system began the process of formalizing its operations. These formal operations include a configuration management system, a change control board, an issue tracking system, and extensive formal testing, for both functionality and reliability. Functionality is tested with formal test cases chosen to fully represent the data types and methods of analysis that will be commonly encountered. Reliability is tested with iterative, concurrent testing where up to five analyses are executed simultaneously for thousands of cycles. Iterative concurrent testing helps ensure that there are no resource conflicts or leaks when multiple system components are in use simultaneously. The goal of this work is to provide a high quality, reliable product, commensurate with the criticality of the application. Testing results will be presented that demonstrate that this goal has been achieved and the impact of the introduction of a formal software engineering framework to the UNARM product will be presented.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Hyperswitch Network For Hypercube Computer

Data-driven dynamic switching enables high speed data transfer. Proposed hyperswitch network based on mixed static and dynamic topologies. Routing header modified in response to congestion or faults encountered as path established. Static topology meets requirement if nodes have switching elements that perform necessary routing header revisions dynamically. Hypercube topology now being implemented with switching element in each computer node aimed at designing very-richly-interconnected multicomputer system. Interconnection network connects great number of small computer nodes, using fixed hypercube topology, characterized by point-to-point links between nodes.

Chow, Edward↗

MEDUSA - An overset grid flow solver for network-based parallel computer systems

Continuing improvement in processing speed has made it feasible to solve the Reynolds-Averaged Navier-Stokes equations for simple three-dimensional flows on advanced workstations. Combining multiple workstations into a network-based heterogeneous parallel computer allows the application of programming principles learned on MIMD (Multiple Instruction Multiple Data) distributed memory parallel computers to the solution of larger problems. An overset-grid flow solution code has been developed which uses a cluster of workstations as a network-based parallel computer. Inter-process communication is provided by the Parallel Virtual Machine (PVM) software. Solution speed equivalent to one-third of a Cray-YMP processor has been achieved from a cluster of nine commonly used engineering workstation processors. Load imbalance and communication overhead are the principal impediments to parallel efficiency in this application.

Smith, Merritt H.↗

ChemComp: Compiling and Computing with Chemical Reaction Networks

The exponential growth in computing demands driven by scientific computing, data analytics, and artificial intelligence is pushing conventional CMOS-based high-performance computing systems to their physical and energy efficiency limits. As we approach the era of post-exascale computing, disruptive approaches are necessary to overcome these barriers and achieve substantial gains in energy efficiency. Analog and hybrid digital-analog computing systems have emerged as promising alternatives, offering the potential for orders-of-magnitude improvements in efficiency. Among these, biochemical computing stands out as a novel paradigm capable of leveraging the natural efficiency of chemical reactions, which have shown promise in solving optimization problems by converging to steady states. By scaling up reaction networks or reaction vessel sizes, biochemical systems present an opportunity to meet the high-performance demands of modern computing tasks. Despite their promise, significant theoretical and practical challenges remain, particularly in formulating and mapping computational problems to chemical reaction networks (CRNs) and designing viable biochemical computing devices. This paper addresses these challenges by introducing new ideas to ChemComp, a compilation and emulation framework for chemical computation. This work describes the mechanisms through which solutions to ordinary differential equations (ODEs) that can be represented as CRN systems can be achieved. Furthermore, we explain the design principles of an ODE dialect implemented as a multi-level intermediate representation (MLIR) compiler extension that will be coupled with existing infrastructure. We demonstrate the potential of our framework through a case study emulating a simplified chemical reservoir computing device. This work establishes foundational tools and methodologies necessary to harness the computational power of chemistry, paving the way for the development of energy-efficient, high-performance computing systems tailored to contemporary and future computational needs.

Bohm Agostini, Nicolas↗

I/O routing in a multidimensional torus network

A method, system and computer program product are disclosed for routing data packet in a computing system comprising a multidimensional torus compute node network including a multitude of compute nodes, and an I/O node network including a plurality of I/O nodes. In one embodiment, the method comprises assigning to each of the data packets a destination address identifying one of the compute nodes; providing each of the data packets with a toio value; routing the data packets through the compute node network to the destination addresses of the data packets; and when each of the data packets reaches the destination address assigned to said each data packet, routing said each data packet to one of the I/O nodes if the toio value of said each data packet is a specified value. In one embodiment, each of the data packets is also provided with an ioreturn value used to route the data packets through the compute node network.

97 MATHEMATICS AND COMPUTING↗

I/O routing in a multidimensional torus network

A method, system and computer program product are disclosed for routing data packet in a computing system comprising a multidimensional torus compute node network including a multitude of compute nodes, and an I/O node network including a plurality of I/O nodes. In one embodiment, the method comprises assigning to each of the data packets a destination address identifying one of the compute nodes; providing each of the data packets with a toio value; routing the data packets through the compute node network to the destination addresses of the data packets; and when each of the data packets reaches the destination address assigned to said each data packet, routing said each data packet to one of the I/O nodes if the toio value of said each data packet is a specified value. In one embodiment, each of the data packets is also provided with an ioreturn value used to route the data packets through the compute node network.

97 MATHEMATICS AND COMPUTING↗

I/O routing in a multidimensional torus network

A method, system and computer program product are disclosed for routing data packet in a computing system comprising a multidimensional torus compute node network including a multitude of compute nodes, and an I/O node network including a plurality of I/O nodes. In one embodiment, the method comprises assigning to each of the data packets a destination address identifying one of the compute nodes; providing each of the data packets with a toio value; routing the data packets through the compute node network to the destination addresses of the data packets; and when each of the data packets reaches the destination address assigned to said each data packet, routing said each data packet to one of the I/O nodes if the toio value of said each data packet is a specified value. In one embodiment, each of the data packets is also provided with an ioreturn value used to route the data packets through the compute node network.

97 MATHEMATICS AND COMPUTING↗

I/O routing in a multidimensional torus network

A method, system and computer program product are disclosed for routing data packet in a computing system comprising a multidimensional torus compute node network including a multitude of compute nodes, and an I/O node network including a plurality of I/O nodes. In one embodiment, the method comprises assigning to each of the data packets a destination address identifying one of the compute nodes; providing each of the data packets with a toio value; routing the data packets through the compute node network to the destination addresses of the data packets; and when each of the data packets reaches the destination address assigned to said each data packet, routing said each data packet to one of the I/O nodes if the toio value of said each data packet is a specified value. In one embodiment, each of the data packets is also provided with an ioreturn value used to route the data packets through the compute node network.

97 MATHEMATICS AND COMPUTING↗

Optimizing High-Throughput Inference on Graph Neural Networks at Shared Computing Facilities with the NVIDIA Triton Inference Server

Abstract With machine learning applications now spanning a variety of computational tasks, multi-user shared computing facilities are devoting a rapidly increasing proportion of their resources to such algorithms. Graph neural networks (GNNs), for example, have provided astounding improvements in extracting complex signatures from data and are now widely used in a variety of applications, such as particle jet classification in high energy physics (HEP). However, GNNs also come with an enormous computational penalty that requires the use of GPUs to maintain reasonable throughput. At shared computing facilities, such as those used by physicists at Fermi National Accelerator Laboratory (Fermilab), methodical resource allocation and high throughput at the many-user scale are key to ensuring that resources are being used as efficiently as possible. These facilities, however, primarily provide CPU-only nodes, which proves detrimental to time-to-insight and computational throughput for workflows that include machine learning inference. In this work, we describe how a shared computing facility can use the NVIDIA Triton Inference Server to optimize its resource allocation and computing structure, recovering high throughput while scaling out to multiple users by massively parallelizing their machine learning inference. To demonstrate the effectiveness of this system in a realistic multi-user environment, we use the Fermilab Elastic Analysis Facility augmented with the Triton Inference Server to provide scalable and high-throughput access to a HEP-specific GNN and report on the outcome.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Dynamic defense and network randomization for computer systems

The various technologies presented herein relate to determining a network attack is taking place, and further to adjust one or more network parameters such that the network becomes dynamically configured. A plurality of machine learning algorithms are configured to recognize an active attack pattern. Notification of the attack can be generated, and knowledge gained from the detected attack pattern can be utilized to improve the knowledge of the algorithms to detect a subsequent attack vector(s). Further, network settings and application communications can be dynamically randomized, wherein artificial diversity converts control systems into moving targets that help mitigate the early reconnaissance stages of an attack. An attack(s) based upon a known static address(es) of a critical infrastructure network device(s) can be mitigated by the dynamic randomization. Network parameters that can be randomized include IP addresses, application port numbers, paths data packets navigate through the network, application randomization, etc.

97 MATHEMATICS AND COMPUTING↗

I/O routing in a multidimensional torus network

A method, system and computer program product are disclosed for routing data packet in a computing system comprising a multidimensional torus compute node network including a multitude of compute nodes, and an I/O node network including a plurality of I/O nodes. In one embodiment, the method comprises assigning to each of the data packets a destination address identifying one of the compute nodes; providing each of the data packets with a toio value; routing the data packets through the compute node network to the destination addresses of the data packets; and when each of the data packets reaches the destination address assigned to said each data packet, routing said each data packet to one of the I/O nodes if the toio value of said each data packet is a specified value. In one embodiment, each of the data packets is also provided with an ioreturn value used to route the data packets through the compute node network.

Chen, Dong↗

Using high-performance networks to enable computational aerosciences applications

One component of the U.S. Federal High Performance Computing and Communications Program (HPCCP) is the establishment of a gigabit network to provide a communications infrastructure for researchers across the nation. This gigabit network will provide new services and capabilities, in addition to increased bandwidth, to enable future applications. An understanding of these applications is necessary to guide the development of the gigabit network and other high-performance networks of the future. In this paper we focus on computational aerosciences applications run remotely using the Numerical Aerodynamic Simulation (NAS) facility located at NASA Ames Research Center. We characterize these applications in terms of network-related parameters and relate user experiences that reveal limitations imposed by the current wide-area networking infrastructure. Then we investigate how the development of a nationwide gigabit network would enable users of the NAS facility to work in new, more productive ways.

Johnson, Marjory J.↗

Network Anomaly Detection in Distributed Edge Computing Infrastructure

As networks continue to grow in complexity and scale, detecting anomalies has become increasingly challenging, particularly in diverse and geographically dispersed environments. Traditional approaches often struggle with managing the computational burden associated with analyzing large-scale network traffic to identify anomalies. This paper introduces a distributed edge computing framework that integrates federated learning with Apache Spark and Kubernetes to address these challenges. We hypothesize that our approach, which enables collaborative model training across distributed nodes, significantly enhances the detection accuracy of network anomalies across different network types. We show that by leveraging distributed computing and containerization technologies, our framework not only improves scalability and fault tolerance but also achieves superior detection performance compared to state-of-the-art methods. Extensive experiments on the UNSW-NB15 and ROAD datasets validate the effectiveness of our approach, demonstrating statistically significant improvements in detection accuracy and training efficiency over baseline models, as confirmed by MannWhitney U and Kolmogorov-Smirnov tests (p<0.05).

Marfo, William [University of Texas at El Paso,Dep↗

Calculating a checksum with inactive networking components in a computing system

Calculating a checksum utilizing inactive networking components in a computing system, including: identifying, by a checksum distribution manager, an inactive networking component, wherein the inactive networking component includes a checksum calculation engine for computing a checksum; sending, to the inactive networking component by the checksum distribution manager, metadata describing a block of data to be transmitted by an active networking component; calculating, by the inactive networking component, a checksum for the block of data; transmitting, to the checksum distribution manager from the inactive networking component, the checksum for the block of data; and sending, by the active networking component, a data communications message that includes the block of data and the checksum for the block of data.

97 MATHEMATICS AND COMPUTING↗