Search NASA⌕ Search

SEARCH · Search NASA

Results for “Network Performance”

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 901 records · Page 50

DTN Applications

Explore the source record for details and available documents.

High data-rate Delay Tolerant Networking↗

Livermore Big Artificial Neural Network Toolkit

LBANN is a toolkit that is designed to train artificial neural networks efficiently on high performance computing architectures. It is optimized to take advantages of key High Performance Computing features to accelerate neural network training. Specifically it is optimized for low-latency, high bandwidth interconnects, node-local NVRAM, node-local GPU accelerators, and high bandwidth parallel file systems. It is built on top of the open source Elemental distributed-memory dense and spars-direct linear algebra and optimization library that is released under the BSD license. The algorithms contained within LBANN are drawn from the academic literature and implemented to work within a distributed-memory framework.

Essen, Brian Van↗

GLM Proxy Data Generation: Methods for Stroke/Pulse Level Inter-Comparison of Ground-Based Lightning Reference Networks

In order to produce useful proxy data for the GOES-R Geostationary Lightning Mapper (GLM) in regions not covered by VLF lightning mapping systems, we intend to employ data produced by ground-based (regional or global) VLF/LF lightning detection networks. Before using these data in GLM Risk Reduction tasks, it is necessary to have a quantitative understanding of the performance of these networks, in terms of CG flash/stroke DE, cloud flash/pulse DE, location accuracy, and CLD/CG classification error. This information is being obtained through inter-comparison with LMAs and well-quantified VLF/LF lightning networks. One of our approaches is to compare "bulk" counting statistics on the spatial scale of convective cells, in order to both quantify relative performance and observe variations in cell-based temporal trends provided by each network. In addition, we are using microsecond-level stroke/pulse time correlation to facilitate detailed inter-comparisons at a more-fundamental level. The current development status of our ground-based inter-comparison and evaluation tools will be presented, and performance metrics will be discussed through a comparison of Vaisala s Global Lightning Dataset (GLD360) with the NLDN at locations within and outside the U.S.

Cummins, Kenneth L.↗

An Evaluation of the Effect of Network Cost Optimization for Leadership Class Supercomputers

Dragonfly-based networks are an extensively deployed network topology in large-scale high-performance computing due to their cost-effectiveness and efficiency. The US will soon have three Exascale supercomputers for leadership class workloads deployed using dragonfly networks. Compared to indirect networks of similar scale, the dragonfly network has considerably reduced cable lengths, cable counts, and switch counts, resulting in significant network cost savings for a given system size, however, these cost reductions result in reduced global minimal paths and more challenging routing. Additionally, large scale dragonfly networks often require a taper at the global link level, resulting in less bisection bandwidth than is achievable in other traditional non-blocking topologies of equivalent scale. While dragonfly networks have been extensively studied, they have yet to be fully evaluated in an extreme scale (i.e., exascale) system that targets capability workloads. In this paper, we present the results of the first large scale evaluation of a dragonfly network on an exascale system (Frontier) and compare its behavior to a similar scale fat-tree network on a previous generation TOP500 system (Summit). This evaluation aims to determine the effect of network cost optimizations by measuring a tapered topology’s impact on capability workloads. Our evaluation is based on a collection of synthetic microbenchmarks, mini-apps, and full scale applications. It compares the scaling efficiencies of each benchmark between the dragonfly-based Frontier and the fat-tree-based Summit systems. Our results show that a dragonfly network is $\sim \mathbf{3 0 \%}$ more cost efficient than a fat-tree topology, which amortizes to $\sim 3 \%$ of an exascale system cost. Furthermore, while tapered dragonfly networks impose significant tradeoffs, the impacts are not as broad as initially thought and are mostly seen in applications with global communication patterns, particularly all-to-all (e.g., FFT-based algorithms), but also local communication patterns (e.g., nearest-neighbor algorithms) that are sensitive to network performance variability.

Khan, Awais↗

Benchmarking a Tunable Quantum Neural Network on Trapped-Ion and Superconducting Hardware

We implement a quantum generalization of a neural network on trapped-ion and IBM superconducting quantum computers to classify MNIST images, a common benchmark in computer vision. The network feedforward involves qubit rotations whose angles depend on the results of measurements in the previous layer. The network is trained via simulation, but inference is performed experimentally on quantum hardware. The classical-to-quantum correspondence is controlled by an interpolation parameter, $a$, which is zero in the classical limit. Increasing $a$ introduces quantum uncertainty into the measurements, which is shown to improve network performance at moderate values of the interpolation parameter. We then focus on particular images that fail to be classified by a classical neural network but are detected correctly in the quantum network. For such borderline cases, we observe strong deviations from the simulated behavior. We attribute this to physical noise, which causes the output to fluctuate between nearby minima of the classification energy landscape. Such strong sensitivity to physical noise is absent for clear images. We further benchmark physical noise by inserting additional single-qubit and two-qubit gate pairs into the neural network circuits. Our work provides a springboard toward more complex quantum neural networks on current devices: while the approach is rooted in standard classical machine learning, scaling up such networks may prove classically non-simulable and could offer a route to near-term quantum advantage.

FOS: Physical sciences↗

A neural-network based estimator to search for primordial non-Gaussianity in Planck CMB maps

We present an upgraded combined estimator, based on Minkowski Functionals and Neural Networks, with excellent performance in detecting primordial non-Gaussianity in simulated maps that also contain a weighted mixture of Galactic contaminations, besides real pixel's noise from Planck cosmic microwave background radiation data. We rigorously test the efficiency of our estimator considering several plausible scenarios for residual non-Gaussianities in the foreground-cleaned Planck maps, with the intuition to optimize the training procedure of the Neural Network to discriminate between contaminations with primordial and secondary non-Gaussian signatures. We look for constraints of primordial local non-Gaussianity at large angular scales in the foreground-cleaned Planck maps. For the SMICA map we found f{sub NL} = 33 ± 23, at 1σ confidence level, in excellent agreement with the WMAP-9yr and Planck results. In addition, for the other three Planck maps we obtain similar constraints with values in the interval f{sub NL} element of [33, 41], concomitant with the fact that these maps manifest distinct features in reported analyses, like having different pixel's noise intensities.

79 ASTRONOMY AND ASTROPHYSICS↗

Doing the impossible: Why neural networks can be trained at all

As deep neural networks grow in size, from thousands to millions to billions of weights, the performance of those networks becomes limited by our ability to accurately train them. A common naive question arises: if we have a system with billions of degrees of freedom, don't we also need billions of samples to train it? Of course, the success of deep learning indicates that reliable models can be learned with reasonable amounts of data. Similar questions arise in protein folding, spin glasses and biological neural networks. With effectively infinite potential folding/spin/wiring configurations, how does the system find the precise arrangement that leads to useful and robust results? Simple sampling of the possible configurations until an optimal one is reached is not a viable option even if one waited for the age of the universe. On the contrary, there appears to be a mechanism in the above phenomena that forces them to achieve configurations that live on a low-dimensional manifold, avoiding the curse of dimensionality. In the current work we use the concept of mutual information between successive layers of a deep neural network to elucidate this mechanism and suggest possible ways of exploiting it to accelerate training.

deep learning, training, curse of dimensionality, ↗

Applying a Space-Based Security Recovery Scheme for Critical Homeland Security Cyberinfrastructure Utilizing the NASA Tracking and Data Relay (TDRS) Based Space Network

Protection of the national infrastructure is a high priority for cybersecurity of the homeland. Critical infrastructure such as the national power grid, commercial financial networks, and communications networks have been successfully invaded and re-invaded from foreign and domestic attackers. The ability to re-establish authentication and confidentiality of the network participants via secure channels that have not been compromised would be an important countermeasure to compromise of our critical network infrastructure. This paper describes a concept of operations by which the NASA Tracking and Data Relay (TDRS) constellation of spacecraft in conjunction with the White Sands Complex (WSC) Ground Station host a security recovery system for re-establishing secure network communications in the event of a national or regional cyberattack. Users would perform security and network restoral functions via a Broadcast Satellite Service (BSS) from the TDRS constellation. The BSS enrollment only requires that each network location have a receive antenna and satellite receiver. This would be no more complex than setting up a DIRECTTV-like receiver at each network location with separate network connectivity. A GEO BSS would allow a mass re-enrollment of network nodes (up to nationwide) simultaneously depending upon downlink characteristics. This paper details the spectrum requirements, link budget, notional assets and communications requirements for the scheme. It describes the architecture of such a system and the manner in which it leverages off of the existing secure infrastructure which is already in place and managed by the NASAGSFC Space Network Project.

Cybersecurity↗

SuperCDMS Event Reconstruction Using Convolutional Neural Networks [Thesis]

The SuperCDMS experiment uses cryogenic silicon and germanium detectors to search for dark matter candidates such as WIMPs (Weakly Interacting Massive Particles) streaming through the Earth. Collisions in the silicon and germanium crystals are expected to produce phonons whose thermal signatures can be measured. This thesis first describes the integration of a new Signal Distribution Unit (SDU) to the SuperCDMS data acquisition system, which allows for synchronization of multiple detectors and electronic/mechanical noise characterization via accelerometer, antenna, and AC phase measurements. From SuperCDMS detector data it is necessary to reconstruct the energies of the particle events. This thesis explores the use of Convolutional Neural Networks (CNNs) to perform this reconstruction and finds that, although they perform well, changing the noise model breaks the model and requires the neural network to be retrained. In order to mitigate this issue, a new CNN model is proposed which includes the noise Power Spectral Density (PSD) of the data as an additional input to the CNN. While it proves to be effective as a denoising algorithm, it still fails for data with a different noise model. However, including data from multiple PSDs in the neural network training sample allows it to handle data with different types of noise while still maintaining the quality of the reconstruction. Nevertheless, neural networks trained even on multiple PSDs do not robustly handle data taken with PSDs dissimilar to those in the training sample, suggesting that CNNs may need to be retrained whenever the noise environment changes in a significant way.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Reconstructing the Hopfield network as an inverse Ising problem

We test four fast mean-field-type algorithms on Hopfield networks as an inverse Ising problem. The equilibrium behavior of Hopfield networks is simulated through Glauber dynamics. In the low-temperature regime, the simulated annealing technique is adopted. Although performances of these network reconstruction algorithms on the simulated network of spiking neurons are extensively studied recently, the analysis of Hopfield networks is lacking so far. For the Hopfield network, we found that, in the retrieval phase favored when the network wants to memory one of stored patterns, all the reconstruction algorithms fail to extract interactions within a desired accuracy, and the same failure occurs in the spin-glass phase where spurious minima show up, while in the paramagnetic phase, albeit unfavored during the retrieval dynamics, the algorithms work well to reconstruct the network itself. This implies that, as an inverse problem, the paramagnetic phase is conversely useful for reconstructing the network while the retrieval phase loses all the information about interactions in the network except for the case where only one pattern is stored. The performances of algorithms are studied with respect to the system size, memory load, and temperature; sample-to-sample fluctuations are also considered.

36 MATERIALS SCIENCE↗

Method and system for pattern analysis using a coarse-coded neural network

A method and system for performing pattern analysis with a neural network coarse-coding a pattern to be analyzed so as to form a plurality of sub-patterns collectively defined by data. Each of the sub-patterns comprises sets of pattern data. The neural network includes a plurality fields, each field being associated with one of the sub-patterns so as to receive the sub-pattern data therefrom. Training and testing by the neural network then proceeds in the usual way, with one modification: the transfer function thresholds the value obtained from summing the weighted products of each field over all sub-patterns associated with each pattern being analyzed by the system.

Spirkovska, Liljana↗

Software defined network inference with evolutionary optimal observation matrices

A key requirement for network management is the accurate and reliable monitoring of relevant network characteristics. In today's large-scale networks, this is a challenging task due to the scarcity of network measurement resources and the hard constraints that this imposes. Here, this paper proposes a new framework, called SNIPER, which leverages the flexibility provided by Software-Defined Networking (SDN) to design the optimal observation or measurement matrix that can lead to the best achievable estimation accuracy using Matrix Completion (MC) techniques. To cope with the complexity of designing large-scale optimal observation matrices, we use the Evolutionary Optimization Algorithms (EOA) which directly target the ultimate estimation accuracy as the optimization objective function. We evaluate the performance of SNIPER using both synthetic and real network measurement traces from different network topologies and by considering two main applications for per-flow size and delay estimations. Our results show that SNIPER can be applied to a variety of network performance measurements under hard resource constraints. For example, by measuring only 8.8% of all per-flow path delays in Harvard network [1], congested paths can be detected with probability of 0.94. Finally, to demonstrate the feasibility of our framework, we also have implemented a prototype of SNIPER in Mininet.

97 MATHEMATICS AND COMPUTING↗

Toward higher-radix switches with co-packaged optics for improved network locality in data center and HPC networks [Invited]

In this work, we study the network locality improvements that can be achieved by using co-packaged optics in data center and high-performance computing (HPC) networks. The increased escape bandwidth offered by co-packaged optics can enable switches with speeds of 51.2 Tb/s and beyond. From a network architecture perspective, the key advantages of introducing co-packaged optics at the switch points include the implementation of large-scale topologies of >12,000 end points with 4× higher bisection bandwidth and the reduction of the required number of switches by >40% compared with state-of-the-art approaches. From a network operation perspective, improved network locality and faster operation can be achieved since the higher-radix switches can mitigate the impact of network contention. Placing applications under fewer leaf switches reduces the number of packets that cross the spine switches in a leaf-spine topology. The proposed scheme is evaluated via discrete-event simulations: we initially evaluate the network locality properties of the system by using virtual-machine traces from a production data center, and we subsequently quantify the performance improvements by simulating an all-to-all pattern for a variety of message sizes over a number of nodes. The results suggest that co-packaged optics form a promising solution for keeping up with bandwidth scaling in future networks. The virtual-machine analysis shows that large-scale applications can be placed under up to 50% fewer first-level switches, while the network analysis shows speedups of up to 7.1, which translates to execution time reductions of up to 26% and 42.7% for applications with communication ratios of 0.3 and 0.5, respectively.

99 GENERAL AND MISCELLANEOUS↗

A collaborative network middleware project by Lambda Station, TeraPaths, and Phoebus

The TeraPaths, Lambda Station, and Phoebus projects, funded by the US Department of Energy, have successfully developed network middleware services that establish on-demand and manage true end-to-end, Quality-of-Service (QoS) aware, virtual network paths across multiple administrative network domains, select network paths and gracefully reroute traffic over these dynamic paths, and streamline traffic between packet and circuit networks using transparent gateways. These services improve network QoS and performance for applications, playing a critical role in the effective use of emerging dynamic circuit network services. They provide interfaces to applications, such as dCache SRM, translate network service requests into network device configurations, and coordinate with each other to setup up end-to-end network paths. The End Site Control Plane Subsystem (ESCPS) builds upon the success of the three projects by combining their individual capabilities into the next generation of network middleware. ESCPS addresses challenges such as cross-domain control plane signalling and interoperability, authentication and authorization in a Grid environment, topology discovery, and dynamic status tracking. The new network middleware will take full advantage of the perfSONAR monitoring infrastructure and the Inter-Domain Control plane efforts and will be deployed and fully vetted in the Large Hadron Collider data movement environment.

99 GENERAL AND MISCELLANEOUS↗

The Global Oscillation Network Group site survey. 1: Data collection and analysis methods

The Global Oscillation Network Group (GONG) Project is planning to place a set of instruments around the world to observe solar oscillations as continuously as possible for at least three years. The Project has now chosen the sites that will comprise the network. This paper describes the methods of data collection and analysis that were used to make this decision. Solar irradiance data were collected with a one-minute cadence at fifteen sites around the world and analyzed to produce statistics of cloud cover, atmospheric extinction, and transparency power spectra at the individual sites. Nearly 200 reasonable six-site networks were assembled from the individual stations, and a set of statistical measures of the performance of the networks was analyzed using a principal component analysis. An accompanying paper presents the results of the survey.

Hill, Frank↗

Scalable Lunar Surface Networks and Adaptive Orbit Access

Teranovi Technologies, Inc., has developed innovative network architecture, protocols, and algorithms for both lunar surface and orbit access networks. A key component of the overall architecture is a medium access control (MAC) protocol that includes a novel mechanism of overlaying time division multiple access (TDMA) and carrier sense multiple access with collision avoidance (CSMA/CA), ensuring scalable throughput and quality of service. The new MAC protocol is compatible with legacy Institute of Electrical and Electronics Engineers (IEEE) 802.11 networks. Advanced features include efficiency power management, adaptive channel width adjustment, and error control capability. A hybrid routing protocol combines the advantages of ad hoc on-demand distance vector (AODV) routing and disruption/delay-tolerant network (DTN) routing. Performance is significantly better than AODV or DTN and will be particularly effective for wireless networks with intermittent links, such as lunar and planetary surface networks and orbit access networks.

Wang, Xudong↗

Validation of Image-Based Neural Network Controllersthrough Adaptive Stress Testing

Neural networks have become state-of-the-art for computer vision problems because of their ability to efficiently model complex functions from large amounts of data. While neural networks can be shown to perform well empirically fora variety of tasks, their performance is difficult to guarantee.Neural network verification tools have been developed that can certify robustness with respect to a given input image; however,for neural network systems used in closed-loop controllers,robustness with respect to individual images does not address multi-step properties of the neural network controller and itsenvironment. Furthermore, neural network systems interacting in the physical world and using natural images are operating in a black-box environment, making formal verification in-tractable. This work combines the adaptive stress testing (AST)framework with neural network verification tools to search for the most likely sequence of image disturbances that cause the neural network controlled system to reach a failure. Anautonomous aircraft taxi application is presented, and results show that the AST method finds failures with more likely image disturbances than baseline methods. Further analysis of AST results revealed an explainable cause of the failure, giving insight into the problematic scenarios that should be addressed.

Adaptive Stress Testing, Marabou, Deep Neural Netw↗