Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “large networks”

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 163 records · Page 9

Stability of Kuramoto networks subject to large and small fluctuations from heterogeneous and spatially correlated noise

Oscillatory networks subjected to noise are broadly used to model physical and technological systems. Due to their nonlinear coupling, such networks typically have multiple stable and unstable states that a network might visit due to noise. In this article, we focus on the assessment of fluctuations resulting from heterogeneous and spatially correlated noise inputs on Kuramoto model networks. We evaluate the typical, small fluctuations near synchronized states and connect the network variance to the overlap between stable modes of synchronization and the input noise covariance. Going beyond small to large fluctuations, we introduce the indicator mode approximation that projects the dynamics onto a single amplitude dimension. Such an approximation allows for estimating rates of fluctuations to saddle instabilities, resulting in phase slips between connected oscillators. Statistics for both regimes are quantified in terms of effective noise amplitudes that are compared and contrasted for several noise models. In conclusion, bridging the gap between small and large fluctuations, we show that a larger network variance does not necessarily lead to higher rates of large fluctuations.

42 ENGINEERING↗

Crystal Structures of SlyA Protein, a Master Virulence Regulator of Salmonella, in Free and DNA-bound States

SlyA is a master virulence regulator that controls the transcription of numerous genes in Salmonella enterica. We present here crystal structures of SlyA by itself and bound to a high-affinity DNA operator sequence in the slyA gene. SlyA interacts with DNA through direct recognition of a guanine base by Arg-65, as well as interactions between conserved Arg-86 and the minor groove and a large network of non-base-specific contacts with the sugar phosphate backbone. Our structures, together with an unpublished structure of SlyA bound to the small molecule effector salicylate (Protein Data Bank code 3DEU), reveal that, unlike many other MarR family proteins, SlyA dissociates from DNA without large conformational changes when bound to this effector. We propose that SlyA and other MarR global regulators rely more on indirect readout of DNA sequence to exert control over many genes, in contrast to proteins (such as OhrR) that recognize a single operator.

59 BASIC BIOLOGICAL SCIENCES↗

Multi-Commodity Traffic Signal Control and Routing With Connected Vehicles

We report a real-time traffic management policy that integrates traffic signal control and multi-commodity routing of connected vehicles in networks with multiple destinations is developed. The proposed policy is based on a multi-commodity formulation of the store-and-forward model and assumes all vehicles are able to exchange information with the infrastructure. Vehicles share information about their current location and final destination. Based on this information, the strategy determines both optimized signal timings at every intersection and vehicle-specific routing information at every link of the network. The control actions, i.e., signal times and routing information, are updated at every cycle and delivered by a finite horizon optimal control problem cast into a rolling horizon framework. The underlying optimization problem is convex, and thus the method is suitable for real-time operation in large networks. The method is validated via a micro-simulation study in networks with up to twenty intersections and, in all simulations, outperforms a real-time traffic-responsive signal control strategy that is based on a single-commodity store-and-forward model. The scalable computation effort for increasing network sizes and prediction horizon confirms the computational efficiency of the method.

42 ENGINEERING↗

Spaceborne ranging system

A spaceborne laser ranging system is described that could survey a large network of ground reflectors, and provide their relative locations to a precision of + or - 1 cm. This performance is believed realizable for networks covering up to 10 to 6th power square kilometers and from only a few days of observations. This system could be used to monitor crustal movements in many areas of the world and has the potential to provide an almost real-time system for detecting precursory ground motions before large earthquakes.

Smith, D. E.↗

GSplit: Scaling Graph Neural Network Training on Large Graphs via Split-Parallelism

Graph neural networks (GNNs), an emerging class of machine learning models for graphs, have gained popularity for their superior performance in various graph analytical tasks. Mini-batch training is commonly used to train GNNs on large graphs, and data parallelism is the standard approach to scale mini-batch training across multiple GPUs. Data parallel approaches contain redundant work as subgraphs sampled by different GPUs contain significant overlap. To address this issue, we introduce a hybrid parallel mini-batch training paradigm called Split parallelism. Split parallelism avoids redundant work by splitting the sampling, loading, and training of each mini-batch across multiple GPUs. Split parallelism, however, introduces communication overheads that can be more than the savings from removing redundant work. We further present a lightweight partitioning algorithm that probabilistically minimizes these overheads. We implement spllit parllelism in GSplit and show that it outperforms state-of-the-art mini-batch training systems like DGL, Quiver, and P3.

Lim, Seung-Hwan [ORNL] (ORCID:0000000194616866)↗

Ten million and one penguins, or, lessons learned from booting millions of virtual machines on HPC systems.

In this paper we describe Megatux, a set of tools we are developing for rapid provisioning of millions of virtual machines and controlling and monitoring them, as well as what we've learned from booting one million Linux virtual machines on the Thunderbird (4660 nodes) and 550,000 Linux virtual machines on the Hyperion (1024 nodes) clusters. As might be expected, our tools use hierarchical structures. In contrast to existing HPC systems, our tools do not require perfect hardware; that all systems be booted at the same time; and static configuration files that define the role of each node. While we believe these tools will be useful for future HPC systems, we are using them today to construct botnets. Botnets have been in the news recently, as discoveries of their scale (millions of infected machines for even a single botnet) and their reach (global) and their impact on organizations (devastating in financial costs and time lost to recovery) have become more apparent. A distinguishing feature of botnets is their emergent behavior: fairly simple operational rule sets can result in behavior that cannot be predicted. In general, there is no reducible understanding of how a large network will behave ahead of 'running it'. 'Running it' means observing the actual network in operation or simulating/emulating it. Unfortunately, this behavior is only seen at scale, i.e. when at minimum 10s of thousands of machines are infected. To add to the problem, botnets typically change at least 11% of the machines they are using in any given week, and this changing population is an integral part of their behavior. The use of virtual machines to assist in the forensics of malware is not new to the cyber security world. Reverse engineering techniques often use virtual machines in combination with code debuggers. Nevertheless, this task largely remains a manual process to get past code obfuscation and is inherently slow. As part of our cyber security work at Sandia National Laboratories, we are striving to understand the global network behavior of botnets. We are planning to take existing botnets, as found in the wild, and run them on HPC systems. We have turned to HPC systems to support the creation and operation of millions of Linux virtual machines as a means of observing the interaction of the botnet and other noninfected hosts. We started out using traditional HPC tools, but these tools are designed for a much smaller scale, typically topping out at one to ten thousand machines. HPC programming libraries and tools also assume complete connectivity between all nodes, with the attendant configuration files and data structures to match; this assumption holds up very poorly on systems with millions of nodes.

99 GENERAL AND MISCELLANEOUS↗

Scaling and Quality of Modularity Optimization Methods for Graph Clustering

Real-world graphs exhibit structures known as “communities” or “clusters” consisting of a group of vertices with relatively high connectivity between them, as compared to the rest of the vertices in the network. Graph clustering or community detection is a fundamental graph operation used to analyze real-world graphs occurring in the areas of computational biology, cybersecurity, electrical grids, etc. Similar to other graph algorithms, owing to irregular memory accesses and inherently sequential nature, current algorithms for community detection are challenging to parallelize. However, in order to analyze large networks, it is important to develop scalable parallel implementations of graph clustering that are capable of exploiting the architectural features of modern supercomputers. In response to the 2019 Streaming Graph Challenge, we present quality and performance analysis of our distributed- memory community detection using Vite, which is our distributed memory implementation of the popular Louvain method, on the ALCF Theta supercomputer. Clustering methods such as Louvain that rely on modularity maximization are known to suffer from the resolution limit problem, preventing identification of clusters of certain sizes. Hence, we also include quality analysis of our shared-memory implementation of the Fast-tracking Resistance method, in comparison with Louvain on the challenge datasets. Furthermore, we introduce an edge-balanced graph distribution for our distributed memory implementation, that significantly reduces communication, offering up to 80% improvement in the overall execution time. In addition to performance/quality analysis, we also include details on the power/energy consumption, and memory traffic of the distributed-memory clustering implementation using real-world graphs with over a billion edges.

Ghosh, Sayan↗

Cascading Failures in Interdependent Networks with Multiple Supply-Demand Links and Functionality Thresholds

Various social, financial, biological and technological systems can be modeled by interdependent networks. It has been assumed that in order to remain functional, nodes in one network must receive the support from nodes belonging to different networks. So far these models have been limited to the case in which the failure propagates across networks only if the nodes lose all their supply nodes. In this paper we develop a more realistic model for two interdependent networks in which each node has its own supply threshold, i.e., they need the support of a minimum number of supply nodes to remain functional. In addition, we analyze different conditions of internal node failure due to disconnection from nodes within its own network. We show that several local internal failure conditions lead to similar nontrivial results. When there are no internal failures the model is equivalent to a bipartite system, which can be useful to model a financial market. We explore the rich behaviors of these models that include discontinuous and continuous phase transitions. Using the generating functions formalism, we analytically solve all the models in the limit of infinitely large networks and find an excellent agreement with the stochastic simulations.

97 MATHEMATICS AND COMPUTING↗

Standardization of the Definitions of Vertical Resolution and Uncertainty in the NDACC-archived Ozone and Temperature Lidar Measurements

The international Network for the Detection of Atmospheric Composition Change (NDACC) is a global network of high-quality, remote-sensing research stations for observing and understanding the physical and chemical state of the Earth atmosphere. As part of NDACC, over 20 ground-based lidar instruments are dedicated to the long-term monitoring of atmospheric composition and to the validation of space-borne measurements of the atmosphere from environmental satellites such as Aura and ENVISAT. One caveat of large networks such as NDACC is the difficulty to archive measurement and analysis information consistently from one research group (or instrument) to another [1][2][3]. Yet the need for consistent definitions has strengthened as datasets of various origin (e.g., satellite and ground-based) are increasingly used for intercomparisons, validation, and ingested together in global assimilation systems.In the framework of the 2010 Call for Proposals by the International Space Science Institute (ISSI) located in Bern, Switzerland, a Team of lidar experts was created to address existing issues in three critical aspects of the NDACC lidar ozone and temperature data retrievals: signal filtering and the vertical filtering of the retrieved profiles, the quantification and propagation of the uncertainties, and the consistent definition and reporting of filtering and uncertainties in the NDACC- archived products. Additional experts from the satellite and global data standards communities complement the team to help address issues specific to the latter aspect.

Detection of Atmospheric Composition Change (NDACC↗

A Simplified Algorithm for Statistical Investigation of Damage Spreading

On the way to simulating adaptive evolution of complex system describing a living object or human developed project, a fitness should be defined on node states or network external outputs. Feedbacks lead to circular attractors of these states or outputs which make it difficult to define a fitness. The main statistical effects of adaptive condition are the result of small change tendency and to appear, they only need a statistically correct size of damage initiated by evolutionary change of system. This observation allows to cut loops of feedbacks and in effect to obtain a particular statistically correct state instead of a long circular attractor which in the quenched model is expected for chaotic network with feedback. Defining fitness on such states is simple. We calculate only damaged nodes and only once. Such an algorithm is optimal for investigation of damage spreading i.e. statistical connections of structural parameters of initial change with the size of effected damage. It is a reversed-annealed method--function and states (signals) may be randomly substituted but connections are important and are preserved. The small damages important for adaptive evolution are correctly depicted in comparison to Derrida annealed approximation which expects equilibrium levels for large networks. The algorithm indicates these levels correctly. The relevant program in Pascal, which executes the algorithm for a wide range of parameters, can be obtained from the author.

99 GENERAL AND MISCELLANEOUS↗

Scalable and Memory-Efficient Algorithms for Controlling Networked Epidemic Processes Using Multiplicative Weights Update Method

We study the problem of designing scalable algorithms to find effective intervention strategies for controlling stochastic epidemic processes on networks. This is a common problem arising in agent based models for epidemic spread. Previous approaches to this problem focus on either heuristics with no guarantees or approximation algorithms that scale only to networks corresponding to county-sized populations, typically, with less than a million nodes. In particular, the mathematical-programming based approaches need to solve the Linear Program (LP) relaxation of the problem using an LP solver, which restricts the scalability of this approach. In this work, we overcome this restriction by designing an algorithm that adapts the multiplicative weights update (MWU) framework, along with the sample average approximation (SAA) technique, to approximately solve the linear program (LP) relaxation for the problem. To scale this approach further, we provide a memory-efficient algorithm that enables scaling to large networks, corresponding to country-size populations, with over 300 million nodes and 30 billion edges. Furthermore, we show that this approach provides near-optimal solutions to the LP in practice.

Sambaturu, Prathyush↗

Simultaneous global and local clustering in multiplex networks with covariate information

Understanding both global and layer-specific group structures is useful for uncovering complex patterns in networks with multiple interaction types. In this work, we introduce a new model, the hierarchical multiplex stochastic blockmodel, which simultaneously detects communities within individual layers of a multiplex network while inferring a global node clustering across the layers. A stochastic blockmodel is assumed in each layer, with probabilities of layer-level group memberships determined by a node’s global group assignment. Our model uses a Bayesian framework, employing a probit stick-breaking process to construct node-specific mixing proportions over a set of shared Griffiths–Engen–McCloseky distributions. These proportions determine layer-level community assignment, allowing for an unknown and varying number of groups across layers, while incorporating nodal covariate information to inform the global clustering. We propose a scalable variational inference procedure with parallelisable updates for application to large networks. Extensive simulation studies demonstrate our model’s ability to accurately recover both global and layer-level clusters in complicated settings, and applications to real data showcase the model’s effectiveness in uncovering interesting latent network structure.

community detection↗

Multifold Insights for Power System Dynamics From Data Assimilation: Meeting Current Challenges

The world’s electric power grids have evolved over the last 120 years from a single power line to today’s large networks. The evolution will continue at an accelerated rate with extensive smart grid development worldwide. Here for example, the U.S. government has set a goal of reaching 100 percent carbon pollution-free electricity by 2035, while the Department of Energy (DOE) shared a goal to deploy 30 GW and 110 GW of offshore wind by 2030 and 2050, respectively. To meet such ambitious goals, in the years to come, a significant percentage of electricity will come from intermittent renewable sources, electric vehicles (EVs), and be supplied to a vast number of loads that will actively respond to grid conditions and incentive signals. This development is largely driven by environmental and economic factors, such as reducing carbon emissions and saving electricity cost for consumers. These energy resources lead to new uncertain behaviors and dynamics which the grid has never seen and were not considered in its design. Operating such a dynamic grid with sufficient reliability and efficiency is a monumental challenge.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Joint Spectrum Access and Power Control in Air-Air Communications - A Deep Reinforcement Learning Based Approach

This paper considers the dynamic spectrum access and power control problem in a single-hop point-to-point Air-Air Communication Network (AACN). Due to spectrum scarcity, we assume the number of Aircraft-to-Aircraft (A2A) communication links is greater than that of the available channels, such that some communication links need to share the same channel, causing co-channel interference. We formulate the joint channel selection and power control optimization problem to maximize the Weighted Sum Spectral Efficiency (WSSE). A distributed and dynamic deep Q learning-based algorithm is proposed to find the optimal solution. Specifically, we design two different policies that are trained by conducting a trial-and-error scheme. Each communication link can achieve the optimal policy by exploiting the local information from its neighbors, and this distributive approach make it scalable to large networks. Finally, our experimental results demonstrate the effectiveness of the proposed solution in various AACN scenarios.

Zhe Wang↗

Joint Spectrum Access and Power Control in Air-Air Communications - A Deep Reinforcement Learning Based Approach

This paper considers the dynamic spectrum access and power control problem in a single-hop point-to-point Air-Air Communication Network (AACN). Due to spectrum scarcity, we assume the number of Aircraft-to-Aircraft (A2A) communication links is greater than that of the available channels, such that some communication links need to share the same channel, causing co-channel interference. We formulate the joint channel selection and power control optimization problem to maximize the Weighted Sum Spectral Efficiency (WSSE). A distributed and dynamic deep Q learning-based algorithm is proposed to find the optimal solution. Specifically, we design two different policies that are trained by conducting a trial-and-error scheme. Each communication link can achieve the optimal policy by exploiting the local information from its neighbors, and this distributive approach make it scalable to large networks. Finally, our experimental results demonstrate the effectiveness of the proposed solution in various AACN scenarios.

Zhe Wang↗

Students Planting Research Of Use To Space (SPROUTS) Networking and Engagement of Interns at Kennedy Space Center

Beginning in 2016, the Kennedy Space Center (KSC) internship program for life sciences was re-envisioned. Rather than hosting interns in the summer only with numerous innovative small projects, the current model is to offer internship opportunities every semester, so that research is continuous and sustained. Since 2016 we have had 64 internship opportunities for undergraduate and graduate students in space crop production and microgravity simulation. 52 separate individuals have participated in this program, with some participating in more than one term. While interns primarily worked with individual mentors on a main project, most also worked on group projects or developed smaller secondary projects on their own initiative. Additionally, interns participated in a weekly Lunch, Learn, and Discuss seminar series where they presented their research, learned from scientists at KSC, and engaged with members of the space life sciences community. When the face-to-face spring 2020 internship ended prematurely, the weekly seminar series continued as a virtual event, which led to the concept of engaging former interns. Current and former interns were invited to participate in the on-going virtual series, and 46 asked to be included on the mailing list. Our current seminar series involves remote seminars from contacts within and outside of the agency, and engages a large network of current scientists and former interns with weekly participation averaging ~40 people. Speakers include experts from industry, academia, and other space agencies working on topics of interest to KSC researchers. This series has also served as a brainstorming platform for KSC scientists to discuss new ideas and concepts and invite feedback. Remote work and virtual platforms have allowed engagement in a new way that will continue as future interns are integrated into this network of individuals interested in space research. This work is funded by NASA’s Space Biology Program.

Space Crop Production↗

A Novel Multi-Agent Deep Reinforcement Learning-enabled Distributed Power Allocation Scheme for mmWave Cellular Networks

We consider the power allocation problem over shared spectrum for millimeter-Wave (mmWave) cellular downlink. Existing approaches usually find sub-optimal solutions by solving a non-convex optimization which leads to scalability issues due to centralized control. Therefore, distributed and adaptive approaches are desirable. Recently, model-free Deep Reinforcement Learning (DRL) has achieved success in such wireless resource management tasks. By modeling the radio environment as a Markov Decision Process (MDP) with the base stations (BSs) being the agents, power allocation can be automated at the agent level with comparable throughput performance to conventional centralized schemes. The multi-agent setting presents new challenges as the radio environment is impacted by the joint actions of the agents and is no longer stationary from any individual agent’s perspective. Existing literature bypasses this non-stationarity violation by ignoring it which may cause performance degradation. To tackle this issue, we propose a distributed continuous power allocation scheme based on a modified version of multi-agent Deep Deterministic Policy Gradient (MADDPG) that is tailored for the distributed multiple-agent setting. The proposed scheme employs a centralized-training distributed-execution framework where Q-functions are trained over subsets of BSs while each BS determines its transmit power based only on its own local observation. It admits constant per-BS communication and computation complexity and is thus scalable to large networks. Numerical evaluation shows that the proposed scheme adapts well to a wide range of interference conditions and can achieve comparable or better performance than several state-of-the-art non-learning approaches.

99 GENERAL AND MISCELLANEOUS↗

Students Planting Research Of Use To Space (SPROUTS) Networking and Engagement of Interns at Kennedy Space Center

While Kennedy Space Center (KSC) has had summer interns in the life sciences for many years, beginning in 2016, the internship program was re-envisioned. Rather than having many interns in the summer with numerous innovative small projects created primarily for this purpose, the current model is to offer internship opportunities every semester (Spring, Summer, and Fall), so that research is continuous and sustained. Since 2016 we have had 64 internship opportunities for undergraduate and graduate students in the areas of space crop production and microgravity simulation support. Some students have participated in more than one opportunity, so 52 separate individuals have participated in this program. While interns primarily worked with individual mentors on a main project, most also worked on group projects with other interns or developed smaller secondary projects on their own initiative. In addition, all interns participated in a weekly Lunch, Learn, and Discuss seminar series where they were encouraged to present their research, learn from scientists at KSC, and engage with members of the space life sciences community. When the face-to-face spring internship of 2020 ended prematurely, we resumed the weekly seminar series as a virtual event, and this led to the concept of engaging former interns. Current and former interns were invited to participate in the on-going virtual series, and 46 asked to be included on the mailing list. Our current seminar series involves remote seminars or hosted discussions from contacts within and outside of the agency, and engages a large network of current scientists and former interns with an weekly participation averaging 40-50 people. Speakers include experts from industry, academia, and other space agencies working on topics of interest to KSC space life science researchers. In addition, this series has served as a brain trust and brainstorming mechanism for KSC scientists to invite and discuss new ideas and concepts. The advent of remote work and virtual platforms has allowed us to engage with interns in a new way that will continue as future interns are integrated into this network of individuals interested in space research. This work is funded by NASA’s Space Biology Program.

Internships↗