Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “networking”

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 1,279 records · Page 71
All sourcesNASA NTRSDOE OSTI
NASA NTRS2018

Use of Nack Oriented Reliable Multicast (NORM) Protocol for Transport of Spacecraft Telemetry in Ground Networks

Frequently it is necessary to distribute spacecraft telemetry to multiple destinations on Internet Protocol (IP) networks. There are various methods of delivering this data, but when the data is high rate and there are many destinations IP multicast provides the best solution from a network perspective. However, IP multicast only supports the User Datagram Protocol (UDP) which is connectionless and unreliable. This nature of UDP and by extension the low priority treatment of UDP packets by some network hardware creates a difficult environment for distribution of high rate telemetry to a number of recipients over IP networks. NACK (Negative ACKnowledgment) Oriented Reliable Multicast (NORM) is defined by IETF (Internet Engineering Task Force) RFC (Request for Comments) 5740 in 2009 by B. Adamson et al. NORM provides a mechanism of retransmission and or forward error correction for UDP packets, potentially overcoming the limits of UDP for the transmission of spacecraft telemetry. Marshall Space Flight Center’s (MSFC) Huntsville Operations Support Center (HOSC) manages the delivery of the International Space Station (ISS) payload science and health data. The ISS IP Ground Routed (IIGoR) network provides the delivery of the data from the ground stations to the control centers. HOSC personnel have been experimenting with the use of NORM on the link between the ground stations and the front end processor equipment in Huntsville, AL.

Baldwin, Patrick A.↗
NASA NTRS2019

Space Communications and Navigation Validation: Extracting Data for the Strategic Center for Networking, Integration, and Communications Scheduling Algorithms

Efficiency in communication system architecture performance between Space Communications and Navigation (SCaN) assets and missions is crucial, as space communication is varied, complex, and often not utilized to its full potential. The SCaN Strategic Center for Networking, Integration, and Communications (SCENIC) new scheduling algorithms, which are designed to simulate the allocation of resources between SCaN assets and missions, have the potential to simulate an increase of this efficiency; however, they require real-world data to be validated against. The purpose of this project was to extract said validation data, which details the frequency and duration of utilized contact windows between missions and assets in the Near Earth Network (NEN), Space Network (SN), and Deep Space Network (DSN). Stored as images in daily operations summaries (DOSs), the tabular data existed in a variety of file formats such as.pdf, .docx, and .doc. Since the tables were stored as images, ABBYY® FineReader® (ABBYY Software Ltd.) optical character recognition (OCR) was implemented, which is a proprietary software that reads images from text. The comma separated value (CSV) output was utilized as input to a series of MATLAB® (The MathWorks, Inc.) methods for reformatting, at which point it was ready to be machine-read. Finally, the results were converted to a Microsoft Excel format for human readability. Along with being used for validation purposes, the data will also be used to map equipment degradation as a function of time to analyze the reliability of network assets.

Kontur, Noah P.↗
NASA NTRS2018

The Way to the Future Has Already Started: ICAO Aeronautical Telecommunication Network (ATN) Using Internet Protocol Suite (IPS) Standards and Protocol Evolution Update

Millions of people are able to exchange information over phones, computers, tablets and an array of new interconnected devices such as sensors, cameras, home appliances and others are increasingly becoming networked. At the core of this sizeable interconnection is a set of underlying protocols that enable millions of devices to seamlessly talk to each other. Aviation, as many other industries have, utilizes the internet protocol for ground-to-ground network communications and efforts are underway to bring it to the cockpit.In 2015, the International Civil Aviation Organization (ICAO) published the second edition of Document 9896 (Doc 9896) titled: Manual on the Aeronautical Telecommunication Network (ATN) using Internet Protocol Suite (IPS) Standards and Protocol. This manual adopted the Internet protocol version 6 (IPv6) for Internet layer interoperability. Document 9896 defined data communications protocols and services to be used for implementing aeronautical telecommunication network (ATN) using the Internet protocol suite (IPS). The document provided technical specifications that addressed security, network and transport protocols, described applications, supported by ATN/IPS and provided communications guidance. The international aviation community working under the guidance of ICAO initiated work on updates to Doc 9896. Under the ICAO Communication Panel, Working Group Internetworking (WG-I) got this challenging task. Key areas of work for WG-I are related to ATN/IPS Security and ANT/IPS Mobility. Two sub-groups have been assembled to work on these elements. The ICAO anticipates the release of updated Doc 9896 in 2020. Additional ATN/IPS standardization efforts include Radio Technical Commission for Aeronautics Special Committee 223 (RTCA SC-223) IPS profile development, Airlines Electronic Engineering Committee (AEEC) A658 Roadmap of IPS future activities and European Organization for Civil Aviation Equipment (EUROCAE) WG-108 development of guidance documentation.This paper describes ICAO WG-I approach to update key areas of the document in closed coordination with AEEC, RTCA and EUROCAE committees supporting the development of ATN/IPS. Additionally, the paper will explore WG-I mobility solutions under investigation, security developments, IPv6 addressing challenges and other proposed updates. Finally, as new operational expectations such as Free Route and Remote Piloted Aircraft System integration are currently pushing the boundaries set by old operational concepts, the paper identifies the emerging new challenges that will drive the ATN/IPS development beyond 2020. They are related to future voice utilization, System Wide Information Management and Remotely Piloted Aircraft System deployment. IPS systems are seen as the only way forward.

Apaza, Rafael↗
NASA NTRS2021

Identifying Planetary Transit Candidates in TESS Full-frame Image Light Curves via Convolutional Neural Networks

The Transiting Exoplanet Survey Satellite(TESS)mission measured light from stars in∼75% of the sky throughout its 2 yr primary mission, resulting in millions of TESS 30-minute-cadence light curves to analyze in the search for transiting exoplanets. To search this vast data trove for transit signals, we aim to provide an approach that both is computationally efficient and produces highly performant predictions. This approach minimizes the required human search effort. We present a convolutional neural network, which we train to identify planetary transit signals and dismiss false positives. To make a prediction for a given light curve, our network requires no prior transit parameters identified using other methods. Our network performs inference on a TESS 30-minute-cadence light curve in∼5 ms on a single GPU, enabling large-scale archival searches. We present 181 new planet candidates identified by our network, which pass subsequent human vetting designed to rule out false positives.Our neural network model is additionally provided as open-source code for public use and extension

Gregory Olmschenk↗
NASA NTRSDate not supplied

NASA’s Evolving Ka-band Network Capabilities to Meet Mission Demand

Space missions are increasingly demanding higher data rates to support the growth in information-intensive mission operations. This growth is reflected in planned and operational missions from low Earth orbit, such as the upcoming NASA-Indian Space Research Organization (ISRO) Synthetic Aperture Radar (NISAR) and Plankton, Aerosol, Cloud and Ocean Ecosystem (PACE) missions, to the future Artemis lunar campaign, and the recently launched James Webb Space Telescope orbiting at the Sun-Earth L2 Lagrange point. JWST was the first L2 mission to be defined as a high data rate mission transmitting at 8 Mbps, or 270 gigabits of science data per day. ISRO and PACE anticipate achieving data throughputs of 5-40 terabits per day. These data rates exceed the capabilities of S-band and X-band frequency allocations and are a key driver for migrating to the 26 GHz Ka-band frequency allocation. The NASA Space Communications and Navigation (SCaN) program has been preparing the networks to support this demand by pursuing critical Ka-band infrastructure. The status of current and evolving network capability, including the Near Space Network’s Initiative for Ka-band Advancement (NIKA), and the Deep Space Network’s Lunar Exploration Upgrades (DLEU), as well as profiling mission usage of Ka-band services, are discussed in detail. The push toward Ka-band, is not only an opportunity for increased performance, but alleviates current challenges with contentious and cluttered spectrum access in S- and X-band. The paper provides an overview of these advantages and advanced techniques that optimize its use before the transition to optical communications becomes an imperative. The challenges and potential mitigations for missions considering selection of Ka-band network services are also discussed.

space communications↗
NASA NTRSDate not supplied

Seven Years of Bright Meteor Data from the NASA All Sky Fireball Network

The NASA All Sky Fireball Network is a network consisting of 18 all-sky meteor cameras across the continental US. We present a data release of 33,660 bright meteors collected between 2013 and 2019. The released data consists of trajectory, orbit, radiant, shower association, and brightness. We discuss the camera hardware and software used to make the observations, the characteristics of the data, and known error sources. The camera hardware consists of Watec 902H2 Ultimate CCD cameras with a fish-eye lens giving an all-sky field of view. The data is complete until an absolute meteor magnitude of approximately -4, but meteors down to a magnitude of -1 are routinely observed. The cameras are spread across the continental United States in five subnetworks, with each network consisting of two to five cameras. The cameras report to a central server where the data is processed each morning. The network is modeled after the University of Western Ontario’s Southern Ontario Meteor Network. The cameras are running All-sky and Guided Real-time Detection (ASGARD) meteor detection software and analysis pipeline.

Aaron Kingery↗
NASA NTRSDate not supplied

Gateway Command and Data Handling Network Implementation and Validation

As initial Lunar Gateway modules approach design maturity, the Artemis Network Validation and Integration Lab (ANVIL) has begun demonstrations to validate avionics network dataflows. The flight architecture design includes utilizing Time-Triggered Ethernet (TTE) and layer-3 switching capabilities to enable greater automation and flexibility of critical and best effort traffic. Critical traffic is considered as Time-Triggered (TT), Rate Constrained (RC) and prioritized Best Effort (BE) traffic classes. End systems, such as mission computers, power control, and robotics, use three planes and all traffic classes while other devices interface via Best Effort. Typical best effort devices include video, laptops, wireless access points, and payloads. Other devices have a various hybrid approach of interfaces including alarms, telemetry/logging, and communication units. In the paper, we will present an update of the Gateway network architecture and how the system will operate nominally and during a stack topology reconfiguration. We will also discuss the network risks, and trade-offs of performance, flexibility, and redundancy. Finally, we will show the process for validation, demonstration, and verification approaches to the vehicle network.

Gateway↗
NASA NTRSDate not supplied

NASA’s Evolving Ka-Band Network Capabilities to Meet Mission Demand

Space missions are increasingly demanding higher data rates to support the growth in information-intensive mission operations. This growth is reflected in planned and operational missions from low Earth orbit, such as the upcoming NASA-Indian Space Research Organization (ISRO) Synthetic Aperture Radar (NISAR) and Plankton, Aerosol, Cloud and Ocean Ecosystem (PACE) missions, to the future Artemis lunar campaign, and the recently launched James Webb Space Telescope (JWST) orbiting at the Sun-Earth L2 Lagrange point. JWST was the first L2 mission to be defined as a high data rate mission transmitting at 28 Megabits per second (Mbps), or 270 Gigabits of science data per day. ISRO and PACE anticipate achieving data throughputs of 5-40 Terabits per day. These data rates exceed the capabilities of S-band and X-band frequency allocations and are a key driver for migrating to the 26 GHz Ka-band frequency allocation. The NASA Space Communications and Navigation (SCaN) program has been preparing the networks to support this demand by pursuing critical Ka-band infrastructure. The status of current and evolving network capability, including the Near Space Network’s Initiative for Ka-band Advancement (NIKA), and the Deep Space Network’s Lunar Exploration Upgrades (DLEU), as well as profiling mission usage of Ka-band services, are discussed in detail. The push toward Kaband, is not only an opportunity for increased performance, but alleviates current challenges with contentious and cluttered spectrum access in S- and X-band. The paper provides an overview of these advantages and advanced techniques that optimize its use before the transition to optical communications becomes an imperative. The challenges and potential mitigations for missions considering selection of Ka-band network services are also discussed.

Ka-band↗
NASA NTRSDate not supplied

Future Lunar Geophysical Mission Opportunities Including the Lunar Geophysical Network and CLPS

In the next few years, several opportunities are underway to take new geophysical observations of the Moon including geodetic and seismic. NASA’s novel Commercial Lunar Payload Services (CLPS) program seeks to acquire delivery services from 14 US companies. Nine funded task orders have been selected with payloads from multiple disciplines. Here we review the upcoming geophysical CLPS payloads and their measurement objectives then we provide a review of the Lunar Geophysical Network mission in development for New Frontiers 5. The Lunar Geophysical Network (LGN) mission is proposed to land on the Moon in the early 2030’s and deploy packages at four locations to enable continuous geophysical measurements for a minimum of 6 and a goal of 10 years. Returning to the lunar surface with a long-lived geophysical network is a key next step to advance lunar and planetary science. LGN will greatly expand our primarily Apollo-based knowledge of the deep lunar interior by identifying and characterizing mantle melt layers, as well as core size and state. To meet the mission objectives, the instrument suite provides complementary seismic, geodetic, heat flow, and electromagnetic (EM) observations. We discuss the network landing site requirements and provide example sites that meet these requirements. Landing sites include the P-5 region within the Procellarum KREEP Terrane (PKT; (lat:15˚; lon:-35˚), Schickard basin (lat:-44.3˚; lon:-55.1˚), Crisium basin (lat:18.5˚; lon:61.8˚), and the farside Korolev basin (lat:-2.4˚; lon:-159.3˚) (Figure 1). Network optimization considers the best locations to observe seismic core phases, e.g., ScS and PKP. Ray path density and proximity to young fault scarps are also analyzed to provide increased opportunities for seismic observations. Geodetic constraints from laser ranging require the LGN to have at least three nearside stations at maximum limb distances. Heat flow and EM measurements should be obtained away from terrane boundaries and from magnetic anomalies at locations representative of global trends. In our recent paper, an in-depth case study is provided for Mare Crisium. We also discuss the consequences for scientific return of less-than-optimal locations or number of stations.

Moons↗
NASA NTRSDate not supplied

Future Lunar Geophysical Mission Opportunities Including the Lunar Geophysical Network, Artemis and CLPS

In the next few years, several opportunities are underway to take new geophysical observations of the Moon including geodetic and seismic. NASA’s novel Commercial Lunar Payload Services (CLPS) program seeks to acquire delivery services from 14 US companies. Nine funded task orders have been selected with payloads from multiple disciplines. Here we review the upcoming geophysical CLPS payloads and their measurement objectives then we provide a review of the Lunar Geophysical Network mission in development for New Frontiers 5. The Lunar Geophysical Network (LGN) mission is proposed to land on the Moon in the early 2030’s and deploy packages at four locations to enable continuous geophysical measurements for a minimum of 6 and a goal of 10 years. Returning to the lunar surface with a long-lived geophysical network is a key next step to advance lunar and planetary science. LGN will greatly expand our primarily Apollo-based knowledge of the deep lunar interior by identifying and characterizing mantle melt layers, as well as core size and state. To meet the mission objectives, the instrument suite provides complementary seismic, geodetic, heat flow, and electromagnetic (EM) observations. We discuss the network landing site requirements and provide example sites that meet these requirements. Landing sites include the P-5 region within the Procellarum KREEP Terrane (PKT; (lat:15˚; lon:-35˚), Schickard basin (lat:-44.3˚; lon:-55.1˚), Crisium basin (lat:18.5˚; lon:61.8˚), and the farside Korolev basin (lat:-2.4˚; lon:-159.3˚) (Figure 1). Network optimization considers the best locations to observe seismic core phases, e.g., ScS and PKP. Ray path density and proximity to young fault scarps are also analyzed to provide increased opportunities for seismic observations. Geodetic constraints from laser ranging require the LGN to have at least three nearside stations at maximum limb distances. Heat flow and EM measurements should be obtained away from terrane boundaries and from magnetic anomalies at locations representative of global trends. In our recent paper, an in-depth case study is provided for Mare Crisium. We also discuss the consequences for scientific return of less-than-optimal locations or number of stations.

Moons↗
NASA NTRS2022

Preparing the Mars Relay Network for the Arrival of the Perseverance Rover at Mars

The Perseverance rover represents NASA’s latest achievement in Mars exploration. Landing successfully on 18 Feb 2021, the rover’s transmitted data during its entry, descent, and landing (EDL) were captured by the Mars Reconnaissance Orbiter (MRO) and the Mars Atmosphere and Volatile Evolution (MAVEN) orbiter. This data, broadcast in near-realtime to the world, allowed everyone to share in the excitement (and “terror”) of the day. The images returned thereafter included the first images of the new landing site, video of the landing itself taken from a variety of vantage points, and eventually the historic images of the first powered flight on another planet. Behind the scenes, the return of that data to Earth was accomplished via Mars orbiters operated by NASA and ESA, using three different ground tracking networks. Considered together, this Mars Relay Network (MRN) enabled the successful, timely, and unobtrusive return of the rover’s data. This paper describes the preparations taken by the participants of the MRN in anticipation of the arrival of Perseverance at Mars. These were not only focused on successfully acquiring the rover’s critical event telemetry during its EDL, but also on readying the network to return the rover’s data on an ongoing basis as it pursued its mission objectives. Included is a brief description of the MRN, which represents a highly successful international collaboration and continues as critical infrastructure for NASA’s and ESA’s ongoing Mars exploration. Also summarized are the activities performed prior to EDL, including landing site reconnaissance and mission test and training activities; those activities performed on EDL day, especially the recording, return, and processing of the rover’s critical event telemetry; and those activities that are now being performed on an ongoing basis during the rover’s surface operations, including an outline of the planning processes that enable relay services. Finally, a description of the performance of the network to-date on behalf of the Perseverance rover is given, summarizing the success of the network to provide support to both it and other spacecraft on the surface of Mars.

Srnka, Evan↗
NASA NTRSDate not supplied

CLAIRE: Enabling Heterogeneous Communication Network Optimization for Robust and Resilient Operations

In this paper, we present the capabilities of the CLAIRE System which provides resilient communications for NASA in presence of interference and congestion for a heterogeneous multi-vendor network. CLAIRE increases mission science data return to improve resource efficiencies and ensures resilience in the unpredictable space environment for NASA missions and communication networks. CLAIRE provides technology / waveform agnostic cognitive control plane that is instantiated at the Application Layer (APP) so that it can ride on NASA’s HDTN bundle protocol or any other protocol stack that is used by the network. The cognitive control plane is instantiated using Heartbeats (HTBTs). CLAIRE is assisted by Wideband UHF-Ka Band RF Sensing that leverages advances in the Direct Digital Transceiver (DDTRX) technology. The Wideband RF Sensing is driven by statistical signal processing and machine learning algorithms. Interference is mitigated using Dynamic Spectrum Access (DSA). Finally, CLAIRE addresses congestion using spectrum aware packet forwarding algorithm. CLAIRE provides an extensible protocol that allows passing of RF spectrum situational awareness, cross-layer sensing, delay tolerant networking and dynamic spectrum access information that can help with network optimization. Cross-Layer Sensing (CLS) and CLAIRE Decision Engine (CDE) enable spectrum and delay aware packet forwarding and Dynamic Spectrum Access during cases of severe interference.

cognitive communications↗
NASA NTRS2024

Exploring New Frontiers in Space Communications: Enhancing Delay Tolerant Networking through Cloud and Containerization

The High-rate Delay Tolerant Networking (HDTN) project at NASA Glenn Research Center has developed software that enables more flexible, reliable, and efficient space internetworking by using modern computing techniques such as cloud services, microservices, network function virtualization, software defined networking, and a distributed architecture. HDTN is built upon the Bundle Protocol and related convergence layers which have been developed to mitigate the challenges of the space networking environment including long delays, asymmetric data rates, and intermittent connectivity. The HDTN implementation employs asynchronous message processing tasks which allow for non-blocking operations as well as deployment in both centralized and distributed architectures. This paper investigates deploying HDTN in a containerized approach on the NASA Goddard’s Mission Cloud Platform using Amazon Web Services Elastic Compute Cloud (EC2). Commercial cloud computing will lower operating costs, provide flexible resource allocation, and allow for interconnectivity between multiple NASA centers as well as external partners. Containerization using Docker will enable greater portability and scalability for HDTN to be deployed into a variety of environments. We discuss possible NASA missions and use-cases such as the Laser Communications Relay Demonstration (LCRD) where the services provided by HDTN (reliable transport, high-rate message processing, and store-and-forward capabilities) will be enhanced through cloud computing and containerization. In addition, we describe the HDTN architecture and possible microservice-based networking approaches that can be obtained via HDTN’s configuration capabilities. Finally, we detail the EC2 specifications needed to achieve data rates greater than 1 Gbps to support optical communication missions such as LCRD.

Blake LaFuente↗
DOE OSTI2020

Dynamic frequency control in power networks

Node controllers in power distribution networks in accordance with embodiments of the invention enable dynamic frequency control. One embodiment includes a node controller comprising a network interface a processor; and a memory containing a frequency control application; and a plurality of node operating parameters describing the operating parameters of a node, where the node is selected from a group consisting of at least one generator node in a power distribution network wherein the processor is configured by the frequency control application to calculate a plurality of updated node operating parameters using a distributed process to determine the updated node operating parameter using the node operating parameters, where the distributed process controls network frequency in the power distribution network; and adjust the node operating parameters.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2020

Embedding global barrier and collective in a torus network

Embodiments of the invention provide a method, system and computer program product for embedding a global barrier and global interrupt network in a parallel computer system organized as a torus network. The computer system includes a multitude of nodes. In one embodiment, the method comprises taking inputs from a set of receivers of the nodes, dividing the inputs from the receivers into a plurality of classes, combining the inputs of each of the classes to obtain a result, and sending said result to a set of senders of the nodes. Embodiments of the invention provide a method, system and computer program product for embedding a collective network in a parallel computer system organized as a torus network. In one embodiment, the method comprises adding to a torus network a central collective logic to route messages among at least a group of nodes in a tree structure.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Systems and methods for monitoring traffic on industrial control and building automation system networks

Technologies relating to monitoring communications traffic to detect potential attacks on industrial control system networks and building automation system networks are described herein. In an embodiment, a monitoring device receives a plurality of communications from a control network. The monitoring device transmits the communications to a computing device. Based on the communications, the computing device generates a listing of devices that communicated by way of the control network over a period of time, and computes a volume of traffic between each pair of devices in the listing of devices. The computing device then outputs a graphical user interface (GUI) by way of display, the GUI comprising data indicative of the computed volumes of traffic, which may be indicative of a potential attack on the control network.

Jenkins, Chris↗
DOE OSTI2020

Thermodynamic Consistent Neural Networks for Learning Material Interfacial Mechanics

For multilayer materials in thin substrate systems, interfacial failure is one of the most challenges. The traction-separation relations (TSR) quantitatively describe the mechanical behavior of a material interface undergoing openings, which is critical to understand and predict interfacial failures under complex loadings. However, existing theoretical models have limitations on enough complexity and flexibility to well learn the real-world TSR from experimental observations. A neural network can fit well along with the loading paths but often fails to obey the laws of physics, due to a lack of experimental data and understanding of the hidden physical mechanism. In this paper, we propose a thermodynamic consistent neural network (TCNN) approach to build a data-driven model of the TSR with sparse experimental data. The TCNN leverages recent advances in physics-informed neural networks (PINN) that encode prior physical information into the loss function and efficiently train the neural networks using automatic differentiation. We investigate three thermodynamic consistent principles, i.e., positive energy dissipation, steepest energy dissipation gradient, and energy conservative loading path. All of them are mathematically formulated and embedded into a neural network model with a novel defined loss function. A real-world experiment demonstrates the superior performance of TCNN, and we find that TCNN provides an accurate prediction of the whole TSR surface and significantly reduces the violated prediction against the laws of physics.

Zhang, Jiaxin↗
DOE OSTI2021

Binary Complex Neural Network Acceleration on FPGA

Being able to learn from complex data with phase information is imperative for many signal processing applications. Today’s real-valued deep neural networks (DNNs) have shown efficiency in latent information analysis but fall short when applied to the complex domain. Deep complex networks (DCN) , in contrast, can learn from complex data, but have high computational costs; therefore, they cannot satisfy the instant decision making requirements of many deployable systems dealing with short observations or short signal bursts. Recent, Binarized Complex Neural Network (BCNN), which integrates DCNs with binarized neural networks (BNN), shows great potential in classifying complex data in real-time. In this paper, we propose a structural pruning based accelerator of BCNN, which is able to provide more than 5000 frames/s inference throughput on edge devices. The high performance comes from both the algorithm and hardware sides. On the algorithm side, we conduct structural pruning to the original BCNN models and obtain 20 × pruning rates with negligible accuracy loss; on the hardware side, we propose a novel 2D convolution operation accelerator for the binary complex neural network. Experimental results show that the proposed design works with over 90% utilization and is able to achieve the inference throughput of 5882 frames/s and 4938 frames/s for complex NIN-Net and ResNet-18 using CIFAR-10 dataset and Alveo U280 Board.

Peng, Hongwu↗