Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Sensor 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 847 records · Page 47

Developing Flexible, Networked Lighting Control Systems That Reliably Save Energy in California Buildings

An important strategy to meet California's ambitious energy efficiency goals is to use innovative wireless communications, embedded sensors, data analytics and controls to significantly reduce lighting energy use in commercial buildings. This project developed a suite of networked lighting solutions to further this goal. The technologies include a platform for low-cost sensing, distributed intelligence and communications, the “PermaMote,” which is a self-powered sensor and controller for lighting applications. The project team also developed a task ambient daylighting system that integrates sensors with data-driven daylighting control using an open communication interface, called the “Readings-At-Desk” (RAD) system. To address the problem of building occupants being confused about how to operate traditional lighting control systems, the research team created content that could be the basis for a user interface standard for lighting controls. Finally, to address the difficulty of ensuring that advanced lighting control systems actually deliver their promised energy savings, the project team developed a new method for evaluating and specifying lighting systems’ performance. The research team validated these technologies in the laboratory, showing significant lighting energy savings, up to 73% for the PermaMote sensor system from occupancy control and daylight dimming features, compared to the same light source (LED replacement lamps) operated via simple on/off scheduling. The project team also developed a proposed standard lighting data model and user interface elements, which were contributed to the ANSI Lighting Systems Committee (C137) for standardization. Existing data models are incomplete and inconsistent, whereas the lighting-specific data model developed here is clear and comprehensive, to serve as a starting point for creating common, universally agreed upon semantic definitions of key lighting parameters, to promote interoperability. For the task on verifiable performance of lighting systems, the project team developed a more effective metric for capturing the actual energy impact of a lighting system over time — the energy usage intensity (kWh/ft2/year). Three commercial lighting systems were tested in FLEXLAB® using this new metric, and the tests show a wide range in the accuracy of the self-reported energy-use metric, from 0.5% to 28% error compared to direct measurement of lighting energy using dedicated submeters. Overall, the project team estimates that these advanced technologies can reduce California office lighting energy use by 20% (above and beyond normal advanced lighting controls mandated by Title 24), resulting in about 1,600 GWh/year in savings.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

An Evaluation of the Network Simulators in Large-Scale Distributed Simulations

This is a survey paper about the state-of-the-art in large-scale network simulation. Networks for the smart grids are characterized by millions of sensor nodes exchanging information about the status of the grid. This information exchange must be realized reliably and efficiently due to the mission critical nature of the power grid. Hence, the applications and the network protocols developed for the smart grid need go through rigorous testing and analysis before deployment. Developers usually do not have access to such a large-scale network that can be used as a controlled test-bed; therefore, network simulation becomes an essential tool for testing. Network simulation is a well studied problem in the literature and there are various widely used network simulators. These simulators can be adopted for testing applications and protocols of the smart grid. Due to the scale of these networks, parallel/distributed simulations need to be conducted. Even though most network simulators support distributed simulations, generating a large-scale network model to simulate can still be a cumbersome task. In this survey, we describe a selection of commonly used network simulators and evaluate them with respect to the following features that can aid users in distributed simulations of large-scale networks: transparency of setting up distributed simulation, automated topology generation, information hiding, lightweight routing protocols, network error simulation, evaluation of the network model during simulation and trace analysis tools. As a complementary result, we identify two issues with network simulators that can be addressed with runtime steering methods.

99 GENERAL AND MISCELLANEOUS↗

Thermoacoustic Thermometry for Nuclear Reactor Monitoring

On Friday, March 11, 2011, at 2:46pm (Japan Standard Trme), the Tohoku region on the east coast of northern Japan experi­enced what would become known as the largest earthquake in the country's history at magnitude 9.0 on the Richter scale. The Fukushima Daiichi nuclear power plant suffered exten­sive and irreversible damage. Six operating units were at the site, each with a boiling water reactor. When the earthquake struck, three of the six reactors were operating and the others were in a periodic inspection outage phase. In one reactor, all of the fuel had been relocated to a spent fuel pool in the reactor building. The seismic acceleration caused by the earthquake brought the three operating units to an automatic shutdown. Since there was damage to the power transmission lines, the emergency diesel generators (EDG) were automat­ically started to ensure continued cooling of the reactors and spent fuel pools. The situation was under control until the tsunami hit about forty-five minutes later with a maximum wave height of approximately 15 meters, which was three times taller than the sea wall of 5m. The influx of water submerged the EDGs, the electrical switchgear, and dc batteries, resulting in the total loss of power to five of the six reactors. The flooding also resulted in the loss of instrumentation that would have other­ wise been used to monitor and control the emergency. The ugly aftermath included high radiation exposure to operators at the nuclear power plants and early contamina­tion of food supplies and water within several restricted areas in Japan, where high radiation levels have rendered them un­safe for human habitation. While the rest of the story will remain a tragic history, it is this part of the series of unfortunate events that has inspired our research. It has indubitably highlighted the need for a novel sensor and instrumentation system that can withstand similar or worse conditions to avoid future catastrophe and assume damage prevention as quickly as possible. This is the question which we are attempting to answer: Is it possible to implement a self-powered sensor that could transmit data independently of electronic networks while taking advantage of the harsh operating environment of the nuclear reactor?

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Data Release Report for the Source Physics Experiment Phase II: Dry Alluvium Geology Experiments (DAG-1 through DAG-4), Nevada National Security Site

The Dry Alluvium Geology (DAG) project was Phase II of the Source Physics Experiment and consisted of a series of four chemical explosive tests conducted in the same source hole on the Nevada National Security Site. This hole is located at 37.1146°N and -116.0693°W, with a surface elevation of 1,285.2 meters (m) (4,216.5 feet [ft]) above sea level. The first test (DAG-1) was conducted on July 20, 2018, at 16:51:52.67838 Coordinated Universal Time (UTC). The explosive source for DAG-1 was nitromethane initiated by a small plastic-bonded explosive (PBX) charge, detonated at the depth of 385.0 m (1,263.2 ft) below ground surface. DAG-1 had a trinitrotoluene (TNT) equivalent yield of 0.908 metric tons (2,002 pounds [lbs]). DAG-2 was conducted on December 19, 2018, at 18:45:56.92115 UTC. This test was the largest in the series, with a TNT equivalent yield of 50.997 metric tons (112,429 lbs). The explosive source for DAG-2 was nitromethane initiated by a small PBX charge, detonated at the depth of 299.8 m (983.6 ft) below ground surface. DAG-3 was conducted on April 27, 2019, at 15:49:01.84183 UTC. The explosive source for this test was nitromethane initiated by a small PBX charge, detonated at the depth of 149.9 m (492.0 ft) below ground surface. DAG-3 had a TNT equivalent yield of 0.908 metric tons (2,002 lbs). The final DAG test (DAG-4) was conducted on June 22, 2019, at 21:06:19.87632 UTC. The explosive source for DAG-4 was nitromethane initiated by a small PBX charge, detonated at the depth of 51.6 m (169.3 ft) below ground surface. DAG-4 had a TNT equivalent yield of 10.357 metric tons (22,833 lbs). The four tests were recorded by an extensive set of instrumentation that included sensors both at near-field (less than 200 m) and far-field (200 m or greater) distances. The near-field instruments consisted of three-component (3C) accelerometers installed at various depths ranging from 51.6 to 385 m (169.3 to 1,263.1 ft) below ground surface in boreholes positioned around the source hole, and arrays of single-component and 3C accelerometers on the surface. The far-field network comprised a variety of seismic and acoustic sensors, including short-period geophones, broadband seismometers, and 3C accelerometers at distances of 200 m to 400 kilometers. In addition, the DAG-2, DAG-3, and DAG-4 explosions were recorded by a temporary array of 496 geophones arranged in a densely spaced grid pattern known as “Large N.” This report coincides with the release of these data for analysts and organizations that are not participants in this program. This report describes the four DAG tests and the various types of near-field, far-field, and other data that are available. Assembled data sets are accessible through: Incorporated Research Institutions for Seismology, Data Management Center 1408 NE 45th Street, Suite 201, Seattle, Washington 98105 USA. www.iris.washington.edu

58 GEOSCIENCES↗

NeuroCoreX: An Open-Source FPGA-Based Spiking Neural Network Emulator with On-Chip Learning

Spiking Neural Networks (SNNs) are computational models inspired by the event-driven communication and connectivity patterns of biological neural circuits. They enable high energy efficiency and natural support for diverse architectures ranging from layered networks to small-world and graphstructured topologies. In this work, we introduce NeuroCoreX, an open-source, FPGA-based spiking neural network emulator that provides real-time, on-chip learning and flexible network organization. NeuroCoreX supports both feedforward sensory inputs streamed directly from sensors or PCs via UART and recurrent on-chip connectivity, enabling simultaneous processing and learning from external stimuli and internal network dynamics-capabilities rarely available in existing FPGA SNN platforms. The system implements a Leaky Integrate-and-Fire (LIF) neuron model with current-based synapses and supports pair-based STDP learning on both feedforward and recurrent synapses. A lightweight Python interface enables interactive configuration, live monitoring, weight read-back, and experiment control. Importantly, NeuroCoreX is tightly integrated with the SuperNeuroMAT simulator, allowing SNN models to be transferred seamlessly from software to hardware for hardware-in-the-loop development. By combining real-time plasticity, flexible connectivity, and an open-source VHDL implementation, NeuroCoreX provides an extensible and accessible platform for neuromorphic research, algorithm-hardware co-design, and energy-efficient edge intelligence.

Gautam, Ashish [ORNL]↗

Deep compressed seismic learning for fast location and moment tensor inferences with natural and induced seismicity

Fast detection and characterization of seismic sources is crucial for decision-making and warning systems that monitor natural and induced seismicity. However, besides the laying out of ever denser monitoring networks of seismic instruments, the incorporation of new sensor technologies such as Distributed Acoustic Sensing (DAS) further challenges our processing capabilities to deliver short turnaround answers from seismic monitoring. In response, this work describes a methodology for the learning of the seismological parameters: location and moment tensor from compressed seismic records. In this method, data dimensionality is reduced by applying a general encoding protocol derived from the principles of compressive sensing. The data in compressed form is then fed directly to a convolutional neural network that outputs fast predictions of the seismic source parameters. Thus, the proposed methodology can not only expedite data transmission from the field to the processing center, but also remove the decompression overhead that would be required for the application of traditional processing methods. An autoencoder is also explored as an equivalent alternative to perform the same job. We observe that the CS-based compression requires only a fraction of the computing power, time, data and expertise required to design and train an autoencoder to perform the same task. Implementation of the CS-method with a continuous flow of data together with generalization of the principles to other applications such as classification are also discussed.

54 ENVIRONMENTAL SCIENCES↗

Insights from application of a hierarchical spatio-temporal model to an intensive urban black carbon monitoring dataset

Existing regulatory pollutant monitoring networks rely on a small number of centrally located measurement sites that are purposefully sited away from major emission sources. While informative of general air quality trends regionally, these networks often do not fully capture the local variability of air pollution exposure within a community. Recent technological advancements have reduced the cost of sensors, allowing air quality monitoring campaigns with high spatial resolution. The 100×100 black carbon (BC) monitoring network deployed 100 low-cost BC sensors across the 15 km 2 West Oakland, CA community for 100 days in the summer of 2017, producing a nearly continuous site-specific time series of BC concentrations which we aggregated to one-hour averages. Leveraging this dataset, we employed a hierarchical spatio-temporal model to accurately predict local spatio-temporal concentration patterns throughout West Oakland, at locations without monitors (average cross-validated hourly temporal R 2 =0.60). Using our model, we identified spatially varying temporal pollution patterns associated with small-scale geographic features and proximity to local sources. In a sub-sampling analysis, here we demonstrated that fine scale predictions of nearly comparable accuracy can be obtained with our modeling approach by using ~30% of the 100×100 BC network supplemented by a shorter-term high-density campaign.

54 ENVIRONMENTAL SCIENCES↗

Synergistic Smart Fuel For In-pile Nuclear Reactor Measurements

In March 2011, an earthquake of magnitude 9.0 on the Richter scale struck Japan with its epicenter on the northeast coast, near the Tohoku region. In addition to the immense physical destruction and casualties across the country, several nuclear power plants (NPP) were affected. It was the Fukushima Daiichi NPP that experienced the most severe and irreversible damage. The earthquake brought the reactors at Fukushima to an automatic shutdown and because the power transmission lines were damaged, emergency diesel generators (EDGs) were activated to ensure that there was continued cooling of the reactors and spent fuel pools. The situation was being successfully managed until the tsunami hit about forty-five minutes later with a maximum wave height of approximately 15 m. The influx of water submerged the EDGs, the electrical switchgear, and dc batteries, resulting in the total loss of power to the reactors.2 At this point, the situation became critical. There was a loss of the sensors and instrumentation within the reactor that could have provided valuable information to guide the operators to make informed decisions and avoid the unfortunate events that followed. In the light of these events, we have developed and tested a potential self-powered thermoacoustic system, which will have the ability to serve as a temperature sensor and can transmit data independently of electronic networks. Such a device is synergistic with the harsh environment of the nuclear reactor as it utilizes the heat from the nuclear fuel to provide the input power.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

High-Performance Computer Modeling of the Cosmos-Iridium Collision

This paper describes the application of a new, integrated modeling and simulation framework, encompassing the space situational awareness (SSA) enterprise, to the recent Cosmos-Iridium collision. This framework is based on a flexible, scalable architecture to enable efficient simulation of the current SSA enterprise, and to accommodate future advancements in SSA systems. In particular, the code is designed to take advantage of massively parallel, high-performance computer systems available, for example, at Lawrence Livermore National Laboratory. We will describe the application of this framework to the recent collision of the Cosmos and Iridium satellites, including (1) detailed hydrodynamic modeling of the satellite collision and resulting debris generation, (2) orbital propagation of the simulated debris and analysis of the increased risk to other satellites (3) calculation of the radar and optical signatures of the simulated debris and modeling of debris detection with space surveillance radar and optical systems (4) determination of simulated debris orbits from modeled space surveillance observations and analysis of the resulting orbital accuracy, (5) comparison of these modeling and simulation results with Space Surveillance Network observations. We will also discuss the use of this integrated modeling and simulation framework to analyze the risks and consequences of future satellite collisions and to assess strategies for mitigating or avoiding future incidents, including the addition of new sensor systems, used in conjunction with the Space Surveillance Network, for improving space situational awareness.

42 ENGINEERING↗

Path Network Recovery Using Remote Sensing Data and Geospatial-Temporal Semantic Graphs

Remote sensing systems produce large volumes of high-resolution images that are difficult to search. The GeoGraphy (pronounced Geo-Graph-y) framework [2, 20] encodes remote sensing imagery into a geospatial-temporal semantic graph representation to enable high level semantic searches to be performed. Typically scene objects such as buildings and trees tend to be shaped like blocks with few holes, but other shapes generated from path networks tend to have a large number of holes and can span a large geographic region due to their connectedness. For example, we have a dataset covering the city of Philadelphia in which there is a single road network node spanning a 6 mile x 8 mile region. Even a simple question such as "find two houses near the same street" might give unexpected results. More generally, nodes arising from networks of paths (roads, sidewalks, trails, etc.) require additional processing to make them useful for searches in GeoGraphy. We have assigned the term Path Network Recovery to this process. Path Network Recovery is a three-step process involving (1) partitioning the network node into segments, (2) repairing broken path segments interrupted by occlusions or sensor noise, and (3) adding path-aware search semantics into GeoQuestions. This report covers the path network recovery process, how it is used, and some example use cases of the current capabilities.

47 OTHER INSTRUMENTATION↗

Smart pixel sensors: towards on-sensor filtering of pixel clusters with deep learning

Highly granular pixel detectors allow for increasingly precise measurements of charged particle tracks. Next-generation detectors require that pixel sizes will be further reduced, leading to unprecedented data rates exceeding those foreseen at the High- Luminosity Large Hadron Collider. Signal processing that handles data incoming at a rate of $\mathcal{O}$(40 MHz) and intelligently reduces the data within the pixelated region of the detector at rate will enhance physics performance at high luminosity and enable physics analyses that are not currently possible. Using the shape of charge clusters deposited in an array of small pixels, the physical properties of the traversing particle can be extracted with locally customized neural networks. In this first demonstration, we present a neural network that can be embedded into the on-sensor readout and filter out hits from low momentum tracks, reducing the detector's data volume by 57.1%–75.7%. The network is designed and simulated as a custom readout integrated circuit with 28 nm CMOS technology and is expected to operate at less than 300 μW with an area of less than 0.2 mm 2 . The temporal development of charge clusters is investigated to demonstrate possible future performance gains, and there is also a discussion of future algorithmic and technological improvements that could enhance efficiency, data reduction, and power per area.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Evaluating the Capability of High-Altitude Infrasound Platforms to Cover Gaps in Existing Networks.

A variety of Earth surface and atmospheric sources generate low frequency sound waves that can travel great distances. Despite a rich history of ground-based sensor studies, very few experiments have investigated the prospects of free floating microphone arrays at high altitudes. However, recent initiatives have shown that such networks have very low background noise and may sample an acoustic wave field that is fundamentally different than that at the Earth's surface. The experiments have been limited to at most two stations at altitude, limiting their utility in acoustic event detection and localization. We describe the deployment of five drifting microphone stations at altitudes between 21 and 24 km above sea level. The stations detected one of two regional ground-based explosions as well as the ocean microbarom while traveling almost 500 km across the American Southwest. The explosion signal consisted of multiple arrivals; signal amplitudes did not correlate with sensor elevation or source range. A sparse network method that employed curved wave front corrections was able to determine the backazimuth from the free flying network to the acoustic source. Episodic broad band signals similar to those seen on previous flights in the same region were noted as well, but their source remains unclear. Background noise levels were commensurate with those on infrasound stations in the International Monitoring System (IMS) below 2 seconds, but sensor self noise appears to dominate at higher frequencies.

47 OTHER INSTRUMENTATION↗

Smart pixel sensors Towards on-sensor filtering of pixel clusters with deep learning

High granularity silicon pixel sensors are at the heart of energy frontier particle physics collider experiments. At an collision rate of 40\,MHz, these detectors create massive amounts of data. Signal processing that handles data incoming at those rate and intelligently reduces the data within the pixelated region of the detector \textit{at rate} will enhance physics performance and enable physics analyses that are not currently possible. Using the shape of charge clusters deposited in an array of small pixels, the physical properties of the traversing particle can be extracted with locally customized neural networks. In this first work, we present a neural network that can be embedded into the on-sensor readout and filter out hits from low momentum tracks, reducing the detector's data volume by 54.4-75.4\%. The network is designed and simulated as a custom readout integrated circuit with 28\,nm CMOS technology and is expected to operate at less than 300\,$\mu W$ with an area of less than 0.2\,mm$^2$.

43 PARTICLE ACCELERATORS↗

Blockchain based Communication Architectures with Applications to Private Security Networks

Existing communication protocols in high consequence security networks are highly centralized. While this naively makes the controls easier to physically secure, external actors require fewer resources to disrupt the system because there are fewer points in the system can be destroyed or interrupted without the entire system failing. We present a solution to this problem using a proof-of-work-based blockchain implementation built on MultiChain. We construct a test-bed network containing two types of data input: visual imagers and microwave sensor information. These data types are ubiquitous in perimeter intrusion detection security systems and allow a realistic representation of a real-world network architecture. The cameras in this system use an object detection algorithm to nd important targets in the scene. The raw data from the camera and the outputs from the detection algorithm are then placed in a transaction on the distributed ledger. Similarly, microwave data is used to detect relevant events and are placed in a transaction. These transactions are then bundled into blocks and broadcast to the rest of the network using the Bitcoin-based MultiChain protocol. We develop five tests to examine the security metrics of our network. We performed the five security metric test using different sized networks from 7 to 39 nodes to determine how the metrics scale with respect to size. We nd that when compared to a centralized architecture our implementation provides a resiliency increase that is expected from a blockchain-based protocol without slowing the system so much that a human operator would notice. Furthermore, our approach is able to detect tampering in real time. Based on these results, we theorize that security networks in general could use a blockchain-based approach in a meaningful way.

97 MATHEMATICS AND COMPUTING↗

Modulation and Modeling of Three-Dimensional Nanowire Assemblies Targeting Gas Sensors with High Response and Reliability

Despite improved sensitivity, simple downsizing of gas-sensing components to randomly arranged nanostructures often faces challenges associated with unpredictable electrical conduction pathways. In the present study, controlled fabrication of three-dimensional (3D) metal oxide nanowire networks is demonstrated that can greatly improve both signal stability and sensor response compared to random nanowire arrays. For example, the highest ever reported H 2 S gas response value, and a 5 times lower relative standard deviation of baseline resistance than that of random nanowires assemblies, are achieved with the ordered 3D nanowire network. Systematic engineering of 3D geometries and their modeling, utilizing equivalent circuit components, provide additional insights into the electrical conduction and gas-sensing response of 3D assemblies, revealing the critical importance of wire-to-wire junction points and their arrangement. Here these findings suggest new design rules for both enhanced performance and reliability of chemical sensors, which may also be extended to other devices based on nanoscale building blocks.

36 MATERIALS SCIENCE↗

CdSe/ZnS quantum dot fluorescence spectra shape-based thermometry via neural network reconstruction

As a system of interest gets small, due to the influence of the sensor mass and heat leaks through the sensor contacts, thermal characterization by means of contact temperature measurements becomes cumbersome. Non-contact temperature measurement offers a suitable alternative, provided a reliable relationship between the temperature and the detected signal is available. In this work, exploiting the temperature dependence of their fluorescence spectrum, the use of quantum dots as thermomarkers on the surface of a fiber of interest is demonstrated. The performance is assessed of a series of neural networks that use different spectral shape characteristics as inputs (peak-based—peak intensity, peak wavelength; shape-based—integrated intensity, their ratio, full-width half maximum, peak normalized intensity at certain wavelengths, and summation of intensity over several spectral bands) and that yield at their output the fiber temperature in the optically probed area on a spider silk fiber. Starting from neural networks trained on fluorescence spectra acquired in steady state temperature conditions, numerical simulations are performed to assess the quality of the reconstruction of dynamical temperature changes that are photothermally induced by illuminating the fiber with periodically intensity-modulated light. Comparison of the five neural networks investigated to multiple types of curve fits showed that using neural networks trained on a combination of the spectral characteristics improves the accuracy over use of a single independent input, with the greatest accuracy observed for inputs that included both intensity-based measurements (peak intensity) and shape-based measurements (normalized intensity at multiple wavelengths), with an ultimate accuracy of 0.29 K via numerical simulation based on experimental observations. The implications are that quantum dots can be used as a more stable and accurate fluorescence thermometer for solid materials and that use of neural networks for temperature reconstruction improves the accuracy of the measurement.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A Control Chart Approach for Representing and Mining Data Streams with Shape Based Similarity

The mining of data streams for online condition monitoring is a challenging task in several domains including (electric) power grid system, intelligent manufacturing, and consumer science. Considering a power grid application in which thousands of sensors, called the phasor measurement units, are deployed on the power grid network to continuously collect streams of digital data for real-time situational awareness and system management. Depending on design, each sensor could stream between ten and sixty data samples per second. The myriad of sensory data captured could convey deeper insights about sequence of events in real-time and before major damages are done. However, the timely processing and analysis of these high-velocity and high-volume data streams is a challenge. Hence, a new data processing and transformation approach, based on the concept of control charts, for representing sequence of data streams from sensors is proposed. In addition, an application of the proposed approach for enhancing data mining tasks such as clustering using real-world power grid data streams is presented. The results indicate that the proposed approach is very efficient for data streams storage and manipulation.

big data↗

A General Spatiotemporal Imputation Framework for Missing Sensor Data

Many applications from precision agriculture, environmental monitoring and transportation networks rely on data collected across space and time over a large geographic area. Missing data poses a significant challenge for any data-driven inference and control tasks. Data imputation or the estimation of missing data can help fill these gaps by utilizing inherent spatial relationships and temporal patterns. A variety of spatiotemporal imputation models have been developed to address missing data in spatiotemporal datasets. However, these classical methods rely on the assumption that the underlying data follows a smooth trend and fail to provide accurate estimates when there is a large number of missing points in the data. Even though there are machine learning driven tensor completion approaches such as convolutional neural network based tensor completion (CoSTCo) that capture the non-linear relationships in the dataset, the transductive nature makes the algorithm less scalable. Thus, existing approaches for estimating the missing information do not effectively capture all dimensions of the spatiotemporal data structure, resulting in erroneous predictions and poor performance. The main contributions of this paper are: (1) We propose a novel inductive framework (G-LSTM) for missing data imputation that integrates a graph neural network with LSTMs to effectively capture both spatial and temporal dependencies. (2) Experimental results on a traffic dataset demonstrate that the proposed GNN integrated with an LSTM framework achieves improved imputation and maintains steady performance even when there are extreme missing conditions in comparison with the state-of-the-art imputation framework (i.e, CoSTCo). (3) The simulation results on a traffic network show up to 69% reduction in mean absolute error and 61% reduction in root mean square error when compared to CoSTCo.

Tharzeen, Aabila↗