System-level Metrics Research Articles

High-Performance Lithium-Sulfur Batteries with Atomic V- and Co-Modified Ketjen Black-Sulfur Composite Cathode

Lithium-sulfur (Li-S) batteries are potential alternatives to lithium-ion batteries, which are known to have certain drawbacks. This has drawn special technological attention. Sulfur’s low electrical conductivity, shuttling of the intermediate polysulfides, the slow conversion reaction kinetics of various lithium polysulfides (LiPS), and the major volume change of sulfur in the lithiation reaction are some of the main issues that still need to be resolved before Li-S batteries can be used in practical applications. Additionally, it is likely because there are not enough prognostic models determining energy densities at the cell and pack level based on materials and cell design, most of the articles do not report system-level metrics, specific energy (Wh kg-1), and energy density (Wh L-1).To overcome these problems, atomic vanadium (V) and cobalt (Co) modified ketjen black-sulfur composite (VCKBS) was prepared and used as a cathode in Li-S batteries. The synthesized composite was characterized by various techniques, e.g., XRD, Raman, SEM, HR-TEM, HAADF-STEM, and XPS spectra. The polysulfide adsorption test showed that V had increased the adsorption ability, and cyclic voltammetry displayed the increased catalytic activity resulting from Co, offering a large number of interfacial active sites and ensuring smooth electron transfer. Cycling performance tests showed an initial specific capacity of 1329 mAh g−1, maintained at 1249 mAh g−1 after 100 cycles. Under a higher sulfur loading (2.4 mg cm−2), at 0.2 C, Li-S cells demonstrated an initial specific capacity of 1201 mAh g−1 and retained 985 mAh g−1 after 300 cycles with a low fading rate of 0.059% per cycle. The rate performance tests showed that the prior capacity is nearly recovered, going from 1C back to 0.1C, demonstrating the superiority of the VCKBS material.In addition to the experimental characterization of the Li-S cell performance, energy densities were also predicted by a proposed system-level performance model. The 1st and 100th discharge capacities were used to forecast the system-level specific energies and energy densities. The system-level specific energies and energy densities based on the 100th discharge capacity of VCKBS cathodes had improvements of 1342 and 568%, respectively. This suggests that an easy route to construct the highly stable cathode material may pave the way for the development of highly efficient Li-S batteries. Acknowledgments Hira Fazal acknowledges support from the Türkiye Scholarship/Research Fellowship Program sponsored by the Prime Ministry of Turkey Presidency for Turks Abroad and Related Communities (21PK070917). Damla Eroglu acknowledges support from the Istanbul Development Agency (Award No: TR10/21/YEP/0001). We also thank the National Natural Science Foundation of China (22171180) and the Science and Technology Commission of Shanghai Municipality (20520710400).

Abstract PO1-10-10: Leveraging technology to optimize symptom management and patient empowerment in routine breast cancer care: implementation of a remote patient monitoring (RPM) pathway across 24 hospitals in Europe

Abstract Background: In clinical trials, RPM has been proven to improve the quality of life and toxicity management for patients receiving systemic treatment and is recommended by international guidelines for routine clinical practice (Di Maio et al., Ann Oncol 2022). Resilience is an oncology-specific RPM system (CE marked, class IIa medical device - mobile application [app] or web interface) which prompts patients to complete a weekly survey including core symptoms from NCI’s PRO-CTCAE questionnaire. Severe or worsening symptoms trigger an alert notification sent to the patient’s care team. In addition to the RPM system, the mobile app offers a large pool of breast cancer-specific educational, self-management, and patient empowerment resources that are personalized according to symptoms reported, disease stage, and treatments received. Here we present the implementation of the Resilience RPM pathway in routine breast cancer care across 24 highly diverse cancer centers. Methods: The implementation process followed a phasic process of Exploration, Preparation, Implementation, and Sustainability that involved close collaboration with technology providers, hospital managers, and care teams at participating cancer centers to coordinate, train, provide technical assistance, and respond to site-specific needs. In addition, monthly webinars offered participating centers an opportunity to share best practices and ways to overcome implementation barriers. A dedicated platform with educational RPM resources for healthcare providers was also available. A process evaluation according to the RE-AIM framework was performed assessing the reach, adoption, implementation, and maintenance of the RPM pathway in routine care. The effectiveness of the pathway was assessed as the ability of the care team to timely respond to alerts. Aggregated system level metric of RPM was collected covering patients with breast cancer under active systemic treatment followed by Resilience from 01-11-2021 to 29-06-2023. Usage data of mobile app resources was tracked from 01-01-2023 to 29-06-2023. Qualitative assessments with focus groups complemented this evaluation. Results: 985 patients, 1% male, 99% female, with a median age of 57y (range: 21-95y; P25-75 38.5-64.5y), were registered across 24 cancer centers. In the last 6 months, the predominant RPM interface chosen by patients was the mobile app (64.1 vs 35.9% for the web platform). Adherence to weekly ePRO reporting was 89.4% in the mobile app and 86.6% in the web platform. The median time from prompt receipt to survey completion was 5.6 hours. The median time for the nurse navigator (NN) to process an alert was 3h50min. Most alerts (91%) were handled directly by NN with 9% requiring the support of a physician. The mean number of individual content consumed per patient was 13 (6744 resources/523 pts, 100% of app users accessed resources). Preferred formats of content delivery were written articles (66.5%), followed by videos (23.1%) and podcasts (10.4%). Most assessed contents referred to: symptom management advice (16%), followed by information about breast cancer disease characteristics (12.5%), supportive care strategies (11.5%), self-guided exercises of mindfulness-based stress reduction (9.4%), and breast cancer treatment journey (8.3%). Qualitative assessments of the app interface, educational content, and self-management programs revealed high levels of satisfaction. The RPM pathway is sustained in the 24 cancer centers and implementation is expanding to additional centers. Conclusion: The Resilience RPM pathway was successfully implemented in routine care across a diverse group of cancer centers with elevated levels of patient engagement and responsiveness from NN. In addition, patients accessed evidence-based educational and empowerment resources, adapted to their medical profile according to their preferences and learning style. Citation Format: Maria Alice Franzoi, Arlindo R Ferreira, Juliette Fanton d'Andon, Thomas Grellety, Eliane Ithurbide-Dachary, Laura Polastro, Laurence Caravella, Joseph Rodriguez, Nathanaelle Leprevier, Elsa Blanchard, Stéphane Remy, Belen d' Ythurbide, Romain Rivoirard, Raoudha Boughzala, Joana M Ribeiro, Ines Vaz Luis. Leveraging technology to optimize symptom management and patient empowerment in routine breast cancer care: implementation of a remote patient monitoring (RPM) pathway across 24 hospitals in Europe [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO1-10-10.

FedMon: A Federated Learning Monitoring Toolkit

Federated learning (FL) is rapidly shaping into a key enabler for large-scale Artificial Intelligence (AI) where models are trained in a distributed fashion by several clients without sharing local and possibly sensitive data. For edge computing, sharing the computational load across multiple clients is ideal, especially when the underlying IoT and edge nodes encompass limited resource capacity. Despite its wide applicability, monitoring FL deployments comes with significant challenges. AI practitioners are required to invest a vast amount of time (and labor) in manually configuring state-of-the-art monitoring tools. This entails addressing the unique characteristics of the FL training process, including the extraction of FL-specific and system-level metrics, aligning metrics to training rounds, pinpointing performance inefficiencies, and comparing current to previous deployments. This work introduces FedMon, a toolkit designed to ease the burden of monitoring FL deployments by seamlessly integrating the probing interface with the FL deployment, automating the metric extraction, providing a rich set of system, dataset, model, and experiment-level metrics, and providing the analytic means to assess trade-offs and compare different model and training configurations.

Assessing stakeholder structure in water governance in the Murray-Darling Basin, a public submission perspective

Transformation from top-down government to collaborative governance for natural resources requires broadening the scope of stakeholders involved. However, the systemic identification and assessment of stakeholder engagement remains a challenge. This paper develops a network-based framework for assessing stakeholder engagement in natural resources governance. The framework proposed five system-level metrics: inclusivity, cohesion, centralization, efficiency, and brokerage, as well as three node-level metrics: dominators, brokers, and communicators. By using public submission data on four government policy initiatives each reflecting a major component of the Murray-Darling Basin Plan, this study assessed the structural aspects of stakeholder engagement relevant to water resource management in the Murray-Darling Basin following legislative reforms in 2007. Findings suggested that current stakeholder engagement structures are neither balanced, nor have adequate maturity to develop a governance approach to resolve issues within the Basin. High variability between patterns of involvement in submissions to key policy initiatives indicated there were haphazard aggregations of public comments and the top-down approach to water management is still dominant. The framework developed in this paper offers a structural and systemic perspective for re-configuring stakeholder engagement networks for improving collaborative governance.

Towards improvement of oncology care through digital technology: A snapshot of Resilience remote patient monitoring (RPM) system implementation across 19 hospitals in Europe.

1518 Background: In clinical trials, RPM based on patient reported outcomes (PROs) has been shown to reduce symptom burden, increase dose intensity, improve quality of life and overall survival, and is recommended by international guidelines in routine oncology practice during systemic treatment (Di Maio et al., Ann Oncol 2022). Resilience developed a digital RPM system for oncology (CE marked, class IIa medical device) which prompts patients weekly with a notification to complete a survey including common symptoms from the NCI’s PRO-CTCAE questionnaire. Severe or worsening symptoms trigger an alert notification to the patient’s cancer care team, as well as personalized self-management advice, empowerment and educational material to patients. In this study we report the real-world implementation of Resilience RPM system in France and Belgium. Methods: Aggregated system-level metrics were collected covering all participating hospitals between Nov 2021 - Dec 2022. Qualitative and quantitative assessments focused on patient and provider utilization of Resilience RPM system was done. The implementation included a continued interaction between Resilience and hospital personnel for training, coordination and technical support. Results: 19 hospitals deployed Resilience in France and Belgium, including two Organization of European Cancer Institutes (OECI)-designated cancer centers and community oncology hospitals. Three integrated RPM with their electronic medical record. Overall, 1262 pts were registered (age range, 20-94), 40% with breast, 20% gastrointestinal and 8% genitourinary cancers. In the last 6 months, the main RPM interface was the mobile app (70%). Patient adherence with weekly RPM surveys was 85%, with a median time from prompt receipt to survey completion of 5h20min. RPM retention for 3 months, regardless of disease status, was 80%. Irrespective of severity, the most commonly reported PROs were performance status decline (94%), pain (89%) and anorexia (82%). The most common severe alert notifications were for nausea (33%), pain (31%) and anorexia (23%). The median time to alert management by the care team was 12h (70% within 24h), with 70% of alerts managed by a nurse navigator follow-up call, 5% by a referral for internal/external appointments, 0.5% with an emergency room referral and 0.4% with a hospitalization referral. 87% (32/38) of providers were satisfied with integrating the solution into their organization and 80% (30/38) felt patients were better managed through use of Resilience RPM. Conclusions: Implementation of the Resilience RPM system was feasible across a diverse group of 19 cancer centers in France and Belgium, with high levels of patient and provider participation and engagement. Resilience RPM offers an evidence-based approach to improve the quality and patient-centeredness of cancer care.

Lifetime Performance Analysis of Imbalanced EV Battery Packs and Small-Signal Cell Modeling for Improved Active Balancing Control

Conventional battery balancing techniques target fast voltage or state-of-charge convergence while improving circuit-level metrics such as power density, efficiency, and component count. System-level metrics, including prolonged pack cycle life, reduced degradation rate, and increased energy capacity, are often overlooked. In this work, the lifetime discharge energy of imbalanced battery packs is quantified and compared using Monte-Carlo-style battery lifetime transient simulations with experimentally captured drive cycles. Computations were carried out across <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$&gt;$</tex-math></inline-formula> 20 000 core-hours on the Niagara supercomputer at the SciNet HPC Consortium. Imbalance was introduced by sampling the cell capacity and impedance values from normal distributions with up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sigma$</tex-math></inline-formula> =5% capacity/impedance variation at the beginning-of-life. Conventional balancing techniques are found to have little lifetime energy benefit over no balancing. A linearized equivalent circuit model (L-ECM) technique is introduced for small-signal analysis of battery packs with arbitrary capacity and impedance imbalance. An L-ECM-based balancing control is found to have a lifetime discharge energy improvement of up to 53.2% in the worst-case pack lifetime energy over no balancing. The L-ECM balancing control is demonstrated experimentally to provide 9.2% increased single-cycle discharge energy compared to no balancing in a Tesla Model S battery module.

Ultra-Dense LEO Satellite-Based Communication Systems: A Novel Modeling Technique

Low Earth orbit (LEO) satellite plays an indispensable role in the equal access network because of its low latency, large capacity, and seamless global coverage. For such an unprecedented extensive irregular system, stochastic geometry (SG) is a suitable research method. The SG model can not only cope with the increasing network scale, but also accurately analyze and estimate the network&#x0027;s performance. Several standard satellite distribution models and satellite-ground channel models are investigated in this article. System-level metrics such as coverage probability and their intermediates are introduced in the non-technical description. Then the influence of gateway density, and the number and height of satellites on latency and coverage probability is studied. Finally, this article presents the possible challenges and corresponding solutions for the SG-based LEO satellite system analysis.

Fault-Tolerance in the Scope of Cloud Computing

Fault-tolerance methods are required to ensure high availability and high reliability in cloud computing environments. In this survey, we address fault-tolerance in the scope of cloud computing. Recently, cloud computing-based environments have presented new challenges to support fault-tolerance and opened new paths to develop novel strategies, architectures, and standards. We provide a detailed background of cloud computing to establish a comprehensive understanding of the subject, from basic to advanced. We then highlight fault-tolerance components and system-level metrics and identify the needs and applications of fault-tolerance in cloud computing. Furthermore, we discuss state-of-the-art proactive and reactive approaches to cloud computing fault-tolerance. We further structure and discuss current research efforts on cloud computing fault-tolerance architectures and frameworks. Finally, we conclude by enumerating future research directions specific to cloud computing fault-tolerance development.

Deep Learning Based Intrusion Detection in Cloud Services for Resilience Management

In the global scenario one of the important goals for sustainable development in industrial field is innovate new technology, and invest in building infrastructure. All the developed and developing countries focus on building resilient infrastructure and promote sustainable developments by fostering innovation. At this juncture the cloud computing has become an important information and communication technologies model influencing sustainable development of the industries in the developing countries. As part of the innovations happening in the industrial sector, a new concept termed as ‘smart manufacturing’ has emerged, which employs the benefits of emerging technologies like internet of things and cloud computing. Cloud services deliver an on-demand access to computing, storage, and infrastructural platforms for the industrial users through Internet. In the recent era of information technology the number of business and individual users of cloud services have been increased and larger volumes of data is being processed and stored in it. As a consequence, the data breaches in the cloud services are also increasing day by day. Due to various security vulnerabilities in the cloud architecture; as a result the cloud environment has become non-resilient. To restore the normal behavior of the cloud, detect the deviations, and achieve higher resilience, anomaly detection becomes essential. The deep learning architectures-based anomaly detection mechanisms uses various monitoring metrics characterize the normal behavior of cloud services and identify the abnormal events. This paper focuses on designing an intelligent deep learning based approach for detecting cloud anomalies in real time to make it more resilient. The deep learning models are trained using features extracted from the system level and network level performance metrics observed in the Transfer Control Protocol (TCP) traces of the simulation. The experimental results of the proposed approach demonstrate a superior performance in terms of higher detection rate and lower false alarm rate when compared to the Support Vector Machine (SVM).

A comprehensive study on software aging across android versions and vendors

This paper analyzes the phenomenon of software aging – namely, the gradual performance degradation and resource exhaustion in the long run – in the Android OS. The study intends to highlight if, and to what extent, devices from different vendors, under various usage conditions and configurations, are affected by software aging and which parts of the system are the main contributors. The results demonstrate that software aging systematically determines a gradual loss of responsiveness perceived by the user, and an unjustified depletion of physical memory. The analysis reveals differences in the aging trends due to the workload factors and to the type of running applications, as well as differences due to vendors’ customization. Moreover, we analyze several system-level metrics to trace back the software aging effects to their main causes. We show that bloated Java containers are a significant contributor to software aging, and that it is feasible to mitigate aging through a micro-rejuvenation solution at the container level.

On Developing Techniques for Sharing Satellite Spectrum with Indoor Small Cells in 5G

In this paper, we present two spectrum sharing techniques for a multisystem, incorporating an integrated satellite-mobile system and an autonomous terrestrial-mobile system (iSMS/aTMS), namely orthogonal spectrum sharing (OSS) and non-orthogonal spectrum sharing (nOSS) techniques. aTMS consists of numerous small cells deployed in several buildings, and iSMS consists of a satellite station integrated with complementary ground component (CGC) stations deployed within buildings. By exploiting the high external wall penetration loss of a building, the iSMS spectrum is shared with small cells per building in OSS, and small cells per 3-dimensional (3D) cluster per building in nOSS. An interference management scheme, to avoid interference in apartments with collocated CGC stations and small cells, was developed and an optimal number of almost blank subframes (ABSs) per ABS pattern period (APP) was defined. System-level capacity, spectral efficiency, and energy efficiency performance metrics were derived. Furthermore, we present an algorithm for both OSS and nOSS techniques. With extensive simulation and numerical analysis, it is shown that the proposed nOSS significantly outperforms OSS in terms of spectral efficiency and energy efficiency, and both techniques can meet the expected spectral efficiency and energy efficiency requirements for the fifth-generation (5G) mobile networks.

A Tactic for Architectural Exploitation of Indoor Small Cells for Dynamic Spectrum Sharing in 5G

This paper presents a tactic to realize numerous in-building small cell base station (SBS) architectures for dynamic spectrum sharing by varying the number of physical transceivers as well as the number, amount, and characteristics of operating spectrums per SBS. Each SBS architecture is defined as a Type and the spectrum sharing mechanism of each Type of architecture is detailed and mathematically analyzed to define existing dynamic spectrum sharing techniques suggested in the literature. The high external wall penetration loss of a building and the eICIC technique are exploited to manage co-channel interference generated due to sharing the spectrum of one system to another. System-level capacity, spectral efficiency, and energy efficiency performance metrics are derived for each SBS architecture to show the relative outperformance of one to another. It is found that, unlike energy efficiency (EE), the spectral efficiency (SE) response of an SBS architecture is directly affected by the channel characteristics as well as the number and amount of operating spectrum bands. However, the number of transceivers of an SBS does not have a noticeable impact on both SE and EE so long as the feature of an operating spectrum is not altered. Finally, we show that each Type of the proposed SBS architectures can achieve the prospective SE and EE requirements for fifth-generation (5G) mobile networks.

Optimal UAV Base Station Trajectories Using Flow-Level Models for Reinforcement Learning

Cellular base stations (BS) and remote radio heads can be mounted on unmanned aerial vehicles (UAV) for flexible, traffic-aware deployment. These UAV base station networks (UAVBSN) promise an unprecendented degree of freedom that can be exploited for spectral efficiency gains as well as optimal network utilization. However, the current literature lacks realistic radio and traffic models for UAVBSN deployment planning and for performance evaluation. In this paper, we propose flow-level models (FLM) for realistically characterizing the UAVBSN performance in terms of a broad range of flow- and system-level metrics. Further, we propose a deep reinforcement learning (DRL) approach that relies on the UAVBSN FLM for learning the optimal traffic-aware UAV trajectories. For a given user traffic density and starting UAV locations, our RL approach learns the optimal UAV trajectories offline that maximizes a cumulative performance metric. We then execute the learned UAV trajectories in a discrete event simulator to evaluate online UAVBSN performance. For ${M}\pmb {=}$ 9 UAVs deployed in a simulated Downtown San Francisco model, where the UAV trajectories are defined by ${N}\pmb {=}$ 20 discrete actions, our approach achieves approximately a three-fold increase in the average user throughput compared to the initial UAV placement, while simultaneously balancing traffic loads across the BSs.

Countrywide Mobile Spectrum Sharing with Small Indoor Cells for Massive Spectral and Energy Efficiencies in 5G and Beyond Mobile Networks

In this paper, we propose a technique to share the licensed spectrums of all mobile network operators (MNOs) of a country with in-building small cells per MNO by exploiting the external wall penetration loss of a building and introducing the time-domain eICIC technique. The proposed technique considers allocating the dedicated spectrum Bop per MNO only its to outdoor macro UEs, whereas the total spectrum of all MNOs of the country Bco to its small cells indoor per building such that technically any small indoor cell of an MNO can have access to Bco instead of merely Bop assigned only to the MNO itself. We develop an interference management strategy as well as an algorithm for the proposed technique. System-level capacity, spectral efficiency, and energy efficiency performance metrics are derived, and a generic model for energy efficiency is presented. An optimal amount of small indoor cell density in terms of the number of buildings L carrying these small cells per MNO to trade-off the spectral efficiency and the energy efficiency is derived. With the system-level numerical and simulation results, we define an optimal value of L for a dense deployment of small indoor cells of an MNO and show that the proposed spectrum sharing technique can achieve massive indoor capacity, spectral efficiency, and energy efficiency for the MNO. Finally, we demonstrate that the proposed spectrum sharing technique could meet both the spectral efficiency and the energy efficiency requirements for 5G mobile networks for numerous traffic arrival rates to small indoor cells per building of an MNO.

STRAM

Various system metrics have been proposed for measuring the quality of computer-based systems, such as dependability and security metrics for estimating their performance and security characteristics. As computer-based systems grow in complexity with many subsystems or components, measuring their quality in multiple dimensions is a challenging task. In this work, we tackle the problem of measuring the quality of computer-based systems based on the four key attributes of trustworthiness we developed: security, trust, resilience, and agility. In addition to conducting a systematic survey on metrics, measurements, attributes of metrics, and associated ontologies, we propose a system-level trustworthiness metric framework that accommodates four submetrics, called STRAM (&lt;u&gt;S&lt;/u&gt;ecurity, &lt;u&gt;T&lt;/u&gt;rust, &lt;u&gt;R&lt;/u&gt;esilience, and &lt;u&gt;A&lt;/u&gt;gility &lt;u&gt;M&lt;/u&gt;etrics). The proposed STRAM framework offers a hierarchical ontology structure where each submetric is defined as a sub-ontology. Moreover, this work proposes developing and incorporating metrics describing key assessment tools, including vulnerability assessment, risk assessment, and red teaming, to provide additional evidence in the measurement and quality of trustworthy systems. We further discuss how assessment tools are related to measuring the quality of computer-based systems and the limitations of the state-of-the-art metrics and measurements. Finally, we suggest future research directions for system-level metrics research toward measuring fundamental attributes of the quality of computer-based systems and improving the current metric and measurement methodologies.

Data-Driven Ranking of Coordinated Traffic Signal Systems for Maintenance and Retiming

Automated traffic signal performance measures (ATSPMs) have been deployed with increasing frequency. At present, the existing ATSPMs are focused on the performance of individual movements or intersections. As the number of ATSPM users has increased, a need for system-level metrics has emerged. This paper proposes a method of evaluating corridor performance at the system level using high-resolution data. The method is demonstrated for eight signalized corridors in Indiana, including 87 intersections. This method develops five subscores for the areas of communication, detection, safety, capacity allocation, and progression; these five interrelated aspects of performance are each given a category subscore based on quantitative performance measures, with scales appropriate to the context of the operation. An overall score for each corridor is determined from the lowest subscore of each of the five areas. This approach simplifies the analysis process, as opposed to examining several hundred individual movements as currently would be required using ATSPM tools that are commonly available at present. The methodology is presented as a prototype for further development and adaptation to individual agency objectives and data sources.

Geomorphic trajectory and landform analysis using graph theory: A panel data approach to quantitative geomorphology

Comparing successive datasets of GIS polygons derived from remote-sensing data is a common approach to quantify morphological change. GIS-derived datasets capture instantaneous observations or “snapshots” of the state of a system at a given time but do not explicitly capture the temporal sequences needed to characterize system processes. Comparisons between these “temporally-naive” datasets can be used to infer properties and trends of the landscape as a whole, but tracking changes in the characteristics of individual landforms (e.g. sandbars, dunes, or other surface features of interest) across snapshots is labor-intensive and infeasible for large or irregular datasets. Using traditional computer-based procedural methods to compare sequences of datasets without knowledge of temporal trajectories introduces several challenges and data artifacts that complicate analysis. We propose a graph-theory approach for processing sequential spatial data to automatically identify and track distinct groups of related landforms or “geomorphic units” across fully or partially overlapping snapshots. This approach allows tracking even in cases where landforms fragment, merge, migrate, or become temporarily obstructed from view. The method promotes new panel data analysis opportunities and overcomes three critical limitations of traditional procedural methods of assessing landscape change from spatial data: (1) it can generate landscape metrics based on geomorphic units, rather than the arbitrary geographic units of the underlying spatial datasets, (2) it distinguishes missing or obstructed observations from changes in the characterization of landforms due to environmental conditions, and (3) it automatically generates panel datasets and discriminates between within-landform change and across-landform variation. The panel datasets can be used to upscale feature-level information to system-level metrics and analysis. Furthermore, a graph-theory approach can yield insight on geomorphic change through analysis of the graph structure, and offers a promising approach for geomorphological analyses which retain information on the spatial configuration of geomorphic units. We demonstrate the method with examples from emergent sandbars on the Missouri River.

System level indicators of changing marine connectivity

Spatial connectivity has long been recognized as a key process for sustaining healthy ecosystems and robust ecosystem services. However, system-level metrics that capture environmentally significant aspects of connectivity at appropriate temporal and spatial scales have not previously been identified. Using a major industrial harbour adjacent to Australia’s Great Barrier Reef as a test case, we developed a consistent and comprehensive set of connectivity indicators associated with waterborne dispersal that transparently relate to water quality, spread of contaminants, and potential for recruitment of planktonic larvae to nursery habitats. Results indicate all measures of connectivity are variable across management zones and likely to influence water quality and breeding success at these scales. Connectivity indicators also reveal environmental and ecological trade-offs. For example, while reduced flushing of creeks and estuaries may negatively impact local water quality, it can benefit ecological connectivity through more effective upstream transport of larvae to nursery habitats.

Genetic Programming for Energy-Efficient and Energy-Scalable Approximate Feature Computation in Embedded Inference Systems

With the increasing interest in deploying embedded sensors in a range of applications, there is also interest in deploying embedded inference capabilities. Doing so under the strict and often variable energy constraints of the embedded platforms requires algorithmic, in addition to circuit and architectural, approaches to reducing energy. A broad approach that has recently received considerable attention in the context of inference systems is approximate computing. This stems from the observation that many inference systems exhibit various forms of tolerance to data noise. While some systems have demonstrated significant approximation-versus-energy knobs to exploit this, they have been applicable to specific kernels and architectures; the more generally available knobs have been relatively weak, resulting in large data noise for relatively modest energy savings (e.g., voltage overscaling, bit-precision scaling). In this work, we explore the use of genetic programming (GP) to compute approximate features. Further, we leverage a method that enhances tolerance to feature-data noise through directed retraining of the inference stage. Previous work in GP has shown that it generalizes well to enable approximation of a broad range of computations, raising the potential for broad applicability of the proposed approach. The focus on feature extraction is deliberate because they involve diverse, often highly nonlinear, operations, challenging general applicability of energy-reducing approaches. We evaluate the proposed methodologies through two case studies, based on energy modeling of a custom low-power microprocessor with a classification accelerator. The first case study is on electroencephalogram-based seizure detection. We find that the choice of two primitive functions (square root, subtraction) out of seven possible primitive functions (addition, subtraction, multiplication, logarithm, exponential, square root, and square) enables us to approximate features in 0.41 <inline-formula><tex-math notation="LaTeX">$mJ$</tex-math></inline-formula> per feature vector (FV), as compared to 4.79 <inline-formula><tex-math notation="LaTeX">$mJ$</tex-math></inline-formula> per FV required for baseline feature extraction. This represents a feature extraction energy reduction of 11.68 <inline-formula><tex-math notation="LaTeX"> $\times$</tex-math></inline-formula> . The important system-level performance metrics for seizure detection are sensitivity, latency, and number of false alarms per hour. Our set of GP models achieves 100 percent sensitivity, 4.37 second latency, and 0.15 false alarms per hour. The baseline performance is 100 percent sensitivity, 3.84 second latency, and 0.06 false alarms per hour. The second case study is on electrocardiogram-based arrhythmia detection. In this case, just one primitive function (multiplication) suffices to approximate features in 1.13 <inline-formula><tex-math notation="LaTeX">$\mu J$</tex-math> </inline-formula> per FV, as compared to 11.69 <inline-formula><tex-math notation="LaTeX">$\mu J$</tex-math></inline-formula> per FV required for baseline feature extraction. This represents a feature extraction energy reduction of 10.35 <inline-formula><tex-math notation="LaTeX"> $\times$</tex-math></inline-formula> . The important system-level metrics in this case are sensitivity, specificity, and accuracy. Our set of GP models achieves 81.17 percent sensitivity, 80.63 percent specificity, and 81.86 percent accuracy, whereas the baseline achieves 82.05 percent sensitivity, 88.12 percent specificity, and 87.92 percent accuracy. These case studies demonstrate the possibility of a significant reduction in feature extraction energy at the expense of a slight degradation in system performance.

Packaging and Miniaturization of a 2–18 GHz UWB Radar for Measurements of Snow and Ice: Initial Results

Abstract We report recent progress towards miniaturizing and packaging an ultra wideband (UWB) radar developed for measurements of ice and snow over the 2–18 GHz range. The initial objective of this work is the close integration of the radar's RF section (composed of a receiver and a low-power transmitter) while maintaining or enhancing its electrical performance. First, we describe a connectorized reference system developed to serve as a performance benchmark for transmit and receive signal chains. Next, we discuss the use of drop-in modules to achieve a significant size reduction of the RF front-end. These modules are based on packaged chips and arranged in a planar configuration. We present system-level performance metrics of the reference system justifying the need for further integration. We briefly discuss other potential applications of the miniaturized radar and the direction of future packaging efforts as we work toward our future goal of integrating the complete radar “from antenna to bits” by using a system-in-package approach.