The proceedings contain 39 papers. The special focus in this conference is on Network and parallelcomputing. The topics include: LCache: Machine Learning-Enabled Cache Management in Near-Data Processing-Based Solid-S...
ISBN:
(纸本)9783030794774
The proceedings contain 39 papers. The special focus in this conference is on Network and parallelcomputing. The topics include: LCache: Machine Learning-Enabled Cache Management in Near-Data Processing-Based Solid-State Disks;security Situation Prediction of Network Based on Lstm Neural Network;Dynamic GMMU Bypass for Address Translation in Multi-GPU Systems;parallel Fast DOA Estimation Algorithm Based on SML and Membrane computing;segmented Merge: A New Primitive for parallel Sparse Matrix Computations;a Hierarchical Model of Control Logic for Simplifying Complex Networks Protocol Design;FPGA-Based Multi-precision Architecture for Accelerating Large-Scale Floating-Point Matrix computing;a Configurable Hardware Architecture for Runtime Application of Network Calculus;FEB3D : An Efficient FPGA-Accelerated Compression Framework for Microscopy Images;A Dynamic Mapping Model for General CNN Accelerator Based on FPGA;NUMA-Aware Optimization of Sparse Matrix-Vector Multiplication on ARMv8-Based Many-Core Architectures;CARAM: A Content-Aware Hybrid PCM/DRAM Main Memory System Framework;Optimization of RDMA-Based HDFS Data Distribution Mechanism;reducing the Time of Live Container Migration in a Workflow;RDMA-Based Apache Storm for High-Performance Stream Data Processing;payment Behavior Prediction and Statistical Analysis for Shared Parking Lots;location-Based Service Recommendation for Cold-Start in Mobile Edge computing;an Adaptive Delay-Limited Offloading Scheme Based on Multi-round Auction Model for Mobile Edge computing;an Efficient Data Transmission Strategy for Edge-computing-Based Opportunistic Social Networks;shadow Data: A Method to Optimize Incremental Synchronization in Data Center;A Dynamic Protection Mechanism for GPU Memory Overflow;DROAllocator: A Dynamic Resource-Aware Operator Allocation Framework in distributed Streaming Processing;a Medical Support System for Prostate Cancer Based on Ensemble Method in Developing Countries.
The proceedings contain 7 papers. The topics discussed include: productive and performant generic lossy data compression with LibPressio;mitigating catastrophic forgetting in deep learning in a streaming setting using...
ISBN:
(纸本)9781728186726
The proceedings contain 7 papers. The topics discussed include: productive and performant generic lossy data compression with LibPressio;mitigating catastrophic forgetting in deep learning in a streaming setting using historical summary;PyParSVD: a streaming, distributed and randomized singular-value-decomposition library;unbalanced parallel I/O: an often-neglected side effect of lossy scientific data compression;TributaryPCA: distributed, streaming PCA for in situ dimension reduction with application to space weather simulations;understanding effectiveness of multi-error-bounded lossy compression for preserving ranges of interest in scientific analysis;and exploring lossy compressibility through statistical correlations of scientific datasets.
Modern power transmission systems are under increasing pressure to ensure reliability and minimize outages in the face of aging infrastructure, growing demand, and integration of renewable energy sources. Traditional ...
详细信息
ISBN:
(数字)9798331528614
ISBN:
(纸本)9798331528621
Modern power transmission systems are under increasing pressure to ensure reliability and minimize outages in the face of aging infrastructure, growing demand, and integration of renewable energy sources. Traditional maintenance strategies, which rely on fixed schedules or reactive repairs, often fail to address the complexities of these evolving networks, leading to inefficient resource allocation, higher operational costs, and unplanned *** proposed solution is a combined implementation of Reliability-Centered Maintenance (RCM) and Predictive Maintenance (PdM). RCM, as implemented by POWERgrid, focuses on optimizing maintenance efforts by prioritizing assets based on the criticality of failure and its impact on system performance. This structured approach helps in reducing downtime, improving system reliability, and making better use of available manpower and *** parallel, Predictive Maintenance introduces data-driven insights through real-time monitoring, advanced analytics, and machine learning algorithms to predict equipment failures before they occur. By identifying early warning signs of equipment wear or performance degradation, this approach ensures timely interventions, thus preventing unexpected breakdowns and extending the lifespan of critical components. The paper highlights key advantages of these methods over traditional strategies, including enhanced system reliability, reduced operational costs, and more efficient asset management. Additionally, it explores future trends such as deep learning, edge computing, and cybersecurity, demonstrating how predictive maintenance is evolving to meet the challenges of modern power transmission systems. Together, RCM and PdM present a comprehensive solution for enhancing the resilience and efficiency of the electrical grid.
The payload in the network traffic contains a variety of information related to the traffic. Identifying anomalous attack behaviors through the payload is a crucial method to protect against network attacks effectivel...
详细信息
ISBN:
(数字)9781510651890
ISBN:
(纸本)9781510651890;9781510651883
The payload in the network traffic contains a variety of information related to the traffic. Identifying anomalous attack behaviors through the payload is a crucial method to protect against network attacks effectively. The payload structure is complex, which contains a large number of contents related to the security field, and these contents have contextual semantics strong relevance. To fully express the relevance of payload contents and better improve the quality of payload feature extraction, this paper proposes a feature extraction algorithm for payload based on tree structure representation, called TSR. The experimental results show that, compared with the existing feature extraction algorithms, the ROC-AUC of TSR increases by 3.32% on average, and the PR-AUC increases by 24.15% on average.
With the development of machine learning stepping into a bottleneck period, quantum machine learning has become a new popular research direction. Quantum computing is built on the principle of quantum mechanics, which...
详细信息
Within the cloud computing scenario, each server usually carries multiple service processes, which intensifies the concurrency pressure of the system. As a result, the process of memory management during page allocati...
详细信息
作者：
Pham, NghiaHoang, Diep ThiVnu
University of Engineering and Technology Vietnam National University Hanoi Faculty of Information Technology Hanoi Viet Nam
Accelerating phylogenetic tree reconstruction and bootstrapping is critical, especially to support the study of the evolution of dangerous viruses. In this paper, we propose the MPBoot-MPI, a distributed algorithm eff...
详细信息
Directed Greybox Fuzzing employs code analysis to predict critical paths or potential vulnerabilities. Multi-objective optimization is a challenge to Directed Greybox Fuzzing. Current approaches employ evolutionary al...
Directed Greybox Fuzzing employs code analysis to predict critical paths or potential vulnerabilities. Multi-objective optimization is a challenge to Directed Greybox Fuzzing. Current approaches employ evolutionary algorithms which may trap in local optima during mutation operation procedure. In this paper, we propose MODGF, a Directed Greybox Fuzzing method based on the elitist multi-objective evolutionary algorithm. The key insight of our approach is to get a more diverse distribution along the Pareto front during evolutionary iterations, which speed the convergence of the test cases. In MODGF, we use function distance, basic block distance, and seed distance as objective functions, and calculate distance metrics of the nearest neighbor solutions to infer the distribution density of solutions. We further introduce convergence measure and diversity measure as fitness values to ensure that the generated test cases are more uniformly distributed with more vulnerability types. In comparison to AFL and AFL-GO, the experiment reveals that MODGF exhibits significant performance enhancements in terms of convergence time, target function coverage, and vulnerability triggering efficiency.
Against the backdrop of the “dual carbon” goals, the country is undergoing a low-carbon transition, with a large amount of clean energy being integrated into the power grid, leading to diversified and decentralized ...
详细信息
ISBN:
(数字)9798350366914
ISBN:
(纸本)9798350366921
Against the backdrop of the “dual carbon” goals, the country is undergoing a low-carbon transition, with a large amount of clean energy being integrated into the power grid, leading to diversified and decentralized energy transactions. Virtual power plants (VPPs) aggregate and coordinate resources such as distributed generation, controllable loads, and energy storage to optimize energy utilization and enhance energy efficiency. However, VPPs face potential risks and uncertainties in market transactions. This paper uses text mining technology to extract preliminary risk factors through word frequency analysis and combines Failure Mode and Effects Analysis (FMEA) to determine key risk indicators, forming a VPP transaction risk assessment index system. Based on game theory and the two-dimensional cloud model, a VPP transaction risk assessment model is constructed to comprehensively evaluate VPP transaction risks from the dimensions of risk occurrence probability and risk consequences. The model's effectiveness is verified through case analysis and comparative analysis, assessing the overall risk level of VPP participation in market transactions and formulating corresponding control measures to ensure the feasibility and reliability of VPP transactions.
The evolving network infrastructure, particularly the 5G core network, is increasingly adopting cloud technologies. This shift brings to the forefront the challenge of meeting the demanding per-packet processing requi...
详细信息
ISBN:
(数字)9798350386059
ISBN:
(纸本)9798350386066
The evolving network infrastructure, particularly the 5G core network, is increasingly adopting cloud technologies. This shift brings to the forefront the challenge of meeting the demanding per-packet processing requirements posed by multi-hundred Gbps Ethernet NICs (network interface cards). While traditional NFV (network function virtualization) platforms are effective on older hardware, the per-packet run-to-completion (RTC) execution model for per-packet processing suffers from stalling on state access due to L1/L2 cache misses. Although previous work applying software prefetching can mitigate the issues, their applications are fundamentally limited by the nature of a single execution stream, hence limiting them to batch lookups, suffering from control-flow divergence, and requiring manual tuning. To address the limitations, we introduce a novel interleaved function stream execution model that exploits the function-level parallelism through memory-level parallelism, targeting feature-rich network functions such as 5G Core. To provide the visibility into network functions, we introduce a novel programming model based on the principle of Granular Decomposition, which provides deep visibility into the state access by decoupling the state in a more fine-grained manner compared to traditional modular approaches. We integrate these two innovative designs into a new open-source NF platform, which we refer to as GuNFu. We have tested GuNFu on widely deployed network functions such as 5G UPF (User Plane Function), 5G AMF (Access Management Function), NAT (Network Address Translator) and others. Extensive evaluations reveal that GuNFu can achieve throughput ranging from 1.5 to 6 times over the traditional modular approach.
暂无评论