Journal of Systems Architecture最新文献

筛选
英文 中文
Back to fundamentals: Low-level visual features guided progressive token pruning 回到基本原理:低级的视觉特征引导逐步的标记修剪
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-10-01 DOI: 10.1016/j.sysarc.2025.103579
Yuanbing Ouyang , Yizhuo Liang , Qingpeng Li , Xinfei Guo , Yiming Luo , Di Wu , Hao Wang , Yushan Pan
{"title":"Back to fundamentals: Low-level visual features guided progressive token pruning","authors":"Yuanbing Ouyang ,&nbsp;Yizhuo Liang ,&nbsp;Qingpeng Li ,&nbsp;Xinfei Guo ,&nbsp;Yiming Luo ,&nbsp;Di Wu ,&nbsp;Hao Wang ,&nbsp;Yushan Pan","doi":"10.1016/j.sysarc.2025.103579","DOIUrl":"10.1016/j.sysarc.2025.103579","url":null,"abstract":"<div><div>Vision Transformers (ViTs) excel in semantic segmentation but demand significant computation, posing challenges for deployment on resource-constrained devices. Existing token pruning methods often overlook fundamental visual data characteristics. This study introduces ‘<strong>LVTP</strong>’, a progressive token pruning framework guided by multi-scale Tsallis entropy and low-level visual features with twice clustering. It integrates high-level semantics and basic visual attributes for precise segmentation. A novel dynamic scoring mechanism using multi-scale Tsallis entropy weighting overcomes limitations of traditional single-parameter entropy. The framework also incorporates low-level feature analysis to preserve critical edge information while optimizing computational cost. As a plug-and-play module, it requires no architectural changes or additional training. Evaluations across multiple datasets show 20%–45% computational reductions with negligible performance loss, outperforming existing methods in balancing cost and accuracy, especially in complex edge regions.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103579"},"PeriodicalIF":4.1,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145220059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Privacy-by-design AIoT vision for intelligent urban environments 智能城市环境的设计隐私AIoT愿景
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-29 DOI: 10.1016/j.sysarc.2025.103586
Marija Ivanovska, Jakob Kreft, Vitomir Štruc, Janez Perš
{"title":"Privacy-by-design AIoT vision for intelligent urban environments","authors":"Marija Ivanovska,&nbsp;Jakob Kreft,&nbsp;Vitomir Štruc,&nbsp;Janez Perš","doi":"10.1016/j.sysarc.2025.103586","DOIUrl":"10.1016/j.sysarc.2025.103586","url":null,"abstract":"<div><div>The recent advancements in AI (Artificial Intelligence) have been instrumental in fostering the development of AIoT (Artificial Intelligence of Things)-enabled urban environments. Machine vision and image analysis, in particular, have become integral to a wide array of AI applications within the field of urban planning and monitoring. Yet, the rapid adoption of AI algorithms in public areas has significantly heightened privacy concerns. In this paper, we introduce a privacy-by-design approach tailored for intelligent urban systems, presenting a holistic approach to the development and deployment of AI-driven systems for privacy-preserving image acquisition and analysis. Specifically, we design an embedded vision system that acquires privacy-protected data, safeguarding sensitive information against unauthorized access and potential misuse. Furthermore, we propose a strategy for developing AI vision methods using data that has been anonymized, ensuring that privacy is maintained throughout the AI application building process. Through experiments on a real-world AIoT-enabled urban environment use case – traffic flow monitoring at a city intersection – we demonstrate that our approach upholds strong privacy guarantees while maintaining the operational performance of modern AI vision systems.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"169 ","pages":"Article 103586"},"PeriodicalIF":4.1,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145236642","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Databelt: A continuous data path for serverless workflows in the 3D compute continuum Databelt:用于3D计算连续体中的无服务器工作流的连续数据路径
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-27 DOI: 10.1016/j.sysarc.2025.103577
Cynthia Marcelino, Leonard Guelmino, Thomas Pusztai, Stefan Nastic
{"title":"Databelt: A continuous data path for serverless workflows in the 3D compute continuum","authors":"Cynthia Marcelino,&nbsp;Leonard Guelmino,&nbsp;Thomas Pusztai,&nbsp;Stefan Nastic","doi":"10.1016/j.sysarc.2025.103577","DOIUrl":"10.1016/j.sysarc.2025.103577","url":null,"abstract":"<div><div>Serverless computing allows for dynamic and flexible execution of FaaS functions while simplifying infrastructure management. Typically, serverless functions rely on remote storage services for managing state, which can result in increased latency and network communication overhead. In a dynamic environment such as the 3D (Edge–Cloud–Space) Compute Continuum, serverless functions face additional challenges due to frequent changes in network topology. As satellites move in and out of the range of ground stations, functions must make multiple hops to access cloud services, leading to high-latency state access and unnecessary data transfers. In this paper, we present Databelt, a state management framework for serverless workflows designed for the dynamic environment of the 3D Compute Continuum. Databelt introduces an SLO-aware state propagation mechanism that enables the function state to move continuously in orbit. Databelt proactively offloads function states to the most suitable node, such that when functions execute, the data is already present on the execution node or nearby, thus minimizing state access latency and reducing the number of network hops. Additionally, Databelt introduces a function state fusion mechanism that abstracts state management for functions sharing the same serverless runtime. When functions are fused, Databelt seamlessly retrieves their state as a group, reducing redundant network and storage operations and improving overall workflow efficiency. Our experimental results show that Databelt reduces workflow execution time by up to 66% and increases throughput by 50% compared to the baselines. Furthermore, our results show that Databelt function state fusion reduces storage operations latency by up to 20%, by reducing repetitive storage requests for functions within the same runtime, ensuring efficient execution of serverless workflows in highly dynamic network environments such as the 3D Continuum.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103577"},"PeriodicalIF":4.1,"publicationDate":"2025-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145220018","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Optimizing utilization in logical execution time system with preserved externally-observable timed I/O semantics 通过保留外部可观察的定时I/O语义来优化逻辑执行时间系统的利用率
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-26 DOI: 10.1016/j.sysarc.2025.103589
Bo Zhang , Caixu Zhao , Yinkang Gao , Yixuan Zhu , Lei Gong , Teng Wang , Wenqi Lou , Xi Li
{"title":"Optimizing utilization in logical execution time system with preserved externally-observable timed I/O semantics","authors":"Bo Zhang ,&nbsp;Caixu Zhao ,&nbsp;Yinkang Gao ,&nbsp;Yixuan Zhu ,&nbsp;Lei Gong ,&nbsp;Teng Wang ,&nbsp;Wenqi Lou ,&nbsp;Xi Li","doi":"10.1016/j.sysarc.2025.103589","DOIUrl":"10.1016/j.sysarc.2025.103589","url":null,"abstract":"<div><div>Complex real-time functions, such as autonomous driving, are modeled as directed acyclic graphs (DAGs) of tasks that communicate with each other, subject to strict end-to-end latency constraints. The Logical Execution Time (LET) task model supports verification of these constraints by providing predictable and composable timed input/output (I/O). LET-based effect chains produce deterministic external behavior, defined through Externally-observable Timed I/O Semantics (ETIOS), where the external output traces are uniquely determined by the external input traces. However, these chains may include ineffective computations that do not impact the external output. Analyzing the effectiveness of these computations is further complicated by whether tasks are stateful (maintaining internal state) or stateless (not maintaining internal state). This paper proposes an offline optimization approach to achieve ETIOS withimproved utilization (computational demand) by deriving multiframe tasks that include only effective computations under loosened constraints. Our method effectively manages DAGs that combine both stateful and stateless tasks. A key contribution of our approach is the explicit modeling and analysis of how internal state influences data propagation within LET-based effect chains. Experiments on synthetic DAGs show utilization reductions of up to 77.79%, maximal for pure-stateless DAGs with unit max in-/out-degree and large period differences.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103589"},"PeriodicalIF":4.1,"publicationDate":"2025-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145220017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Toward TinyDPFL systems for real-time cardiac healthcare: Trends, challenges, and system-level perspectives on AI algorithms, hardware, and edge intelligence 面向实时心脏医疗的TinyDPFL系统:AI算法、硬件和边缘智能的趋势、挑战和系统级观点
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-25 DOI: 10.1016/j.sysarc.2025.103587
Muhammad Shakeel Akram , Bogaraju Sharatchandra Varma , Aqib Javed , Jim Harkin , Dewar Finlay
{"title":"Toward TinyDPFL systems for real-time cardiac healthcare: Trends, challenges, and system-level perspectives on AI algorithms, hardware, and edge intelligence","authors":"Muhammad Shakeel Akram ,&nbsp;Bogaraju Sharatchandra Varma ,&nbsp;Aqib Javed ,&nbsp;Jim Harkin ,&nbsp;Dewar Finlay","doi":"10.1016/j.sysarc.2025.103587","DOIUrl":"10.1016/j.sysarc.2025.103587","url":null,"abstract":"<div><div>Despite rapid advances in medical technology, cardiac diseases remain the leading cause of global mortality, with arrhythmias that pose significant diagnostic and treatment challenges. This survey presents a comprehensive review of 176 state-of-the-art contributions in machine learning (ML), federated learning (FL), TinyML, and hardware acceleration for efficient, real-time, and privacy-preserving cardiac diagnosis and care. Explores both software and hardware advancements, including differential privacy (DP), quantized neural networks, and FPGA (Field Programmable Gate Array)-based implementations optimized for edge devices and wearable devices. Key challenges, such as latency, energy constraints, adversarial robustness, and personalization, are systematically examined. The survey synthesizes solutions across algorithmic innovations, secure and adaptive FL frameworks, and neuromorphic and sparse architectures, especially FPGA-based solutions, for resource-aware inference and training. Informed by original research, it highlights emerging directions: AI-driven data mining, DP for quantized models, continual learning (CL) on the edge, FPGA-accelerators including quantized DNN, SNN, and Sparse architectures, tuneable/reconfigurable FPGA-based TinyDPFL, Multimodal heterogeneous FL, real-time adversarial detection via model watermarking. This work offers a unified system-level perspective bridging ML algorithms and edge AI hardware, guiding the development of scalable, adaptive, and trustworthy cardiac healthcare systems. Beyond surveying existing literature, it proposes forward-looking design principles to advance intelligent, secure, and practical digital cardiology.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103587"},"PeriodicalIF":4.1,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145220061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Scheduling virtual machines and containers: A comparative review of techniques, performance, and future trends 调度虚拟机和容器:技术、性能和未来趋势的比较回顾
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-25 DOI: 10.1016/j.sysarc.2025.103583
Nannan Zhao , Hao Wang , Fan Yang , Jiameng Zhang , Ruofei Wu , Taoyu Zhong , Shujie Han , Zhijie Huang , Xiao Zhang , Xiaonan Zhao
{"title":"Scheduling virtual machines and containers: A comparative review of techniques, performance, and future trends","authors":"Nannan Zhao ,&nbsp;Hao Wang ,&nbsp;Fan Yang ,&nbsp;Jiameng Zhang ,&nbsp;Ruofei Wu ,&nbsp;Taoyu Zhong ,&nbsp;Shujie Han ,&nbsp;Zhijie Huang ,&nbsp;Xiao Zhang ,&nbsp;Xiaonan Zhao","doi":"10.1016/j.sysarc.2025.103583","DOIUrl":"10.1016/j.sysarc.2025.103583","url":null,"abstract":"<div><div>Virtual machines (VMs) and containers underpin modern cloud infrastructures through distinct virtualization mechanisms, yet their scheduling strategies remain theoretically disconnected. Existing studies often address each paradigm in isolation, overlooking shared system-level challenges and conceptual overlaps. In this survey, we offers a virtualization-agnostic perspective by introducing a layered abstraction of scheduling, encompassing task dispatch, instance placement, and resource provisioning. We use this taxonomy to systematically compare five popular scheduling paradigms – mathematical modeling, heuristics, meta-heuristics, machine learning, and hybrids – on VM and container domains. Instead of enumerating techniques, our comparison emphasizes methodological convergence, indicating shared trade-offs guided by deployment granularity, orchestration flexibility, and performance isolation.</div><div>By projecting scheduling classes onto workload sensitivities, we provide context-specific guidance that connects theoretical research to practical configurations. Further, the research explains high-level algorithmic trends which indicate a shift away from isolated techniques to integrated orchestration systems. In this comparative viewpoint, the paper redefines VM and container scheduling as structurally equivalent problems with equivalent constraints and objectives. The outcomes are designed to help system designers and researchers design scheduling solutions that are both platform-adaptive and workload-responsive in heterogeneous computing systems.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103583"},"PeriodicalIF":4.1,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145220060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
A survey of optimization algorithms for differential privacy in Federated Learning 联邦学习中差分隐私优化算法综述
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-23 DOI: 10.1016/j.sysarc.2025.103582
Fangfang Shan , Yuhang Liu , Lulu Fan , Yifan Mao , Zhuo Chen , Shuaifeng Li
{"title":"A survey of optimization algorithms for differential privacy in Federated Learning","authors":"Fangfang Shan ,&nbsp;Yuhang Liu ,&nbsp;Lulu Fan ,&nbsp;Yifan Mao ,&nbsp;Zhuo Chen ,&nbsp;Shuaifeng Li","doi":"10.1016/j.sysarc.2025.103582","DOIUrl":"10.1016/j.sysarc.2025.103582","url":null,"abstract":"<div><div>Federated Learning (FL), as a distributed machine learning approach, enables joint model training without sharing raw data, but during the model update process, the transmission of information still poses potential risks that may lead to the leakage of user privacy. In recent years, differential privacy (DP) techniques have been widely applied to federated learning to enhance data privacy protection. However, the introduction of differential privacy often has a negative impact on model performance, such as reducing model accuracy and increasing training time. Therefore, how to effectively balance privacy protection and model performance in federated learning has become an urgent problem to address. This paper first introduces the basic principles of federated learning and differential privacy, and then focuses on reviewing optimization algorithms for Differential Privacy in Federated Learning (DPFL). Unlike existing reviews on DPFL, we categorize the optimization algorithms into three types: noise mechanism optimization, privacy budget management, and model update optimization. By referencing a large number of related studies, we elaborate on the basic ideas, key innovations, and other aspects of various optimization methods, showcasing their performance and advantages in balancing privacy protection and model performance. Finally, we provide an outlook on future research directions, including further integrating DPFL with other advanced technologies to provide stronger support for applications in complex scenarios, enhancing visualization, and exploring the application of DPFL optimization algorithms in more practical fields.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103582"},"PeriodicalIF":4.1,"publicationDate":"2025-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145157981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Efficient two-party Private Set Union and circuit-based version for large-scale data set based on novel oblivious filter 基于新型遗忘滤波器的大规模数据集的高效两方私有集联合和基于电路的版本
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-20 DOI: 10.1016/j.sysarc.2025.103581
Qian Xu, Huajie Shen, Yuhan Yang, Bo Yu, Wei He
{"title":"Efficient two-party Private Set Union and circuit-based version for large-scale data set based on novel oblivious filter","authors":"Qian Xu,&nbsp;Huajie Shen,&nbsp;Yuhan Yang,&nbsp;Bo Yu,&nbsp;Wei He","doi":"10.1016/j.sysarc.2025.103581","DOIUrl":"10.1016/j.sysarc.2025.103581","url":null,"abstract":"<div><div>Private Set Union (PSU) protocol enables secure computation of set unions across multiple parties while preserving input privacy — a critical capability for large-scale data aggregation. To date, two-party PSU schemes predominantly adopt the “split-then-execute” paradigm, which as demonstrated by Jia et al. (2024), inherently suffers from <em>during-execution leakage</em>. While their proposed symmetric-key approach addresses this vulnerability, the dependency on private equality test (PET) protocol additionally introduces non-negligible computational and communication overheads. Meanwhile, research on circuit-PSU protocol capable of securely computing a function (merge, sort or federated learning) over the union set remains blank.</div><div>In this paper, we focus on two main objectives: (1) We present an optimized two-party PSU protocol that eliminates during-execution leakage while achieving enhanced efficiency; (2) We propose the <em><strong>first</strong></em> circuit-PSU. Unlike conventional PSU where participant obtains plaintext result, circuit-PSU provides secret-shared outputs that inherently prevent both during-execution leakage and intersection cardinality exposure. Furthermore, circuit-PSU removes the requirement for predefined result recipient and thus can offer superior flexibility. Meanwhile, our technical innovations include an optimized Share Translation (ShareTrans) protocol and a novel Oblivious Filter (OF) primitive that synergistically enhance both PSU and circuit-PSU efficiency. The experimental results indicate that under constrained computational resources (8-core CPU, 16GB RAM), the proposed circuit-PSU achieves a comparatively better improvement in communication complexity compared to existing state-of-the-art schemes, and our two-party PSU also exhibits comparable performance, for example, achieving approximately <span><math><mrow><mn>4</mn><mo>×</mo></mrow></math></span> efficiency improvement over Jia et al. (2024).</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103581"},"PeriodicalIF":4.1,"publicationDate":"2025-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145117616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
SADDLE: A runtime feedback control architecture for adaptive distributed deep learning in heterogeneous GPU clusters 鞍:一种用于异构GPU集群中自适应分布式深度学习的运行时反馈控制架构
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-19 DOI: 10.1016/j.sysarc.2025.103573
HyungJun Kim , Eunyoung Lee , Heonchang Yu
{"title":"SADDLE: A runtime feedback control architecture for adaptive distributed deep learning in heterogeneous GPU clusters","authors":"HyungJun Kim ,&nbsp;Eunyoung Lee ,&nbsp;Heonchang Yu","doi":"10.1016/j.sysarc.2025.103573","DOIUrl":"10.1016/j.sysarc.2025.103573","url":null,"abstract":"<div><div>Adaptive training in heterogeneous GPU clusters requires more than isolated heuristics—it demands a real-time, feedback-driven control system. <em>SADDLE</em> is a self-adaptive framework that unifies global batch scaling, local throughput balancing, and transient straggler mitigation into a fully coordinated runtime. It combines scaling guided by the Gradient Noise Scale (GNS), z-score detection over Exponentially Weighted Moving Average (EWMA)-smoothed iteration times, and responsiveness tuned via a Proportional–Integral–Derivative (PID) controller into a single, event-driven control loop. Across vision and language tasks, <em>SADDLE</em> improves training time by up to 2.84×and accuracy by up to 5.26% over strong baselines, while maintaining under 6% runtime overhead. This reframing positions adaptive training as dynamic system regulation, enabling deep learning frameworks to self-optimize under real-world heterogeneity.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103573"},"PeriodicalIF":4.1,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145157980","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
ARES: Adaptive robust object detection framework for enhancing real-time performance in autonomous vehicle systems ARES:用于增强自动驾驶车辆系统实时性的自适应鲁棒目标检测框架
IF 4.1 2区 计算机科学
Journal of Systems Architecture Pub Date : 2025-09-18 DOI: 10.1016/j.sysarc.2025.103574
Sunghwan Park , Hyeongboo Baek , Jaewoo Lee
{"title":"ARES: Adaptive robust object detection framework for enhancing real-time performance in autonomous vehicle systems","authors":"Sunghwan Park ,&nbsp;Hyeongboo Baek ,&nbsp;Jaewoo Lee","doi":"10.1016/j.sysarc.2025.103574","DOIUrl":"10.1016/j.sysarc.2025.103574","url":null,"abstract":"<div><div>In contemporary autonomous vehicles, object detection must provide both robust detection against threats like adversarial patch attacks and timely execution to meet real-time deadlines. Certifiably robust detection, also known as patch-agnostic approach, meets the first requirement. However, it introduces significant computational overhead, thereby compromising its real-time performance. To resolve this conflict, we propose ARES, a novel framework inspired by mixed-criticality systems. ARES introduces a security-driven paradigm. By default, the framework operates in a high-performance, low-security mode. However, it transitions to a high-security mode, utilizing a computationally intensive and robust detector, only when an active attack is detected. This selective activation is managed by the ARES transition manager, which captures the attack timing and handles tasks during mode transition. The ARES scheduling framework, on the other hand, guarantees formal schedulability analysis and optimal priority assignment. In our experiments, ARES demonstrated an increase of up to 8.9<span><math><mo>×</mo></math></span> in overall FPS detection over baseline. Furthermore, when evaluating the acceptance ratio with randomly generated task sets, ARES exhibited a 40.8–62.9% enhancement in schedulability compared to baseline.</div></div>","PeriodicalId":50027,"journal":{"name":"Journal of Systems Architecture","volume":"168 ","pages":"Article 103574"},"PeriodicalIF":4.1,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145117617","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
相关产品
×
本文献相关产品
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信