Future Generation Computer Systems-The International Journal of Escience最新文献

{"title":"VeriTrac: Verifiable and traceable cross-silo federated learning","authors":"Yanxin Xu ,&nbsp;Hua Zhang ,&nbsp;Zhenyan Liu ,&nbsp;Fei Gao ,&nbsp;Lei Qiao","doi":"10.1016/j.future.2025.107780","DOIUrl":"10.1016/j.future.2025.107780","url":null,"abstract":"<div><div>Cross-silo federated learning enables many clients to train a machine learning model collaboratively, while keeping the raw training data locally. It faces the risks of privacy leakage and malicious participants. In this paper, we introduce a new security risk that malicious clients may disrupt the training process of cross-silo federated learning by falsifying the verification evidences. The verification failure caused by this malicious behavior is not easily distinguishable from that caused by the malicious server falsifying the aggregated model. To address this issue, we design VeriTrac, the first privacy-preserving cross-silo federated learning scheme that supports verifiability and traceability. Before performing the aggregation, the server can utilize the non-private information of clients to verify messages submitted by them to avoid being framed. When the proportion of malicious clients is less than 50%, malicious participants causing the verification error can be traced. In addition, to verify the correctness of the aggregated models, a model vector with a verification factor is constructed and encrypted. The vector is confidential for the server, and the factor is part of the verification evidence and recoverable for clients. Security analysis shows that VeriTrac can guarantee the tracing of malicious participants and the data security of clients. Experimental evaluation shows that computation efficiency and communication efficiency of VeriTrac are acceptable.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"168 ","pages":"Article 107780"},"PeriodicalIF":6.2,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548536","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}

{"title":"Alleviating straggler impacts for data parallel deep learning with hybrid parameter update","authors":"Hongliang Li ,&nbsp;Qi Tian ,&nbsp;Dong Xu ,&nbsp;Hairui Zhao ,&nbsp;Zhewen Xu","doi":"10.1016/j.future.2025.107775","DOIUrl":"10.1016/j.future.2025.107775","url":null,"abstract":"<div><div>Data parallelism in distributed clusters faces challenges due to costly global parameter updates and performance imbalances, leading to stragglers that negatively impact training speed and accuracy. This paper proposes Cooperate Grouping Parallel (CGP), a hybrid parameter update scheme to alleviate the problem. CGP supports dynamic grouping among parallel workers and utilizes both intra-group synchronous update and inter-group asynchronous update. CGP treats straggler as an opportunity for worker groups to cooperatively reduce the global parameter update cost. We give the theoretical upper bound of model accuracy deviation caused by inter-group asynchronous updates, which proves the convergence property of the proposed CGP. Extensive testbed experiments on different workloads shows that CGP achieves 1.94<span><math><mo>×</mo></math></span> speedup compared to the other methods on average in different scenarios, and CGP improves accuracy by 16.8% over the asynchronous methods.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"168 ","pages":"Article 107775"},"PeriodicalIF":6.2,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548053","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}

{"title":"Entropy-based genetic feature engineering and multi-classifier fusion for anomaly detection in vehicle controller area networks","authors":"Mohammad Fatahi ,&nbsp;Danial Sadrian Zadeh ,&nbsp;Behzad Moshiri ,&nbsp;Otman Basir","doi":"10.1016/j.future.2025.107779","DOIUrl":"10.1016/j.future.2025.107779","url":null,"abstract":"<div><div>Technological advances in mobile computing, wireless communications, and remote sensing have provided the foundation for expanding and improving intelligent transportation systems (ITS), making modern vehicles susceptible to cyberattacks due to their evolved functionality and connectivity. In-vehicle networks, such as controller area networks (CAN), are highly vulnerable to attacks due to the lack of security architecture. Considering the temporal and spatial aspects of attacks and the need to develop lightweight models, this study develops a flexible and lightweight anomaly detection model for CAN bus with normal and sensitive duty cycles. To achieve optimal performance and consider spatio-temporal information, the feature space is optimized by extracting new features based on a two-parameter genetic algorithm (2P-GA) and Shannon entropy. Next, a synergistic combination of different supervised machine learning classifiers based on the ordered weighted averaging (OWA) operators is leveraged to optimize the results and achieve better performance. Also, to show the effectiveness of the proposed method in the present study, a comprehensive and unique comparative analysis with previous works and state-of-the-art models is presented. The results show that the proposed framework achieves the highest performance in terms of accuracy and F1-score and the lowest computational cost compared with previous works.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"169 ","pages":"Article 107779"},"PeriodicalIF":6.2,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143548635","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}

{"title":"Dynamic FPGA reconfiguration for scalable embedded artificial intelligence (AI): A co-design methodology for convolutional neural networks (CNN) acceleration","authors":"Jalil Boudjadar ,&nbsp;Saif Ul Islam ,&nbsp;Rajkumar Buyya","doi":"10.1016/j.future.2025.107777","DOIUrl":"10.1016/j.future.2025.107777","url":null,"abstract":"<div><div>In recent years, FPGA platforms have shown significant potential for accelerating artificial intelligence (AI) applications, particularly in Embedded AI. While various studies have explored adaptive AI deployment on FPGAs, there remains a gap in methodologies fully integrating software adaptability with FPGA hardware reconfigurability. This article presents a novel end-to-end co-design methodology for deploying adaptable and scalable Convolutional Neural Networks (CNNs) on FPGA platforms. The framework enhances computational performance and reduces latency by dynamically modifying hardware acceleration units by combining CNN architecture adaptability with dynamic partial reconfiguration of FPGA hardware. The proposed methodology enables automated synthesis and runtime customization of both hardware accelerators and CNN architectures, eliminating the need for iterative synthesis. This approach has been implemented and tested on a Xilinx XC7020 FPGA board for a CNN-based image classifier, achieving superior computation performance (0.68s/image) and accuracy (97%) compared to state-of-the-art alternatives.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"169 ","pages":"Article 107777"},"PeriodicalIF":6.2,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143620559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A distributed monitoring architecture for JointCloud computing","authors":"Yadi Wu ,&nbsp;Lina Wang ,&nbsp;Rongwei Yu ,&nbsp;Xiuwen Huang ,&nbsp;Jiatong Liu","doi":"10.1016/j.future.2025.107773","DOIUrl":"10.1016/j.future.2025.107773","url":null,"abstract":"<div><div>JointCloud computing supports large-scale resource consolidation and collaboration among multiple cloud service providers to provide users with powerful performance and adequate services. In the face of exponential scaling of resources, monitoring is an indispensable part of effective resource management. Monitoring provides methods for reviewing and managing the performance status of JointCloud resources and services to better characterize the overall operating status of JointCloud system. However, the collaboration between cloud service providers and the scale of resources in JointCloud are dynamically changing, and it is not easy to perform monitoring in a flexible and scalable way. In order to cover all aspects related to resource monitoring in JointCloud environments, we propose a distributed monitoring architecture for JointCloud computing. The architecture focuses on the ability to obtain information, organizes monitoring components in a modular way, and supports on-demand startup to provide dynamic monitoring capabilities. The proposed distributed monitoring approach provides load balancing and fault tolerance services to ensure reliability and performance of monitoring. The architecture also considers the JointCloud quality of service (QoS) and designs a virtual resource orchestration approach aimed at improving the efficiency of resource utilization. We have developed a prototype architecture and presented experimental results to evaluate our design. The prototype architecture can be easily deployed in public or private JointCloud infrastructures for flexible and scalable monitoring. The evaluation results show that our architecture is feasible in terms of performance and scalability.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"168 ","pages":"Article 107773"},"PeriodicalIF":6.2,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143527563","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}

{"title":"Solving online resource-constrained scheduling for follow-up observation in astronomy: A reinforcement learning approach","authors":"Yajie Zhang,&nbsp;Ce Yu,&nbsp;Chao Sun,&nbsp;Jizeng Wei,&nbsp;Junhan Ju,&nbsp;Shanjiang Tang","doi":"10.1016/j.future.2025.107781","DOIUrl":"10.1016/j.future.2025.107781","url":null,"abstract":"<div><div>In the astronomical observation field, determining the allocation of observation resources of the telescope array and planning follow-up observations for targets of opportunity (ToOs) are indispensable components of astronomical scientific discovery. This problem is computationally challenging, given the online observation setting and the abundance of time-varying factors that can affect whether an observation can be conducted. This paper presents <span>ROARS</span>, a reinforcement learning approach for online astronomical resource-constrained scheduling. To capture the structure of the astronomical observation scheduling, we depict every schedule using a directed acyclic graph (DAG), illustrating the dependency of timing between different observation tasks within the schedule. Deep reinforcement learning is used to learn a policy that can improve the feasible solution by iteratively local rewriting until convergence. It can solve the challenge of obtaining a complete solution directly from scratch in astronomical observation scenarios, due to the high computational complexity resulting from numerous spatial and temporal constraints. A simulation environment is developed based on real-world scenarios for experiments, to evaluate the effectiveness of our proposed scheduling approach. The experimental results show that <span>ROARS</span> surpasses 5 popular heuristics, adapts to various observation scenarios and learns effective strategies with hindsight.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"169 ","pages":"Article 107781"},"PeriodicalIF":6.2,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143563730","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}

{"title":"An incomplete three-way consensus algorithm for unmanned aerial vehicle purchase using optimization-driven sentiment analysis","authors":"Chao Zhang ,&nbsp;Yating Wang ,&nbsp;Arun Kumar Sangaiah ,&nbsp;Mohammed J.F. Alenazi ,&nbsp;Majed Aborokbah","doi":"10.1016/j.future.2025.107761","DOIUrl":"10.1016/j.future.2025.107761","url":null,"abstract":"<div><div>As a novel productive force in the low-altitude economy, the unmanned aerial vehicle (UAV) industry has emerged as a crucial engine of the digital economy growth. However, high-dimensional online reviews, incomplete information systems (IISs), and coordination among numerous sellers may influence the purchasing decision for UAVs. To address these challenges, first, the sentiment analysis (SA) of UAV online reviews is conducted using BiLSTM and BiGRU models optimized by the hippopotamus optimization (HO) algorithm. Meanwhile, the K-nearest neighbor (KNN) algorithm that combines the Jensen–shannon (JS) divergence with the Hellinger distance is applied to construct a complete information system (CIS). Second, three-way clustering (TWC) is performed on sellers, followed by the calculation of seller weights and group weights using the full consistency method. Third, to closely align with the behavior of sellers, a two-stage consensus reaching process (CRP) model based on TWC and the dual fine-tuning (DFT) theory is proposed, referred to as TWC-DFT-CCRP. In the first stage, the behavior of sellers is adjusted based on the TWC result. In the second stage, optimization-based rules are used to reduce the conflict degree among sellers to reach consensus. Fourth, integrating the TWD process with prospect regret theory (P-RT) can reduce potential decision risks and identify the optimal solution. Finally, the model’s feasibility is demonstrated via a case study of UAV online reviews. In summary, the method not only addresses the challenge of handling high-dimensional data but also optimizes large-scale group decision-making (LSGDM), thereby providing effective decision support for purchasing UAVs.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"168 ","pages":"Article 107761"},"PeriodicalIF":6.2,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143474372","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}

{"title":"Multigrain: Adaptive multilevel hot data identifier with a stack distance-based prefilter","authors":"Hyerim Lee ,&nbsp;Dongchul Park","doi":"10.1016/j.future.2025.107762","DOIUrl":"10.1016/j.future.2025.107762","url":null,"abstract":"<div><div>Many computer system applications, such as data caching and Not AND (NAND) flash memory-based storage systems, employ a hot data identification scheme. However, regardless of the workload characteristics, most existing studies have adopted only a fine-grained (i.e., block-level) hot data decision policy, causing high computational overhead and error rates. Different workloads mandate different treatments to achieve effective hot data identification. Based on our comprehensive workload studies, this paper proposes Multigrain, an <em>adaptive multilevel</em> hot data identification scheme that dynamically selects a coarse-grained (i.e., subrequest-level) policy or coarser-grained (i.e., request-level) policy based on the workload. The proposed Multigrain employs multiple effective bloom filters to capture frequency and recency information. Moreover, it adopts a simple and smart <em>prefilter mechanism</em> leveraging workload stack distance information. To our knowledge, the proposed scheme is the <em>first multilevel coarse-grained hot data identification scheme</em> that judiciously selects an optimal hot data decision granularity to achieve effective and accurate identification. Our extensive experiments with many realistic workloads demonstrate that our adaptive multilevel scheme significantly reduces the execution time (by an average of up to 6.9<span><math><mo>×</mo></math></span>) and error rate (by an average of up to 2.27<span><math><mo>×</mo></math></span>) using the effective coarse-grained policies and a prefiltering mechanism.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"167 ","pages":"Article 107762"},"PeriodicalIF":6.2,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143465148","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ESDI: An efficient and secure data integrity verification scheme for indoor navigation","authors":"Mingxi Liu ,&nbsp;Tailong Yang ,&nbsp;Wenbo Shi ,&nbsp;Athanasios V. Vasilakos ,&nbsp;Ning Lu","doi":"10.1016/j.future.2025.107759","DOIUrl":"10.1016/j.future.2025.107759","url":null,"abstract":"<div><div>Currently, indoor navigation software is typically embedded in smartphones as mobile applications. These apps enable users to access cloud-based data while retrieving indoor navigation information. However, cloud data faces risks of tampering and deletion, necessitating verification of its integrity by users. While smartphones possess certain computational capabilities, prolonged execution of computationally intensive tasks can lead to rapid battery depletion. Additionally, excessive storage demands may prompt users to frequently close or even uninstall the apps to free up memory. This paper presents a blockchain-based data integrity verification scheme tailored for indoor navigation. To address storage overhead, we introduce a user-frequency-based selection technique that designates certain blockchain nodes as light nodes. We further propose a Merkle Hash Tree-based proof extraction method to facilitate efficient proof transfer between different types of nodes. Our approach incorporates an efficient Zhang-Safavi-Susilo (ZSS) signature-based data auditing protocol. By leveraging a data label placement mechanism during signature generation, our scheme supports tamper-proof batch verification, significantly reducing computational overhead. To enable dynamic data updates, we design a novel dynamic data structure, the Red-Black Hash Table, which enhances efficiency in handling updates. Through rigorous security analysis, we demonstrate that our scheme effectively defends against forgery, replay, and replacement attacks. We implemented and simulated our solution on smartphones and indoor navigation apps, conducting experimental evaluations using indoor positioning data. We take audit initialization overhead, audit verification computation overhead, evidence storage overhead, consensus computation overhead, etc. as important experimental indicators. Performance results indicate that our scheme, Efficient and Secure Data Integrity (ESDI), improves auditing efficiency by approximately 54% on average compared to existing approaches.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"168 ","pages":"Article 107759"},"PeriodicalIF":6.2,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143519546","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}

{"title":"Energy-aware scheduling for reliability-oriented real-time parallel applications allocation on heterogeneous computing systems","authors":"Rui She ,&nbsp;Yuting Wu ,&nbsp;Enfang Cui","doi":"10.1016/j.future.2025.107738","DOIUrl":"10.1016/j.future.2025.107738","url":null,"abstract":"<div><div>Heterogeneous computing systems (HCSs) have rapidly developed and been widely applied due to their high performance and low cost characteristics. However, HCSs face trade-offs and conflicts among the three core indicators: energy consumption, reliability, and scheduling length. How to balance the three core indicators to achieve optimal performance is the core issue faced by HCSs. In this paper, we propose an energy-aware scheduling model for reliability-oriented real-time parallel applications on heterogeneous computing systems. The problem of minimum system-centric energy efficiency problem is studied. In terms of problem solving, minimum schedule time length (MSTL) algorithm is proposed, which provides a baseline for assessing feasibility and ensuring compliance with both response time and reliability criteria. To further enhance reliability, this paper considers both transient faults and permanent faults, and proposes the primary–secondary backup (PSB) algorithm to improve the fault tolerance, with dynamic power management (DPM) and dynamic voltage and frequency scaling (DVFS) to reduce energy consumption. Furthermore, the dynamic voltage and frequency scaling (DVFS) algorithm is proposed, within the deadline, redistributing tasks that have not been executed on failed processors to reduce energy consumption caused by excessively long redundant backups. Extensive experimental results on real-world and randomly generated applications demonstrate the effectiveness of the proposed algorithms under various conditions.</div></div>","PeriodicalId":55132,"journal":{"name":"Future Generation Computer Systems-The International Journal of Escience","volume":"168 ","pages":"Article 107738"},"PeriodicalIF":6.2,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480287","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}