IEEE Transactions on Sustainable Computing最新文献

{"title":"Multi-Task Deep Learning Framework for Anomaly Detection and Carbon Emission Prediction in Precision Agriculture","authors":"T. C. Jermin Jeaunita;T. Ramesh;A. V. Kalpana;P. T. Shanthala","doi":"10.1109/TSUSC.2025.3619577","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3619577","url":null,"abstract":"Precision agriculture faces persistent challenges, including unpredictable weather, fluctuating carbon emissions, noisy sensor data, and security vulnerabilities, which hinder the optimization of crop yield and environmental sustainability. This study proposes a multi-task deep learning framework for precision agriculture that targets carbon emission prediction and anomaly detection using sensor and drone imagery data. Initially, the data is acquired using a testbed and stored in Interplanetary File System storage for secure transmission. The sensor-based statistical features were extracted via CoStat-Principal Component Analysis, and spatial features for drone images through the Visual Geometry Group 16 to extract vital patterns. These features are selected using Mutual Chi-squared Recursive Feature Elimination for feature refinement. Additionally, a feature fusion approach combines sensor and image data for a holistic analysis. Finally, anomaly patterns are identified through a deep autoencoder, while a feedforward neural network predicts the average carbon emission. This classified output integrates explainable artificial intelligence to provide transparent decision-making. A blockchain-based decentralized alert system is used for real-time anomaly detection. The proposed system achieves an accuracy of 99.20% in carbon emissions forecasting, 99.73% in anomaly detection, and a blockchain-triggered alerts rate of 99%, providing sustainable farming practices and a secure solution for precision agriculture.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 6","pages":"1337-1350"},"PeriodicalIF":5.3,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148479775","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"EMC-LSP: A Novel Lightweight Architecture for Edge Multi-Node Long Sequence Prediction","authors":"Chuanyue Xiong;Jing Yang;Jiahao Zhong;Zirui He;Pu Pang;Minyi Guo","doi":"10.1109/TSUSC.2025.3620441","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3620441","url":null,"abstract":"Edge device traffic prediction is crucial for autonomous network control and management. However, the rapid proliferation of smart 5G networks results in increasingly heterogeneous, dynamic, and complex traffic loads on edge nodes, rendering traditional short-term prediction methods insufficient for medium- and long-term network resource scheduling. To address this, we propose a novel multi-node lightweight long-sequence deep learning-based prediction architecture (EMC-LSP) to effectively capture complex long- and short-term correlations in edge environments. Specifically, EMC-LSP employs frequency-domain hard-attention decomposition to separately model non-stationary high-frequency and low-frequency traffic, utilizes two-layer null frequency-domain convolution for long-term low-frequency similarity, and designs a high-frequency interpolation-based prediction method. In extensive tests on 18 datasets, EMC-LSP demonstrated superior performance, reducing the average prediction error of MSE and MAE by 15.20% while decreasing model parameters by 50 times.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 6","pages":"1351-1365"},"PeriodicalIF":5.3,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148480795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"Linguistically-Informed Dataset Curation for Efficient LLM Fine-Tuning: Balancing Performance and Efficiency","authors":"Sultan Alshamrani","doi":"10.1109/TSUSC.2025.3605796","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3605796","url":null,"abstract":"The rapid growth of AI systems, particularly large language models, has raised significant concerns about their environmental impact due to excessive energy consumption and carbon emissions. Despite these concerns, the trend in AI development continues to prioritize performance gains through increasingly resource-intensive approaches, such as utilizing more powerful hardware and larger datasets, often at the expense of efficiency. This study contributes to the efforts towards more efficient AI by proposing and empirically evaluating two dataset curation strategies, DS-1 and DS-2, which are linguistically-informed by syntactic features like Part-of-Speech (POS) tags to prune lexical content, for fine-tuning LLMs. The focus of this work is the empirical demonstration of how these linguistically-motivated curation approaches can create a balance between computational efficiency gains and performance maintenance during the LLM fine-tuning process. This approach represents a promising step in addressing the critical issue of AI’s environmental impact. The evaluation is conducted on sentiment analysis tasks, serving as a focused case study. We evaluate two curation approaches, DS-1 and DS-2, applied to sentiment analysis tasks across three datasets, achieving substantial dataset size reductions while preserving essential linguistic information. Our assessment of six state-of-the-art LLMs—RoBERTa, ALBERT, ERNIE, DeBERTa, BERT, and GPT-2—on both curated and original datasets reveals that curated datasets can yield comparable performance to uncurated ones, with efficiency gains of up to 70% in tokenization time and up to 80% in training energy. Notably, the study uncovers varying resilience of different model architectures to curation, with the GPT-based model demonstrating perfect performance adaptation, while other advanced models like ERNIE and DeBERTa also showed strong resilience. This research is crucial in promoting the development of more sustainable and resource-efficient NLP systems, challenging the prevailing notion that larger datasets invariably lead to better performance in fine-tuning LLMs, and paving the way for more environmentally conscious AI development practices. The generalizability of these specific curation strategies to other NLP tasks is an avenue for future investigation.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 6","pages":"1267-1277"},"PeriodicalIF":5.3,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148480800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"FedFusionQuant (FFQ): Federated Learning With Feature Fusion and Model Quantisation for Human Activity Recognition Using CSI","authors":"Ahsan Raza Khan;Rao Naveed Bin Rais;Sarmad Sohaib;Sajjad Hussain;Ahmed Zoha","doi":"10.1109/TSUSC.2025.3627484","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3627484","url":null,"abstract":"Human Activity Recognition (HAR) using Channel State Information (CSI) enables energy-efficient and non-invasive healthcare monitoring. However, conventional HAR systems rely on centralised model training, which requires the sharing of raw data, leading to privacy risks, excessive bandwidth usage, and high communication latency that limit scalability. This paper proposes <bold>FedFusionQuant (FFQ)</b>, a federated learning (FL) framework that jointly performs feature fusion, adaptive aggregation, and quantisation-aware compression during training. A novel <bold>federated distance (FedDist)</b> mechanism dynamically adjusts parameter updates using neuron dissimilarity metrics, enhancing generalisation across heterogeneous clients. Meanwhile, <bold>quantisation-aware training (QAT)</b> reduces model size and transmission cost while preserving accuracy. Extensive experiments on real CSI data from 30 participants demonstrate that FFQ improves multi-class HAR accuracy by <bold>4.29%</b> and binary fall detection by <bold>5.55%</b> compared to raw fusion models. Furthermore, model compression with QAT achieves a <bold>47% reduction in communication overhead</b> while maintaining accuracy comparable to state-of-the-art methods.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 6","pages":"1421-1434"},"PeriodicalIF":3.9,"publicationDate":"2025-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145712547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"OSPDP: One-Sided Personalized Differential Privacy","authors":"Jiajun Chen;Chunqiang Hu;Huijun Zhuang;Ruifeng Zhao;Jiguo Yu","doi":"10.1109/TSUSC.2025.3626773","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3626773","url":null,"abstract":"Differential privacy has received considerable attention as a privacy concept for releasing statistical information from datasets. While differential privacy provides strict statistical guarantees, it is equally crucial to investigate how these guarantees interact with individual privacy preferences and privacy policies. Existing solutions, such as one-sided differential privacy, treat all sensitive records equally in terms of privacy protection, although datasets can be classified based on predetermined privacy policies that differentiate between sensitive and insensitive records. In this paper, we present a novel concept of privacy termed One-sided Personalized Differential Privacy (OSPDP), offering verifiable privacy assurances at the user level for sensitive records derived from privacy policies. Specifically, OSPDP enables data owners to articulate their privacy needs more flexibly, avoiding a one-size-fits-all approach to privacy protection and potentially establishing a dichotomous privacy policy regarding the sensitivity of records. Furthermore, the truthful release or legitimate disclosure of non-sensitive records reduces unnecessary privacy consumption and can be utilized to significantly enhance data utility. Additionally, we present several well-performing mechanisms for achieving OSPDP. Finally, we evaluate and analyze the trade-off between privacy and utility of the proposed mechanisms through extensive experiments.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 6","pages":"1435-1448"},"PeriodicalIF":3.9,"publicationDate":"2025-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145712554","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"FedUP: Federated Unlearning With Prototypes","authors":"Yuhong Huang;Xue Li;Songle Chen;Siguang Chen","doi":"10.1109/TSUSC.2025.3612138","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3612138","url":null,"abstract":"As an extension of machine unlearning in distributed scenarios, federated unlearning gains significant attention. However, federated unlearning remains challenging, as many studies require additional resources, such as auxiliary dataset or storage, to achieve high-quality models. These requirements incur extra costs and are often difficult to satisfy in practical applications. To address these issues, we propose a flexible client-level federated unlearning algorithm with prototypes, called FedUP. Specifically, our algorithm consists of two components: prototype-based unlearning and model recovering. First, we design a prototype-based unlearning strategy that uses prototypes of the erased client to guide the unlearning process, and maximizes the prototype loss between the remaining and erased clients to unlearn the information. It does not rely on historical storage updates or additional standard datasets, making the unlearning process more streamlined. To mitigate performance degradation from the unlearning process, we develop a brief model recovering approach guided by global prototypes to swiftly and efficiently restore models’ accuracy on the remaining datasets. Unlike other unlearning algorithms, our approach exchanges prototypes instead of model parameters, significantly reducing communication overhead. Finally, we empirically evaluate the proposed algorithm from multiple perspectives on two datasets, demonstrating that our algorithm can achieve high-quality unlearned models with minimal communication cost.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 6","pages":"1461-1467"},"PeriodicalIF":3.9,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145712546","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"Unsupervised Multi-Target Cross-Service Log Anomaly Detection","authors":"Shiming He;Rui Liu;Bowen Chen;Kun Xie;Jigang Wen","doi":"10.1109/TSUSC.2025.3578517","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3578517","url":null,"abstract":"Log analysis, especially log anomaly detection, can help debug systems and analyze root causes to provide reliable services. Deep learning is a promising technology for log anomaly detection. However, deep learning methods need a large amount of training data, which is hard for a newly deployed system to collect sufficient logs. Transfer learning becomes a possible method to solve the problem that can apply the knowledge from a long-term deployed system (source) to a newly deployed system (target). Existing transfer learning methods focus on transferring the knowledge from a source system to a single target system within the same service, in which the source and the target belong to the same service (e.g. operating system, supercomputer, or distributed system). They achieve low performance when applied to multiple target and different services systems because of the obvious differences in log format, syntax, semantics, and component call between different services and the individual training of multiple models for each target system. To tackle the problems, we propose an unsupervised multi-target cross-service log anomaly detection method based on transfer learning and contrastive learning (LogMTC). LogMTC exploits contrastive learning to learn a single model on combined data from the source and multiple target systems, which can fit multiple target systems simultaneously and improve efficiency. LogMTC exploits a hypersphere loss and two contrastive losses to minimize the feature differences crossing different services. Our experiments on two services (supercomputer and distributed system) and three log datasets show that our method is superior to the existing transfer learning methods in the same service, cross-service, and multi-target log anomaly detection. Compared with the best peer accurate transfer learning algorithm LogTAD, LogMTC improves 1.14%–8.28<inline-formula><tex-math>$%$</tex-math></inline-formula> F1 score in multi-target transfer and is 1.12–1.22 times faster.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 5","pages":"1056-1069"},"PeriodicalIF":3.9,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145248086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 Efficient Scheduling Approach for Target Coverage in Solar Powered Internet of Things","authors":"Dipak Kumar Sah;Abhishek Hazra;Nabajyoti Mazumdar;Annavarapu Chandra Sekhara Rao;Tarachand Amgoth","doi":"10.1109/TSUSC.2025.3578433","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3578433","url":null,"abstract":"The Internet of Things (IoT) has been increasingly applied in various applications in recent years. In IoT, many tasks are performed for a load operation, such as creating a cluster, preserving convergence/connectivity issues, etc. However, the energy consumption rate is also high due to more traffic in dense networks. Generally, a traditional IoT node’s battery power capacity is limited due to a short-range cycle. To address the energy shortage problem, researchers have tackled it through the Solar Powered (SP) energy harvesting technique. This method provides abundant energy to the IoT nodes at a lower cost. One issue arises from the target coverage area, which requires that each target must have at least one node for continuous monitoring in a given area. To address these issues, we have designed an effective solution called the Efficient Scheduling Target Coverage (ESTC) algorithm. This approach consists of various cover sets that work in an interleaving way. If only some node sets need to be active to satisfy coverage constraints, then there is no need to activate all sets simultaneously. ESTC provides robust coverage awareness with a perpetual network lifetime using scheduling techniques. Furthermore, the proposed work also promotes a green IoT network.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 5","pages":"1043-1055"},"PeriodicalIF":3.9,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145248062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"GreeNX: An Energy-Efficient and Sustainable Approach to Sparse Graph Convolution Networks Accelerators Using DVFS","authors":"Siqin Liu;Prakash Chand Kuve;Avinash Karanth","doi":"10.1109/TSUSC.2025.3577218","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3577218","url":null,"abstract":"Graph convolutional networks (GCNs) have emerged as an effective approach to extend deep learning algorithms for graph-based data analytics. However, GCNs implementation over large, sparse datasets presents challenges due to irregular computation and dataflow patterns. Specialized GCN accelerators have emerged to deliver superior performance over generic processors. However, prior techniques that include specialized datapaths, optimized sparse computation, and memory access patterns, handle different phases of GCNs differently which results in excess energy consumption and reduced throughput due to sub-optimal dataflows. In this paper, we propose GreeNX, a computation and communication-aware GCN accelerator that uniformly applies three complementary techniques to all phases of GCN. First, we abstract two cascaded sparse-dense matrix multiplications that uniformly process the computation in both aggregation and combination phases of GCNs to improve throughput. Second, to mitigate the overheads of processing irregular sparse data, we develop a dynamic-voltage-and-frequency-scaling (DVFS) scheme by grouping a row of processing elements (PEs) that dynamically changes the applied voltage/frequency (V/F) to improve energy efficiency. Third, we conduct a comprehensive carbon footprint evaluation, analyzing both embodied and operational emissions for GCNs. Extensive simulation and experiments validate that our GreeNX consistently reduces memory accesses and energy consumption leading to an average 7.3× speedup and 5.6× energy savings on six real-world graph datasets over several state-of-the-art GCN accelerators including HyGCN, AWB-GCN, GCoD, GRIP, IGCN, and LW-GCN.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 5","pages":"1031-1042"},"PeriodicalIF":3.9,"publicationDate":"2025-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145248071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"HS-GCN: A High-Performance, Sustainable, and Scalable Chiplet-Based Accelerator for Graph Convolutional Network Inference","authors":"Yingnan Zhao;Ke Wang;Ahmed Louri","doi":"10.1109/TSUSC.2025.3575285","DOIUrl":"https://doi.org/10.1109/TSUSC.2025.3575285","url":null,"abstract":"Graph Convolutional Networks (GCNs) have been proposed to extend machine learning techniques for graph-related applications. A typical GCN model consists of multiple layers, each including an aggregation phase, which is communication-intensive, and a combination phase, which is computation-intensive. As the size of real-world graphs increases exponentially, current customized accelerators face challenges in efficiently performing GCN inference due to limited on-chip buffers and other hardware resources for both data computation and communication, which degrades performance and energy efficiency. Additionally, scaling current monolithic designs to address the aforementioned challenges will introduce significant cost-effectiveness issues in terms of power, area, and yield. To this end, we propose HS-GCN, a high-performance, sustainable, and scalable chiplet-based accelerator for GCN inference with much-improved energy efficiency. Specifically, HS-GCN integrates multiple reconfigurable chiplets, each of which can be configured to perform the main computations of either the aggregation phase or the combination phase, including Sparse-dense matrix multiplication (SpMM) and General matrix-matrix multiplication (GeMM). HS-GCN implements an active interposer with a flexible interconnection fabric to connect chiplets and other hardware components for efficient data communication. Additionally, HS-GCN introduces two system-level control algorithms that dynamically determine the computation order and corresponding dataflow based on the input graphs and GCN models. These selections are used to further configure the chiplet array and interconnection fabric for much-improved performance and energy efficiency. Evaluation results using real-world graphs demonstrate that HS-GCN achieves significant speedups of 26.7×, 11.2×, 3.9×, 4.7×, 3.1×, along with substantial memory access savings of 94%, 89%, 64%, 85%, 54%, and energy savings of 87%, 84%, 49%, 78%, 41% on average, as compared to HyGCN, AWB-GCN, GCNAX, I-GCN, and SGCN, respectively.","PeriodicalId":13268,"journal":{"name":"IEEE Transactions on Sustainable Computing","volume":"10 5","pages":"1019-1030"},"PeriodicalIF":3.9,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145256014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}