IEEE Transactions on Services Computing最新文献

{"title":"A Novel Approach to Temporal QoS Estimation via Extended Kalman Filter-Incorporated Latent Feature Analysis","authors":"Ye Yuan;Song Wang;Hongxun Zhou;Ling Wang;Xin Luo","doi":"10.1109/TSC.2026.3697552","DOIUrl":"10.1109/TSC.2026.3697552","url":null,"abstract":"Predicting temporal Quality of Service (QoS) data is critical for optimizing network services and rationalizing resource allocation in cloud computing and service-oriented systems. Existing mainstream methods have achieved promising predictive performance. However, their purely data-driven manner limits their ability to capture non-stationary temporal patterns, thereby leading to accuracy degradation when temporal QoS data exhibits fluctuations. To tackle this limitation, we propose a novel <underline>E</u>xtended <underline>K</u>alman Filter-Enhanced <underline>L</u>atent Feature Analysis (EKL) model to perform efficient and accurate temporal QoS prediction from the perspective of bidirectional model–data–driven. Its main idea is three-fold: a) designing a model-driven feature producer to obtain the temporal latent features to capture the intricate temporal pattern following the principle of an Extended Kalman Filter; b) building a data-driven feature producer based on the alternating least squares algorithm to identify time-invariant latent features describing intrinsic user-service characteristics; c) exploiting a density-oriented parallel strategy that achieves workload balancing by sorting users in accordance with their service invocation density, which effectively elevates computational efficiency. In addition, we provide a rigorous theoretical analysis to formally prove the convergence of the proposed EKL. Experimental evaluations conducted on real-world temporal QoS datasets reveal that our proposed EKL surpasses existing state-of-the-art models with respect to both computational efficiency and prediction accuracy for missing temporal QoS data.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"2634-2645"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148288582","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":"A Spatio-Temporal Graph Neural Network for Process Remaining Time Prediction","authors":"Xinzheng Cui;Miao Liu;Lichun Li;Wenbin Chen","doi":"10.1109/TSC.2026.3706241","DOIUrl":"10.1109/TSC.2026.3706241","url":null,"abstract":"Predictive process monitoring aims to anticipate the future behavior of ongoing business processes by leveraging historical execution data and real-time event streams. Among its core tasks, remaining time prediction plays a crucial role in enabling proactive decision-making and efficient resource allocation. In this paper, we propose STG-TempNet, a spatio-temporal graph neural network designed to predict the remaining time of process instances. The model integrates three key components: (1) a GINE-based spatial feature extraction module that captures complex control-flow dependencies among activities; (2) a GRU-based short-term dependency module that models local temporal dynamics; and (3) a Transformer-based long-term dependency module that learns global temporal correlations across event sequences. Event logs are transformed into sequences of process instance graphs enriched with structural, temporal, and resource-related attributes. Extensive experiments on four real-world datasets (Sepsis, BPIC2012, BPIC2017, and Traffic Fines) demonstrate that STG-TempNet consistently outperforms state-of-the-art baseline models. It achieves up to a 14.8% improvement over the second-best method and a 32.1% reduction in mean absolute error (MAE) compared with the baseline average, confirming the effectiveness of jointly modeling structural and temporal information for predictive accuracy.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"2660-2673"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148299504","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":"BOTVPA: An SLO-Aware and Efficient Resource Scheduling Method for Microservice","authors":"Xiaoming Ye;Weiwei Lin;Xiaoxuan Luo;Mengheng Li;Qi Mu","doi":"10.1109/TSC.2026.3709898","DOIUrl":"10.1109/TSC.2026.3709898","url":null,"abstract":"Complex microservices topology poses significant challenges to resource scheduling. Consequently, most schedulers rely on extensive offline analysis to prevent potential SLO violations resulting from inaccurate online decisions, leading to high deployment costs. Furthermore, regarding metric selection that is important to bridge the scheduling decision to actual resource allocation, although CPU throttling-based auto-scaling has emerged as a promising approach to capture latency sensitivity, existing methods lack a quantitative mechanism to map throttling signals to resource quota, relying instead on coarse heuristics that potentially waste resources. To address these issues, we propose Bayesian Optimization and Throttling based Vertical Pod Autoscaling (BOTVPA), a sample-efficient scheduling method. BOTVPA features a two-tier architecture: 1) Central Scheduler: We employ Bayesian Optimization (BO) to enable rapid decision-making with minimal samples, eliminating offline training overhead. 2) Local Scheduler: We introduce a novel CPU throttling model that mathematically bridges the gap between throttling rates and CPU allocation. Leveraging this model, the local scheduler dynamically translates BO’s decision into precise CPU resource limits at runtime. Results show that BOTVPA reduces CPU consumption by up to 19.4% over the best-performing baseline while satisfying SLO.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"2716-2729"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148383985","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":"OSGS: A Framework for Online Scheduling of Satellite-Ground Collaborative Inference With Space Edge Computing","authors":"Kongyange Zhao;Yuanming Wang;Zhi Zhou;Ruiting Zhou;Xiaoxi Zhang;Xu Chen;Dechao Ran;Fei Zhang;Lu Cao","doi":"10.1109/TSC.2026.3688483","DOIUrl":"10.1109/TSC.2026.3688483","url":null,"abstract":"With the rapid deployment of camera-equipped nanosatellites in Low Earth Orbit (LEO) for high-resolution Earth observation, traditional data processing pipelines that rely solely on ground stations are struggling to manage the escalating data volume and stringent latency requirements. Space Edge Computing (SEC) has emerged as a transformative solution that enables data processing directly on satellites, alleviating the burden on ground infrastructure and facilitating real-time inference for critical applications. However, SEC brings forth new challenges in online scheduling for satellite-ground collaborative inference, especially when optimizing the balance between resource allocation and inference accuracy across large LEO constellations. This article introduces the Online Satellite-Ground Scheduling (OSGS) framework, a novel approach designed to address these challenges by minimizing both system-wide cost and quality of service (QoS) degradation under resource and latency constraints. OSGS dynamically adjusts the selection of ground stations, configuration, and resource provisioning decisions in response to real-time satellite and ground station status, including the amount of collected data, ground station resource pricing, and satellite-ground connection latency. OSGS employs an innovative combination of primal-dual approximation techniques and lazy-switching method to handle the NP-hard, time-coupled challenges inherent in long-term optimization for satellite-ground collaborative inference. We rigorously prove the approximation ratio of the proposed algorithm and further derive the competitive analysis of the online algorithm framework. Extensive trace-driven simulations verify the effectiveness of the proposed framework, where OSGS demonstrates significant improvements over baseline methods.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"2970-2983"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148290406","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":"PlanTwin: Privacy-Preserving Planning Abstractions for Cloud-Assisted LLM Agents","authors":"Guangsheng Yu;Qin Wang;Rui Lang;Shuai Su;Xu Wang","doi":"10.1109/TSC.2026.3695494","DOIUrl":"10.1109/TSC.2026.3695494","url":null,"abstract":"Cloud-hosted large language models (LLMs) have become the de facto planners in agentic systems, coordinating tools and guiding execution over local environments. In many deployments, however, the environment being planned over is private, containing source code, files, credentials, and metadata that cannot be exposed to the cloud. Existing solutions address adjacent concerns, such as execution isolation, access control, or confidential inference, but they do not control what cloud planners observe during planning: within the permitted scope, <i>raw environment state is still exposed</i>. We introduce <small>PlanTwin</small>, a schema-constrained projection-based architecture for cloud-assisted planning that prevents raw local context from leaving the local boundary. The key idea is to project the real environment into a <i>planning-oriented digital twin:</i> a schema-constrained and de-identified abstract graph that preserves planning-relevant structure while removing reconstructable details. The cloud planner operates solely on this sanitized twin through a bounded capability interface, while a local gatekeeper enforces safety policies and cumulative disclosure budgets. We further formalize the privacy–utility trade-off as a capability granularity problem, define architectural privacy goals using <inline-formula><tex-math>$(k,delta )$</tex-math></inline-formula>-anonymity and <inline-formula><tex-math>$epsilon$</tex-math></inline-formula>-unlinkability, and mitigate compositional leakage through multi-turn disclosure control. We implement <small>PlanTwin</small> as middleware between local agents and cloud planners and evaluate it on 60 agentic tasks across ten domains with four cloud planners. <small>PlanTwin</small> achieves SND <inline-formula><tex-math>$=1.0$</tex-math></inline-formula> against passive-observer adversaries, while maintaining planning quality close to full-context systems: three of four cloud planners achieve PQS <inline-formula><tex-math>$&gt; 0.79$</tex-math></inline-formula>, within <inline-formula><tex-math>$sim$</tex-math></inline-formula>4% of the no-privacy Raw Context baseline; the privacy-hardening pipeline stages add less than 2.2 percentage points of further PQS variation. Residual identifiability under stronger structural-fingerprint adversaries persists and is bounded by deployment-side controls rather than architecturally eliminated.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"2984-2997"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148290573","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":"STAR: A Sequential Transformer-Based Adaptive Reinforcement Learning Framework for Service Recommendation","authors":"Dunlei Rong;Mingyi Liu;Yang Zhang;Quan Z. Sheng;Lina Yao;Zhongjie Wang;Xiaofei Xu","doi":"10.1109/TSC.2026.3704136","DOIUrl":"10.1109/TSC.2026.3704136","url":null,"abstract":"Service requirements in service computing often exhibit semantic decomposability, where complex demands can be fulfilled progressively through multiple functional aspects. Motivated by this observation, We propose STAR, a Sequential Transformer-based Adaptive Reinforcement Learning framework. STAR adopts a fulfillment-oriented modeling paradigm that reformulates service recommendation as a progressive demand-satisfaction process rather than a one-shot matching task. It integrates graph-based service representation learning with a residual-requirement updating mechanism to capture both general requirement-related semantics and the dynamically preserved unmet semantics in complex service demands. To further support stable sequential recommendation, STAR employs a two-stage optimization strategy with warm-up, feedback-driven refinement, and retained ground-truth supervision to mitigate requirement drift caused by progressive demand updating. Experiments on the ProgrammableWeb dataset show that STAR outperforms competitive baselines across multiple evaluation metrics, demonstrating its effectiveness.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"3099-3111"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148288510","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":"AgenticOTA: An Agentic AI Pipeline for Context-Aware Over-the-Air Update Deployments in Connected Autonomous Vehicles","authors":"Shikhar Vashistha;Neetesh Kumar","doi":"10.1109/TSC.2026.3700604","DOIUrl":"10.1109/TSC.2026.3700604","url":null,"abstract":"Connected Autonomous Vehicles (CAVs) rely on software code for functioning, which requires continuous maintenance and real-time over-the-air (OTA) updates to ensure optimal performance in dynamically changing environments. However, current OTA update systems operate on pre-defined schedules or reactive diagnostics, failing to adapt deployment strategies based on contextual insights such as subsystem performance degradation and user demands. Furthermore, existing approaches lack support for on-demand decision-making across heterogeneous CAV stacks, treating vehicle software as a monolith and limiting fine-grained interventions at individual functional modules. To address these limitations, this work presents AgenticOTA, a novel agentic AI pipeline for context-aware OTA update deployments in CAVs. To the best of the authors’ knowledge, this is the first method that enables context-aware, fine-grained, and autonomous OTA deployments in CAVs. AgenticOTA utilizes LLM-driven agent integration to autonomously select validated container images and determine deployment viability based on real-time contextual information. The pipeline implements a modular, fine-grained software management layer that enables targeted updates and the selective replacement of individual functional modules in the event of failures, through containerized microservices, without impacting the entire software system. The effectiveness of AgenticOTA is demonstrated through comprehensive experiments against the state-of-the-art baselines. Experimental results demonstrate significant improvements, including a higher update success rate with fewer failures and rollbacks, as well as an orchestration initiation latency of 0.008 seconds, with a full end-to-end latency of 2.34 s, across baseline methods.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"2688-2701"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148288547","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":"SecDiv: Privacy-Preserving Diversity-Constrained Top-$k$k Query Processing in the Cloud","authors":"Yinxing Zhang;Guang Tang;Qingwang Wang;Songlei Wang;Zhiquan Liu;Zhongyun Hua","doi":"10.1109/TSC.2026.3700282","DOIUrl":"10.1109/TSC.2026.3700282","url":null,"abstract":"With the proliferation of cloud computing, outsourcing databases has become a common strategy for reducing on premise storage and computation costs. However, this approach raises serious privacy concerns, as sensitive data and query information may be exposed to the cloud. While existing top-<inline-formula><tex-math>$k$</tex-math></inline-formula> query methods have made progress in performance and privacy protection, they offer limited support for the more advanced requirement of diversity-constrained queries. In light of this, we present SecDiv, the first privacy-preserving query system that supports diversity constraints over ciphertext in the cloud. SecDiv is built on a two-server distributed trust model and lightweight additive secret sharing, and hides data contents under an honest-but-curious, non-colluding adversary model to ensure that cloud servers learn no sensitive information. SecDiv comprises three customized secure components: SecDMap maps the structured database of the data owner into two secret-shared tables; SecQMap translates each SQL statement and its diversity constraints into vectors whose lengths match the database attributes, thereby hiding targeted attributes and literal values; and SecCQ performs secure filtering, ordering, and top-<inline-formula><tex-math>$k$</tex-math></inline-formula> selection in the cloud, centered on a secure most significant bit comparison implemented by a parallel-prefix adder. A formal security analysis is conducted to provide theoretical guarantees for the security of SecDiv. SecDiv is evaluated on three real datasets, with diversity constraints configured using top-<inline-formula><tex-math>$k$</tex-math></inline-formula> and category count conditions to emulate practical ranking scenarios. Compared with a plaintext baseline, SecDiv achieves identical results with 100% accuracy. Query latency remains practical, with second-level response times in typical settings, and communication overhead increases as expected. Overall, experimental results demonstrate that SecDiv attains a balanced trade-off among privacy, accuracy, and efficiency in real-world cloud service environments.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"3029-3041"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148288549","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":"Online Multi-Task Business Process Prediction Using Dynamic Representation","authors":"Xiwei Zhang;Xianwen Fang;Wei Bao;Cong Liu","doi":"10.1109/TSC.2026.3706819","DOIUrl":"10.1109/TSC.2026.3706819","url":null,"abstract":"Existing Predictive Process Monitoring (PPM) methods typically rely on static offline methods, limiting their ability to adapt to dynamic process evolution driven by emerging activities and shifting behavioral patterns. This constraint is particularly critical for multi-task prediction, such as the simultaneous forecasting of the next activity and the remaining process time. To address this challenge, we propose an online multi-task prediction framework based on dynamic graph representations. The framework enables a Graph Neural Network (GNN) to incrementally learn newly emerging activities by leveraging dynamic graph snapshots and an architecture expansion strategy. For efficient online adaptation, the framework incorporates two update strategies, a standard periodic update and a drift-aware adaptive update triggered by the Maximum Mean Discrepancy (MMD<bold><inline-formula><tex-math>${^{2}}$</tex-math><alternatives><mml:math><mml:msup><mml:mrow/><mml:mn>2</mml:mn></mml:msup></mml:math><inline-graphic></alternatives></inline-formula></b>) between subgraph embeddings. Both strategies are integrated with a Prioritized Experience Replay (PER) mechanism, augmented with a rarity-aware bonus, to ensure rapid and robust model adjustments in non-stationary environments. Comprehensive experiments on multiple real-world event logs demonstrate that our framework, when combined with different GNN backbones such as GCN, GAT, and GIN, significantly outperforms state-of-the-art baselines in both next-activity and remaining-time prediction. Notably, under concept drift, the proposed drift-aware strategy exhibits strong adaptability, highlighting the framework’s effectiveness and potential for addressing complex online process prediction challenges.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"2956-2969"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148314112","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":"SpaceMeet: Bringing Conferencing Closer Through In-Orbit Conferencing Services","authors":"Songshi Dou;Feihu Jin;Jinxian Wu;A-Long Jin;Kwan L. Yeung","doi":"10.1109/TSC.2026.3707012","DOIUrl":"10.1109/TSC.2026.3707012","url":null,"abstract":"Real-time conferencing services have significantly reshaped global communication. These services typically require low-latency network conditions, which have traditionally been supported by high-speed terrestrial networks. However, emerging Low Earth Orbit (LEO) satellite constellations offer opportunities to provide Internet access to remote or underdeveloped regions that lack terrestrial connectivity. Despite this potential, users in these areas may still experience high latency issues due to rigid routing patterns and insufficient terrestrial facility deployment. Existing works either optimize terrestrial-based media server allocation without being tailored for these satellite-based Internet users, or rely on satellite-cloud cooperative relays but still depend on terrestrial facilities. Consequently, neither fundamentally resolves the inherent high-latency issue, especially for users without reliable terrestrial connectivity. To fill the research gap, we propose <sc>SpaceMeet</small> to leverage LEO satellites as direct hosts for conferencing sessions, thereby minimizing latency by avoiding terrestrial detours. Considering the dynamic energy availability caused by varying satellite idle periods and illumination conditions, we formulate a satellite-based conferencing service assignment problem to strategically allocate satellites for session hosting. Since solving this problem for real-time decision-making is infeasible due to its high complexity, we introduce a heuristic algorithm enhanced through iterative refinement using large language models. Extensive evaluations utilizing realistic LEO satellite constellation configurations demonstrate that <sc>SpaceMeet</small> significantly reduces latency by up to 7.78% compared to the baseline while efficiently utilizing idle satellite energy resources.","PeriodicalId":13255,"journal":{"name":"IEEE Transactions on Services Computing","volume":"19 4","pages":"3085-3098"},"PeriodicalIF":6.2,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148314113","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}