Pub Date : 2025-07-04DOI: 10.26599/TST.2023.9010097
Shaolin Zhu;Dong Jian;Deyi Xiong
Recent research has shown a burgeoning interest in exploring sparse models for massively Multilingual Neural Machine Translation (MNMT). In this paper, we present a comprehensive survey of this emerging topic. Massively MNMT, when based on sparse models, offers significant improvements in parameter efficiency and reduces interference compared to its dense model counterparts. Various methods have been proposed to leverage sparse models for enhancing translation quality. However, the lack of a thorough survey has hindered the identification and further investigation of the most promising approaches. To address this gap, we provide an exhaustive examination of the current research landscape in massively MNMT, with a special emphasis on sparse models. Initially, we categorize the various sparse model-based approaches into distinct classifications. We then delve into each category in detail, elucidating their fundamental modeling principles, core issues, and the challenges they face. Wherever possible, we conduct comparative analyses to assess the strengths and weaknesses of different methodologies. Moreover, we explore potential future research avenues for MNMT based on sparse models. This survey serves as a valuable resource for both newcomers and established experts in the field of MNMT, particularly those interested in sparse model applications.
{"title":"A Survey of Multilingual Neural Machine Translation Based on Sparse Models","authors":"Shaolin Zhu;Dong Jian;Deyi Xiong","doi":"10.26599/TST.2023.9010097","DOIUrl":"https://doi.org/10.26599/TST.2023.9010097","url":null,"abstract":"Recent research has shown a burgeoning interest in exploring sparse models for massively Multilingual Neural Machine Translation (MNMT). In this paper, we present a comprehensive survey of this emerging topic. Massively MNMT, when based on sparse models, offers significant improvements in parameter efficiency and reduces interference compared to its dense model counterparts. Various methods have been proposed to leverage sparse models for enhancing translation quality. However, the lack of a thorough survey has hindered the identification and further investigation of the most promising approaches. To address this gap, we provide an exhaustive examination of the current research landscape in massively MNMT, with a special emphasis on sparse models. Initially, we categorize the various sparse model-based approaches into distinct classifications. We then delve into each category in detail, elucidating their fundamental modeling principles, core issues, and the challenges they face. Wherever possible, we conduct comparative analyses to assess the strengths and weaknesses of different methodologies. Moreover, we explore potential future research avenues for MNMT based on sparse models. This survey serves as a valuable resource for both newcomers and established experts in the field of MNMT, particularly those interested in sparse model applications.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2399-2418"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072061","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557888","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
The rapid growth of marine applications leads to a significant increase in Maritime Devices (MDs). Traditional shore-based maritime communication networks face limitations, such as overloaded and transmission distance to provide network services for MDs. Unmanned Aerial Vehicles (UAVs) act as relays that can expand coverage and enhance the quality of service for offshore communication networks. We consider a multi-UAV-assisted Offshore Internet of Things (mUAV-OloT), and formulate a throughput maximization problem by jointly optimizing channel allocation, Leader MD (LMD) selection, UAV-LMD association, and LMD-MD association. Firstly, we propose the Hypergraph-based Two-Stage Matching (HTSM) algorithm where a Hypergraph-based LMD Selection (HLMDS) strategy is employed to identify the set of LMDs. Secondly, the Kuhn-Munkres algorithm is used to optimize the UAV-LMD association and a Weighted Three-dimensional Hypergraph Matching (WTHM) algorithm is designed to solve the LMD-MD association and channel allocation. Numerical results show that the HTSM algorithm outperforms benchmark algorithms regarding throughput.
{"title":"Throughput Optimization for Multi-UAV-Assisted Offshore Internet of Things: A Hypergraph Approach","authors":"Shuang Qi;Bin Lin;Xu Hu;Chaoyue Zhang;Luyao Zheng;Liping Qian;Yuan Wu","doi":"10.26599/TST.2024.9010087","DOIUrl":"https://doi.org/10.26599/TST.2024.9010087","url":null,"abstract":"The rapid growth of marine applications leads to a significant increase in Maritime Devices (MDs). Traditional shore-based maritime communication networks face limitations, such as overloaded and transmission distance to provide network services for MDs. Unmanned Aerial Vehicles (UAVs) act as relays that can expand coverage and enhance the quality of service for offshore communication networks. We consider a multi-UAV-assisted Offshore Internet of Things (mUAV-OloT), and formulate a throughput maximization problem by jointly optimizing channel allocation, Leader MD (LMD) selection, UAV-LMD association, and LMD-MD association. Firstly, we propose the Hypergraph-based Two-Stage Matching (HTSM) algorithm where a Hypergraph-based LMD Selection (HLMDS) strategy is employed to identify the set of LMDs. Secondly, the Kuhn-Munkres algorithm is used to optimize the UAV-LMD association and a Weighted Three-dimensional Hypergraph Matching (WTHM) algorithm is designed to solve the LMD-MD association and channel allocation. Numerical results show that the HTSM algorithm outperforms benchmark algorithms regarding throughput.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2452-2466"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072113","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Pub Date : 2025-07-04DOI: 10.26599/TST.2024.9010094
Guangshun Li;Tielin Wang;Junhua Wu;Zhiyun Guan
Due to their versatility and ease of movement, Unmanned Aerial Vehicles (UAVs) have become crucial tools in data collection for Wireless Sensor Networks (WSNs). While numerous UAV-based solutions exist, the focus often needs to be on optimizing flight trajectories and managing energy use, sometimes neglecting key factors affecting channel quality. In this article, we introduce a collaborative design framework designed to alleviate channel quality degradation caused by UAV flight distance in three-dimensional spaces. Our approach jointly optimizes UAV power schemes, positions, and flight trajectories. Firstly, we start by introducing a novel enhancing power model developed explicitly for rotary-wing UAVs gathering data, utilizing an alternating optimization method to achieve locally optimal solutions. Next, we frame an optimization problem aimed at maximizing the total average collection rate while achieving approximate optimal position relationships among UAVs. Additionally, we propose a new trajectory optimization model based on the Steiner Minimal Tree (SMT) concept, which is called the Circumcircle Steiner Minimal Tree Problem with Neighborhood (CSMTPN). Finally, we confirm our theoretical insights and numerical outcomes through extensive simulations demonstrating our framework's effectiveness.
{"title":"Co-Design Enhanced Power Scheme and Trajectory Optimization of UAV-Enabled Data Collection from WSNs","authors":"Guangshun Li;Tielin Wang;Junhua Wu;Zhiyun Guan","doi":"10.26599/TST.2024.9010094","DOIUrl":"https://doi.org/10.26599/TST.2024.9010094","url":null,"abstract":"Due to their versatility and ease of movement, Unmanned Aerial Vehicles (UAVs) have become crucial tools in data collection for Wireless Sensor Networks (WSNs). While numerous UAV-based solutions exist, the focus often needs to be on optimizing flight trajectories and managing energy use, sometimes neglecting key factors affecting channel quality. In this article, we introduce a collaborative design framework designed to alleviate channel quality degradation caused by UAV flight distance in three-dimensional spaces. Our approach jointly optimizes UAV power schemes, positions, and flight trajectories. Firstly, we start by introducing a novel enhancing power model developed explicitly for rotary-wing UAVs gathering data, utilizing an alternating optimization method to achieve locally optimal solutions. Next, we frame an optimization problem aimed at maximizing the total average collection rate while achieving approximate optimal position relationships among UAVs. Additionally, we propose a new trajectory optimization model based on the Steiner Minimal Tree (SMT) concept, which is called the Circumcircle Steiner Minimal Tree Problem with Neighborhood (CSMTPN). Finally, we confirm our theoretical insights and numerical outcomes through extensive simulations demonstrating our framework's effectiveness.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2343-2365"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072068","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557581","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Pub Date : 2025-07-04DOI: 10.26599/TST.2025.9010014
Jun Qian;Jinwei Miao;Xiao Sun
Currently, various rapidly developing information technologies are gradually transforming traditional social systems into Complex Social Systems (CSS). On the one hand, individuals' ability to make decisions and access information is increasing, making their behaviors more unpredictable. On the other hand, technology is facilitating an increase in the intensity and scope of individual interactions, with cascade effects making the outcomes of interactions difficult to estimate. To improve the performance of CSS, it is essential to examine the causal laws that determine what kind of performance the system exhibits. However, researches on the causal laws of CSS remain scarce, leading to the lack of foundations for analyzing such systems. Inspired by computational experiments and causal analysis, this paper proposes a Hierarchical Causal Model (HCM) with three layers, each of which presents, extracts, and applies the causality. We apply the proposed model to enhance the system performance in a typical CSS, a software-enabled small-scale plant. Experimental results show that 98.38% of the working days have better system performance than the actual performance after applying our proposed model, and the mean of the median improvement reaches 41.38%. These results validate the proposed model, demonstrating that this work provides a feasible method for the analysis of CSS.
{"title":"Hierarchical Causal Model for Analysis of Complex Social System","authors":"Jun Qian;Jinwei Miao;Xiao Sun","doi":"10.26599/TST.2025.9010014","DOIUrl":"https://doi.org/10.26599/TST.2025.9010014","url":null,"abstract":"Currently, various rapidly developing information technologies are gradually transforming traditional social systems into Complex Social Systems (CSS). On the one hand, individuals' ability to make decisions and access information is increasing, making their behaviors more unpredictable. On the other hand, technology is facilitating an increase in the intensity and scope of individual interactions, with cascade effects making the outcomes of interactions difficult to estimate. To improve the performance of CSS, it is essential to examine the causal laws that determine what kind of performance the system exhibits. However, researches on the causal laws of CSS remain scarce, leading to the lack of foundations for analyzing such systems. Inspired by computational experiments and causal analysis, this paper proposes a Hierarchical Causal Model (HCM) with three layers, each of which presents, extracts, and applies the causality. We apply the proposed model to enhance the system performance in a typical CSS, a software-enabled small-scale plant. Experimental results show that 98.38% of the working days have better system performance than the actual performance after applying our proposed model, and the mean of the median improvement reaches 41.38%. These results validate the proposed model, demonstrating that this work provides a feasible method for the analysis of CSS.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2606-2624"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072060","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Pub Date : 2025-07-04DOI: 10.26599/TST.2024.9010135
Guillermo Kemper;Kevin Guerra-Huamán
This work proposes a computational algorithm to improve the determination of the timing of the respiratory phases. The algorithm was developed using a database of breathing sound signals acquired through properly positioned face masks and electret microphones. Most of the proposed works use the frequency domain and decimation in time to detect the respiratory period and phases, as well as some specific pathology. In this work the processing applied is only in time without applying decimation, thus improving the detection of a greater number of respiratory periods. The segmentation is very important since it allows the isolation of phases of the signal to later detect some pathology or to estimate the volume of inspired and exhaled air. The proposed algorithm involves the extraction of signal envelopes with the use of high selectivity filters without decimation and adaptive normalization processes that aim to achieve an adequate detection. In the validation process, the algorithm detection results were compared with the timing of respiratory periods and phases marked by visual inspection. The results show a maximum error of 4.36% for the respiratory period and 3.23% and 3.09% for the expiration and inspiration times, respectively.
{"title":"An Improved Algorithm for Segmentation and Determination of Respiratory Phase Times Based on Temporal Processing of Nasal and Mouth Breathing Sound Signals","authors":"Guillermo Kemper;Kevin Guerra-Huamán","doi":"10.26599/TST.2024.9010135","DOIUrl":"https://doi.org/10.26599/TST.2024.9010135","url":null,"abstract":"This work proposes a computational algorithm to improve the determination of the timing of the respiratory phases. The algorithm was developed using a database of breathing sound signals acquired through properly positioned face masks and electret microphones. Most of the proposed works use the frequency domain and decimation in time to detect the respiratory period and phases, as well as some specific pathology. In this work the processing applied is only in time without applying decimation, thus improving the detection of a greater number of respiratory periods. The segmentation is very important since it allows the isolation of phases of the signal to later detect some pathology or to estimate the volume of inspired and exhaled air. The proposed algorithm involves the extraction of signal envelopes with the use of high selectivity filters without decimation and adaptive normalization processes that aim to achieve an adequate detection. In the validation process, the algorithm detection results were compared with the timing of respiratory periods and phases marked by visual inspection. The results show a maximum error of 4.36% for the respiratory period and 3.23% and 3.09% for the expiration and inspiration times, respectively.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2522-2543"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072062","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557915","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Knowledge graphs (KGs) have garnered significant attention in recommender systems as auxiliary information. Most existing studies consider an item as an entity of a KG and utilize graph neural networks to learn item representations. However, two challenges exist regarding these algorithms: 1) they provide recommended results but fail to explain the reason for which they are preferred by users; 2) user vector representations are concentrated in a small area, thus resulting in similar mass recommendations. In this study, we focus on learning fine-grained user preferences (LFUP) via user-item interactions and using KGs that can capture the reason for which users interact with items. Additionally, a personalized recommendation task is achieved by optimizing the distribution of users in the vector space. User preferences are modeled by using historical interaction items pertaining to users and important relations within the KG. Subsequently, information from two views is aggregated to reduce the semantic differences between them. Finally, user preferences are personalized by maximizing the spatial distance between various user representations via contrastive learning. Experiments on public datasets prove that LFUP significantly benefits user-preference modeling and personalized recommendations.
{"title":"Learning Fine-Grained User Preference for Personalized Recommendation","authors":"Mingxing Zhang;Xiaoxiong Zhang;Witold Pedrycz;Shuai Wang;Guohua Wu","doi":"10.26599/TST.2024.9010216","DOIUrl":"https://doi.org/10.26599/TST.2024.9010216","url":null,"abstract":"Knowledge graphs (KGs) have garnered significant attention in recommender systems as auxiliary information. Most existing studies consider an item as an entity of a KG and utilize graph neural networks to learn item representations. However, two challenges exist regarding these algorithms: 1) they provide recommended results but fail to explain the reason for which they are preferred by users; 2) user vector representations are concentrated in a small area, thus resulting in similar mass recommendations. In this study, we focus on learning fine-grained user preferences (LFUP) via user-item interactions and using KGs that can capture the reason for which users interact with items. Additionally, a personalized recommendation task is achieved by optimizing the distribution of users in the vector space. User preferences are modeled by using historical interaction items pertaining to users and important relations within the KG. Subsequently, information from two views is aggregated to reduce the semantic differences between them. Finally, user preferences are personalized by maximizing the spatial distance between various user representations via contrastive learning. Experiments on public datasets prove that LFUP significantly benefits user-preference modeling and personalized recommendations.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2544-2556"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072064","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557877","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Pub Date : 2025-07-04DOI: 10.26599/TST.2024.9010177
Linna Zhao;Jianqiang Li;Xi Xu;Chujie Zhu;Wenxiu Cheng;Suqin Liu;Mingming Zhao;Lei Zhang;Jing Zhang;Jian Yin;Jijiang Yang
Myasthenia Gravis (MG) is an autoimmune neuromuscular disease. Given that extraocular muscle manifestations are the initial and primary symptoms in most patients, ocular muscle assessment is regarded necessary early screening tool. To overcome the limitations of the manual clinical method, an intuitive idea is to collect data via imaging devices, followed by analysis or processing using Deep Learning (DL) techniques (particularly image segmentation approaches) to enable automatic MG evaluation. Unfortunately, their clinical applications in this field have not been thoroughly explored. To bridge this gap, our study prospectively establishes a new DL-based system to promote the diagnosis of MG disease, with a complete workflow including facial data acquisition, eye region localization, and ocular structure segmentation. Experimental results demonstrate that the proposed system achieves superior segmentation performance of ocular structure. Moreover, it markedly improves the diagnostic accuracy of doctors. In the future, this endeavor can offer highly promising MG monitoring tools for healthcare professionals, patients, and regions with limited medical resources.
{"title":"A Deep Learning-Based Ocular Structure Segmentation for Assisted Myasthenia Gravis Diagnosis from Facial Images","authors":"Linna Zhao;Jianqiang Li;Xi Xu;Chujie Zhu;Wenxiu Cheng;Suqin Liu;Mingming Zhao;Lei Zhang;Jing Zhang;Jian Yin;Jijiang Yang","doi":"10.26599/TST.2024.9010177","DOIUrl":"https://doi.org/10.26599/TST.2024.9010177","url":null,"abstract":"Myasthenia Gravis (MG) is an autoimmune neuromuscular disease. Given that extraocular muscle manifestations are the initial and primary symptoms in most patients, ocular muscle assessment is regarded necessary early screening tool. To overcome the limitations of the manual clinical method, an intuitive idea is to collect data via imaging devices, followed by analysis or processing using Deep Learning (DL) techniques (particularly image segmentation approaches) to enable automatic MG evaluation. Unfortunately, their clinical applications in this field have not been thoroughly explored. To bridge this gap, our study prospectively establishes a new DL-based system to promote the diagnosis of MG disease, with a complete workflow including facial data acquisition, eye region localization, and ocular structure segmentation. Experimental results demonstrate that the proposed system achieves superior segmentation performance of ocular structure. Moreover, it markedly improves the diagnostic accuracy of doctors. In the future, this endeavor can offer highly promising MG monitoring tools for healthcare professionals, patients, and regions with limited medical resources.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2592-2605"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072111","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Pub Date : 2025-07-04DOI: 10.26599/TST.2024.9010096
Hang Li;Wusheng Chou
In this article, a finite-time adaptive dynamic surface synchronization tracking controller with guaranteed transient performance is proposed for bilateral teleoperation manipulators. To achieve this objective, we establish a comprehensive model of the teleoperation system incorporating asymmetric time-varying delays, external disturbances, joint frictions, and additive uncertainties. Subsequently, the dynamic surface control approach is introduced to reduce computational complexity by avoiding repeated differentiation of virtual signals in traditional backstepping algorithms. Moreover, this law address the passivity issue associated with time-delayed channels by substituting joint frictions and environmental parameter uncertainties with non-power approximate signals generated using fuzzy logic algorithms. Additionally, through the utilization of the finite-time performance function, assurance is provided for the transient performance of the system. The synchronization errors can converge to a small neighborhood around zero in a finite time which can be arbitrarily set. Theoretically, the Semi-Global Practical Finite-Time Stability (SGPFTS) of the closed-loop signals is derived from the Lyapunov function. The simulation and practical experiment are both performed, and the results verify the effectiveness of the proposed control approach. In the future, the work will consider the teleoperation system where the initial error is not within the constraints of the finite-time performance function, and simplify the adaptive updating law.
{"title":"Dynamic Surface Control for Nonlinear Bilateral Teleoperation Manipulators with Guaranteed Transient Performance","authors":"Hang Li;Wusheng Chou","doi":"10.26599/TST.2024.9010096","DOIUrl":"https://doi.org/10.26599/TST.2024.9010096","url":null,"abstract":"In this article, a finite-time adaptive dynamic surface synchronization tracking controller with guaranteed transient performance is proposed for bilateral teleoperation manipulators. To achieve this objective, we establish a comprehensive model of the teleoperation system incorporating asymmetric time-varying delays, external disturbances, joint frictions, and additive uncertainties. Subsequently, the dynamic surface control approach is introduced to reduce computational complexity by avoiding repeated differentiation of virtual signals in traditional backstepping algorithms. Moreover, this law address the passivity issue associated with time-delayed channels by substituting joint frictions and environmental parameter uncertainties with non-power approximate signals generated using fuzzy logic algorithms. Additionally, through the utilization of the finite-time performance function, assurance is provided for the transient performance of the system. The synchronization errors can converge to a small neighborhood around zero in a finite time which can be arbitrarily set. Theoretically, the Semi-Global Practical Finite-Time Stability (SGPFTS) of the closed-loop signals is derived from the Lyapunov function. The simulation and practical experiment are both performed, and the results verify the effectiveness of the proposed control approach. In the future, the work will consider the teleoperation system where the initial error is not within the constraints of the finite-time performance function, and simplify the adaptive updating law.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2505-2521"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072116","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557918","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Pub Date : 2025-07-04DOI: 10.26599/TST.2024.9010100
Zhiyu Chen;Jian Mao;Qixiao Lin;Liran Ma;Jianwei Liu
Incremental search provides real-time suggestions as users type their queries. However, recent studies demonstrate that its encrypted search traffic can disclose privacy-sensitive data through side channels. Specifically, attackers can derive information about user keystrokes from observable traffic features, like packet sizes, timings, and directions, thereby inferring the victim's entered search query. This vulnerability is known as a remote keystroke inference attack. While various attacks leveraging different traffic features have been developed, accompanied by obfuscation-based countermeasures, there is still a lack of overall and in-depth understanding regarding these attacks and defenses. To fill this gap, we conduct the first comprehensive evaluation of existing remote keystroke inference attacks and defenses. We carry out extensive experiments on five well-known incremental search websites, all listed in Alexa's top 50, to evaluate and compare their real-world performance. The results demonstrate that attacks utilizing multidimensional request features pose the greatest risk to user privacy, and random padding is currently considered the optimal defense balancing both efficacy and resource demands. Our work sheds light on the real-world implications of remote keystroke inference attacks and provides developers with guidelines to enhance privacy protection strategies.
{"title":"Empirical Analysis of Remote Keystroke Inference Attacks and Defenses on Incremental Search","authors":"Zhiyu Chen;Jian Mao;Qixiao Lin;Liran Ma;Jianwei Liu","doi":"10.26599/TST.2024.9010100","DOIUrl":"https://doi.org/10.26599/TST.2024.9010100","url":null,"abstract":"Incremental search provides real-time suggestions as users type their queries. However, recent studies demonstrate that its encrypted search traffic can disclose privacy-sensitive data through side channels. Specifically, attackers can derive information about user keystrokes from observable traffic features, like packet sizes, timings, and directions, thereby inferring the victim's entered search query. This vulnerability is known as a remote keystroke inference attack. While various attacks leveraging different traffic features have been developed, accompanied by obfuscation-based countermeasures, there is still a lack of overall and in-depth understanding regarding these attacks and defenses. To fill this gap, we conduct the first comprehensive evaluation of existing remote keystroke inference attacks and defenses. We carry out extensive experiments on five well-known incremental search websites, all listed in Alexa's top 50, to evaluate and compare their real-world performance. The results demonstrate that attacks utilizing multidimensional request features pose the greatest risk to user privacy, and random padding is currently considered the optimal defense balancing both efficacy and resource demands. Our work sheds light on the real-world implications of remote keystroke inference attacks and provides developers with guidelines to enhance privacy protection strategies.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 6","pages":"2434-2451"},"PeriodicalIF":6.6,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11072067","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
Pub Date : 2025-04-29DOI: 10.26599/TST.2024.9010155
Zhen Guo;Ying Zhou;Jun Ye;Yongxu Hou
Privacy-sensitive data encounter immense security and usability challenges in processing, analyzing, and sharing. Meanwhile, traditional privacy data desensitization methods suffer from issues such as poor quality and low usability after desensitization. Therefore, a text data desensitization model that combines Transformer and Wasserstein Text convolutional Generative Adversarial Network (Trans-WTGAN) is proposed. Transformer as the generator and its self-attention mechanism can handle long-range dependencies, enabling the generated of higher-quality text; Text Convolutional Neural Network (TextCNN) integrates the idea of Wasserstein as the discriminator to enhance the stability of model training; and the strategy gradient scheme of reinforcement learning is employed. Reinforcement learning utilizes the policy gradient scheme as the updating method of generator parameters, ensuring the generated data retains the original key features and maintains a certain level of usability. The experimental results indicate that the proposed model scheme holds a greater advantage over existing methods in terms of text quality and structural consistency, can guarantee the desensitization effect, and ensures the usability of the privacy-sensitive data to a certain extent after desensitization, facilitates the simulation of the development environment for the use of real data and the analysis and sharing of data.
{"title":"Desensitization of Private Text Dataset Based on Gradient Strategy Trans-WTGAN","authors":"Zhen Guo;Ying Zhou;Jun Ye;Yongxu Hou","doi":"10.26599/TST.2024.9010155","DOIUrl":"https://doi.org/10.26599/TST.2024.9010155","url":null,"abstract":"Privacy-sensitive data encounter immense security and usability challenges in processing, analyzing, and sharing. Meanwhile, traditional privacy data desensitization methods suffer from issues such as poor quality and low usability after desensitization. Therefore, a text data desensitization model that combines Transformer and Wasserstein Text convolutional Generative Adversarial Network (Trans-WTGAN) is proposed. Transformer as the generator and its self-attention mechanism can handle long-range dependencies, enabling the generated of higher-quality text; Text Convolutional Neural Network (TextCNN) integrates the idea of Wasserstein as the discriminator to enhance the stability of model training; and the strategy gradient scheme of reinforcement learning is employed. Reinforcement learning utilizes the policy gradient scheme as the updating method of generator parameters, ensuring the generated data retains the original key features and maintains a certain level of usability. The experimental results indicate that the proposed model scheme holds a greater advantage over existing methods in terms of text quality and structural consistency, can guarantee the desensitization effect, and ensures the usability of the privacy-sensitive data to a certain extent after desensitization, facilitates the simulation of the development environment for the use of real data and the analysis and sharing of data.","PeriodicalId":48690,"journal":{"name":"Tsinghua Science and Technology","volume":"30 5","pages":"2081-2096"},"PeriodicalIF":6.6,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10979794","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143888337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：