In response to the lack of effective means for detecting and locating malicious exchange nodes in data flow transmission links within the Internet of Things (IoT), this paper proposes a zero-trust management method for data flow between edge nodes based on software defined networking (SDN) communication and control of cyber-physical systems (CPS). To detect and prevent anomalous behaviors like data tampering, forwarding path anomalies, and malicious packet drops through forwarding verification by exchange nodes, SDN-ZTM applies SDN to the data transmission process between IoT edge nodes. This approach applies the SDN architecture to the transmission process of data flows between edge nodes, utilizing a fixed-length header overhead for zero-trust management of data flows, nodes, and paths, thereby enabling lightweight packet forwarding verification and malicious exchange node localization. Simulation studies and theoretical research show that SDN-ZTM offers more extensive security features than similar methods. Additionally, SDN-ZTM is a lightweight, useful solution appropriate for IoT application scenarios since it introduces a fixed-length header and has a smaller performance overhead. Experimental results show that the method introduces less than 10% forwarding delay and less than 8% throughput loss.
{"title":"Zero Trust Management Over Consumer Technology-Based IoT Edge Node for SDN Communication and Control of Cyber–Physical Systems","authors":"Haewon Byeon;Mahmood Alsaadi;Sachin Gupta;Jagdish Chandra Patni;Tariq Ahamed Ahanger;Brajesh Kumar Singh;Ajeet Kumar Srivastava;Pardaeva Shakhnoza Abdinabievna;Santhosh Boddupalli","doi":"10.1109/TCE.2025.3563408","DOIUrl":"https://doi.org/10.1109/TCE.2025.3563408","url":null,"abstract":"In response to the lack of effective means for detecting and locating malicious exchange nodes in data flow transmission links within the Internet of Things (IoT), this paper proposes a zero-trust management method for data flow between edge nodes based on software defined networking (SDN) communication and control of cyber-physical systems (CPS). To detect and prevent anomalous behaviors like data tampering, forwarding path anomalies, and malicious packet drops through forwarding verification by exchange nodes, SDN-ZTM applies SDN to the data transmission process between IoT edge nodes. This approach applies the SDN architecture to the transmission process of data flows between edge nodes, utilizing a fixed-length header overhead for zero-trust management of data flows, nodes, and paths, thereby enabling lightweight packet forwarding verification and malicious exchange node localization. Simulation studies and theoretical research show that SDN-ZTM offers more extensive security features than similar methods. Additionally, SDN-ZTM is a lightweight, useful solution appropriate for IoT application scenarios since it introduces a fixed-length header and has a smaller performance overhead. Experimental results show that the method introduces less than 10% forwarding delay and less than 8% throughput loss.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"4849-4858"},"PeriodicalIF":10.9,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144868112","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}
Pub Date : 2025-04-22DOI: 10.1109/TCE.2025.3563240
Weiwen Zhang;Shuo Yang;Yifeng Jiang
Federated learning has shown its great applicability in intelligent transportation systems, where prediction models can be trained across regions or cities without leaking raw data. However, current federated learning approaches often ignore energy consumption, while energy consumption is playing a pivotal role in sustainability of transportation systems. In this paper, we propose a Two-Phase Client Selection strategy for federated learning (FedTPCS) in traffic flow prediction, aiming to minimize the total energy consumption of clients for participation while considering device and data heterogeneity. First, to tackle device heterogeneity, we leverage K-means clustering to group clients based on their computing power and geographic distance. We strategically select the clustered group with the lowest average cost that is the combination of energy consumption and latency. Second, to tackle data heterogeneity, we leverage affinity propagation clustering based on cosine similarity of model update vectors to divide the selected clients into several subgroups of similar clients. We evaluate the performance of the proposed FedTPCS algorithm on two public datasets. Compared to FedAvg, FedAEB and Greedy algorithms, the FedTPCS algorithm reduces cost by up to 56%, 30%, and 20% under the PeMS dataset, and 50%, 28%, and 18% under the Highways England dataset, respectively.
{"title":"A Two-Phase Client Selection Strategy for Cost-Optimal Federated Learning in Traffic Flow Prediction","authors":"Weiwen Zhang;Shuo Yang;Yifeng Jiang","doi":"10.1109/TCE.2025.3563240","DOIUrl":"https://doi.org/10.1109/TCE.2025.3563240","url":null,"abstract":"Federated learning has shown its great applicability in intelligent transportation systems, where prediction models can be trained across regions or cities without leaking raw data. However, current federated learning approaches often ignore energy consumption, while energy consumption is playing a pivotal role in sustainability of transportation systems. In this paper, we propose a Two-Phase Client Selection strategy for federated learning (FedTPCS) in traffic flow prediction, aiming to minimize the total energy consumption of clients for participation while considering device and data heterogeneity. First, to tackle device heterogeneity, we leverage K-means clustering to group clients based on their computing power and geographic distance. We strategically select the clustered group with the lowest average cost that is the combination of energy consumption and latency. Second, to tackle data heterogeneity, we leverage affinity propagation clustering based on cosine similarity of model update vectors to divide the selected clients into several subgroups of similar clients. We evaluate the performance of the proposed FedTPCS algorithm on two public datasets. Compared to FedAvg, FedAEB and Greedy algorithms, the FedTPCS algorithm reduces cost by up to 56%, 30%, and 20% under the PeMS dataset, and 50%, 28%, and 18% under the Highways England dataset, respectively.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"2955-2964"},"PeriodicalIF":10.9,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144868034","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}
The deployment of electric vehicles (EVs) in different grid domains, such as microgrids (MGs), has increased considerably. To fully realize the advantages of EV ecosystems and integrate them with the MG control schemes, the use of information and communication technologies is required, making the EV ecosystem prone to data manipulation and malware injection. On this basis, the potential vulnerabilities of MGs, such as the load frequency control (LFC) model, that plays an important role in keeping a balance between generation and demand, will be discussed. Then, a switching attack vector originating from EV ecosystems is leveraged to launch coordinated EV-based load-altering attacks (EV-LAAs) based on the frequency of lightly damped modes in MGs. A multi-agent cooperative reinforcement learning (RL) control framework based on the actor-critic proximal policy optimization (PPO) model is designed to mitigate the switching attack vectors. A Lyapunov function is developed using the PPO to provide monotonic policies and guarantee MG’s stability. The performance and robustness of the proposed method are compared with a model-based controller and a centralized RL framework for several attack scenarios during disturbances in wind speed, solar irradiation, and parametric uncertainties under a testbed that integrates a virtual sphere (vSphere) of an EV ecosystem with an islanded MG simulated in OPAL-RT 5650.
{"title":"Data-Driven Framework for Mitigating EV-Based Load-Altering Attacks on LFC Model of Microgrid","authors":"Ahmadreza Abazari;Mohsen Ghafouri;Danial Jafarigiv;Ribal Atallah;Chadi Assi","doi":"10.1109/TCE.2025.3563392","DOIUrl":"https://doi.org/10.1109/TCE.2025.3563392","url":null,"abstract":"The deployment of electric vehicles (EVs) in different grid domains, such as microgrids (MGs), has increased considerably. To fully realize the advantages of EV ecosystems and integrate them with the MG control schemes, the use of information and communication technologies is required, making the EV ecosystem prone to data manipulation and malware injection. On this basis, the potential vulnerabilities of MGs, such as the load frequency control (LFC) model, that plays an important role in keeping a balance between generation and demand, will be discussed. Then, a switching attack vector originating from EV ecosystems is leveraged to launch coordinated EV-based load-altering attacks (EV-LAAs) based on the frequency of lightly damped modes in MGs. A multi-agent cooperative reinforcement learning (RL) control framework based on the actor-critic proximal policy optimization (PPO) model is designed to mitigate the switching attack vectors. A Lyapunov function is developed using the PPO to provide monotonic policies and guarantee MG’s stability. The performance and robustness of the proposed method are compared with a model-based controller and a centralized RL framework for several attack scenarios during disturbances in wind speed, solar irradiation, and parametric uncertainties under a testbed that integrates a virtual sphere (vSphere) of an EV ecosystem with an islanded MG simulated in OPAL-RT 5650.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"6093-6108"},"PeriodicalIF":10.9,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144868298","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}
Pub Date : 2025-04-22DOI: 10.1109/TCE.2025.3563272
Xinyu Liang;Hao Wang
The scarcity of high-quality residential load data can pose obstacles for decarbonizing the residential sector as well as effective grid planning and operation. The above challenges have motivated research into generating synthetic load data, but existing methods faced limitations in terms of scalability, diversity, and similarity. This paper proposes a Generative Adversarial Network-based Synthetic Residential Load Pattern (RLP-GAN) generation model, a novel weakly-supervised GAN framework, leveraging an over-complete autoencoder to capture dependencies within complex and diverse load patterns and learn household-level data distribution at scale. We incorporate a model weight selection method to address the mode collapse problem and generate load patterns with high diversity. We develop a holistic evaluation method to validate the effectiveness of RLP-GAN using real-world data of 417 households. The results demonstrate that RLP-GAN outperforms state-of-the-art models in capturing temporal dependencies and generating load patterns with higher similarity to real data. Furthermore, we have publicly released the RLP-GAN generated synthetic dataset, which comprises one million synthetic residential load pattern profiles.
{"title":"Learning and Generating Diverse Residential Load Patterns Using GAN With Weakly-Supervised Training and Weight Selection","authors":"Xinyu Liang;Hao Wang","doi":"10.1109/TCE.2025.3563272","DOIUrl":"https://doi.org/10.1109/TCE.2025.3563272","url":null,"abstract":"The scarcity of high-quality residential load data can pose obstacles for decarbonizing the residential sector as well as effective grid planning and operation. The above challenges have motivated research into generating synthetic load data, but existing methods faced limitations in terms of scalability, diversity, and similarity. This paper proposes a Generative Adversarial Network-based Synthetic Residential Load Pattern (RLP-GAN) generation model, a novel weakly-supervised GAN framework, leveraging an over-complete autoencoder to capture dependencies within complex and diverse load patterns and learn household-level data distribution at scale. We incorporate a model weight selection method to address the mode collapse problem and generate load patterns with high diversity. We develop a holistic evaluation method to validate the effectiveness of RLP-GAN using real-world data of 417 households. The results demonstrate that RLP-GAN outperforms state-of-the-art models in capturing temporal dependencies and generating load patterns with higher similarity to real data. Furthermore, we have publicly released the RLP-GAN generated synthetic dataset, which comprises one million synthetic residential load pattern profiles.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"2837-2848"},"PeriodicalIF":10.9,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144868339","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}
Pub Date : 2025-04-22DOI: 10.1109/TCE.2025.3563339
Xiaoshan Bai;Haoyu Jiang;Chao Li;Inam Ullah;Maryam M. Al Dabel;Ali Kashif Bashir;Zongze Wu;Shuzhi Sam Ge
In recent years, as people’s living standards have improved and consumption concepts have been transformed, the demand for purchasing consumer electronics online has continued to grow, further stimulating the development of the logistics industry. Consequently, how to deliver consumer electronics to households at minimal cost has become a crucial factor that limits the development of the consumer technology industry. To tackle this problem, this paper studies the task assignment problem for multiple initially dispersed UAVs to deliver products to target locations while minimizing their total operation time. Each UAV can continuously provide delivery services to multiple target locations within its limited loading capacity and operation time. To solve this problem, we propose several hybrid multi-population genetic algorithms. First, a novel crossover operator for the genetic algorithms is designed, through which a single parent chromosome can generate offspring individually. Second, two mutation mechanisms are performed to increase gene diversity. Third, multiple local search strategies are employed to enhance the populations’ fitness during each iteration of evolution. An improved 2-opt local search strategy is applied to optimize individual chromosomes when their similarity with the current best chromosome falls below a prescribed threshold. Alternatively, local search strategies are utilized for 1-opt, 2h-opt, and interchange processes. Combining local search strategies, genetic operators, and the multi-population mechanism leads to several hybrid multi-population genetic algorithms. Numerical simulations and experimental tests demonstrate that the hybrid multi-population genetic algorithm, integrated with the improved 2-opt and 1-opt local search strategies, exhibits superior performance among the designed hybrid genetic algorithms, the minimum marginal cost algorithm (MMA), and the existing popular Co-evolutionary Multi-population Genetic Algorithm (CMGA). In experimental scenarios, the hybrid multi-population genetic algorithm significantly improves CMGA and MMA, reducing UAVs’ total operation time by 4.8% and 13.8%, respectively. This demonstrates its efficiency in meeting the growing demand for low-cost delivery of consumer electronics. This method ensures that logistics operations remain agile and approachable to growing market needs, reinforcing the consumer technology industry’s capability to meet customer expectations in a viable landscape.
{"title":"Efficient Hybrid Multi-Population Genetic Algorithm for Multi-UAV Task Assignment in Consumer Electronics Applications","authors":"Xiaoshan Bai;Haoyu Jiang;Chao Li;Inam Ullah;Maryam M. Al Dabel;Ali Kashif Bashir;Zongze Wu;Shuzhi Sam Ge","doi":"10.1109/TCE.2025.3563339","DOIUrl":"https://doi.org/10.1109/TCE.2025.3563339","url":null,"abstract":"In recent years, as people’s living standards have improved and consumption concepts have been transformed, the demand for purchasing consumer electronics online has continued to grow, further stimulating the development of the logistics industry. Consequently, how to deliver consumer electronics to households at minimal cost has become a crucial factor that limits the development of the consumer technology industry. To tackle this problem, this paper studies the task assignment problem for multiple initially dispersed UAVs to deliver products to target locations while minimizing their total operation time. Each UAV can continuously provide delivery services to multiple target locations within its limited loading capacity and operation time. To solve this problem, we propose several hybrid multi-population genetic algorithms. First, a novel crossover operator for the genetic algorithms is designed, through which a single parent chromosome can generate offspring individually. Second, two mutation mechanisms are performed to increase gene diversity. Third, multiple local search strategies are employed to enhance the populations’ fitness during each iteration of evolution. An improved 2-opt local search strategy is applied to optimize individual chromosomes when their similarity with the current best chromosome falls below a prescribed threshold. Alternatively, local search strategies are utilized for 1-opt, 2h-opt, and interchange processes. Combining local search strategies, genetic operators, and the multi-population mechanism leads to several hybrid multi-population genetic algorithms. Numerical simulations and experimental tests demonstrate that the hybrid multi-population genetic algorithm, integrated with the improved 2-opt and 1-opt local search strategies, exhibits superior performance among the designed hybrid genetic algorithms, the minimum marginal cost algorithm (MMA), and the existing popular Co-evolutionary Multi-population Genetic Algorithm (CMGA). In experimental scenarios, the hybrid multi-population genetic algorithm significantly improves CMGA and MMA, reducing UAVs’ total operation time by 4.8% and 13.8%, respectively. This demonstrates its efficiency in meeting the growing demand for low-cost delivery of consumer electronics. This method ensures that logistics operations remain agile and approachable to growing market needs, reinforcing the consumer technology industry’s capability to meet customer expectations in a viable landscape.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"2395-2406"},"PeriodicalIF":10.9,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144868214","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}
Under the “dual carbon” background, consumer electronics consumption has become deeply ingrained in people’s minds. However, consumers often distrust the sustainability claims of consumer electronics products. Artificial intelligence (AI) and blockchain technology can address this trust deficit through transparency and traceability mechanisms. This study integrates blockchain technology into traditional consumer electronics supply chains, considering consumers’ preferences and trust in these products. An AI-based game model is proposed to analyze the interactions among supply chain members before and after implementing blockchain technology, under varying Edge Computing-based power structures. This model quantitatively evaluates emission reduction and pricing strategies, aiming to optimize consumer surplus and total social welfare. By leveraging Lightweight AI and blockchain, smart wholesale and cost-sharing contracts are designed to establish reasonable ranges for wholesale prices and optimal cost-sharing ratios, enhancing enterprise operational efficiency and achieving supply chain coordination. Results demonstrate that when consumers exhibit a stronger preference for consumer electronics products, the adoption of Lightweight AI and blockchain delivers greater benefits across the supply chain. Furthermore, as consumer willingness to purchase these products increases, the advantages become more pronounced. Numerical analysis highlights that smart contracts can better coordinate the supply chain, particularly in retailer-dominated scenarios. Finally, empirical cases validate the effectiveness of the proposed strategies and models.
{"title":"Lightweight AI and Blockchain Optimization for Enhancing Consumer Electronics Decision-Making","authors":"Haewon Byeon;Mahmood Alsaadi;Ismail Keshta;Tariq Ahamed Ahanger;Nodira Safarova;Hamad Aldawsari;Lucia Cascone;Mohammad Shabaz","doi":"10.1109/TCE.2025.3563412","DOIUrl":"https://doi.org/10.1109/TCE.2025.3563412","url":null,"abstract":"Under the “dual carbon” background, consumer electronics consumption has become deeply ingrained in people’s minds. However, consumers often distrust the sustainability claims of consumer electronics products. Artificial intelligence (AI) and blockchain technology can address this trust deficit through transparency and traceability mechanisms. This study integrates blockchain technology into traditional consumer electronics supply chains, considering consumers’ preferences and trust in these products. An AI-based game model is proposed to analyze the interactions among supply chain members before and after implementing blockchain technology, under varying Edge Computing-based power structures. This model quantitatively evaluates emission reduction and pricing strategies, aiming to optimize consumer surplus and total social welfare. By leveraging Lightweight AI and blockchain, smart wholesale and cost-sharing contracts are designed to establish reasonable ranges for wholesale prices and optimal cost-sharing ratios, enhancing enterprise operational efficiency and achieving supply chain coordination. Results demonstrate that when consumers exhibit a stronger preference for consumer electronics products, the adoption of Lightweight AI and blockchain delivers greater benefits across the supply chain. Furthermore, as consumer willingness to purchase these products increases, the advantages become more pronounced. Numerical analysis highlights that smart contracts can better coordinate the supply chain, particularly in retailer-dominated scenarios. Finally, empirical cases validate the effectiveness of the proposed strategies and models.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"6007-6015"},"PeriodicalIF":10.9,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144868019","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}
Pub Date : 2025-04-22DOI: 10.1109/TCE.2025.3563150
Chuanjun Zhao;Lu Kang;Xuzhuang Sun;Xiaoxiong Xi;Lihua Shen;Jing Gao;Yanjie Wang
In recent years, the explosive growth of user-generated review texts has underscored the academic and societal significance of sentiment analysis. Although deep learning has achieved remarkable progress in this field, existing aspect-based sentiment classification (ABSC) methods face challenges in capturing the dynamic nature of sentiment categories. Furthermore, these methods often lack explicit modeling of category information, limiting their ability to adapt attention distributions based on sentiment categories. To address these challenges, this paper proposes a BERT-based model with a category-aware multi-head attention mechanism. The model introduces an aspect projection layer that maps aspect word embeddings into a feature space aligned with the context, thereby enhancing interaction between aspect words and the surrounding text. Additionally, a category-aware multi-head attention mechanism combines category weights and dynamic content weights to effectively fuse sentiment category information. This design significantly improves the model’s ability to capture sentiment features of multiple categories. Experimental evaluations on SemEval public datasets demonstrate that the proposed method outperforms state-of-the-art techniques, and ablation studies further confirm the effectiveness of its design.
{"title":"Aspect-Level Sentiment Classification of Consumer Reviews Utilizing BERT and Category-Aware Multi-Head Attention","authors":"Chuanjun Zhao;Lu Kang;Xuzhuang Sun;Xiaoxiong Xi;Lihua Shen;Jing Gao;Yanjie Wang","doi":"10.1109/TCE.2025.3563150","DOIUrl":"https://doi.org/10.1109/TCE.2025.3563150","url":null,"abstract":"In recent years, the explosive growth of user-generated review texts has underscored the academic and societal significance of sentiment analysis. Although deep learning has achieved remarkable progress in this field, existing aspect-based sentiment classification (ABSC) methods face challenges in capturing the dynamic nature of sentiment categories. Furthermore, these methods often lack explicit modeling of category information, limiting their ability to adapt attention distributions based on sentiment categories. To address these challenges, this paper proposes a BERT-based model with a category-aware multi-head attention mechanism. The model introduces an aspect projection layer that maps aspect word embeddings into a feature space aligned with the context, thereby enhancing interaction between aspect words and the surrounding text. Additionally, a category-aware multi-head attention mechanism combines category weights and dynamic content weights to effectively fuse sentiment category information. This design significantly improves the model’s ability to capture sentiment features of multiple categories. Experimental evaluations on SemEval public datasets demonstrate that the proposed method outperforms state-of-the-art techniques, and ablation studies further confirm the effectiveness of its design.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"3329-3339"},"PeriodicalIF":10.9,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144867740","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}
Pub Date : 2025-04-21DOI: 10.1109/TCE.2025.3562865
Yu Qiao;Hao Ji
The rapid development of Internet of Things technology is changing our daily life and the global industrial pattern at an unprecedented speed. It has brought a revolution of intelligence and personalization to personal consumption. The research on multi trajectory privacy protection has a positive impact on the security of consumer data. This paper focuses on the correlations between multiple trajectories. To streamline trajectory data, the quad-tree method is employed to partition the road network area and segment trajectories into discrete units. Subsequently, we quantify the correlation between the original trajectory and others using visit probability vectors, aiming to reduce their similarity. Within specified constraints, refining visit probability vectors via an optimized particle swarm optimization approach tailored for differential privacy. Experiments conducted on real datasets attest to the solution’s robustness and its ability to achieve a better trade-off between privacy protection and data utility effectively.
{"title":"A Multi Trajectory Privacy Protection Method for IoT Based on Particle Swarm Optimization","authors":"Yu Qiao;Hao Ji","doi":"10.1109/TCE.2025.3562865","DOIUrl":"https://doi.org/10.1109/TCE.2025.3562865","url":null,"abstract":"The rapid development of Internet of Things technology is changing our daily life and the global industrial pattern at an unprecedented speed. It has brought a revolution of intelligence and personalization to personal consumption. The research on multi trajectory privacy protection has a positive impact on the security of consumer data. This paper focuses on the correlations between multiple trajectories. To streamline trajectory data, the quad-tree method is employed to partition the road network area and segment trajectories into discrete units. Subsequently, we quantify the correlation between the original trajectory and others using visit probability vectors, aiming to reduce their similarity. Within specified constraints, refining visit probability vectors via an optimized particle swarm optimization approach tailored for differential privacy. Experiments conducted on real datasets attest to the solution’s robustness and its ability to achieve a better trade-off between privacy protection and data utility effectively.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"5216-5223"},"PeriodicalIF":10.9,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144868263","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}
Pub Date : 2025-04-21DOI: 10.1109/TCE.2025.3562767
Qiong Li;Rongsheng Cai;Yizhao Zhu
As Artificial Intelligence (AI) progresses, the application of federated learning in areas such as consumer app security and intelligent transportation systems is increasing rapidly. Federated learning allows model training without necessitating the sharing of local data, yet security issues present obstacles to its advancement. This paper presents a federated learning method that merges gradient compression with homomorphic encryption. Firstly, a unique gradient compression technique is proposed to reduce data transfer by compressing the model parameters exchanged among clients. Then, homomorphic encryption is utilized to prevent breaches of gradient privacy. Experimental results demonstrate that our proposed approach has a minimal impact on the accuracy of the global model, while it reduces data transmission and improves the privacy and security of federated learning.
{"title":"GHPPFL: A Privacy Preserving Federated Learning Based on Gradient Compression and Homomorphic Encryption in Consumer App Security","authors":"Qiong Li;Rongsheng Cai;Yizhao Zhu","doi":"10.1109/TCE.2025.3562767","DOIUrl":"https://doi.org/10.1109/TCE.2025.3562767","url":null,"abstract":"As Artificial Intelligence (AI) progresses, the application of federated learning in areas such as consumer app security and intelligent transportation systems is increasing rapidly. Federated learning allows model training without necessitating the sharing of local data, yet security issues present obstacles to its advancement. This paper presents a federated learning method that merges gradient compression with homomorphic encryption. Firstly, a unique gradient compression technique is proposed to reduce data transfer by compressing the model parameters exchanged among clients. Then, homomorphic encryption is utilized to prevent breaches of gradient privacy. Experimental results demonstrate that our proposed approach has a minimal impact on the accuracy of the global model, while it reduces data transmission and improves the privacy and security of federated learning.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"5090-5099"},"PeriodicalIF":10.9,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10970749","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144867551","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}
Pub Date : 2025-04-18DOI: 10.1109/TCE.2025.3562388
Deepak Kumar;Udit Satija
Electroencephalograms (EEGs) are effective and patient-friendly for diagnosing, monitoring, and preventing mental disorders. However, due to their low voltage, EEG signals often contain noise that obscures critical features, risking misdiagnosis. Current denoising methods typically address one or two noise types and struggle with memory limitations on edge devices. To overcome these challenges, we introduce a quantization-based compressed denoising autoencoder (DAE) model using a PbP-QLP, a low-rank approximation (LRA) technique, for multi-noise removal (15 types, including power-line, baseline wander, ocular, muscle artifacts, and combinations) in EEGs on low-memory edge devices. Our compression technique reduces the model size from 8 to 1.51 MB, achieving 81% weight compression with minimal loss.
{"title":"Optimized EEG Multi-Noise Removal and Compression: Deploying a PbP-QLP Enhanced Autoencoder on STM32 Microcontroller","authors":"Deepak Kumar;Udit Satija","doi":"10.1109/TCE.2025.3562388","DOIUrl":"https://doi.org/10.1109/TCE.2025.3562388","url":null,"abstract":"Electroencephalograms (EEGs) are effective and patient-friendly for diagnosing, monitoring, and preventing mental disorders. However, due to their low voltage, EEG signals often contain noise that obscures critical features, risking misdiagnosis. Current denoising methods typically address one or two noise types and struggle with memory limitations on edge devices. To overcome these challenges, we introduce a quantization-based compressed denoising autoencoder (DAE) model using a PbP-QLP, a low-rank approximation (LRA) technique, for multi-noise removal (15 types, including power-line, baseline wander, ocular, muscle artifacts, and combinations) in EEGs on low-memory edge devices. Our compression technique reduces the model size from 8 to 1.51 MB, achieving 81% weight compression with minimal loss.","PeriodicalId":13208,"journal":{"name":"IEEE Transactions on Consumer Electronics","volume":"71 2","pages":"3218-3228"},"PeriodicalIF":10.9,"publicationDate":"2025-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144867810","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}
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：