Computer Communications最新文献

Multi-layer guided reinforcement learning task offloading based on Softmax policy in smart cities

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-17 DOI: 10.1016/j.comcom.2025.108105

Bin Wu , Liwen Ma , Yu Ji , Jia Cong , Min Xu , Jie Zhao , Yue Yang

Edge computing is an effective measure for addressing the high demand for computing power on the end-side due to dense task distribution in the mobile Internet. In the case of limited device resources and computing power, how to optimize the task offloading decision has become an important issue for improving computing efficiency. We improve the heuristic algorithm by combining the characteristics of intensive tasks, and optimize the task offloading decision at a lower cost. To overcome the limitation of requiring a large amount of real-time information, we utilize the RL algorithm and design a new reward function to enable the agent to learn from its interactions with the environment. Aiming at the poor performance of the system in the uncertain initial environment, we propose a Q-learning scheme based on the Softmax strategy for the multi-layer agent RL framework. The offloading process is optimized by coordinating agents with different views of the environment between each layer, while balancing the exploration and utilization relationship to improve the performance of the algorithm in a more complex dynamic environment. The experimental results show that in the mobile environment with high device density and diverse tasks, the proposed algorithm achieves significant improvements in key indicators such as task success rate, waiting time, and energy consumption. In particular, it exhibits excellent robustness and efficiency advantages in complex dynamic environments, far exceeding the current benchmark algorithm.

{"title":"Multi-layer guided reinforcement learning task offloading based on Softmax policy in smart cities","authors":"Bin Wu , Liwen Ma , Yu Ji , Jia Cong , Min Xu , Jie Zhao , Yue Yang","doi":"10.1016/j.comcom.2025.108105","DOIUrl":"10.1016/j.comcom.2025.108105","url":null,"abstract":"<div><div>Edge computing is an effective measure for addressing the high demand for computing power on the end-side due to dense task distribution in the mobile Internet. In the case of limited device resources and computing power, how to optimize the task offloading decision has become an important issue for improving computing efficiency. We improve the heuristic algorithm by combining the characteristics of intensive tasks, and optimize the task offloading decision at a lower cost. To overcome the limitation of requiring a large amount of real-time information, we utilize the RL algorithm and design a new reward function to enable the agent to learn from its interactions with the environment. Aiming at the poor performance of the system in the uncertain initial environment, we propose a Q-learning scheme based on the Softmax strategy for the multi-layer agent RL framework. The offloading process is optimized by coordinating agents with different views of the environment between each layer, while balancing the exploration and utilization relationship to improve the performance of the algorithm in a more complex dynamic environment. The experimental results show that in the mobile environment with high device density and diverse tasks, the proposed algorithm achieves significant improvements in key indicators such as task success rate, waiting time, and energy consumption. In particular, it exhibits excellent robustness and efficiency advantages in complex dynamic environments, far exceeding the current benchmark algorithm.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"235 ","pages":"Article 108105"},"PeriodicalIF":4.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143453713","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

DFFL: A dual fairness framework for federated learning

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-17 DOI: 10.1016/j.comcom.2025.108104

Kaiyue Qi, Tongjiang Yan, Pengcheng Ren, Jianye Yang, Jialin Li

Federated learning (FL) is an emerging paradigm of distributed machine learning that facilitates collaborative training of a global model across multiple clients while preserving client-side data privacy. However, current equality fairness methodologies aim to maintain a more uniform performance distribution across clients, but they fail to consider the varying contributions of different clients. In contrast, collaboration fairness takes into account the contributions of clients but may exclude low-contributing clients in pursuit of the interests of high-contributing clients. To address these concerns, this paper proposes a novel Dual Fair Federated Learning (DFFL) framework. Specifically, we combine the concept of cosine annealing to evaluate each client’s contribution from two perspectives. Then, we utilize client’s contribution as the aggregation weight of the global model to improve the global model accuracy. Additionally, we introduce a personalized design and utilize client’s contribution as a regularization coefficient to achieve dual fairness. Furthermore, we conduct a theoretical analysis of the convergence of the global model. Finally, through comprehensive experiments on benchmark datasets, we demonstrate that our method achieves competitive predictive accuracy and dual fairness.

{"title":"DFFL: A dual fairness framework for federated learning","authors":"Kaiyue Qi, Tongjiang Yan, Pengcheng Ren, Jianye Yang, Jialin Li","doi":"10.1016/j.comcom.2025.108104","DOIUrl":"10.1016/j.comcom.2025.108104","url":null,"abstract":"<div><div>Federated learning (FL) is an emerging paradigm of distributed machine learning that facilitates collaborative training of a global model across multiple clients while preserving client-side data privacy. However, current equality fairness methodologies aim to maintain a more uniform performance distribution across clients, but they fail to consider the varying contributions of different clients. In contrast, collaboration fairness takes into account the contributions of clients but may exclude low-contributing clients in pursuit of the interests of high-contributing clients. To address these concerns, this paper proposes a novel Dual Fair Federated Learning (DFFL) framework. Specifically, we combine the concept of cosine annealing to evaluate each client’s contribution from two perspectives. Then, we utilize client’s contribution as the aggregation weight of the global model to improve the global model accuracy. Additionally, we introduce a personalized design and utilize client’s contribution as a regularization coefficient to achieve dual fairness. Furthermore, we conduct a theoretical analysis of the convergence of the global model. Finally, through comprehensive experiments on benchmark datasets, we demonstrate that our method achieves competitive predictive accuracy and dual fairness.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"235 ","pages":"Article 108104"},"PeriodicalIF":4.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143444666","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Secrecy performance optimization for UAV-based relay NOMA systems with friendly jamming

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-17 DOI: 10.1016/j.comcom.2025.108086

Thanh Trung Nguyen , Tran Manh Hoang , Phuong T. Tran

Friendly jamming and relay are effective schemes in physical layer security (PLS) for enhancing security in wireless communication. By deploying unmanned aerial vehicle (UAV)-assisted Non-Orthogonal Multiple Access (NOMA) transmission can extend coverage and enhancing spectrum efficiency. This paper studies the physical layer security of an UAV-based relay NOMA system, consisting of a source, multiple users, and an eavesdropper. To enhance secrecy performance, an additional UAV is employed to transmit jamming signals to the eavesdropper. Moreover, for a more practical approach, we also consider the imperfect collaboration between the jammer device and the legitimate user. The minimum average secrecy rate (MASR) of the users is maximized, assuming that the eavesdropper is capable of intercepting signals both from the source and from the relay UAV. An efficient iterative algorithm is proposed to solve the MASR maximum problem by optimizing UAV trajectories, transmit power, and power allocation coefficients. Simulation results demonstrate that the proposed system achieves 238% better MASR than the system without friendly jamming signals and 633% better than the non-optimal system. In addition, the ability to decode the received signal using successive interference cancellation also significantly affects the MASR of users in the system.

{"title":"Secrecy performance optimization for UAV-based relay NOMA systems with friendly jamming","authors":"Thanh Trung Nguyen , Tran Manh Hoang , Phuong T. Tran","doi":"10.1016/j.comcom.2025.108086","DOIUrl":"10.1016/j.comcom.2025.108086","url":null,"abstract":"<div><div>Friendly jamming and relay are effective schemes in physical layer security (PLS) for enhancing security in wireless communication. By deploying unmanned aerial vehicle (UAV)-assisted Non-Orthogonal Multiple Access (NOMA) transmission can extend coverage and enhancing spectrum efficiency. This paper studies the physical layer security of an UAV-based relay NOMA system, consisting of a source, multiple users, and an eavesdropper. To enhance secrecy performance, an additional UAV is employed to transmit jamming signals to the eavesdropper. Moreover, for a more practical approach, we also consider the imperfect collaboration between the jammer device and the legitimate user. The minimum average secrecy rate (MASR) of the users is maximized, assuming that the eavesdropper is capable of intercepting signals both from the source and from the relay UAV. An efficient iterative algorithm is proposed to solve the MASR maximum problem by optimizing UAV trajectories, transmit power, and power allocation coefficients. Simulation results demonstrate that the proposed system achieves 238% better MASR than the system without friendly jamming signals and 633% better than the non-optimal system. In addition, the ability to decode the received signal using successive interference cancellation also significantly affects the MASR of users in the system.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"235 ","pages":"Article 108086"},"PeriodicalIF":4.5,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143437960","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Incentive mechanisms for non-proprietary vehicles in vehicular crowdsensing with budget constraints 有预算限制的车辆众包中的非专有车辆激励机制

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-14 DOI: 10.1016/j.comcom.2025.108083

Zhirui Feng, Yantao Yu, Guojin Liu, Yang Jiang, TianCong Huang

Vehicular crowdsensing (VCS) utilizes the onboard sensors and computational capabilities of smart vehicles to collect data across diverse regions. Non-dedicated vehicles, due to their lower cost and broad distribution, have emerged as a central focus in VCS research. However, their trajectories are often concentrated in urban areas, resulting in uneven data coverage. Existing incentive mechanisms primarily rely on platforms to dynamically adjust task allocation based on vehicle trajectory predictions. Yet, they frequently neglect the influence of geographic locations on vehicle routing choices and fail to incentivize proactive route planning. To address this, we propose a novel two-phase incentive mechanism that, for the first time, incorporates a willingness to traverse factor. This mechanism aims to maximize spatial coverage within a limited budget by encouraging vehicles to voluntarily traverse remote areas to complete tasks. In the initial phase, a multi-agent deep reinforcement learning algorithm dynamically adjusts each vehicle’s route and quote price, which is then reported to the platform. In the second phase, the platform allocates tasks and adjusts compensation based on the provided routes and quotes to optimize overall platform benefits. Experimental results show that our mechanism effectively balances platform and vehicle benefits, achieving optimal outcomes even under budget constraints.

{"title":"Incentive mechanisms for non-proprietary vehicles in vehicular crowdsensing with budget constraints","authors":"Zhirui Feng, Yantao Yu, Guojin Liu, Yang Jiang, TianCong Huang","doi":"10.1016/j.comcom.2025.108083","DOIUrl":"10.1016/j.comcom.2025.108083","url":null,"abstract":"<div><div>Vehicular crowdsensing (VCS) utilizes the onboard sensors and computational capabilities of smart vehicles to collect data across diverse regions. Non-dedicated vehicles, due to their lower cost and broad distribution, have emerged as a central focus in VCS research. However, their trajectories are often concentrated in urban areas, resulting in uneven data coverage. Existing incentive mechanisms primarily rely on platforms to dynamically adjust task allocation based on vehicle trajectory predictions. Yet, they frequently neglect the influence of geographic locations on vehicle routing choices and fail to incentivize proactive route planning. To address this, we propose a novel two-phase incentive mechanism that, for the first time, incorporates a <em><strong>willingness to traverse</strong></em> factor. This mechanism aims to maximize spatial coverage within a limited budget by encouraging vehicles to voluntarily traverse remote areas to complete tasks. In the initial phase, a multi-agent deep reinforcement learning algorithm dynamically adjusts each vehicle’s route and quote price, which is then reported to the platform. In the second phase, the platform allocates tasks and adjusts compensation based on the provided routes and quotes to optimize overall platform benefits. Experimental results show that our mechanism effectively balances platform and vehicle benefits, achieving optimal outcomes even under budget constraints.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"235 ","pages":"Article 108083"},"PeriodicalIF":4.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143437959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Load-balanced multi-user mobility-aware task offloading in multi-access edge computing

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-13 DOI: 10.1016/j.comcom.2025.108102

Shanchen Pang, Meng Zhou, Haiyuan Gui, Xiao He, Nuanlai Wang, Luqi Wang

In scenarios with dense user network service requests, multi-access edge computing demonstrates significant advantages in reducing user device load and decreasing service response time. However, the dynamic changes in user trajectories cause edge server load fluctuations, inevitably impacting the overall service processing performance. To tackle this problem, this paper introduces a load-balanced multi-user mobility-aware service request offloading method, achieving efficient service request offloading in mobile user scenarios. Specifically, this paper divides the service request offloading problem into two stages: dynamic edge server allocation and real-time offloading decision generation. In the first stage, users are allocated edge servers based on their location distribution, implementing an adaptive decreasing variance optimization server load balancing algorithm to achieve edge server load balancing. In the second stage, based on the edge server allocation results from the first stage, a latency performance self-optimizing task offloading decision-making algorithm is employed to minimize the processing latency of user requests, utilizing dueling double deep Q-network to generate real-time decisions on whether to offload service requests to the corresponding edge servers. According to experimental results, the proposed algorithm markedly decreases the processing latency of user network service requests in scenarios of different scales, with an average task completion rate of 99.94%. This effectively addresses the problem of inefficient processing requests caused by load fluctuations due to user movement in multi-access edge computing.

{"title":"Load-balanced multi-user mobility-aware task offloading in multi-access edge computing","authors":"Shanchen Pang, Meng Zhou, Haiyuan Gui, Xiao He, Nuanlai Wang, Luqi Wang","doi":"10.1016/j.comcom.2025.108102","DOIUrl":"10.1016/j.comcom.2025.108102","url":null,"abstract":"<div><div>In scenarios with dense user network service requests, multi-access edge computing demonstrates significant advantages in reducing user device load and decreasing service response time. However, the dynamic changes in user trajectories cause edge server load fluctuations, inevitably impacting the overall service processing performance. To tackle this problem, this paper introduces a load-balanced multi-user mobility-aware service request offloading method, achieving efficient service request offloading in mobile user scenarios. Specifically, this paper divides the service request offloading problem into two stages: dynamic edge server allocation and real-time offloading decision generation. In the first stage, users are allocated edge servers based on their location distribution, implementing an adaptive decreasing variance optimization server load balancing algorithm to achieve edge server load balancing. In the second stage, based on the edge server allocation results from the first stage, a latency performance self-optimizing task offloading decision-making algorithm is employed to minimize the processing latency of user requests, utilizing dueling double deep Q-network to generate real-time decisions on whether to offload service requests to the corresponding edge servers. According to experimental results, the proposed algorithm markedly decreases the processing latency of user network service requests in scenarios of different scales, with an average task completion rate of 99.94%. This effectively addresses the problem of inefficient processing requests caused by load fluctuations due to user movement in multi-access edge computing.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"235 ","pages":"Article 108102"},"PeriodicalIF":4.5,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143420581","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Cloud-edge-end integrated Artificial intelligence based on ensemble learning

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-13 DOI: 10.1016/j.comcom.2025.108103

Zhen Gao , Daning Su , Shuang Liu , Yuqi Zhang , Chenyang Wang , Cheng Zhang , Xiaofei Wang , Tarik Taleb

Deep neural networks (DNNs) have been extensively used in the domains of artificial intelligence (AI) applications. Their inherent complexity primarily drives the deployment of DNN models in cloud environments. However, the geographical distance between the cloud and the end-users fails to meet the low-latency requirements of time-sensitive applications. Edge computing has emerged as a viable way to address this issue, nevertheless, the inherent constraints of limited resources on edge servers pose challenges in supporting intricate models. Solutions relying on network compression or model segmentation often fall short in meeting both performance and reliability needs. For the few ensemble-based solutions, the diversity between base models is not fully explored, and the low-latency advantage of edge computing is not fully utilized. In this paper, we propose a cloud–edge-end integrated approach for building an efficient and reliable DNN inference platform based on ensemble learning. In this design, heterogeneous models are trained on the cloud according to the resource constraints of edge servers, and the inference process is performed independently on each edge server, whose outputs are combined at the end-user side to get the final result. Furthermore, a diversity-based deployment scheme is proposed to build a user-centric network for edge AI. The generation of base models is explored, and the effectiveness of the proposed approach is demonstrated through two case studies.

{"title":"Cloud-edge-end integrated Artificial intelligence based on ensemble learning","authors":"Zhen Gao , Daning Su , Shuang Liu , Yuqi Zhang , Chenyang Wang , Cheng Zhang , Xiaofei Wang , Tarik Taleb","doi":"10.1016/j.comcom.2025.108103","DOIUrl":"10.1016/j.comcom.2025.108103","url":null,"abstract":"<div><div>Deep neural networks (DNNs) have been extensively used in the domains of artificial intelligence (AI) applications. Their inherent complexity primarily drives the deployment of DNN models in cloud environments. However, the geographical distance between the cloud and the end-users fails to meet the low-latency requirements of time-sensitive applications. Edge computing has emerged as a viable way to address this issue, nevertheless, the inherent constraints of limited resources on edge servers pose challenges in supporting intricate models. Solutions relying on network compression or model segmentation often fall short in meeting both performance and reliability needs. For the few ensemble-based solutions, the diversity between base models is not fully explored, and the low-latency advantage of edge computing is not fully utilized. In this paper, we propose a cloud–edge-end integrated approach for building an efficient and reliable DNN inference platform based on ensemble learning. In this design, heterogeneous models are trained on the cloud according to the resource constraints of edge servers, and the inference process is performed independently on each edge server, whose outputs are combined at the end-user side to get the final result. Furthermore, a diversity-based deployment scheme is proposed to build a user-centric network for edge AI. The generation of base models is explored, and the effectiveness of the proposed approach is demonstrated through two case studies.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"235 ","pages":"Article 108103"},"PeriodicalIF":4.5,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143420580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

5GMap: Enabling external audits of access security and attach procedures in real-world cellular deployments

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-10 DOI: 10.1016/j.comcom.2025.108091

Andrea Paci , Matteo Chiacchia , Giuseppe Bianchi

In cellular networks, security vulnerabilities often arise from misconfigurations and improper implementations of protection mechanisms. Typically, ensuring proper security configurations is the responsibility of network operators. The tool described in this paper, called 5GMap, empowers legitimate subscribers, equipped with software-defined radios (Ettus B210 or X310), with innovative means and methodologies for auditing security configurations of the access networks they are connecting to. Specifically, 5GMap allows to evaluate negotiable ciphers, predictability of temporary identifiers (TMSI), resilience against disclosure of privacy-sensitive identifiers (IMSI, IMEI), and susceptibility to downgrade attacks. 5GMap achieves this by iterating access and attach primitives using either carefully crafted signaling messages requiring specific cryptographic configuration, as well as custom methodologies such as using predictable TMSIs and querying the network with non-standard signaling message sequences to detect potential departures from the expected protocol specification. Extensive testing over four mobile network operators and three virtual network operators reveals significant security and privacy issues: many networks allow unencrypted or even unauthenticated communication, TMSI randomness and IMSI concealment are not consistently ensured across all operators tested, and many other fine-grained concerns emerge among different operators. We believe that our findings highlight the usefulness of tools like 5GMap to assess (and ultimately improve, through responsible disclosure) the security posture of 4G and 5G cellular networks in the wild.

{"title":"5GMap: Enabling external audits of access security and attach procedures in real-world cellular deployments","authors":"Andrea Paci , Matteo Chiacchia , Giuseppe Bianchi","doi":"10.1016/j.comcom.2025.108091","DOIUrl":"10.1016/j.comcom.2025.108091","url":null,"abstract":"<div><div>In cellular networks, security vulnerabilities often arise from misconfigurations and improper implementations of protection mechanisms. Typically, ensuring proper security configurations is the responsibility of network operators. The tool described in this paper, called 5GMap, empowers legitimate subscribers, equipped with software-defined radios (Ettus B210 or X310), with innovative means and methodologies for auditing security configurations of the access networks they are connecting to. Specifically, 5GMap allows to evaluate negotiable ciphers, predictability of temporary identifiers (TMSI), resilience against disclosure of privacy-sensitive identifiers (IMSI, IMEI), and susceptibility to downgrade attacks. 5GMap achieves this by iterating access and attach primitives using either carefully crafted signaling messages requiring specific cryptographic configuration, as well as custom methodologies such as using predictable TMSIs and querying the network with non-standard signaling message sequences to detect potential departures from the expected protocol specification. Extensive testing over four mobile network operators and three virtual network operators reveals significant security and privacy issues: many networks allow unencrypted or even unauthenticated communication, TMSI randomness and IMSI concealment are not consistently ensured across all operators tested, and many other fine-grained concerns emerge among different operators. We believe that our findings highlight the usefulness of tools like 5GMap to assess (and ultimately improve, through responsible disclosure) the security posture of 4G and 5G cellular networks in the wild.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"234 ","pages":"Article 108091"},"PeriodicalIF":4.5,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143394626","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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}

引用次数: 0

Efficient post-quantum attribute-based access control scheme for blockchain-empowered metaverse data management

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-09 DOI: 10.1016/j.comcom.2025.108092

Yuxuan Pan , Rui Jin , Yu Liu , Lin Zhang

Driven by recent advances in mobile networks and distributed computing, the Metaverse provides photorealistic services where humans can experience different virtual landscapes through avatars derived from abundant personal user data. To address significant concerns about privacy breaches in private digital assets, access control based on cryptography systems has become the focus of common research. However, existing designs have scalability and efficiency issues, appealing to more investigation in real-world implementation. In response to this security challenge using the prevalent cryptography tool, this paper proposes an Attribute-Based Access Control mechanism for Metaverse Data management (ABAC-MD). It provides a flexible and secure data-sharing framework that integrates the ciphertext-policy attribute-based encryption scheme with the polynomial function technique on lattice. Reliable outsourcing decryption based on blockchain facilitates efficient data processing by employing an attribute-associated access tree. It exploits a pragmatic solution to guarantee fine-grained data privacy control and fortify resilience against quantum attacks. Simulated experiment with relevant schemes based on custom-made avatars proves that the proposed scheme reduces ciphertext size by 43.6% and improve efficiency by at least 25.4%. With higher security, lower storage costs, and reduced computational complexity, the ABAC-MD is more practical for privacy preservation in Metaverse services.

{"title":"Efficient post-quantum attribute-based access control scheme for blockchain-empowered metaverse data management","authors":"Yuxuan Pan , Rui Jin , Yu Liu , Lin Zhang","doi":"10.1016/j.comcom.2025.108092","DOIUrl":"10.1016/j.comcom.2025.108092","url":null,"abstract":"<div><div>Driven by recent advances in mobile networks and distributed computing, the Metaverse provides photorealistic services where humans can experience different virtual landscapes through avatars derived from abundant personal user data. To address significant concerns about privacy breaches in private digital assets, access control based on cryptography systems has become the focus of common research. However, existing designs have scalability and efficiency issues, appealing to more investigation in real-world implementation. In response to this security challenge using the prevalent cryptography tool, this paper proposes an Attribute-Based Access Control mechanism for Metaverse Data management (ABAC-MD). It provides a flexible and secure data-sharing framework that integrates the ciphertext-policy attribute-based encryption scheme with the polynomial function technique on lattice. Reliable outsourcing decryption based on blockchain facilitates efficient data processing by employing an attribute-associated access tree. It exploits a pragmatic solution to guarantee fine-grained data privacy control and fortify resilience against quantum attacks. Simulated experiment with relevant schemes based on custom-made avatars proves that the proposed scheme reduces ciphertext size by 43.6% and improve efficiency by at least 25.4%. With higher security, lower storage costs, and reduced computational complexity, the ABAC-MD is more practical for privacy preservation in Metaverse services.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"234 ","pages":"Article 108092"},"PeriodicalIF":4.5,"publicationDate":"2025-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143394631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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}

引用次数: 0

Modeling and predicting starlink throughput with fine-grained burst characterization

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-08 DOI: 10.1016/j.comcom.2025.108090

Johan Garcia , Matthias Beckerle , Simon Sundberg , Anna Brunstrom

Leveraging a dataset of almost half a billion packets with high-precision packet times and sizes, we extract characteristics of the bursts emitted over Starlink’s Ethernet interface. The structure of these bursts directly reflects the physical layer reception of OFDMA frames on the satellite link. We study these bursts by analyzing their rates, and thus indirectly also the transition between different physical layer rates. The results highlight that there is definitive structure in the transition behavior, and we note specific behaviors such as particular transition steps associated with rate switching, and that rate switching occurs mainly to neighboring rates. We also study the joint burst rate and burst duration transitions, noting that transitions occur mainly within the same rate, and that changes in burst duration are often performed with an intermediate short burst in-between. Furthermore, we examine the configurations of the three factors burst rate, burst duration, and inter-burst silent time, which together determine the effective throughput of a Starlink connection. We perform pattern mining on these three factors, and we use the patterns to construct a dynamic N-gram model predicting the characteristics of the next upcoming burst, and by extension, the short-term future throughput. We further train a Deep Learning time-series model which shows improved prediction performance.

{"title":"Modeling and predicting starlink throughput with fine-grained burst characterization","authors":"Johan Garcia , Matthias Beckerle , Simon Sundberg , Anna Brunstrom","doi":"10.1016/j.comcom.2025.108090","DOIUrl":"10.1016/j.comcom.2025.108090","url":null,"abstract":"<div><div>Leveraging a dataset of almost half a billion packets with high-precision packet times and sizes, we extract characteristics of the bursts emitted over Starlink’s Ethernet interface. The structure of these bursts directly reflects the physical layer reception of OFDMA frames on the satellite link. We study these bursts by analyzing their rates, and thus indirectly also the transition between different physical layer rates. The results highlight that there is definitive structure in the transition behavior, and we note specific behaviors such as particular transition steps associated with rate switching, and that rate switching occurs mainly to neighboring rates. We also study the joint burst rate and burst duration transitions, noting that transitions occur mainly within the same rate, and that changes in burst duration are often performed with an intermediate short burst in-between. Furthermore, we examine the configurations of the three factors burst rate, burst duration, and inter-burst silent time, which together determine the effective throughput of a Starlink connection. We perform pattern mining on these three factors, and we use the patterns to construct a dynamic N-gram model predicting the characteristics of the next upcoming burst, and by extension, the short-term future throughput. We further train a Deep Learning time-series model which shows improved prediction performance.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"234 ","pages":"Article 108090"},"PeriodicalIF":4.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143402442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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}

引用次数: 0

An anomaly-based approach for cyber–physical threat detection using network and sensor data

IF 4.5 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications

Pub Date : 2025-02-06 DOI: 10.1016/j.comcom.2025.108087

Roberto Canonico, Giovanni Esposito, Annalisa Navarro, Simon Pietro Romano, Giancarlo Sperlí, Andrea Vignali

Integrating physical and cyber realms, Cyber–Physical Systems (CPSs) expand the potential attack surface for intruders. Given their deployment in critical infrastructures like Industrial Control Systems (ICSs), ensuring robust security is imperative. Current research has developed various Intrusion Detection techniques to identify and counter malicious activities. However, traditional methods often encounter challenges in detecting several attack types due to reliance on a single data source such as time series data from sensors and actuators. In this study, we meticulously design advanced Deep Learning (DL) anomaly-based techniques trained on either sensor/actuator data or network traffic statistics in an unsupervised setting. We evaluate these techniques on network and physical data collected concurrently from a real-world CPS. Through meticulous hyperparameter tuning, we identify the optimal parameters for each model and compare their efficiency and effectiveness in detecting different types of attacks. In addition to demonstrating superior performance compared to various baselines, we showcase the best model for each data source. Eventually, we show how utilizing diverse data sources can enhance cyber-threat detection, recognizing different kinds of attacks.

{"title":"An anomaly-based approach for cyber–physical threat detection using network and sensor data","authors":"Roberto Canonico, Giovanni Esposito, Annalisa Navarro, Simon Pietro Romano, Giancarlo Sperlí, Andrea Vignali","doi":"10.1016/j.comcom.2025.108087","DOIUrl":"10.1016/j.comcom.2025.108087","url":null,"abstract":"<div><div>Integrating physical and cyber realms, Cyber–Physical Systems (CPSs) expand the potential attack surface for intruders. Given their deployment in critical infrastructures like Industrial Control Systems (ICSs), ensuring robust security is imperative. Current research has developed various Intrusion Detection techniques to identify and counter malicious activities. However, traditional methods often encounter challenges in detecting several attack types due to reliance on a single data source such as time series data from sensors and actuators. In this study, we meticulously design advanced Deep Learning (DL) anomaly-based techniques trained on either sensor/actuator data or network traffic statistics in an unsupervised setting. We evaluate these techniques on network and physical data collected concurrently from a real-world CPS. Through meticulous hyperparameter tuning, we identify the optimal parameters for each model and compare their efficiency and effectiveness in detecting different types of attacks. In addition to demonstrating superior performance compared to various baselines, we showcase the best model for each data source. Eventually, we show how utilizing diverse data sources can enhance cyber-threat detection, recognizing different kinds of attacks.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"234 ","pages":"Article 108087"},"PeriodicalIF":4.5,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143387701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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}

引用次数: 0