Journal of Network and Computer Applications最新文献

英文中文

SDN-AAA: Towards the standard management of AAA infrastructures

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-26 DOI: 10.1016/j.jnca.2025.104114

Francisco Lopez-Gomez , Rafa Marin-Lopez , Oscar Canovas , Gabriel Lopez-Millan , Fernando Pereniguez-Garcia

Software Defined Networking (SDN) is a widely adopted technology that enables agile and flexible management of networks and services. This paradigm is a strong candidate for addressing the dynamic and secure management of large and complex Authentication, Authorization and Accounting (AAA) infrastructures. In those infrastructures, multiple nodes must securely exchange information to interconnect different realms, and the manual configuration of these nodes represents a significant point of failure and a challenge for administrators. This paper presents a novel SDN-based framework, named SDN-AAA, that follows a data model-driven approach using the YANG standard. This framework enables the dynamic management of routing and security configurations in AAA scenarios. Additionally, empirical results demonstrate that the proposed framework can handle increasing numbers of nodes without significant performance degradation in mesh and star topologies, with configuration and routing times that linearly or exponentially scale depending on the topology used. This validates the feasibility of the solution in real-world scenarios.

{"title":"SDN-AAA: Towards the standard management of AAA infrastructures","authors":"Francisco Lopez-Gomez , Rafa Marin-Lopez , Oscar Canovas , Gabriel Lopez-Millan , Fernando Pereniguez-Garcia","doi":"10.1016/j.jnca.2025.104114","DOIUrl":"10.1016/j.jnca.2025.104114","url":null,"abstract":"<div><div>Software Defined Networking (SDN) is a widely adopted technology that enables agile and flexible management of networks and services. This paradigm is a strong candidate for addressing the dynamic and secure management of large and complex Authentication, Authorization and Accounting (AAA) infrastructures. In those infrastructures, multiple nodes must securely exchange information to interconnect different realms, and the manual configuration of these nodes represents a significant point of failure and a challenge for administrators. This paper presents a novel SDN-based framework, named SDN-AAA, that follows a data model-driven approach using the YANG standard. This framework enables the dynamic management of routing and security configurations in AAA scenarios. Additionally, empirical results demonstrate that the proposed framework can handle increasing numbers of nodes without significant performance degradation in mesh and star topologies, with configuration and routing times that linearly or exponentially scale depending on the topology used. This validates the feasibility of the solution in real-world scenarios.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104114"},"PeriodicalIF":7.7,"publicationDate":"2025-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049839","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}

引用次数: 0

GridFL: A 3D-Grid-based Federated Learning framework

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-24 DOI: 10.1016/j.jnca.2025.104115

Jiagao Wu, Yudong Jiang, Zhouli Fan, Linfeng Liu

Federated Learning (FL) is an emerging distributed machine learning framework that enables a large number of devices to train machine learning models collaboratively without sharing local data. Despite the extensive potential of FL, in practical scenarios, different characteristics of clients lead to the presence of different heterogeneity in resources, data distribution, and data quantity, which poses a challenge for the training of FL. To address this problem, in this paper, we first conduct an exhaustive experimental study on all three kinds of heterogeneity in FL and provide insights into the specific impact of heterogeneity on training performance. Subsequently, we propose GridFL, a 3D-grid-based FL framework, where the three kinds of heterogeneity are defined as three dimensions (i.e., dimensions of training speed, data distribution, and data quantity) independently, and all clients in FL training are assigned to corresponding cells of the 3D grid by a gridding algorithm based on K-means clustering. In addition, we propose a grid scheduling algorithm with a dynamic selection strategy, which can select an optimal subset of clients to participate in FL training per round by adopting different strategies for different dimensions and cells. The simulation experiments show that GridFL exhibits superior performance in heterogeneous environments and outperforms several related state-of-the-art FL algorithms. Thus, the effectiveness of the proposed algorithms and strategies in GridFL are verified.

{"title":"GridFL: A 3D-Grid-based Federated Learning framework","authors":"Jiagao Wu, Yudong Jiang, Zhouli Fan, Linfeng Liu","doi":"10.1016/j.jnca.2025.104115","DOIUrl":"10.1016/j.jnca.2025.104115","url":null,"abstract":"<div><div>Federated Learning (FL) is an emerging distributed machine learning framework that enables a large number of devices to train machine learning models collaboratively without sharing local data. Despite the extensive potential of FL, in practical scenarios, different characteristics of clients lead to the presence of different heterogeneity in resources, data distribution, and data quantity, which poses a challenge for the training of FL. To address this problem, in this paper, we first conduct an exhaustive experimental study on all three kinds of heterogeneity in FL and provide insights into the specific impact of heterogeneity on training performance. Subsequently, we propose GridFL, a 3D-grid-based FL framework, where the three kinds of heterogeneity are defined as three dimensions (i.e., dimensions of training speed, data distribution, and data quantity) independently, and all clients in FL training are assigned to corresponding cells of the 3D grid by a gridding algorithm based on K-means clustering. In addition, we propose a grid scheduling algorithm with a dynamic selection strategy, which can select an optimal subset of clients to participate in FL training per round by adopting different strategies for different dimensions and cells. The simulation experiments show that GridFL exhibits superior performance in heterogeneous environments and outperforms several related state-of-the-art FL algorithms. Thus, the effectiveness of the proposed algorithms and strategies in GridFL are verified.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104115"},"PeriodicalIF":7.7,"publicationDate":"2025-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143083156","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}

引用次数: 0

STARNeT: Multidimensional spatial–temporal attention recall network for accurate encrypted traffic classification

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-22 DOI: 10.1016/j.jnca.2025.104109

Xinjie Guan, Shuyan Zhu, Xili Wan, Yaping Wu

Network traffic classification is crucial for monitoring network health, detecting malicious activities, and ensuring Quality-of-Service (QoS). The use of dynamic ports and encryption complicates the process, rendering traditional port-based or payload-based classification methods ineffective. Conventional machine learning and statistical approaches often depend on manual feature or pattern extraction by experts, leading to inefficiencies and potential inaccuracies. Deep learning offers a promising alternative, with its inherent capability to autonomously extract patterns and features from data. Nonetheless, the design of existing deep learning models often limits them to high-level semantic feature extraction, neglecting the rich multidimensional spatial and temporal information in network traffic. To address these limitations, this paper introduces STARNet, a deep learning-based model for encrypted traffic classification. STARNet incorporates a dual-stream pathway network architecture that optimizes feature extraction from each pathway. It also features a novel spatiotemporal multidimensional semantic feature recall mechanism, designed to enrich the model’s analytical depth by retaining important information that might be missed when focusing solely on high-level features. Evaluated on two public network traffic datasets, STARNet demonstrates superior accuracy in traffic classification tasks, highlighting its potential to enhance network monitoring and security.

{"title":"STARNeT: Multidimensional spatial–temporal attention recall network for accurate encrypted traffic classification","authors":"Xinjie Guan, Shuyan Zhu, Xili Wan, Yaping Wu","doi":"10.1016/j.jnca.2025.104109","DOIUrl":"10.1016/j.jnca.2025.104109","url":null,"abstract":"<div><div>Network traffic classification is crucial for monitoring network health, detecting malicious activities, and ensuring Quality-of-Service (QoS). The use of dynamic ports and encryption complicates the process, rendering traditional port-based or payload-based classification methods ineffective. Conventional machine learning and statistical approaches often depend on manual feature or pattern extraction by experts, leading to inefficiencies and potential inaccuracies. Deep learning offers a promising alternative, with its inherent capability to autonomously extract patterns and features from data. Nonetheless, the design of existing deep learning models often limits them to high-level semantic feature extraction, neglecting the rich multidimensional spatial and temporal information in network traffic. To address these limitations, this paper introduces STARNet, a deep learning-based model for encrypted traffic classification. STARNet incorporates a dual-stream pathway network architecture that optimizes feature extraction from each pathway. It also features a novel spatiotemporal multidimensional semantic feature recall mechanism, designed to enrich the model’s analytical depth by retaining important information that might be missed when focusing solely on high-level features. Evaluated on two public network traffic datasets, STARNet demonstrates superior accuracy in traffic classification tasks, highlighting its potential to enhance network monitoring and security.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104109"},"PeriodicalIF":7.7,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049872","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}

引用次数: 0

Quality-aware multi-task allocation based on location importance in mobile crowdsensing

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-22 DOI: 10.1016/j.jnca.2025.104113

Yuping Liu , Honglong Chen , Xiang Liu , Wentao Wei , Guoqi Ma , Xiaolong Liu , Duannan Ye

Mobile crowdsensing (MCS) is a new data acquisition mode, which recruits the appropriate mobile users to complete the sensing tasks based on each task’s relevant attributes. With the budget constraints, each task can only be allocated to a limited number of users. To improve the total sensing quality, the MCS platform should employ more users for important sensing tasks. Location information is a crucial parameter for evaluating the task’s importance. Previous works have only considered location as an attribute of tasks without fully examining the impact of location information on task allocation, which is extremely significant. In this paper, we study the problem of quality-aware multi-task allocation based on location importance (QMLI) in mobile crowdsensing, which considers the impact of location information on task allocation to maximize the sensing quality. Moreover, we convert the analysis of location importance into a graph theory problem and propose a location importance evaluation method, which can analyze the importance of each subarea based on different location information. The QMLI problem is proved to be NP-hard, and two task allocation algorithms are proposed to obtain near-optimal solutions. We conduct the performance evaluation based on both the simulation and real-world dataset to illustrate the effectiveness of the proposed approaches.

{"title":"Quality-aware multi-task allocation based on location importance in mobile crowdsensing","authors":"Yuping Liu , Honglong Chen , Xiang Liu , Wentao Wei , Guoqi Ma , Xiaolong Liu , Duannan Ye","doi":"10.1016/j.jnca.2025.104113","DOIUrl":"10.1016/j.jnca.2025.104113","url":null,"abstract":"<div><div>Mobile crowdsensing (MCS) is a new data acquisition mode, which recruits the appropriate mobile users to complete the sensing tasks based on each task’s relevant attributes. With the budget constraints, each task can only be allocated to a limited number of users. To improve the total sensing quality, the MCS platform should employ more users for important sensing tasks. Location information is a crucial parameter for evaluating the task’s importance. Previous works have only considered location as an attribute of tasks without fully examining the impact of location information on task allocation, which is extremely significant. In this paper, we study the problem of quality-aware multi-task allocation based on location importance (QMLI) in mobile crowdsensing, which considers the impact of location information on task allocation to maximize the sensing quality. Moreover, we convert the analysis of location importance into a graph theory problem and propose a location importance evaluation method, which can analyze the importance of each subarea based on different location information. The QMLI problem is proved to be NP-hard, and two task allocation algorithms are proposed to obtain near-optimal solutions. We conduct the performance evaluation based on both the simulation and real-world dataset to illustrate the effectiveness of the proposed approaches.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104113"},"PeriodicalIF":7.7,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049840","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}

引用次数: 0

FastTSS: Accelerating tuple space search for fast packet classification in virtual SDN switches

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-22 DOI: 10.1016/j.jnca.2025.104112

Bing Xiong , Jing Wu , Guanglong Hu , Jin Zhang , Baokang Zhao , Keqin Li

The increasing tendency of network virtualization gives rise to extensive deployments of virtual switches in various virtualized platforms. However, virtual switches are encountered with severe performance bottlenecks with regards to their packet classification especially in the paradigm of Software-Defined Networking (SDN). This paper is thus motivated to design a fast packet classification scheme based on accelerated tuple space search, named as FastTSS, for virtual SDN switches. In particular, we devise a well-exploited cache with active exact flows to directly retrieve respective flow entries for most incoming packets, in virtue of the temporal locality of network traffic. Furthermore, we propose a novel hash algorithm to resolve the hash collisions of the cache, by providing three candidate locations for each inserted flow and making room for conflicting flow through kicking operation. As for the case of cache miss, we utilize the spatial locality of packet traffic to accelerate tuple space search, by dynamically sorting all tuples in terms of their reference frequencies and load factors. Eventually, we evaluate our designed packet classification scheme with physical network traffic traces by experiments. Extensive experimental results indicate that our designed FastTSS scheme outperforms the state-of-the-art ones with stable cache hit rates around 85% and the speedup of average search length up to 2.3, significantly promoting the packet classification performance of virtual SDN switches.

{"title":"FastTSS: Accelerating tuple space search for fast packet classification in virtual SDN switches","authors":"Bing Xiong , Jing Wu , Guanglong Hu , Jin Zhang , Baokang Zhao , Keqin Li","doi":"10.1016/j.jnca.2025.104112","DOIUrl":"10.1016/j.jnca.2025.104112","url":null,"abstract":"<div><div>The increasing tendency of network virtualization gives rise to extensive deployments of virtual switches in various virtualized platforms. However, virtual switches are encountered with severe performance bottlenecks with regards to their packet classification especially in the paradigm of Software-Defined Networking (SDN). This paper is thus motivated to design a fast packet classification scheme based on accelerated tuple space search, named as FastTSS, for virtual SDN switches. In particular, we devise a well-exploited cache with active exact flows to directly retrieve respective flow entries for most incoming packets, in virtue of the temporal locality of network traffic. Furthermore, we propose a novel hash algorithm to resolve the hash collisions of the cache, by providing three candidate locations for each inserted flow and making room for conflicting flow through kicking operation. As for the case of cache miss, we utilize the spatial locality of packet traffic to accelerate tuple space search, by dynamically sorting all tuples in terms of their reference frequencies and load factors. Eventually, we evaluate our designed packet classification scheme with physical network traffic traces by experiments. Extensive experimental results indicate that our designed FastTSS scheme outperforms the state-of-the-art ones with stable cache hit rates around 85% and the speedup of average search length up to 2.3, significantly promoting the packet classification performance of virtual SDN switches.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104112"},"PeriodicalIF":7.7,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143083157","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}

引用次数: 0

DRaft: A double-layer structure for Raft consensus mechanism

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-20 DOI: 10.1016/j.jnca.2025.104111

Jiaze Shang, Tianbo Lu, Yingjie Cai, Yanfang Li

The Raft consensus algorithm is based on the design of the leader, which simplifies the replication of logs and node changes. Unfortunately, the heavy responsibility of system interaction, including receiving requests from clients, transmitting heartbeats and entries, falls solely on the leader. A design with a strong leader can lead to an imbalance in the workload of nodes, thereby causing limited scalability. Additionally, the replication of a sole entry imposes constraints on the throughput.

To alleviate these bottlenecks, we put forward a new solution, DRaft, which employs a double-layer architecture and multi-entry replication. To enable DRaft, we revamp the leader change mechanism by introducing Fi-leader and Se-leaders. Moreover, we incorporate a cache-buffer module into DRaft to enable concurrent entry replication. We present a theoretical framework composed of the CPF and CNF models to analyze the consensus success probability of DRaft. We expand DRaft to multi-layer Raft, and discover that the relationship between communication complexity and the number of nodes is proportional. Finally, we implement and evaluate DRaft, comparing its throughput and latency with those of BRaft and Engraft. We show that when 12K TPS is achieved, the latency of BRaft is twice that of DRaft.

{"title":"DRaft: A double-layer structure for Raft consensus mechanism","authors":"Jiaze Shang, Tianbo Lu, Yingjie Cai, Yanfang Li","doi":"10.1016/j.jnca.2025.104111","DOIUrl":"10.1016/j.jnca.2025.104111","url":null,"abstract":"<div><div>The Raft consensus algorithm is based on the design of the leader, which simplifies the replication of logs and node changes. Unfortunately, the heavy responsibility of system interaction, including receiving requests from clients, transmitting heartbeats and entries, falls solely on the leader. A design with a strong leader can lead to an imbalance in the workload of nodes, thereby causing limited scalability. Additionally, the replication of a sole entry imposes constraints on the throughput.</div><div>To alleviate these bottlenecks, we put forward a new solution, DRaft, which employs a double-layer architecture and multi-entry replication. To enable DRaft, we revamp the leader change mechanism by introducing Fi-leader and Se-leaders. Moreover, we incorporate a cache-buffer module into DRaft to enable concurrent entry replication. We present a theoretical framework composed of the CPF and CNF models to analyze the consensus success probability of DRaft. We expand DRaft to multi-layer Raft, and discover that the relationship between communication complexity and the number of nodes is proportional. Finally, we implement and evaluate DRaft, comparing its throughput and latency with those of BRaft and Engraft. We show that when 12K TPS is achieved, the latency of BRaft is twice that of DRaft.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104111"},"PeriodicalIF":7.7,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049626","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}

引用次数: 0

Internet of Things botnets: A survey on Artificial Intelligence based detection techniques

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-17 DOI: 10.1016/j.jnca.2025.104110

Moemedi Lefoane, Ibrahim Ghafir, Sohag Kabir, Irfan-Ullah Awan

The Internet of Things (IoT) is a game changer when it comes to digitisation across industries. The Fourth Industrial Revolution (4IR), brought about a paradigm shift indeed, unlocking possibilities and taking industries to greater heights never reached before in terms of cost saving and improved performance leading to increased productivity and profits, just to mention a few. While there are more benefits provided by IoT, there are challenges arising from the complexities, limitations and requirements of IoT and key enabling technologies. Distributed Denial of Service (DDoS) attacks are among the most prevalent and dominant cyber-attacks that have been making headlines repeatedly in recent years. IoT technology has increasingly become the preferred technology for delivering these cyber-attacks. It does not come as a surprise that IoT devices are an attractive target for adversaries, as they are easy to compromise due to inherent limitations and given that they are deployed in large numbers. This paper reviews IoT botnet detection approaches proposed in recent years. Furthermore, IoT ecosystem components are outlined, revealing their challenges, limitations and key requirements that are vital to securing the whole ecosystem. These include cloud computing, Machine Learning (ML) and emerging wireless technologies: 5G and 6G.

{"title":"Internet of Things botnets: A survey on Artificial Intelligence based detection techniques","authors":"Moemedi Lefoane, Ibrahim Ghafir, Sohag Kabir, Irfan-Ullah Awan","doi":"10.1016/j.jnca.2025.104110","DOIUrl":"10.1016/j.jnca.2025.104110","url":null,"abstract":"<div><div>The Internet of Things (IoT) is a game changer when it comes to digitisation across industries. The Fourth Industrial Revolution (4IR), brought about a paradigm shift indeed, unlocking possibilities and taking industries to greater heights never reached before in terms of cost saving and improved performance leading to increased productivity and profits, just to mention a few. While there are more benefits provided by IoT, there are challenges arising from the complexities, limitations and requirements of IoT and key enabling technologies. Distributed Denial of Service (DDoS) attacks are among the most prevalent and dominant cyber-attacks that have been making headlines repeatedly in recent years. IoT technology has increasingly become the preferred technology for delivering these cyber-attacks. It does not come as a surprise that IoT devices are an attractive target for adversaries, as they are easy to compromise due to inherent limitations and given that they are deployed in large numbers. This paper reviews IoT botnet detection approaches proposed in recent years. Furthermore, IoT ecosystem components are outlined, revealing their challenges, limitations and key requirements that are vital to securing the whole ecosystem. These include cloud computing, Machine Learning (ML) and emerging wireless technologies: 5G and 6G.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104110"},"PeriodicalIF":7.7,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049878","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}

引用次数: 0

Fuzzy neural network based access selection in satellite–terrestrial integrated networks

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-17 DOI: 10.1016/j.jnca.2025.104108

Weiwei Jiang , Yafeng Zhan , Xin Fang

Access selection has become a significant problem in satellite–terrestrial integrated networks (STINs) to determine the most suitable network. Existing solutions fail to solve the complexity and diversity challenges when user preferences are considered. In this study, the access selection problem in satellite–terrestrial integrated networks is considered, and user preferences for different network types are incorporated into the access selection decision-making process. This paper introduces fuzzy neural network (FNN) for access selection in STINs and contributes an improved FNN model with the African Vulture optimization algorithm to solve the access selection problem, which is proven to be better than the three sophisticated baselines in terms of convergence speed, blocking rate, system throughput, and user satisfaction. Compared with the traditional Fuzzy-Logic baseline, the proposed FNN model achieves an approximate 8% lower blocking rate, an approximate 40% higher system throughput, and an approximate 8% higher user satisfaction with an arrival rate of 10 requests per second in numerical experiments.

{"title":"Fuzzy neural network based access selection in satellite–terrestrial integrated networks","authors":"Weiwei Jiang , Yafeng Zhan , Xin Fang","doi":"10.1016/j.jnca.2025.104108","DOIUrl":"10.1016/j.jnca.2025.104108","url":null,"abstract":"<div><div>Access selection has become a significant problem in satellite–terrestrial integrated networks (STINs) to determine the most suitable network. Existing solutions fail to solve the complexity and diversity challenges when user preferences are considered. In this study, the access selection problem in satellite–terrestrial integrated networks is considered, and user preferences for different network types are incorporated into the access selection decision-making process. This paper introduces fuzzy neural network (FNN) for access selection in STINs and contributes an improved FNN model with the African Vulture optimization algorithm to solve the access selection problem, which is proven to be better than the three sophisticated baselines in terms of convergence speed, blocking rate, system throughput, and user satisfaction. Compared with the traditional Fuzzy-Logic baseline, the proposed FNN model achieves an approximate 8% lower blocking rate, an approximate 40% higher system throughput, and an approximate 8% higher user satisfaction with an arrival rate of 10 requests per second in numerical experiments.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104108"},"PeriodicalIF":7.7,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049876","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}

引用次数: 0

A novel staged training strategy leveraging knowledge distillation and model fusion for heterogeneous federated learning

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-17 DOI: 10.1016/j.jnca.2025.104104

Debao Wang, Shaopeng Guan, Ruikang Sun

Client-side data heterogeneity poses a significant challenge in Federated Learning (FL), limiting the effectiveness of global models. To address this, we propose a staged training approach combining Knowledge Distillation and model fusion. First, a regularized KD technique trains a robust teacher model on the server, transferring knowledge to student models to enhance convergence and reduce overfitting. Then, an adaptive parameter assignment mechanism intelligently combines the local and global models, enabling clients to integrate global knowledge with local features for improved accuracy. Experimental results on multiple image classification datasets demonstrate that our approach outperforms existing algorithms in both convergence speed and accuracy, particularly in highly heterogeneous scenarios. It effectively balances the global model’s generalization and local personalization, providing a robust solution for FL.

引用次数: 0

Efficient privacy-preserving ML for IoT: Cluster-based split federated learning scheme for non-IID data

IF 7.7 2区计算机科学 Q1 COMPUTER SCIENCE, HARDWARE & ARCHITECTURE

Journal of Network and Computer Applications

Pub Date : 2025-01-17 DOI: 10.1016/j.jnca.2025.104105

Mohamad Arafeh , Mohamad Wazzeh , Hani Sami , Hakima Ould-Slimane , Chamseddine Talhi , Azzam Mourad , Hadi Otrok

In this paper, we propose a solution to address the challenges of varying client resource capabilities in the IoT environment when using the SplitFed architecture for training models without compromising user privacy. Federated Learning (FL) and Split Learning (SL) are technologies designed to maintain privacy in distributed machine learning training. While FL generally offers faster training, it requires clients to train the entire neural network model, which may not be feasible for resource-limited IoT devices. Additionally, FL’s performance is heavily impacted by client data distribution and struggles with non-Independent and Identically Distributed (non-IID) data. In parallel, SL offloads part of the training to a server, enabling weak devices to participate by training only portions of the model. However, SL performs slower due to forced synchronization between the server and clients. Combining FL and SL can mitigate each approach’s limitations but also introduce new challenges. For instance, integrating FL’s parallelism into SL brings issues such as non-IID data and stragglers, where faster devices must wait for slower ones to complete their tasks. To address these challenges, we propose a novel two-stage clustering scheme: the first stage addresses non-IID clients by grouping them based on their weights, while the second stage clusters clients with similar capabilities to ensure that faster clients do not have to wait excessively for slower ones. To further optimize our approach, we develop a multi-objective client selection solution, which is solved using a genetic algorithm to select the most suitable clients for each training round based on their model contribution and resource availability. Our experimental evaluations demonstrate the superiority of our approach, achieving higher accuracy in less time compared to several benchmarks.

{"title":"Efficient privacy-preserving ML for IoT: Cluster-based split federated learning scheme for non-IID data","authors":"Mohamad Arafeh , Mohamad Wazzeh , Hani Sami , Hakima Ould-Slimane , Chamseddine Talhi , Azzam Mourad , Hadi Otrok","doi":"10.1016/j.jnca.2025.104105","DOIUrl":"10.1016/j.jnca.2025.104105","url":null,"abstract":"<div><div>In this paper, we propose a solution to address the challenges of varying client resource capabilities in the IoT environment when using the SplitFed architecture for training models without compromising user privacy. Federated Learning (FL) and Split Learning (SL) are technologies designed to maintain privacy in distributed machine learning training. While FL generally offers faster training, it requires clients to train the entire neural network model, which may not be feasible for resource-limited IoT devices. Additionally, FL’s performance is heavily impacted by client data distribution and struggles with non-Independent and Identically Distributed (non-IID) data. In parallel, SL offloads part of the training to a server, enabling weak devices to participate by training only portions of the model. However, SL performs slower due to forced synchronization between the server and clients. Combining FL and SL can mitigate each approach’s limitations but also introduce new challenges. For instance, integrating FL’s parallelism into SL brings issues such as non-IID data and stragglers, where faster devices must wait for slower ones to complete their tasks. To address these challenges, we propose a novel two-stage clustering scheme: the first stage addresses non-IID clients by grouping them based on their weights, while the second stage clusters clients with similar capabilities to ensure that faster clients do not have to wait excessively for slower ones. To further optimize our approach, we develop a multi-objective client selection solution, which is solved using a genetic algorithm to select the most suitable clients for each training round based on their model contribution and resource availability. Our experimental evaluations demonstrate the superiority of our approach, achieving higher accuracy in less time compared to several benchmarks.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"236 ","pages":"Article 104105"},"PeriodicalIF":7.7,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143049877","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}

引用次数: 0

首页上一页

下一页尾页

类型

全部化学•材料生命科学医学物理工程技术环境•农林材料科学地球科学法学管理学化学环境科学与生态学计算机科学教育学经济学农林科学人文科学生物学数学物理与天体物理心理学综合性期刊其他工业工程理学历史学农学文学信息工程

数据库

全部 ACS Publications Elsevier ieeexplore Springer The Royal Society of Chemistry Wiley

期刊

Journal of Network and Computer Applications

全部 Acc. Chem. Res. ACS Applied Bio Materials ACS Appl. Electron. Mater. ACS Appl. Energy Mater. ACS Appl. Mater. Interfaces ACS Appl. Nano Mater. ACS Appl. Polym. Mater. ACS BIOMATER-SCI ENG ACS Catal. ACS Cent. Sci. ACS Chem. Biol. ACS Chemical Health & Safety ACS Chem. Neurosci. ACS Comb. Sci. ACS Earth Space Chem. ACS Energy Lett. ACS Infect. Dis. ACS Macro Lett. ACS Mater. Lett. ACS Med. Chem. Lett. ACS Nano ACS Omega ACS Photonics ACS Sens. ACS Sustainable Chem. Eng. ACS Synth. Biol. Anal. Chem. BIOCHEMISTRY-US Bioconjugate Chem. BIOMACROMOLECULES Chem. Res. Toxicol. Chem. Rev. Chem. Mater. CRYST GROWTH DES ENERG FUEL Environ. Sci. Technol. Environ. Sci. Technol. Lett. Eur. J. Inorg. Chem. IND ENG CHEM RES Inorg. Chem. J. Agric. Food. Chem. J. Chem. Eng. Data J. Chem. Educ. J. Chem. Inf. Model. J. Chem. Theory Comput. J. Med. Chem. J. Nat. Prod. J PROTEOME RES J. Am. Chem. Soc. LANGMUIR MACROMOLECULES Mol. Pharmaceutics Nano Lett. Org. Lett. ORG PROCESS RES DEV ORGANOMETALLICS J. Org. Chem. J. Phys. Chem. J. Phys. Chem. A J. Phys. Chem. B J. Phys. Chem. C J. Phys. Chem. Lett. Analyst Anal. Methods Biomater. Sci. Catal. Sci. Technol. Chem. Commun. Chem. Soc. Rev. CHEM EDUC RES PRACT CRYSTENGCOMM Dalton Trans. Energy Environ. Sci. ENVIRON SCI-NANO ENVIRON SCI-PROC IMP ENVIRON SCI-WAT RES Faraday Discuss. Food Funct. Green Chem. Inorg. Chem. Front. Integr. Biol. J. Anal. At. Spectrom. J. Mater. Chem. A J. Mater. Chem. B J. Mater. Chem. C Lab Chip Mater. Chem. Front. Mater. Horiz. MEDCHEMCOMM Metallomics Mol. Biosyst. Mol. Syst. Des. Eng. Nanoscale Nanoscale Horiz. Nat. Prod. Rep. New J. Chem. Org. Biomol. Chem. Org. Chem. Front. PHOTOCH PHOTOBIO SCI PCCP Polym. Chem.

﹀