Computer NetworksPub Date : 2025-06-18DOI: 10.1016/j.comnet.2025.111461
Chen Li , Tingwei Ye , Tongyu Zong , Liyang Sun , Houwei Cao , Yong Liu
{"title":"Coffee: Cost-effective edge caching for live 360 degree video streaming","authors":"Chen Li , Tingwei Ye , Tongyu Zong , Liyang Sun , Houwei Cao , Yong Liu","doi":"10.1016/j.comnet.2025.111461","DOIUrl":"10.1016/j.comnet.2025.111461","url":null,"abstract":"<div><div>While live 360 degree video streaming delivers immersive viewing experience, it poses significant bandwidth and latency challenges for content delivery networks. Edge servers are expected to play an important role in facilitating live streaming of 360 degree videos. In this paper, we propose a novel predictive edge caching framework (Coffee) for live 360 degree video that employs collaborative FoV prediction and predictive tile prefetching to reduce bandwidth consumption and streaming cost, and improve streaming quality and robustness. By utilizing the viewers’ playback latency gaps and exploiting the unique tile consumption patterns of live 360 degree video streaming, our efficient caching algorithms achieve substantial tile caching gains. Through extensive experiments driven by real 360 degree video streaming traces, we demonstrate that edge caching algorithms specifically designed for live 360 degree video streaming can achieve high streaming cost savings with small edge cache space consumption. Coffee, guided by viewer FoV predictions, significantly reduces backhaul traffic by 76% compared to state-of-the-art live 360 edge caching algorithms. In addition, we design a transcoding-aware edge caching variant, called TransCoffee. We assess TransCoffee through extensive experiments, which reveal that it can reduce costs by 63% compared to cutting-edge transcoding-aware methods.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111461"},"PeriodicalIF":4.4,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144502233","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}
Computer NetworksPub Date : 2025-06-17DOI: 10.1016/j.comnet.2025.111442
Yang Li , Jun Xu , Dejun Yang
{"title":"FLCom: Robust federated learning against strong model poisoning attacks","authors":"Yang Li , Jun Xu , Dejun Yang","doi":"10.1016/j.comnet.2025.111442","DOIUrl":"10.1016/j.comnet.2025.111442","url":null,"abstract":"<div><div>Federated learning (FL) is an emerging distributed machine learning framework that enables models to be trained on multiple decentralized devices or servers without transferring data to a centralized server. However, due to its distributed nature, FL is vulnerable to attacks from malicious clients. Although most Byzantine-robust FL methods are designed against model poisoning attacks, they lose effectiveness as the intensity of attacks increases or when new attack strategies emerge. To address these challenges, we propose a novel robust FL method, called FLCom, which leverages outlier detection to defend against model poisoning attacks. FLCom enhances the robustness of FL and outperforms the state-of-the-art methods in accuracy. Additionally, we propose an improved model poisoning attack, called vector-scaling attack (VSA), which exhibits stronger stealthiness against robust aggregation methods. We evaluate both our defense and attack methods under IID and Non-IID settings across three different datasets. The results demonstrate that FLCom achieves higher accuracy than other methods under various attacks, particularly in the Non-IID case. Furthermore, FLCom effectively defends against our proposed VSA, while VSA successfully breaches existing defense mechanisms.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111442"},"PeriodicalIF":4.4,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144322618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"DCDTS: Deterministic cross-domain transmission and scheduling for large-scale deterministic networks","authors":"Xu Huang , Jia Chen , Deyun Gao , Shang Liu , Shangbing Qiao , Hongke Zhang","doi":"10.1016/j.comnet.2025.111434","DOIUrl":"10.1016/j.comnet.2025.111434","url":null,"abstract":"<div><div>Deterministic Networking (DetNet) is emerging to support the deterministic transmission in the Large-scale Deterministic Networks (LDNs). Recent proposals focus on the cooperation of different shaping mechanisms in Time Sensitive Networks (TSN) and DetNet to achieve the deterministic cross-domain in LDNs. However, the queue overflow caused by multi-domain access and the uncontrolled micro bursts during cross-domain remain challenges. To address this dilemma, a deterministic cross-domain scheme for LDNs is essential. In this paper, we design the Deterministic Cross-Domain Transmission and Scheduling (DCDTS), where the deterministic cross-domain cycle mapping scheme is proposed to fulfill the one-to-one cycle mapping between different domains and the cooperation of mechanisms in TSN and DetNet. In addition, we propose the Cyclic Queuing Clusters Forwarding (CQCF) mechanism to solve the effect of micro bursts during transmission in cross-domain and DetNet domains. Furthermore, we design the Hybrid Greedy-DDQN-based Traffic Scheduling (HGDTS) algorithm, which integrates the greedy and double deep Q-network into a two-step scheduling. The prototype and simulation experiments show that CQCF outperforms DIP in the number of successfully scheduled flows by approximately 22.4%. Moreover, HGDTS improves the schedulability, resource utilization, and convergence speed compared to the six baseline algorithms.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111434"},"PeriodicalIF":4.4,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144480669","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}
Computer NetworksPub Date : 2025-06-17DOI: 10.1016/j.comnet.2025.111435
Ronghao Ma , Jianhong Zhou , Maode Ma
{"title":"5G-GRAKA: An efficient group based authentication and key agreement protocol for machine-type communication in 5G networks","authors":"Ronghao Ma , Jianhong Zhou , Maode Ma","doi":"10.1016/j.comnet.2025.111435","DOIUrl":"10.1016/j.comnet.2025.111435","url":null,"abstract":"<div><div>Massive machine type communication (mMTC) is one of the important parts of the fifth-generation (5G) cellular wireless network. In order to meet the security requirements of 5G wireless networks, 3GPP has introduced an authentication and key agreement (AKA) protocol named 5G-AKA; however, it is still inefficient for the mMTC scenario where numerous devices attempt to connect to the network simultaneously. In this paper, we propose a new group-based AKA protocol, which authenticates multiple MTC devices (MTCDs) simultaneously while maintaining consistency with the 5G-AKA framework to ensure security. Specifically, we design a group authentication and key negotiation algorithm based on the challenge-response mechanism used in 5G-AKA protocol and dynamically managed group members to facilitate group authentication. This approach effectively reduces the volume of interactive messages, alleviates signaling congestion, and simultaneously completes key negotiation for multiple MTCDs. The ability of the proposed protocol against significant malicious attacks has been rigorously validated by the deviation of BAN logic and formally verified by the Random Oracle Model (ROM) and the Scyther tool, highlighting its robust security attributes. Extensive simulation experimental results have demonstrated the security, efficiency, and effectiveness of proposed protocol.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"270 ","pages":"Article 111435"},"PeriodicalIF":4.4,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144517252","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}
Computer NetworksPub Date : 2025-06-17DOI: 10.1016/j.comnet.2025.111449
Huaijie Jiang , Guang Cheng , Li Deng
{"title":"TAO: A real-time network traffic analysis task orchestration framework with optimized filtering and scheduling","authors":"Huaijie Jiang , Guang Cheng , Li Deng","doi":"10.1016/j.comnet.2025.111449","DOIUrl":"10.1016/j.comnet.2025.111449","url":null,"abstract":"<div><div>Efficient real-time network traffic analysis is vital for ensuring security and operational effectiveness. Existing traffic analysis frameworks, including holistic and fully decoupled designs, struggle to provide both optimal logical reuse and fine-grained resource allocation, resulting in inefficiencies. To address these challenges, we introduce TAO, a high-performance task orchestration framework that merges the benefits of holistic and decoupled designs for real-time network traffic analysis. TAO separates analysis targets from processing logic, facilitating flexible task scheduling and optimized resource allocation. By generating directed acyclic task graph, TAO minimizes forwarding of shared traffic and employs an innovative packet filtering optimization method using statistical features from a prioritized tree. Additionally, we develop a heuristic scheduling approach that leverages pipeline-based scheduling to achieve comprehensive congestion control. Experimental results show that under 10 Gbps trace replay and 40 Gbps real-world traffic, TAO reduces resource consumption by up to 55% in the lab and 48% in deployment compared with baseline methods. These findings underscore TAO’s potential to significantly enhance the efficiency and scalability of network traffic processing frameworks in high-throughput environments.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"270 ","pages":"Article 111449"},"PeriodicalIF":4.4,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144571250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Dependent task offloading in multi-access edge computing: A GCN augmented deep reinforcement learning approach","authors":"Liqiong Chen, Xinyuan Yang, Huaiying Sun, Xiuchao Yu, Kaiwen Zhi","doi":"10.1016/j.comnet.2025.111445","DOIUrl":"10.1016/j.comnet.2025.111445","url":null,"abstract":"<div><div>Multi-access edge computing (MEC) is a promising distributed computing paradigm that reduces the delayed energy cost (DECC) of users by offloading user-generated application tasks to the network edge. Most of the user-generated tasks contain a series of subtasks with dependencies, how to effectively offload these interdependent tasks and reduce DECC is a key issue. In addition, most existing learning-based approaches are inadequate in dynamically modeling MEC environments and cannot effectively characterize heterogeneous MEC environments. To this end, this paper models the multiuser task offloading problem as a Markov Decision Process (MDP). First, to address the challenges of heterogeneous MEC environments, we propose a multi-dependent task offloading algorithm with state embeddings. This algorithm uses commonality and dissimilarity components to capture the interactions between user and the MEC environment, providing robust state representation. Secondly, we introduce the strategy gradient theorem of the Stackelberg game to optimize the offloading decision. Finally, extensive experiments show that our proposed method significantly reduces DECC compared to existing methods.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111445"},"PeriodicalIF":4.4,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144337929","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}
Computer NetworksPub Date : 2025-06-16DOI: 10.1016/j.comnet.2025.111482
Khaled M. Matrouk , Arunmozhi Selvi , Ahmad Yahiya Ahmad Bani Ahmad , Dhurgadevi M
{"title":"An adaptive deep learning-based user authentication with digitally signed optimal key-aided encryption for secured mobile edge computing data storage in blockchain","authors":"Khaled M. Matrouk , Arunmozhi Selvi , Ahmad Yahiya Ahmad Bani Ahmad , Dhurgadevi M","doi":"10.1016/j.comnet.2025.111482","DOIUrl":"10.1016/j.comnet.2025.111482","url":null,"abstract":"<div><div>The development of Internet of Things (IoT) gadgets boosted the need for a task computing system that is reliable and efficient. Mobile Edge Computing (MEC) is growing and has become a viable tool for proximate and dependent-on latency jobs. Edge technology is well suited to IoT applications that demand minimal latency, position understanding, and large numbers of interconnections. It certainly compensates for some deficiencies in the cloud in the fields of electrical power and immediate analysis. However, ensuring the security of data in an application context remains a significant concern. Furthermore, privacy is an issue for any computing system that contains dispersed and diverse equipment. Blockchain represents a relatively new technology that emerged as an intriguing option for ensuring integrity, safety, uniformity, and authenticity. Yet, blockchain is unable to ensure adequate privacy for information on its own. So, to prevent the privacy of IoT-based data blockchain technology was developed. The data are collected from the IoT devices. The user authentication is verified, and the data are stored in the network. Also, the Adaptive Deep Markov Random Field (ADMRF) model was used for getting the verified data. The parameters in the ADMRF are tuned with the Help of the Improved Equilibrium Optimized (IEO). Once the user authentication is verified, then the data are encrypted with the help of the Attribute-Based Encryption (ABE) technique. The encryption keys are optimally generated with the help of the IEO. The encrypted data are then digitally signed by the authorized user, and then it is stored in the blockchain. The security of the model is measured by comparing it with other existing models.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111482"},"PeriodicalIF":4.4,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144480668","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}
Computer NetworksPub Date : 2025-06-16DOI: 10.1016/j.comnet.2025.111437
Naina Said , Olaf Landsiedel
{"title":"EdgeBoost: Confidence boosting for resource constrained inference via selective offloading","authors":"Naina Said , Olaf Landsiedel","doi":"10.1016/j.comnet.2025.111437","DOIUrl":"10.1016/j.comnet.2025.111437","url":null,"abstract":"<div><div>Deploying large Deep Neural Networks with state-of-the-art accuracy on edge devices is often impractical due to their limited resources. This paper introduces <span>EdgeBoost</span>, a selective input offloading system designed to overcome the challenges of limited computational resources on edge devices. <span>EdgeBoost</span> trains and calibrates a lightweight model for deployment on the edge and, in addition, deploys a large, complex model on the cloud. During inference, the edge model makes initial predictions for input samples, and if the confidence of the prediction is low, the sample is sent to the cloud model for further processing, otherwise, we accept the local prediction. Through careful calibration, <span>EdgeBoost</span> reduces the communication cost by 55%, 27% and 20% for the CIFAR-100, ImageNet-1k and Stanford Cars datasets, respectively, when compared to an cloud-only solution while achieving on-par classification accuracy. Furthermore, <span>EdgeBoost</span> reduces the total inference latency from 148 ms to 123.84 ms per inference compared to a cloud-only solution. Our evaluation also shows that calibrating the edge model for such a collaborative edge–cloud setup results in accuracy gains of up to 8 percent point, compared to an uncalibrated edge model. Additionally, EdgeBoost, when used as an abstaining classifier, can improve accuracy by up to 9 percent points over an uncalibrated model. Finally, <span>EdgeBoost</span> outperforms the Early Exit and Entropy thresholding baselines and achieves comparable accuracy to state-of-the-art routing-based methods without the need for hosting the router on the edge.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111437"},"PeriodicalIF":4.4,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144337928","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}
Computer NetworksPub Date : 2025-06-15DOI: 10.1016/j.comnet.2025.111440
Jintao Zhao , Jialiang Yan , Siyao Cheng , Jie Liu
{"title":"BLAW: BLE Assisted Wi-Fi in idle listening","authors":"Jintao Zhao , Jialiang Yan , Siyao Cheng , Jie Liu","doi":"10.1016/j.comnet.2025.111440","DOIUrl":"10.1016/j.comnet.2025.111440","url":null,"abstract":"<div><div>Wi-Fi is commonly used in IoT devices but its high energy consumption poses a significant challenge particularly during idle listening. We propose BLAW (BLE-Assisted Wi-Fi), a novel approach that leverages the coexistence of Bluetooth Low Energy (BLE) and Wi-Fi in integrated combo modules to reduce power consumption. BLAW enables Wi-Fi devices to enter sleep state while BLE monitors incoming traffic through cross-technology communication. By embedding information via the RTS/CTS mechanism and Wi-Fi MAC address payloads, BLAW improves energy efficiency without modifying Wi-Fi protocols. Our physical layer evaluations using USRP demonstrate BLAW’s reliability. Additionally, we assess the impact of various factors on BLE recognition. Our results demonstrate that BLAW reduces energy consumption, achieving a 64.5% reduction in typical Wi-Fi idle listening energy and a 39.5% improvement over the Wi-Fi PSM mechanism. BLAW offers excellent transparency and deployability making it a promising solution for energy-constrained IoT applications.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111440"},"PeriodicalIF":4.4,"publicationDate":"2025-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144337930","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}
Computer NetworksPub Date : 2025-06-14DOI: 10.1016/j.comnet.2025.111439
YuKuan Tu, Tengyao Li, Meng Zhang, Xiangyang Luo
{"title":"MVC-Corr: An accurate and efficient flow correlation method based on multi-view fusion and contrast augmentation","authors":"YuKuan Tu, Tengyao Li, Meng Zhang, Xiangyang Luo","doi":"10.1016/j.comnet.2025.111439","DOIUrl":"10.1016/j.comnet.2025.111439","url":null,"abstract":"<div><div>Flow correlation attacks determine the access relationship between clients and services under conditions of multi-relay encrypted traffic by analyzing the behavior similarity of their data flows. However, the effectiveness of current flow correlation attacks is susceptible to traffic obfuscation and suffers from low training efficiency, which greatly restricts their application on Tor. To address this, the paper proposes an accurate and efficient flow correlation method named MVC-Corr, which is based on multi-view fusion and contrast augmentation. Firstly, a multi-view fusion feature extraction network (MVF) is designed. The network integrates three types of views: uplink–downlink interaction view, local view, and global view, to achieve precise feature extraction. Secondly, an offset intersection contrast augmentation mechanism (ICA) is developed. The mechanism generates non-correlated flow feature pairs with abundant contrast information, improving the efficiency of correlation analysis. Finally, to enhance the broad applicability of the proposed method across different target scales, MVC-Corr is designed with two operational modes, each tailored for the user tracking scenario and the user discovery scenario. The experimental results show that, in user tracking scenario, MVC-Corr outperforms three existing typical methods—DeepCorr, FlowTracker, and ResTor—in terms of accuracy, achieving improvements ranging from 13.3% to 30.9% under traffic obfuscation conditions. In user discovery scenario, experimental results demonstrate that MVC-Corr’s correlation capability surpasses that of the current state-of-the-art method, DeepCoFFEA, achieving a maximum true positive rate improvement of 2.9%.</div></div>","PeriodicalId":50637,"journal":{"name":"Computer Networks","volume":"269 ","pages":"Article 111439"},"PeriodicalIF":4.4,"publicationDate":"2025-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144471255","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}