Journal of Network and Computer Applications最新文献

{"title":"HERALD: Hybrid Ensemble Approach for Robust Anomaly Detection in encrypted DNS traffic","authors":"Umar Sa’ad ,&nbsp;Demeke Shumeye Lakew ,&nbsp;Nhu-Ngoc Dao ,&nbsp;Sungrae Cho","doi":"10.1016/j.jnca.2025.104342","DOIUrl":"10.1016/j.jnca.2025.104342","url":null,"abstract":"<div><div>The proliferation of encrypted Domain Name System (DNS) traffic through protocols like DNS over Hypertext Transfer Protocol Secure presents significant privacy advantages but creates new challenges for anomaly detection. Traditional security mechanisms that rely on payload inspection become ineffective, necessitating advanced strategies capable of detecting threats in encrypted traffic. This study introduces the Hybrid Ensemble Approach for Robust Anomaly Detection (HERALD), a novel framework designed to detect anomalies in encrypted DNS traffic. HERALD combines unsupervised base detectors, including Isolation Forest (IF), One-Class Support Vector Machine (OCSVM), and Local Outlier Factor (LOF), with a supervised Random Forest meta-model, leveraging the strengths of both paradigms. Our comprehensive evaluation demonstrates HERALD’s exceptional performance, achieving 99.99 percent accuracy, precision, recall, and F1-score on the CIRA-CIC-DoHBrw-2020 dataset, while maintaining competitive computational efficiency with 110s training time and 2.2ms inference time. HERALD also demonstrates superior generalization capabilities on cross-dataset evaluations, exhibiting minimal performance degradation of only 2-4 percent when tested on previously unseen attack patterns, outperforming purely supervised models, which showed 5-8 percent degradation. The interpretability analysis, incorporating feature importance, accumulated local effects, and local interpretable model-agnostic explanations, provides insights into the relative contributions of each base detector, with OCSVM emerging as the most influential component, followed by IF and LOF. This study advances the field of network security by offering a robust, interpretable, and adaptable solution for detecting anomalies in encrypted DNS traffic that balances a high detection rate with a low false-positive rate.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104342"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145223445","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":"A profit-effective function service pricing approach for serverless edge computing function offloading","authors":"Siyuan Liu ,&nbsp;Li Pan ,&nbsp;Shijun Liu","doi":"10.1016/j.jnca.2025.104338","DOIUrl":"10.1016/j.jnca.2025.104338","url":null,"abstract":"<div><div>In recent years, edge computing services have continued to develop and have been better integrated with serverless computing, leading to the improvement of the performance and concurrent request handling capabilities of edge servers. Therefore, an increasing number of IoT devices are willing to pay a certain amount of service processing fees to offload some computing tasks to edge servers for execution, with the aim of meeting their latency requirements. However, the computing capacity and storage space of edge servers at a single base station are still limited. Therefore, base stations must decide which task images to cache for future execution and price these computing services to control the computing offloading of IoT devices, so as to maximize their expected profit under the constraints of limited computing capacity and memory space. In this paper, we stand from the perspective of base stations and formulate the caching and pricing of function images at a base station, as well as the function offloading process of IoT devices, as a Markov Decision Process (MDP). We adopt a Proximal Policy Optimization (PPO)-based function service pricing adjustment algorithm to optimize the profit of base stations. Finally, we evaluate our approach through simulation experiments and compare it with baseline methods. The results show that our approach can significantly improve base stations’ expected profit in various scenarios.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104338"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145160067","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":"DS-RAM: A dynamic sharding and reputation-based auditing mechanisms for blockchain consensus in IIoT","authors":"Jiali Zheng,&nbsp;Jinhui Chen,&nbsp;Shuainan Liu","doi":"10.1016/j.jnca.2025.104362","DOIUrl":"10.1016/j.jnca.2025.104362","url":null,"abstract":"<div><div>Sharding is an effective strategy to improve the scalability of blockchain, especially in the context of massive data processing in Industrial Internet of Things (IIoT) scenarios. However, existing sharding schemes often overlook factors such as node reputation, resource capacity, and historical behavior, leading to imbalanced resource allocation, which in turn causes delays in real-time data processing and compromises system security. The blockchain consensus mechanism determines how nodes reach consensus, serving as the core of system efficiency and security. However, traditional consensus mechanisms lack effective detection of malicious nodes and insufficient supervision of consensus nodes, making the system vulnerable to attacks and malicious actions. To address these issues, this paper proposes DS-RAM (Dynamic Sharding and Reputation-based Auditing Mechanism), a dynamic sharding mechanism based on the weighted K-Medoids and Canopy algorithms. It comprehensively considers factors such as node geographical location, reputation, interaction frequency, and historical behavior to optimize node allocation, ensuring balanced distribution of sharding resources, thus improving system throughput and security. Additionally, DS-RAM introduces an auditing node module, which provides additional supervision of consensus nodes based on the reputation mechanism, enabling timely detection and isolation of potential malicious nodes, thereby effectively enhancing the fault tolerance of the consensus mechanism and system security. Simulation results demonstrate that, compared to traditional sharding schemes and reputation-based blockchains, the proposed method can effectively improve sharding security and blockchain sharding performance.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104362"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145261939","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":"Optimizing service level agreement in cloud computing with smart virtual machine scheduling using clustered differential evolution and deep learning","authors":"Tassawar Ali ,&nbsp;Hikmat Ullah Khan ,&nbsp;Babar Nazir ,&nbsp;Fawaz Khaled Alarfaj ,&nbsp;Mohammed Alreshoodi","doi":"10.1016/j.jnca.2025.104361","DOIUrl":"10.1016/j.jnca.2025.104361","url":null,"abstract":"<div><div>Cloud computing is expanding rapidly due to the increasing demand for scalable and efficient services. This growth necessitates more extensive physical infrastructure to accommodate the growing workload. However, managing these workloads effectively presents issues, particularly in optimizing virtual machine (VM) scheduling. Traditional reactive scheduling methods respond to workload changes only after they occur. These approaches struggle in dynamic cloud environments, leading to performance inefficiencies, frequent VM migrations, and service-level agreement (SLA) violations. The purpose of this study is to introduce IntelliSchNet, a novel VM scheduling approach designed to address these challenges. IntelliSchNet uses a deep learning model in which the feature weights of its neurons are optimized using agglomerative clustering-based differential evolution to accurately predict future workloads. Based on these predictions, an intelligent scheduling plan is created to allocate VMs to suitable hosts. The strategy prioritizes non-overloaded hosts to maximize resource utilization and reduce VM migrations, and hence minimizes SLA violations. The basic methodology includes integrating a clustered adaptation of the differential evolution algorithm to fine-tune deep neural network parameters. Real-world data from Google's datacenters is used for training, consisting of traces collected from a production cluster with over 11,000 machines and more than 650,000 jobs, ensuring reliable and practical workload predictions. The effectiveness of IntelliSchNet is evaluated using nine different performance metrics on actual cloud workload datasets. The major findings highlight a significant improvement in VM scheduling efficiency. IntelliSchNet reduces SLA violations by up to 44 %, ensuring more stable and reliable cloud services. This reduction enhances service dependability and increases customer satisfaction. In conclusion, IntelliSchNet outperforms traditional scheduling methods by optimizing workload placement and resource allocation. Its proactive approach enhances cloud system stability, efficiency, and scalability. These improvements contribute to a more sustainable and high-performing cloud computing environment.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104361"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314970","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":"3D UAV path planning based on an improved TD3 deep reinforcement learning for data collection in an urban environment","authors":"Mohammad Nazemi Jenabi,&nbsp;Hadi Asharioun,&nbsp;Mahdi Pourgholi","doi":"10.1016/j.jnca.2025.104336","DOIUrl":"10.1016/j.jnca.2025.104336","url":null,"abstract":"<div><div>With the rapid growth in the number of users and services in communication networks, unmanned aerial vehicles (UAVs) are expected to play a significant role in future wireless communication systems. One of the key applications of UAVs is data collection in Internet of Things (IoT) networks. This paper addresses a three-dimensional (3D) UAV path planning optimization problem aimed at minimizing the completion time of data collection in urban environments, taking into account real-world constraints such as frequent communication link blockages between UAVs and sensors caused by buildings. To tackle this challenge, we propose an improved Deep Reinforcement Learning (DRL) algorithm, referred to as the Dropout-Based Prioritized TD3 Algorithm (DPTD3). This method integrates the TD3 algorithm with the Prioritized Experience Replay Buffer (PER) strategy and introduces a new Actor network architecture incorporating the Dropout technique. Simulation results demonstrate that the proposed 3D UAV path planning approach reduces both data collection time and UAV energy consumption compared to a two-dimensional (2D) path planning method. Furthermore, the results indicate that during training, the DPTD3 algorithm outperforms other state-of-the-art DRL approaches in terms of both stability and performance.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104336"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145160068","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":"Trace-distance based end-to-end entanglement fidelity with information preservation in quantum networks","authors":"Pankaj Kumar,&nbsp;Binayak Kar,&nbsp;Shan-Hsiang Shen","doi":"10.1016/j.jnca.2025.104366","DOIUrl":"10.1016/j.jnca.2025.104366","url":null,"abstract":"<div><div>Quantum networks have the potential to revolutionize communication and computation by outperforming their classical counterparts. Many quantum applications depend on the reliable distribution of high-fidelity entangled pairs between distant nodes. However, due to decoherence and channel noise, entanglement fidelity degrades exponentially with distance, posing a significant challenge to maintaining robust quantum communication. To address this, we propose two strategies to enhance end-to-end (E2E) fidelity and information preservation in quantum networks. First, we employ closeness centrality to identify optimal intermediary nodes that minimize average path length. Second, we introduce the Trace-Distance based Path Purification (TDPP) algorithm, which fuses topological and quantum state information to support fidelity-aware routing decisions. TDPP leverages closeness centrality and trace-distance to identify paths that optimize both network efficiency and entanglement fidelity. Simulation results demonstrate that our approach significantly improves network throughput and E2E entanglement fidelity, outperforming existing routing methods while enhancing information preservation.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104366"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364206","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":"Lightweight verifiable privacy preserving federated learning","authors":"Li Zhang ,&nbsp;Bing Tang ,&nbsp;Jianbo Xu","doi":"10.1016/j.jnca.2025.104335","DOIUrl":"10.1016/j.jnca.2025.104335","url":null,"abstract":"<div><div>Federated learning (FL) has garnered considerable attention owing to its capability of accomplishing model training through the sharing local models without accessing training datasets. Nevertheless, it has been demonstrated that the shared models still possess sensitive information related to the training data. Moreover, there is a possibility that malicious aggregation servers can return manipulated global models. While the verification problem in FL has been explored in existing schemes, most of these schemes employ bilinear pairing operations and homomorphic hash computations dependent on the model’s dimension, leading to substantial computational costs. Additionally, some schemes necessitate multiple parties to collectively manage one or more sets of confidential keys for privacy preservation and validation, which renders them vulnerable to collusion attacks between certain clients and servers. Consequently, we propose a privacy-preserving federated learning mechanism under a dual-server architecture. This mechanism adopts a coding matrix computation-based approach to ensure the privacy security of local models at the client side and achieves the aggregation of local models through collaborative efforts between two servers situated at the server side. To verify the correctness of the aggregated model, a Model Verification Code (MVC) mechanism is designed. By effectively combining the MVC mechanism with the coded matrix computation, there is no requirement for all clients to possess identical sets of confidential keys during the privacy preservation and verification process. Meanwhile, this ensures the fulfillment of security requirements under the malicious threat posed by the server. The computational overhead of this mechanism remains low since it avoids the application of complex cryptographic primitives. We perform extensive experiments on real datasets, and the experimental results further demonstrate the proposed scheme exhibits lightweight characteristics while ensuring the validity and usability of the model.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104335"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145160066","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":"Cost-effective container elastic scaling and scheduling under multi-resource constraints","authors":"Hongjian Li ,&nbsp;Yu Tian ,&nbsp;Yuzheng Cui ,&nbsp;Xiaolin Duan","doi":"10.1016/j.jnca.2025.104359","DOIUrl":"10.1016/j.jnca.2025.104359","url":null,"abstract":"<div><div>Recent advancements in containerization and Kubernetes have solidified their status as mainstream paradigms for service delivery. However, existing Kubernetes scaling mechanisms often suffer from limitations, such as suboptimal utilization of multi-dimensional resources, reliance on historical workload patterns, and inability to adapt quickly to real-time workload fluctuations. To overcome these limitations, this study introduces two cost-effective resource scheduling strategies. First, a hybrid control-theoretic vertical scaling algorithm is proposed, operating under multi-resource constraints. This algorithm leverages Prometheus monitoring data encompassing diverse resource metrics. It facilitates dynamic resource optimization through a hierarchical decision-making model that combines feedforward prediction with feedback correction mechanisms. Second, a synergistic vertical–horizontal elastic scaling framework, namely the MR-CEHA framework proposed in this work, is developed. This framework classifies resource states using multi-level thresholds and integrates a cost-sensitive optimization model to balance instance-level resource allocation with cluster-level scaling operations. Experimental evaluations demonstrate substantial improvements: under surge load conditions, the SLA violation rate decreased by 16.5%; during load reduction scenarios, energy consumption dropped by 39.4%; and in mixed workload environments, energy usage declined by 16.6% while simultaneously achieving a 37.8% reduction in SLA violation rate. These findings contribute both to the theoretical understanding and the practical advancement of efficient resource utilization and service stability in Kubernetes-based cloud deployments, offering meaningful value for academic exploration and industrial implementation.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104359"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364129","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":"FedCoRE: Effective Federated Learning for constrained RESTful environments in the Artificial Intelligence of Things","authors":"Badis Djamaa,&nbsp;Habib Yekhlef,&nbsp;Mohamed Amine Kouda,&nbsp;Abbas Bradai","doi":"10.1016/j.jnca.2025.104357","DOIUrl":"10.1016/j.jnca.2025.104357","url":null,"abstract":"<div><div>Federated Learning (FL) empowers Internet-of-Things (IoT) devices to train intelligent models without sharing sensitive data, facilitating the transition to an Artificial Intelligence of Things (AIoT) ecosystem. However, FL demands significant storage, computation, and communication resources, which often exceed the capabilities of resource-constrained IoT devices. In this work, we introduce FedCoRE, an effective and practical FL architecture tailored for IoT environments. FedCoRE leverages standards for constrained RESTful environments, such as the Constrained Application Protocol (CoAP), to optimize communication and applies model quantization to address computation and storage limitations. FedCoRE has been implemented on resource-constrained IoT devices with 256 KB of RAM and evaluated on a human activity recognition task using a deep neural network. Extensive evaluations conducted in a real-world IoT environment, comprising 10 Thunderboard Sense 2 nodes, demonstrate the feasibility and effectiveness of our proposal. Notably, compared to FL, FedCoRE achieves up to a 60% reduction in communication cost, while maintaining model accuracy and requiring only approximately 75<!--> <!-->KB of RAM and 438<!--> <!-->KB of ROM.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104357"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145311718","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":"Enhanced spreading factor allocation and backscatter communication via membership based tuna swarm optimization for LoRa protocol","authors":"Swathika R.,&nbsp;Dilip Kumar S.M.","doi":"10.1016/j.jnca.2025.104360","DOIUrl":"10.1016/j.jnca.2025.104360","url":null,"abstract":"<div><div>With spread spectrum modulation, LoRa (Long-Range), an Internet of Things (IoT) communication method, enables ultra-long-distance transmission in the recent times. Data conflicts occur frequently in networks with many nodes, and the equivalent rate often suffers in ultra-long-distance transmissions. This work examines several kinds of data collisions in LoRa wireless networks, most of which are influenced by the assignment of the Spreading Factor (SF). The study also explores the integration of Membership based Tuna Swarm Optimization (MTSO) with LoRa modulation into Backscatter Communications (BackCom). An analytical structure is established to examine the error rate efficiency of the network simulation under consideration. With restricted network resources, MTSO is employed to implement an SF redistribution mechanism, thereby increasing the terminal capacity of the LoRa gateway. Without increasing network or gateway capacity, the proposed technique reduces the frequency of data collisions. This paper addresses the reallocation of SF as the number of terminals increases, presenting an SF selection mechanism and an iterative SF inspection method to ensure independent data rates for each communication link. Specifically, assuming canceled Radio-Frequency Interference (RFI), this paper derives new precise and estimated closed-form equations for the Bit Error Rate (BER), Symbol Error Rate (SER), and Frame Error Rate (FER). The findings show that as the Signal-To-Noise Ratio (SNR) increases, the system’s BER, FER, and SER efficiency also improve when the SF variables are tuned.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"244 ","pages":"Article 104360"},"PeriodicalIF":8.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145314969","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}