Journal of Network and Computer Applications最新文献

{"title":"A piecewise chaotic starfish optimization algorithm for energy-efficient coverage in wireless sensor networks","authors":"Muhammad Suhail Shaikh ,&nbsp;Shuwei Qiu ,&nbsp;Xiaoqing Dong ,&nbsp;Chang Wang ,&nbsp;Wulfran Fendzi Mbasso","doi":"10.1016/j.jnca.2025.104410","DOIUrl":"10.1016/j.jnca.2025.104410","url":null,"abstract":"<div><div>Enhancing coverage and reducing energy consumption are fundamental challenges in wireless sensor networks (WSNs) for high-volume and data-intensive deployment. WSNs play an important role in emerging technologies and face practical limitations, particularly related to coverage and energy consumption. Strategical placement of these sensor nodes is important to ensure service quality; however, many existing optimization algorithms for sensor node placement struggle with low coverage rate and high energy consumption. A significant issue lies in determining the optimal sensor node locations, as these significantly influence the network's coverage and energy consumption. This work presented a Piecewise Chaotic Starfish Optimization Algorithm (CSFOA) for addressing the challenge of optimizing the sensor node placement to maximize coverage and minimize energy consumption in WSNs. The integration of the piecewise chaotic map enhances the convergence and exploration capacity of the algorithm in identifying better solutions. The effectiveness of CSFOA is confirmed by a range of diverse benchmark functions as unimodal, multimodal, fixed, and variable, proving its excellence in optimization performance. CSFOA obtained better results for sensor node deployment in real test cases. For instance, in Test System 1 with 20 nodes, the coverage rate is 97.4757 % and the energy consumption is 0.29967 nJ/bit. In Test System 2 with 30 nodes, the coverage is 99.9713 % and the energy consumption is 3.2193 nJ/bit. Test System 3 with 40 nodes has a 98.8690 % coverage rate and energy consumption of 5.1107 nJ/bit. Compared to CMFO, CSSA, CPSO, SFOA, MFO, SSA, and PSO algorithms, CSFOA realizes an average improvement of 16.41 %, 5.36 %, 3.45 %, 2.371 %, 2.80 %, and 2.18 % on various evaluation metrics. These results underscore the algorithm's capability in balancing coverage and energy efficiency enhancement, and they confirm the algorithm's value as a more effective solution to sensor node deployment issues in different applications.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"247 ","pages":"Article 104410"},"PeriodicalIF":8.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145845509","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":"Seamless service migration for the Internet of Vehicles in edge computing: A dynamic dirty page filtering and two-stages compression technique","authors":"Kaifeng Hua ,&nbsp;Shengchao Su ,&nbsp;Nannan Zhang","doi":"10.1016/j.jnca.2025.104412","DOIUrl":"10.1016/j.jnca.2025.104412","url":null,"abstract":"<div><div>With the growing demand for dynamic resources in the Internet of Vehicles, service migration has become essential for maintaining user service continuity. However, existing techniques often transfer redundant dirty page data during operation state file transfers, leading to high network traffic and significant migration delays, which are unsuitable for the low latency and low traffic requirements of intelligent transportation scenarios. To overcome this issue, this paper proposes a <u>I</u>ntelligent <u>A</u>daptive <u>C</u>ontainer <u>M</u>igration <u>T</u>echnique called IACMT, which is based on dynamic filtering of dirty pages with two-stage compression. IACMT features a dirty page filtering mechanism that intelligently identifies active dirty pages by monitoring the frequency of page accesses and modification patterns in real time. This mechanism facilitates the delayed transmission of less critical dirty pages, effectively reducing the data size during the iterative transmission phase. Furthermore, it incorporates a two-stage data compression algorithm that employs run-length encoding (RLE) followed by dynamic Huffman coding. In the initial stage, RLE eliminates redundant byte sequences in the state file. The subsequent output is then adaptively compressed using a dynamic Huffman tree, improving compression efficiency while managing computational overhead. Experimental results show that IACMT reduces data transmission volume by approximately 35 % for typical in-vehicle workloads, while cutting migration time and service interruption duration by around 24 % and 34 %, respectively.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"247 ","pages":"Article 104412"},"PeriodicalIF":8.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145785053","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":"Multi-objective parrot optimizer with improved Lévy flight and adaptive elliptical segmentation - based screening mechanism for layout optimization of wireless sensor networks","authors":"Yun-Hao Zhang,&nbsp;Jie-Sheng Wang,&nbsp;Yu-Xuan Xing,&nbsp;Yu-Feng Sun,&nbsp;Si-Wen Zhang,&nbsp;Xue-Lian Bai","doi":"10.1016/j.jnca.2025.104413","DOIUrl":"10.1016/j.jnca.2025.104413","url":null,"abstract":"<div><div>With the rapid development of science and technology, wireless sensor networks (WSN) are increasingly applied in environmental monitoring, industrial control, and smart cities. However, WSN deployment faces three core challenges that existing algorithms fail to address comprehensively. (1) Insufficient coverage precision. Traditional optimization algorithms (e.g., NSGA-II, MOPSO) often leave local coverage holes due to limited fine-grained search capability. (2) High resource redundancy. Fixed grid or weight-based methods (e.g., MOEA/D) cannot dynamically adjust node distribution according to environmental density, leading to redundant deployment. (3) Unbalanced energy consumption. Single-objective or simplified multi-objective approaches ignore the trade-off between coverage, waste rate, and energy consumption, shortening network lifetime. To tackle these issues, a multi-objective parrot optimizer (MOPO) based on improved Lévy flight and an adaptive elliptical segmentation screening mechanism is proposed for WSN deployment optimization. The randomness of original Lévy flight causes large step-length jumps, making fine-grained searches difficult. Thus, a Sigmoid perturbation mechanism is integrated into Lévy flight to enhance local search accuracy while preserving global exploration. Based on this improvement, an elite non-dominated strategy is combined with an adaptive grid (dynamic adjustment by solution density) and elliptical segmentation selection—this ensures retention of optimal individuals in high-density areas, maintains population diversity, and accelerates exploration of sparse regions. An external archive further preserves a uniform and diverse Pareto solution set. MOPO is tested in obstacle-free/obstacle WSN models with coverage, waste rate, and energy consumption rate as objectives. Comparative experiments with NSGA-II, MOPSO, and MOGWO in different monitoring areas show MOPO ranks first in all Friedman tests. A real-world test (41°10′20″N, 29°04′30″E, 1320 × 610 m²) achieves 94 % target coverage. This proves MOPO effectively solves the three core challenges of WSN deployment, providing a practical and efficient optimization method for large-scale, resource-constrained WSN scenarios.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"247 ","pages":"Article 104413"},"PeriodicalIF":8.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145813864","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":"CASLO: Joint scaling and deployment for microservices leveraging context-aware SLO assignment","authors":"Shuang Zeng,&nbsp;Haitao Zhang,&nbsp;Zezhong Yan","doi":"10.1016/j.jnca.2025.104417","DOIUrl":"10.1016/j.jnca.2025.104417","url":null,"abstract":"<div><div>Service Level Objective (SLO) assignment involves distributing an application’s end-to-end latency SLO among its microservices, guiding each microservice’s resource allocation based on its assigned sub-SLO. However, existing SLO assignment methods primarily focus on horizontal scaling, overlooking the significant impact of varying resource usage contexts across nodes and container configurations on microservice latency characteristics. Moreover, these methods fail to consider how scaling decisions affect node resource usage. This oversight creates discrepancies between decision-time and runtime latency characteristics, which can lead to SLO violations or resource wastage. This paper proposes CASLO, a joint scaling and deployment method based on context-aware SLO assignment that aims to meet application SLOs with minimal resource usage. It characterizes microservice latency by categorizing influencing factors into node and container contexts, which enables the model to capture dynamic performance under varying conditions. Building on this characterization, CASLO employs Particle Swarm Optimization (PSO) to iteratively estimate each microservice’s tolerance to contextual resource conditions. For each tolerance, it determines the resource usage of each node post-scaling and deployment, addressing discrepancies of latency characteristics between decision-time and runtime and distinguishing latency characteristics across nodes. Based on the determined resource context, CASLO assigns SLOs to each microservice, dynamically configuring container resources to derive scaling and deployment decisions. Resource usage is then calculated to provide feedback to PSO for iterative optimization. Compared to state-of-the-art methods, CASLO achieves 32% reduction in resource usage and decreases the frequency of SLO violations by 61%.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"247 ","pages":"Article 104417"},"PeriodicalIF":8.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145785050","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":"DACS: Distributed adjustable computation scheme in highly scalable data center networks based on multi-protection routing","authors":"Wanling Lin ,&nbsp;Jou-Ming Chang ,&nbsp;Xiao-Yan Li","doi":"10.1016/j.jnca.2025.104425","DOIUrl":"10.1016/j.jnca.2025.104425","url":null,"abstract":"<div><div>In data center networks (DCNs), many value-added services involving multiple tenants and distributed sites can be naturally modeled as multi-party communication (MPC) processes, where participants exchange information across infrastructures to support collaborative computation. For MPC, it usually ensures that the private data of the multiple parties involved in the collaborative computation are safe, the computational results maintain acceptable accuracy, and all participants are in the same fair position in a distributed environment. This article considers an unexplored application based on MPC called the distributed adjustable computation scheme (DACS), which allows computation to be invoked when the collected data reaches a specified threshold in the communication. We developed a distributed algorithm using secure multi-protection routing to enable DACS. The proposed algorithm guarantees that each private data can be successfully delivered to the desired recipient even if any faulty component (including server or link) exists in the network. Also, no other than the destination can receive the complete private data. We implement DACS on highly scalable data center networks. Through simulation, experimental results show that DACS is highly reliable and achieves high security efficiency.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"247 ","pages":"Article 104425"},"PeriodicalIF":8.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145823154","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":"The DPC-LSTM-MD scheme for detecting selective forwarding attack under variable environment in event-driven wireless sensor networks","authors":"Yilun Ma,&nbsp;Yuanming Wu","doi":"10.1016/j.jnca.2025.104411","DOIUrl":"10.1016/j.jnca.2025.104411","url":null,"abstract":"<div><div>In an event-driven wireless sensor network (EWSN), events occur randomly, prompting sensor nodes within the event area to detect and transmit data packets to a sink via router nodes (RNs) through multi-hop communication. Some RNs, referred to as malicious nodes, may launch selective forwarding attacks by selectively dropping part or all of the received packets. Additionally, harsh environmental conditions can degrade channel quality, sometimes forcing RNs to abandon forwarding data packets. Under these conditions, it becomes challenging to distinguish malicious nodes from normal nodes solely based on their packet forwarding rates. To address this issue, we propose the DPC-LSTM-MD scheme to detect selective forwarding attacks. This approach utilizes the time series of nodes’ packet forwarding behaviors as a dataset. The Density Peaks Clustering (DPC) method is employed to extract features representative of normal node behavior. Subsequently, a Long Short-Term Memory (LSTM) network predicts the single round forwarding rate (SFR) of nodes in the next time series interval. Based on the prediction error, we apply the minimum density (MD) method combined with the 3-sigma rule to identify and isolate malicious nodes. Our results demonstrate that the DPC-LSTM-MD scheme achieves a low false detection rate (FDR) of 2% and a low missed detection rate (MDR) of 3%, significantly improving network throughput.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"247 ","pages":"Article 104411"},"PeriodicalIF":8.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145893775","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 digital twin-based reputation assessment model for JointCloud computing","authors":"Yadi Wu ,&nbsp;Lina Wang ,&nbsp;Rongwei Yu ,&nbsp;Xiuwen Huang ,&nbsp;Jiatong Liu","doi":"10.1016/j.jnca.2025.104395","DOIUrl":"10.1016/j.jnca.2025.104395","url":null,"abstract":"<div><div>JointCloud Computing (JCC) supports a collaborative model of multiple cloud service providers to provide users with robust performance and adequate services. Reputation is an important aspect for the stable development of the JCC system, affecting the cooperation among service providers and users’ choice of services. Most of the existing reputation assessment solutions only consider a single factor of user feedback or service quality, and cannot provide an accurate reputation assessment for the complex system of JCC. In addition, JointCloud services are provided by multiple service providers in cooperation, and existing solutions do not consider this service characteristic, making it difficult to accurately measure the reputation of the service. In order to provide a comprehensive reputation assessment for JCC, we proposed a reputation assessment model based on digital twins. A reputation calculation module is embedded in the digital twin, and a hybrid subjective–objective-based reputation assessment method and a split-integration-based reputation assessment method are designed for different JointCloud subjects to achieve a comprehensive and accurate reputation assessment. We conducted a series of experiments to evaluate the performance of the proposed reputation evaluation model and present the experimental results. The proposed method achieves a reputation assessment bias of 0.0112, which reduces the average bias by 0.2184 compared to existing researches. In real-world scenarios, the proposed model incurs a communication overhead of 93.7735 ms, with a digital twin data acquisition frequency of 36.4273 ms. The evaluation results show that our reputation evaluation model is feasible in terms of performance and accuracy.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"246 ","pages":"Article 104395"},"PeriodicalIF":8.0,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145593076","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 robust eclipse attack detection framework for Ethereum networks","authors":"Zubaida Rehman ,&nbsp;Iqbal Gondal ,&nbsp;Hai Dong ,&nbsp;Mengmeng Ge ,&nbsp;Mark A. Gregory ,&nbsp;Ikram ul Haq","doi":"10.1016/j.jnca.2025.104416","DOIUrl":"10.1016/j.jnca.2025.104416","url":null,"abstract":"<div><div>Eclipse attacks, which isolate victim nodes by monopolizing their peer connections, remain a critical threat to Ethereum’s consensus mechanism. To address this, we present a principled framework for detecting Eclipse attacks in Ethereum peer-to-peer networks, grounded in a formal adversarial model. Existing defenses are either ad-hoc or lack provable guarantees, leaving open questions about their reliability under adaptive adversaries. Our work aims to bridge this gap by formally defining eclipse attack detection as a security property. We specify soundness, completeness, and robustness theorems under bounded adversarial drift, and derive formal guarantees within false positive and false negative bounds, resilience to adversarial manipulation, and multi-node compositional reliability. We then instantiate a lightweight detection framework that maps packet-level traffic features to predictions using ensemble classifiers (Random Forest, XGBoost). The system was validated using a controlled Ethereum testbed and extended with CTGAN-generated synthetic traces to emulate networks of up to 100 nodes. Empirical evaluation shows that our framework achieves up to 96% F1-score with sub-second inference latency, well within Ethereum’s 12-second Proof-of-Stake validator time slots. These findings demonstrate that lightweight statistical features, when coupled with formal analysis, enable accurate, efficient, and scalable detection of network-level partitioning attacks. Our work establishes a deployable and theoretically grounded defense foundation for securing modern blockchain systems against eclipse adversaries.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"246 ","pages":"Article 104416"},"PeriodicalIF":8.0,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145785055","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":"BlackoutADR: Exploiting adaptive data rate vulnerabilities in LoRaWAN-based FANETs","authors":"Khaoula Hidawi ,&nbsp;Sabrine Ennaji ,&nbsp;Elena Ferrari","doi":"10.1016/j.jnca.2025.104409","DOIUrl":"10.1016/j.jnca.2025.104409","url":null,"abstract":"<div><div>This paper introduces <em>BlackoutADR</em>, a novel adversarial attack exploiting LoRaWAN’s Adaptive Data Rate (ADR) mechanism in cellular-connected UAV networks, with applicability to other IoT systems as well. By subtly manipulating Received Signal Strength Indicator (RSSI) and Signal-to-Noise Ratio (SNR), <em>BlackoutADR</em> increases UAV transmission power, causing 45% faster battery depletion within 100 s of simulation time and disrupting network operations. Using NS-3 simulations with a 20-UAV FANET, we evaluate its evasion of multiple ML-based IDSs (CNN, LSTM, BiLSTM, FNN, LoRaWAN-specific). Results show <em>BlackoutADR</em> remains undetected due to its subtle manipulations evading even dynamic thresholds, outperforming traditional jamming attacks. To address the identified vulnerability, we outline reactive measures, including dynamic threshold-based IDSs, secure ADR mechanisms, and recommendations for drone manufacturers.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"246 ","pages":"Article 104409"},"PeriodicalIF":8.0,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145705027","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":"Network-assisted relay selection in mmWave D2D communication under presence of dynamic obstacles with unknown orientation","authors":"Durgesh Singh ,&nbsp;Sasthi C. Ghosh","doi":"10.1016/j.jnca.2025.104394","DOIUrl":"10.1016/j.jnca.2025.104394","url":null,"abstract":"<div><div>Millimeter wave (mmWave) D2D communication is highly vulnerable to blockages from dynamic obstacles leading to severe packet loss and deterioration in quality of service. User equipments (UEs) acting as relays are chosen to divert the communication path in case of blockage. However, relays themselves are vulnerable to be blocked due to their and obstacles’ motion. Thus mobility parameters of dynamic obstacles must be captured effectively, but it is challenging as they might not be connected to the network. Additionally, instantaneous change in their orientation in motion is difficult to measure which might vary abruptly compared to their speed which can be sensed with high accuracy. A probabilistic model is developed considering the obstacle’s orientation in motion is <em>unknown</em>. Later the geometrical structure of the problem is analyzed and then exploited to devise a novel technique to derive closed form blockage expression for a given pair of UE. The proposed technique can be applied across exhaustive scenarios of motion of a given UE pair which may get blocked due to dynamic obstacles. Through extensive simulations, we have observed that our proposed approach outperforms both classical received signal strength (RSS) based approach and two recent state of the art approaches. We have also validated our results against an oracle which has complete speed and orientation information regarding UEs and obstacles.</div></div>","PeriodicalId":54784,"journal":{"name":"Journal of Network and Computer Applications","volume":"246 ","pages":"Article 104394"},"PeriodicalIF":8.0,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145760433","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}