Computers

22 pages, 4270 KB

Open AccessArticle

A Coordinated Bidirectional Data Fusion Processing System for Meteorological Applications

by Feifei Yang, Lei Cao, Jinghua Chen and Qiang Zhang

Computers 2026, 15(7), 467; https://doi.org/10.3390/computers15070467 - 22 Jul 2026

Viewed by 409

This study proposes a coordinated bidirectional meteorological data fusion processing system for controlled data exchange across the intranet, demilitarized zone (DMZ), and Internet. The system adopts a three-layer isolation architecture and integrates Apache MiNiFi, Apache NiFi, and Apache Kafka to coordinate edge preprocessing, [...] Read more.

This study proposes a coordinated bidirectional meteorological data fusion processing system for controlled data exchange across the intranet, demilitarized zone (DMZ), and Internet. The system adopts a three-layer isolation architecture and integrates Apache MiNiFi, Apache NiFi, and Apache Kafka to coordinate edge preprocessing, DMZ-based fusion processing, asynchronous message buffering, and Internet service publication. Kerberos authentication, access control, and operational monitoring support outbound data-product services and inbound user-request-driven workflows. Operational evaluation at the National Meteorological Science Data Center showed that edge preprocessing reduced cross-domain data volume by an average of 95%; representative bidirectional workflows were completed within minutes; and the average Kafka message-processing success rate over three consecutive months was 99.77%. The system has supported the automated generation and external publication of human comfort index products, while its core mechanisms have been generalized into a reusable software stack for cross-domain scientific data applications. The results indicate that the proposed architecture provides a practical and deployable approach to efficient and controlled bidirectional meteorological data fusion without changing existing network security boundaries. Full article

► Show Figures

Figure 1

31 pages, 682 KB

Open AccessArticle

Generative AI in Technology-Oriented Higher Education: A Systematized Review and Survey on Students’ Perceptions of Performance, Autonomy, and Ethical Implications

by Mayra Álvarez-Jiménez, Geovanny Cudco, Diego Gamboa and Danny Páez

Computers 2026, 15(7), 466; https://doi.org/10.3390/computers15070466 - 22 Jul 2026

Viewed by 784

Abstract

Generative Artificial Intelligence (GenAI) is rapidly reshaping higher education, especially in technology-oriented programs where critical thinking and complex problem solving are core outcomes. This study triangulates global and local evidence on performance/efficiency, usage, autonomy, critical-thinking engagement, and ethics by combining a systematized review [...] Read more.

Generative Artificial Intelligence (GenAI) is rapidly reshaping higher education, especially in technology-oriented programs where critical thinking and complex problem solving are core outcomes. This study triangulates global and local evidence on performance/efficiency, usage, autonomy, critical-thinking engagement, and ethics by combining a systematized review informed by Kitchenham and structured using selected PRISMA 2020 elements (2020–2025; last search: May 2025; 49 studies; Scopus, ACM Digital Library, IEEE Xplore, and SpringerLink; not prospectively registered) with an anonymous survey of 302 computing and engineering students from a single university in Ecuador. The expert-reviewed instrument showed acceptable internal consistency for most scale-based dimensions (McDonald’s

ω

), whereas institutional and ethics-related items were analyzed primarily at the item level. Results showed near-universal academic GenAI use (96%), with 47% of students reporting weekly use and 26% daily use. Research-related work was the most frequent application (81.5%), followed by homework (48.3%), report writing (43.7%), and exam preparation (41.7%). Although students reported perceived efficiency gains, concerns persisted about reduced analytical engagement and technological dependence (84.1%). Ethical concerns centered on dependence, authenticity, and data privacy, while institutional responses pointed to the need for formal training (96.7%) and clearer guidance. Based on this triangulation, we propose a context-bounded interpretive framework suggesting that GenAI’s educational value depends on instructional and governance conditions that preserve autonomy, critical thinking, integrity, and equity. Full article

(This article belongs to the Topic AI Trends in Teacher and Student Training)

► Show Figures

Figure 1

25 pages, 7249 KB

Open AccessArticle

An SEU-Tolerant Cache for a RISC-V Core

by Ariel David Santana Gil, Salvador Ibarra Delgado, Julio Villalba Moreno, Remberto Sandoval Arechiga, Viktor Iván Rodriguez Abdalá and Manuel Hernández Calviño

Computers 2026, 15(7), 465; https://doi.org/10.3390/computers15070465 - 22 Jul 2026

Viewed by 557

Abstract

The open RISC-V Instruction Set Architecture (ISA) is a versatile architecture with a growing number of implementations, including the aerospace sector, where tolerance to Single-Event Upset (SEU) faults is critical. This paper presents the FPGA-based design and implementation of an SEU-tolerant cache memory [...] Read more.

The open RISC-V Instruction Set Architecture (ISA) is a versatile architecture with a growing number of implementations, including the aerospace sector, where tolerance to Single-Event Upset (SEU) faults is critical. This paper presents the FPGA-based design and implementation of an SEU-tolerant cache memory for integration into RISC-V cores as part of a memory subsystem protection strategy. The proposed architecture incorporates separate instruction and data caches, organized as four-way set-associative with a write-through policy, complemented by Hamming H(39, 32) encoding for single-bit error detection and correction. Two experimental platforms were developed on a ZYNQ-7000 SoC: one dedicated to functional validation of the Hamming modules through controlled error injection and another for the evaluation of the complete system integrated with an internally developed RV32IMAFE core. The results confirm the correction of single errors and the detection of multiple errors while demonstrating performance improvements between 1.42× and 1.67× compared to the baseline RISC-V core. Full article

► Show Figures

Figure 1

40 pages, 830 KB

Open AccessSystematic Review

Adaptive Gamification and Game-Based Learning in Preschool and Early Primary Education: A Systematic Literature Review

by Alkinoos-Ioannis Zourmpakis

Computers 2026, 15(7), 464; https://doi.org/10.3390/computers15070464 - 22 Jul 2026

Cited by 1 | Viewed by 1454

Abstract

In recent years, adaptive gamification and adaptive game-based learning (GBL) have attracted the interest of researchers and educators as a response to the “one-size-fits-all” approach of conventional gamified applications. However, their effectiveness has shown mixed results, and the literature concerning preschool and early [...] Read more.

In recent years, adaptive gamification and adaptive game-based learning (GBL) have attracted the interest of researchers and educators as a response to the “one-size-fits-all” approach of conventional gamified applications. However, their effectiveness has shown mixed results, and the literature concerning preschool and early primary education remains scattered. Therefore, we performed a systematic literature review of 19 empirical studies published between 2016 and 2026, following the PRISMA model, from a total of 5069 records identified across nine electronic databases. This review examines the methodological approaches and assessment tools employed, the content areas, educational levels, and educational contexts addressed, the theoretical frameworks and adaptive mechanisms utilised, and the learning and motivational outcomes reported for young learners. Our findings revealed a strong concentration on mathematics, a heavy reliance on researcher-developed platforms, and limited explicit theoretical grounding. Moreover, most studies adapted only the learning content, while the game elements themselves remained fixed. Although most studies reported positive learning and motivational outcomes, the results were not uniform, with prior knowledge being the most common moderating variable. Benefits are most visible when adaptive systems support children’s pacing, prior knowledge, or task difficulty, with some studies showing improvement in learning efficiency rather than learning gains. Overall, this review reveals the emerging trends and challenges in this field and provides a framework and insight for future researchers regarding the design of adaptive learning environments for young children. Full article

(This article belongs to the Special Issue Transformative Approaches in Education: Harnessing AI, Augmented Reality, and Virtual Reality for Innovative Teaching and Learning (2nd Edition))

► Show Figures

Figure 1

26 pages, 1673 KB

Open AccessArticle

CERO: Cascade-Emergency Resilient Offloading for IIoT Edge Computing via Adversarial Deep Reinforcement Learning

by Zhining Wang, Haibin Yu, Hongfei Bai and Dong Li

Computers 2026, 15(7), 463; https://doi.org/10.3390/computers15070463 - 21 Jul 2026

Viewed by 445

Abstract

Industrial Internet of Things (IIoT) edge computing supports latency-sensitive services through task offloading to distributed edge resources. However, large-scale emergencies such as node failures and traffic surges may trigger cascading failures, leading to severe performance degradation and poor post-crisis recovery. Existing offloading methods [...] Read more.

Industrial Internet of Things (IIoT) edge computing supports latency-sensitive services through task offloading to distributed edge resources. However, large-scale emergencies such as node failures and traffic surges may trigger cascading failures, leading to severe performance degradation and poor post-crisis recovery. Existing offloading methods mainly optimize operational efficiency under normal conditions while overlooking resilience against cascading disruptions. To address this issue, we propose Cascade-Emergency Resilient Offloading (CERO), an adversarial deep reinforcement learning framework for resilient task offloading in IIoT edge computing. Distinct from existing works, CERO introduces a structure-aware shared node encoder to capture heterogeneous topological roles of edge nodes, providing critical structural information for cascade-aware decision making, and incorporates cascade-oriented adversarial training to enhance robustness against compound disturbances. CERO integrates structure-aware state representation, minimax adversarial training, and potential-based reward shaping to learn resource-allocation policies balancing task efficiency and system resilience. By interacting with dynamically generated crisis scenarios, the agent learns resilient offloading policies and achieves high post-crisis recovery performance after cascading disruptions. All performance evaluations are conducted via discrete-event simulation experiments. Simulation results for normal, single-crisis, and compound-crisis scenarios show that CERO achieves comparable task efficiency under normal conditions and significantly superior post-crisis recovery performance compared to conventional rule-based strategies. In the hardest compound-crisis case involving simultaneous node failures and load surges, CERO achieves a post-recovery task-completion rate of 97.8%, surpassing the best rule-based baseline by more than 63 percentage points. Statistical significance is confirmed by the Wilcoxon signed-rank test with Bonferroni correction over 10 independent runs. These results demonstrate that CERO effectively improves the robustness and recoverability of IIoT edge-computing systems under cascading emergency scenarios. Full article

(This article belongs to the Section Internet of Things (IoT) and Industrial IoT)

► Show Figures

Figure 1

32 pages, 6300 KB

Open AccessArticle

An Autonomous AI-Driven Framework for Adaptive Cyber Deception with Real-Time Threat Detection and Behaviour-Based Attribution

by Muhammad Shahzad, Muhsin Hassanu Saleh and Raja Ujjan

Computers 2026, 15(7), 462; https://doi.org/10.3390/computers15070462 - 21 Jul 2026

Viewed by 806

Abstract

Contemporary cyber threats increasingly employ multi-stage and behaviourally adaptive strategies that challenge static intrusion detection and non-adaptive deception mechanisms. Existing approaches typically treat threat detection, deception deployment, and adversarial attribution as separate functions, limiting timely response and underusing the behavioural evidence generated during [...] Read more.

Contemporary cyber threats increasingly employ multi-stage and behaviourally adaptive strategies that challenge static intrusion detection and non-adaptive deception mechanisms. Existing approaches typically treat threat detection, deception deployment, and adversarial attribution as separate functions, limiting timely response and underusing the behavioural evidence generated during attacker interaction. This study develops and evaluates a theory-informed computational and operational framework for autonomous cyber deception. The principal research artefact is a reusable closed-loop architecture rather than a single predictive model: it specifies the interacting components, interfaces, data and control flows, decision rules, and feedback mechanisms that connect detection, deception, telemetry, and attribution. Methodologically, the study follows an engineering design-and-evaluation approach comprising problem and requirement identification from the literature, architectural synthesis, component-level mathematical modelling, prototype implementation, and controlled cyber-range evaluation. In this context, modelling refers to the distinct computational models embedded within the framework: a hybrid detection model combining supervised classification, anomaly detection, and temporal sequence analysis; a Markov Decision Process and reinforcement-learning policy model for selecting and reconfiguring deception actions under engagement, intelligence-gain, resource, and containment objectives; and similarity-based and Bayesian attribution models for estimating MITRE ATT&CK techniques from incomplete behavioural evidence. The component models were developed offline using the NSL-KDD, CICIDS2017, UNSW-NB15, and ToN-IoT datasets, while the integrated prototype was evaluated separately in a controlled enterprise-like cyber range using reconnaissance, brute-force, exploitation, and multi-stage attack scenarios. The reported classification metrics were calculated from the labelled cyber-range evaluation events, not by pooling the four benchmark datasets. On this integrated cyber-range evaluation set, the system achieved 95.4% detection accuracy, 93.6% precision, 94.7% recall, and a 94.1% F1-score, with a mean detection latency of 85 ms. It also achieved 100% honeypot deployment reliability, 92% dynamic reconfiguration success, 88% fingerprinting resistance, and attacker engagement durations of up to 280 s. The attribution component demonstrated end-to-end generation of ATT&CK-aligned technique hypotheses from deception-derived telemetry; however, the present archived evaluation does not support per-technique or baseline-comparative performance claims. These findings show that specialised models and operational services can be coordinated within a unified adaptive defence process, while also identifying the additional class-level and ablation evidence required for rigorous attribution validation. Full article

(This article belongs to the Special Issue Next-Generation Cyber Defense: AI, Automation and Adaptive Security)

► Show Figures

Figure 1

39 pages, 5847 KB

Open AccessArticle

Regularized Multi-Backbone Ensembles for Video-Level Deepfake Detection on Celeb-DF v2: Accuracy–Efficiency and Generalization Limits

by Mohammed Alshalfi, Abdulrahman Alshehri, Qazi Emad Ul Haq and Tariq M. Khan

Computers 2026, 15(7), 461; https://doi.org/10.3390/computers15070461 - 21 Jul 2026

Viewed by 1063

Abstract

The increasing realism of deepfake videos has intensified the need for reliable video-level detection systems, but benchmark performance must be interpreted together with generalization, temporal-modeling, and efficiency limits. This study presents a reproducible multi-backbone framework for detecting manipulated videos on the official Celeb-DF [...] Read more.

The increasing realism of deepfake videos has intensified the need for reliable video-level detection systems, but benchmark performance must be interpreted together with generalization, temporal-modeling, and efficiency limits. This study presents a reproducible multi-backbone framework for detecting manipulated videos on the official Celeb-DF v2 benchmark. Five pretrained image-classification architectures—ResNet-50, EfficientNet-B4, ConvNeXt-Small, ViT-Base, and Swin-Base—are fine-tuned under a unified protocol using uniform frame sampling, class-balanced training, RandAugment, MixUp, CutMix, random erasing, label smoothing, AdamW optimization, cosine learning-rate scheduling, and exponential moving average weights. During inference, frame-level fake probabilities are stabilized using horizontal-flip test-time augmentation and aggregated into video-level predictions by mean probability pooling. This aggregation is a fixed probability-pooling rule rather than an explicit temporal model. A probability-level ensemble of ResNet-50, ConvNeXt-Small, and Swin-Base combines convolutional and attention-based representations. On the official 518-video Celeb-DF v2 test set, the top-three ensemble achieves a video-level AUC of 99.967% and an average precision of 99.983%, while ResNet-50 provides the strongest single-model accuracy–efficiency trade-off. Additional analyses examine frame-to-video aggregation, ROC and precision–recall behavior, probability distributions, frame-probability stability over sampled frames, a proof-of-concept temporal-splice sensitivity test, and inference efficiency. The results demonstrate highly competitive in-dataset performance on Celeb-DF v2 only. Because no external benchmark testing, learned temporal baseline, confidence-gated cascade, or broad partial-manipulation benchmark is included, the results should not be interpreted as evidence of cross-dataset robustness, in-the-wild deployment readiness, learned temporal reasoning, or general localization capability. Cross-dataset evaluation on FaceForensics++, DFDC, WildDeepfake, and related benchmarks, probability calibration, false-positive control, explicit temporal modeling, cascade-based inference, and expanded localization evaluation are identified as priority future-work directions. Full article

(This article belongs to the Section AI-Driven Innovations)

► Show Figures

Graphical abstract

50 pages, 6320 KB

Open AccessArticle

SafeBoundary-LLM: Measuring Safety Boundary Stability in Local Open-Weight LLMs Through Single-Turn Baselines and Multi-Turn Escalation

by Andreea Alexandra Anghel, Catalin Anghel, Emilia Pecheanu, Antonio Stefan Balau, Marian Viorel Craciun, Adina Cocu and Cristian Sandu

Computers 2026, 15(7), 460; https://doi.org/10.3390/computers15070460 - 21 Jul 2026

Cited by 1 | Viewed by 498

Abstract

Local open-weight large language models (LLMs) are increasingly used in privacy-sensitive settings, yet isolated prompts may not reveal whether safety boundaries remain stable during conversation. SafeBoundary-LLM evaluated seven local models across 14 sensitive domains, 84 boundary sets, 672 single-turn prompts, and 84 five-turn [...] Read more.

Local open-weight large language models (LLMs) are increasingly used in privacy-sensitive settings, yet isolated prompts may not reveal whether safety boundaries remain stable during conversation. SafeBoundary-LLM evaluated seven local models across 14 sensitive domains, 84 boundary sets, 672 single-turn prompts, and 84 five-turn escalation conversations; the same models were evaluated separately on XSTest and JBB-Behaviors. Evaluators R1 and R2 independently classified all 12,194 responses, with R2 labels used for primary outcomes and unreconciled labels used for reliability analysis. Exact agreement exceeded 93% in each dataset. In SafeBoundary-LLM, 456 out of 7644 responses (5.97%) were confirmed-or-mixed failures. The multi-turn failure rate was 14.69% versus 0.51% for single-turn prompts, yielding a rate ratio of 28.80 (95% CI [21.20, 43.80]; Holm-adjusted p = 0.0006); boundary collapse occurred only at Turns 4–5, and role-play bypass accounted for 299 out of 456 failures. On answer-expected items, over-refusal was 4.34% in XSTest and 17.29% in JBB-Behaviors, whereas unsafe compliance on refusal-expected items was 0.36% and 0.86%, respectively. These findings support an evaluation strategy that includes public single-turn benchmarks, controlled multi-turn escalation, independent human review, and traceable audit records for locally deployed LLMs. Full article

► Show Figures

Figure 1

11 pages, 281 KB

Open AccessArticle

Application of LoRA+ in Fine-Tuning Large Models for Construction Process and Its Synergy with RAG

by Weitang Zhang and Lang Liu

Computers 2026, 15(7), 459; https://doi.org/10.3390/computers15070459 - 20 Jul 2026

Viewed by 445

Abstract

Addressing the resource constraints of a single NVIDIA RTX 5000 (16 GB) GPU, this applied study takes DeepSeek-LLM-7B-Base as the base model and systematically compares four parameter-efficient fine-tuning methods: LoRA, QLoRA, DoRA, and LoRA+. It also validates a Retrieval-Augmented Generation (RAG) architecture tailored [...] Read more.

Addressing the resource constraints of a single NVIDIA RTX 5000 (16 GB) GPU, this applied study takes DeepSeek-LLM-7B-Base as the base model and systematically compares four parameter-efficient fine-tuning methods: LoRA, QLoRA, DoRA, and LoRA+. It also validates a Retrieval-Augmented Generation (RAG) architecture tailored for zero-tolerance engineering specifications. Experiments are conducted on a private construction process dataset. Theoretical analysis shows that the low-rank assumption of LoRA originates from the intrinsic dimensionality property of pre-trained models; LoRA+ adopts an asymmetric learning rate strategy (with the optimal ratio

\frac{η_{A}}{η_{B}}

= 0.05 determined via grid search), effectively solving the suboptimal training dynamics problem of standard LoRA caused by a uniform learning rate; DoRA decomposes weight updates into magnitude and direction components on a spherical manifold and a positive real manifold; RAG guarantees hallucination suppression through the conditional entropy inequality H(Y|Q,D,θ) ≤ H(Y|Q,θ). Experimental results demonstrate that LoRA+ outperforms other baseline methods in BLEU-4 (0.5609), ROUGE-L (0.5387), and PPL (2.1433), with a training time of 1.8 h and memory usage of 13.1 GB. After introducing RAG on top of LoRA+, BLEU-4 further improves to 0.5814, ROUGE-L to 0.5557, and the hallucination rate(HR) drops from 1.71% to 0.08%, achieving an Exact Match (EM) score of 0.2778 and high traceability (Recall@3 = 0.9961). This study provides a technical pathway and empirical evidence for deploying large models in the construction domain under resource-constrained conditions through the synergy of fine-tuning and RAG. Full article

► Show Figures

Graphical abstract

49 pages, 2081 KB

Open AccessReview

Enhancing the Kubernetes Scheduler: A State-of-the-Art Review from Cloud to Edge

by Mohammed Alhakimi and Rohaya Latip

Computers 2026, 15(7), 458; https://doi.org/10.3390/computers15070458 - 19 Jul 2026

Viewed by 939

Abstract

The rapid expansion of the cloud–edge continuum requires containerized applications to scale dynamically across highly heterogeneous and resource-constrained environments. As the de facto standard for container orchestration, Kubernetes (K8s for short) relies heavily on its scheduling subsystem to manage these complex distributed environments. [...] Read more.

The rapid expansion of the cloud–edge continuum requires containerized applications to scale dynamically across highly heterogeneous and resource-constrained environments. As the de facto standard for container orchestration, Kubernetes (K8s for short) relies heavily on its scheduling subsystem to manage these complex distributed environments. However, default scheduling methodologies are inherently designed for homogeneous cloud data centers and bring substantial deployment challenges when used in edge topologies. While numerous custom schedulers, plugins, and extensions have been put forward to bridge this gap, a review of the state of the art is needed to evaluate existing approaches and capture recent trends. In this survey, we present a comprehensive review of Kubernetes scheduling strategies published between January 2023 and January 2026. We establish a multi-dimensional taxonomy that categorizes scheduling approaches based on common objectives, modification methods, optimization methodologies, targeted workloads, evaluation methods, scheduling scopes, and performance metrics. We investigate open challenges arising across different computing paradigms and highlight recent trends and possible directions for future research in Kubernetes scheduling. Full article

► Show Figures

Graphical abstract

24 pages, 1382 KB

Open AccessArticle

A Multi-Scale Convolutional Neural Network with Residual Blocks and LSTM for Multi-Step Forecasting of Electricity Load

by Yuhang Zhang, Yiting Zhao, Yujing Meng, Jingqi Li, Tianze Zhang and Ying Zhang

Computers 2026, 15(7), 457; https://doi.org/10.3390/computers15070457 - 18 Jul 2026

Viewed by 414

Abstract

Electricity load forecasting is essential for balancing energy supply and demand, reducing energy waste, and maintaining power grid stability. Accurate forecasts enable power utilities to optimize energy dispatch and mitigate the risk of supply shortages or outages. However, conventional forecasting methods often struggle [...] Read more.

Electricity load forecasting is essential for balancing energy supply and demand, reducing energy waste, and maintaining power grid stability. Accurate forecasts enable power utilities to optimize energy dispatch and mitigate the risk of supply shortages or outages. However, conventional forecasting methods often struggle to capture highly nonlinear local fluctuations in electricity consumption and long-term temporal dependencies. To address these challenges, this study proposes MSCNN-ResLSTM, a hybrid model for multi-step electricity load forecasting. The proposed model integrates Multi-Scale Convolutional Neural Networks (MSCNNs) to extract local time-series features at multiple temporal scales, residual blocks (ResBlocks) to enhance feature representation through residual connections, and Long Short-Term Memory (LSTM) networks to model long-range temporal dependencies. To comprehensively evaluate its effectiveness, a cross-paradigm experimental framework is established in which MSCNN-ResLSTM is compared with seven representative benchmark models from three methodological categories: traditional machine learning (Extreme Gradient Boosting-XGBoost), classical recurrent and convolutional neural networks (LSTM, Temporal Convolutional Network-TCN, CNN-LSTM, MSCNN-LSTM, and Direct LSTM (Seq2Seq)), and self-attention-based architectures (Transformer). Experimental results show that MSCNN-ResLSTM achieves higher forecasting accuracy and greater stability across the full 24-step prediction horizon, consistently outperforming all competing baselines while effectively suppressing recursive error propagation. Full article

(This article belongs to the Special Issue Machine Learning: Techniques, Industry Applications, Code Sharing, and Future Trends)

► Show Figures

Figure 1

27 pages, 2050 KB

Open AccessArticle

Intelligent Attack Detection in Blockchain-Enabled Multi-Cloud Systems: A Systematic Review and SOC-LLM-Augmented Architecture Proposal

by Adam Koty Abbass Ahmat and Habiba Chaoui

Computers 2026, 15(7), 456; https://doi.org/10.3390/computers15070456 - 17 Jul 2026

Viewed by 504

Abstract

This paper presents a systematic literature review examining how blockchain technologies can enhance the security and performance of multi-cloud systems. Multi-cloud architectures offer resilience, scalability, and flexibility; however, they also pose complex security challenges related to APIs, service-level agreements (SLAs), orchestration, and authentication. [...] Read more.

This paper presents a systematic literature review examining how blockchain technologies can enhance the security and performance of multi-cloud systems. Multi-cloud architectures offer resilience, scalability, and flexibility; however, they also pose complex security challenges related to APIs, service-level agreements (SLAs), orchestration, and authentication. The promise of blockchain technology to improve the security and transparency of numerous applications, including cloud storage systems, has attracted considerable attention in recent years. Much research has focused on decentralized storage in cloud environments, spanning supply chains, FinTech, healthcare, and education. Still, the integration of blockchain with the cloud and its potential to enhance security and performance warrant an in-depth study. Using the PRISMA methodology, a structured search was conducted across six major scientific databases, including IEEE, ACM Digital Library, ScienceDirect, Scopus, Web of Science, and IJIMAI. Twenty-four primary papers published between 2019 and 2025 were selected for analysis after clear inclusion and exclusion criteria were applied. This review examines the security dimensions in multi-cloud environments—architectural vulnerabilities, API security, authentication, orchestration and automation vulnerabilities, SLAs, and cybersecurity compliance issues—in relation to blockchain technology. Based on the identified gaps, we propose a SOC-LLM-augmented security architecture that integrates blockchain-based evidence integrity, statistical anomaly detection, machine learning, large language models, and autonomous AI agents to enable intelligent attack detection and response. The proposed framework introduces specialized agents for detection, correlation, threat intelligence retrieval, blockchain evidence validation, explanation generation, and response planning. The analysis shows that integrating SOC-LLM capabilities with blockchain can move multi-cloud security from passive auditability toward proactive, explainable, and human-in-the-loop cyber defense. Finally, this paper discusses open challenges, including LLM hallucination, data scarcity, real-time scalability, evaluation standardization, and trustworthy deployment in critical multi-cloud infrastructures. The study’s conclusion highlights research gaps and suggests future lines of inquiry concerning scalable blockchain architectures and the incorporation of AI for proactive cloud security monitoring. Full article

(This article belongs to the Section Blockchain Infrastructures and Enabled Applications)

► Show Figures

Figure 1

21 pages, 1353 KB

Open AccessArticle

An SLA-Aware Priority Management System for HTTP/2 Based on RFC 9218: Design, Implementation, and Performance Evaluation in Service-Based Architectures

by Ahmed Lateef Salih Al-Karawi and Rafet Akdeniz

Computers 2026, 15(7), 455; https://doi.org/10.3390/computers15070455 - 17 Jul 2026

Viewed by 412

Abstract

Service-Based Architectures (SBAs) in 5G core and cloud-native deployments require differentiated treatment for service classes with heterogeneous latency, reliability, and throughput expectations. Although HTTP/3 over QUIC is an important evolution of the HTTP ecosystem, HTTP/2 remains operationally relevant in SBA environments where TCP/TLS-based [...] Read more.

Service-Based Architectures (SBAs) in 5G core and cloud-native deployments require differentiated treatment for service classes with heterogeneous latency, reliability, and throughput expectations. Although HTTP/3 over QUIC is an important evolution of the HTTP ecosystem, HTTP/2 remains operationally relevant in SBA environments where TCP/TLS-based infrastructures and 3GPP service-based interfaces continue to rely on HTTP/2 communication. This paper therefore focuses on HTTP/2 priority signaling and the problem of translating application-level Service Level Agreement (SLA) policies into protocol-level priority metadata. To address this problem, the paper presents an SLA-aware priority management system built around the RFC 9218 extensible prioritization scheme, specifically its urgency and incremental parameters. The system integrates three coordinated subsystems: a rule-based Priority Classification Engine (PCE), a feedback-driven Dynamic Priority Mapping Algorithm (DPMA), and a runtime priority-update manager that applies bounded priority adjustments under changing network and load conditions. The revised evaluation reports a 7200-observation baseline campaign covering four operating modes, ten service classes, nine network profiles, and twenty repetitions per service–profile–mode combination, together with a 14,880-observation scalability and overhead campaign across increasing concurrent-stream levels. Compared with the unmanaged HTTP/2 baseline, DPMA reduced mean latency by 24.8%, P95 latency by 35.1%, P99 latency by 38.0%, and SLA violations by 19.9 percentage points. Compared with the legacy RFC 7540 baseline, DPMA reduced mean latency by 39.0%, P95 latency by 49.3%, P99 latency by 49.9%, and SLA violations by 21.1 percentage points. Compared with the static RFC 9218 baseline, DPMA reduced mean latency by 38.7%, P95 latency by 48.1%, P99 latency by 50.6%, and SLA violations by 21.4 percentage points. The scalability analysis shows that DPMA maintained P95 latency between 126.8 ms and 128.2 ms over the tested 1–100 concurrent-stream range, with priority-update decision overhead below 0.004 ms per request. The results indicate that SLA-aware use of RFC 9218 priority metadata can improve latency and SLA-compliance behavior in controlled SBA-like HTTP/2 environments while preserving a transparent and auditable prioritization policy. Full article

(This article belongs to the Section Cloud Continuum and Enabled Applications)

► Show Figures

Graphical abstract

17 pages, 3745 KB

Open AccessArticle

Multi-Source Domain Adaptive EEG Emotion Recognition Based on Dendrite Net

by Shuang Liu, Huifeng Guo, Rongyu Han, Yajing Pang and Gang Liu

Computers 2026, 15(7), 454; https://doi.org/10.3390/computers15070454 - 17 Jul 2026

Viewed by 498

Abstract

Accurate emotion recognition is crucial for enhancing human–computer interaction, and brain–computer interface (BCI) technology offers an efficient means for emotion detection using the EEG signal. However, existing methods face significant challenges due to the inherent inter-individual differences and temporal variability of EEG data. [...] Read more.

Accurate emotion recognition is crucial for enhancing human–computer interaction, and brain–computer interface (BCI) technology offers an efficient means for emotion detection using the EEG signal. However, existing methods face significant challenges due to the inherent inter-individual differences and temporal variability of EEG data. To address these limitations, this paper introduces a multi-source domain adaptive algorithm based on dendrite net (DD-MSDA). The proposed model employs the dendrite network as a shared feature extractor to align feature distributions across multiple source domains, thereby capturing common features among diverse datasets. Experimental validation on cross-subject and cross-session tasks using the SEED and SEED-IV datasets demonstrates that DD-MSDA achieves highly competitive performance, outperforming all compared single-modal EEG-based domain adaptation methods. Moreover, the algorithm demonstrates statistically significant advantages over existing domain adaptation baselines in cross-dataset settings. These results highlight the consistent competitiveness of DD-MSDA across various cross-domain scenarios, and its unsupervised nature underscores its potential for practical online EEG emotion recognition applications. Full article

(This article belongs to the Special Issue AI/ML-Driven EEG Signal Processing)

► Show Figures

Graphical abstract

30 pages, 6851 KB

Open AccessArticle

Reusing Policy-as-Code Across CI/CD and Kubernetes Admission Control: An Empirical Assessment of Governance Consistency

by Luís Nogueira and Alice Resende

Computers 2026, 15(7), 453; https://doi.org/10.3390/computers15070453 - 17 Jul 2026

Viewed by 745

Abstract

Cloud-native software-delivery pipelines increasingly rely on Policy-as-Code (PaC) to automate security, compliance, and governance enforcement. Although Policy-as-Code is widely adopted within Continuous Integration (CI) pipelines and Kubernetes admission-control frameworks, governance requirements are often implemented independently, potentially increasing maintenance effort and creating opportunities for [...] Read more.

Cloud-native software-delivery pipelines increasingly rely on Policy-as-Code (PaC) to automate security, compliance, and governance enforcement. Although Policy-as-Code is widely adopted within Continuous Integration (CI) pipelines and Kubernetes admission-control frameworks, governance requirements are often implemented independently, potentially increasing maintenance effort and creating opportunities for policy drift. Despite the growing adoption of Policy-as-Code, comparatively little empirical evidence exists regarding the reuse of a shared policy-definition layer across complementary enforcement stages within the software-delivery lifecycle. This paper presents and empirically evaluates a reusable multi-stage Policy-as-Code enforcement model based on a shared policy-definition layer implemented using the Open Policy Agent (OPA) framework and its Rego policy language. Rather than proposing a new Policy-as-Code technology, the study investigates whether a shared policy-definition layer can support consistent policy enforcement across Continuous Integration validation and Kubernetes admission control. The model was evaluated using Conftest and OPA Gatekeeper through a structured experimental study comprising 29 Kubernetes manifests, 37 experimental scenarios, eight Kubernetes resource types, and 261 policy assertions covering representative cloud-native workload-governance requirements. Within the evaluated dataset, all intentionally introduced insecure configurations were correctly identified without observed false positives or false negatives. The shared policy-definition layer was successfully reused across both validation stages, while Kubernetes admission control mitigated all evaluated CI bypass scenarios by providing an independent deployment-time enforcement boundary. The results demonstrate that a shared policy-definition layer can support consistent policy enforcement across complementary enforcement stages while enabling policy reuse without requiring duplicate policy implementations within the evaluated environment. More broadly, the study contributes empirical evidence supporting policy reuse as a governance strategy for cloud-native software delivery and provides a reproducible foundation for future investigations involving larger datasets, broader governance-policy portfolios, alternative Policy-as-Code ecosystems, and production-scale deployments. Full article

► Show Figures

Graphical abstract

49 pages, 3220 KB

Open AccessArticle

The Painted Wolf Decision Optimizer

by Shervin Zakeri, Dimitri Konstantas and Prasenjit Chatterjee

Computers 2026, 15(7), 452; https://doi.org/10.3390/computers15070452 - 16 Jul 2026

Viewed by 475

Abstract

This study introduces the Painted Wolf Decision Optimizer (PWO), the first deterministic, bio-inspired decision framework for discrete multi-criteria decision making (MCDM) derived from specific observed decision behaviors of African wild dogs, including quorum sensing, dominance hierarchy, collective voting, and experience-based learning. Unlike conventional [...] Read more.

This study introduces the Painted Wolf Decision Optimizer (PWO), the first deterministic, bio-inspired decision framework for discrete multi-criteria decision making (MCDM) derived from specific observed decision behaviors of African wild dogs, including quorum sensing, dominance hierarchy, collective voting, and experience-based learning. Unlike conventional nature-inspired metaheuristics that rely on stochastic search across continuous domains, PWO defines a new class of Discrete Bio-Inspired Decision Operators. It formalizes key ethological mechanisms of Lycaon pictus: quorum sensing, hierarchical dominance, and reinforcement-based learning. Additionally, it encodes the principle of survival-through-precision, demonstrating how coordinated strategic alignment can outperform structural dominance under resource constraints, inspired by the high hunting efficiency of African wild dogs. PWO integrates three cognitive weighting components: subjective collective preferences (sneeze-based voting), objective data variability (entropy weighting), and experiential reinforcement (pack memory). These are fused via the Mathematical Compromiser, a convex operator that assigns internal trust based on signal stability rather than fixed weighting rules. Applied to European EV gigafactory location selection, PWO reconciled tensions between cost-driven executive preferences and sustainability-based performance indicators, identifying Spain as the most robust alternative. Sensitivity analysis across the dominance spectrum (

D = 0 \to 1

) and multiple episodes confirmed ranking stability without rank reversal. The Markovian update formalizes longitudinal learning for future multi-episode applications. Beyond discrete selection, PWO functions as a diagnostic and competitive resilience mechanism, revealing whether decisions are shaped by leadership authority, structural necessity, historical trends, or precision-based survival logic. It provides a transparent and strategically adaptive architecture for sustainable governance and high-stakes competitive decision environments. Full article

(This article belongs to the Special Issue Intelligence for Complex Data: From Retrieval and Understanding to Decision-Making)

► Show Figures

Figure 1

19 pages, 1648 KB

Open AccessArticle

A Secure Lightweight SMS Spam Detection Framework with Robustness to Text Obfuscation Attacks

by Baraa Tareq Hammad, Ismail Taha Ahmed, Mohamed A. Hafez and Betty Wan Niu Voon

Computers 2026, 15(7), 451; https://doi.org/10.3390/computers15070451 - 16 Jul 2026

Viewed by 434

Abstract

The proliferation of mobile communications has led to a significant increase in SMS spam, posing challenges related to security, privacy, and user experience. Although numerous machine-learning-based spam detection approaches have been proposed, developing systems that are simultaneously lightweight and resilient to adversarial manipulation [...] Read more.

The proliferation of mobile communications has led to a significant increase in SMS spam, posing challenges related to security, privacy, and user experience. Although numerous machine-learning-based spam detection approaches have been proposed, developing systems that are simultaneously lightweight and resilient to adversarial manipulation remains an open problem. This paper proposes an SMS spam detection framework that incorporates multiple feature extraction methods, including bag-of-words (BoW), Term Frequency–Inverse Document Frequency (TF-IDF), and N-gram models with dimensionality reduction using principal component analysis (PCA), followed by classification using decision tree (DT) and Logistic Regression (LogReg) models. Experimental evaluations on the UCI SMS Spam Collection dataset demonstrate that the TF-IDF-PCA-DT pipeline achieves a detection accuracy of 99% while reducing model size by 77% and inference time by 75%. Robustness evaluation under adversarial text perturbations indicates minimal performance degradation, maintaining an accuracy of 96.5%. These findings demonstrate the practicality of the proposed framework for real-world deployment in resource-constrained environments. Full article

(This article belongs to the Special Issue Advanced Cryptographic Techniques for Digital Watermarking, Encryption, and Steganography)

► Show Figures

Figure 1

36 pages, 3496 KB

Open AccessArticle

A Dual-Track Specialist Feature Fusion and Meta-Learning Stacking Ensemble for Cervical Transformation Zone Classification in Colposcopy

by Edgar Fabián Rivera-Guzmán, Vladimir Espartaco Robles-Bykbaev, Bernardo J. Vega-Crespo and Veronique Verhoeven

Computers 2026, 15(7), 450; https://doi.org/10.3390/computers15070450 - 16 Jul 2026

Viewed by 460

Abstract

Cervical cancer remains a major global public health challenge, and the accurate classification of cervical transformation zones (TZs) constitutes a critical step in early detection and clinical decision-making. However, distinguishing between Type 2 and Type 3 transformation zones remains particularly challenging due to [...] Read more.

Cervical cancer remains a major global public health challenge, and the accurate classification of cervical transformation zones (TZs) constitutes a critical step in early detection and clinical decision-making. However, distinguishing between Type 2 and Type 3 transformation zones remains particularly challenging due to their high morphological similarity and the inherent interobserver variability associated with colposcopic assessment. In this study, we propose a novel Dual-Track Specialist Feature Fusion and Meta-Learning Stacking Ensemble architecture for the automated classification of cervical transformation zones using the Intel & MobileODT Cervical Cancer Screening dataset. The proposed framework integrates a global feature extractor based on ResNet50 (Gatekeeper) with a visual specialist based on InceptionResNetV2, trained exclusively on the most diagnostically ambiguous cases (Type 2 and Type 3). The extracted features are fused and processed through a multi-level stacking scheme composed of Multilayer Perceptron (MLP), Support Vector Machine (SVM), Gradient Boosting (GB), XGBoost, and LightGBM classifiers at the base level, followed by an XGBoost meta-learner and a clinically guided probability calibration strategy designed to maximize diagnostic sensitivity. Experimental results demonstrate a peak overall accuracy of 91.22%, substantially outperforming the baseline ResNet50 model (70%). Furthermore, the proposed system achieved Recall values of 0.90, 0.90, and 0.94 for Type 1, Type 2, and Type 3 transformation zones, respectively, highlighting its ability to accurately identify diagnostically challenging cases. Ablation studies, Grad-CAM visualizations, and external-image validation experiments confirm that the proposed architecture improves discrimination between ambiguous categories, learns clinically meaningful representations, and maintains strong generalization capability across heterogeneous scenarios. These findings demonstrate the potential of visual specialization and calibrated meta-learning strategies for the development of artificial intelligence-assisted colposcopic decision-support systems. Full article

(This article belongs to the Special Issue Machine and Deep Learning in the Health Domain (3rd Edition))

► Show Figures

Figure 1

28 pages, 9062 KB

Open AccessFeature PaperArticle

An Optimized Leakage-Aware YOLO-Based Deep Learning Framework for Instance Segmentation and Environmental Impact Assessment of Mixed Metal and Glass Waste

by Andrew N. Shafik, Mohamed H. Khafagy, Alber S. Aziz and Shereen A. Hussein

Computers 2026, 15(7), 449; https://doi.org/10.3390/computers15070449 - 15 Jul 2026

Viewed by 465

Abstract

Accurate instance segmentation of recyclable waste is important for automated sorting and circular economy applications. Mixed metal and glass municipal waste is challenging because metal surfaces are often reflective, while glass objects have transparent boundaries and high visual variability. This paper presents a [...] Read more.

Accurate instance segmentation of recyclable waste is important for automated sorting and circular economy applications. Mixed metal and glass municipal waste is challenging because metal surfaces are often reflective, while glass objects have transparent boundaries and high visual variability. This paper presents a controlled computer vision framework that uses instance segmentation as an object-level perception layer for mixed metal and glass waste and links eligible detections to screening-level WARM-based environmental interpretation. A seven-class annotated dataset of 1667 images was evaluated using a fixed train/validation/test split. Model development used the validation split for architecture screening, optimization tuning, transfer learning analysis, seed selection, and NMS selection, and the held-out test set was used only for final evaluation. Using mask mAP@50:95 as the primary segmentation metric, the selected YOLOv8m-seg model achieved 0.9435 on the locked test set and obtained a higher mask mAP@50:95 than Mask R-CNN and RF-DETR-Seg under the same locked test protocol. Image-level bootstrap 95% confidence intervals were also used to characterize uncertainty around the locked test segmentation comparison. For environmental reporting, mass-bearing detections were mapped to class-specific mass priors and EPA WARM v16 factors rather than direct mask area-to-mass conversion; the masks were used for instance-level separation, localization, and visual verification, not as direct physical mass measurements. Under the stated assumptions, the framework produced screening-level environmental estimates of 1113.57 MJ of energy savings and 70.20 kg CO₂e of avoided emissions, decreasing to 445.67 MJ and 28.62 kg CO₂e after applying recycling rate priors. Overall, the framework provides a leakage-aware workflow linking instance segmentation to material-specific WARM-based environmental screening. Full article

(This article belongs to the Special Issue Advances in Computer Vision: Models, Learning, and Inference)

► Show Figures

Figure 1

34 pages, 13771 KB

Open AccessArticle

Rehabilitation Engineering Approach to Frozen Shoulder Treatment: Performance Analysis Using Landmark-Based Motion Detection and Assistive Feedback Systems

by Thanawat Srikaewsiew, Sarunya Kanjanawattana, Nuntawut Kaoungku, Parin Sornlertlamvanich and Komsan Srivisut

Computers 2026, 15(7), 448; https://doi.org/10.3390/computers15070448 - 15 Jul 2026

Viewed by 901

Abstract

This paper presents a preliminary technical feasibility study of a landmark-based motion analysis system designed for potential future application in home-based rehabilitation monitoring for frozen shoulder (adhesive capsulitis), developed using computer vision (CV) and human–computer interaction (HCI) principles. The proposed system utilizes real-time [...] Read more.

This paper presents a preliminary technical feasibility study of a landmark-based motion analysis system designed for potential future application in home-based rehabilitation monitoring for frozen shoulder (adhesive capsulitis), developed using computer vision (CV) and human–computer interaction (HCI) principles. The proposed system utilizes real-time body landmark detection to quantify shoulder joint kinematics and provide rule-based automated feedback on exercise execution. The system combines automated and manual components: while shoulder angle assessment, cosine similarity analysis, and keyframe matching are automated, manual researcher input is required to define keyframes corresponding to movement states (start, midpoint, peak) for each therapeutic pose. The CV-driven perception is translated into HCI output, including quantitative movement scores and rule-based feedback indicators, demonstrating the technical potential for objective evaluation of rehabilitation exercise execution without specialized wearable sensors. Technical validation was conducted with 14 healthy volunteers (not frozen shoulder patients) executing standardized shoulder rehabilitation activities, demonstrating shoulder angle measurement with an overall mean absolute error (MAE) of 7.03° against general goniometry and 6.61° against clinical goniometry (RMSE: 8.50° and 8.79°, respectively). Movement similarity classification achieved F1-scores ranging from 0.870 (flexion) to 1.0 (internal rotation) when compared against expert evaluation, though these results are based on a controlled and largely imbalanced dataset with limited incorrect movement examples. The system additionally incorporates a facial expression recognition (FER) module, previously developed and validated in the authors’ prior work, as a supplementary component to support future integration of pain monitoring; this module was not independently validated in the present study. This preliminary technical feasibility study contributes to rehabilitation engineering by demonstrating the potential of semi-automated CV-based motion analysis and rule-based HCI feedback for shoulder movement assessment. The findings indicate technical feasibility for future investigation in home-based exercise monitoring; however, clinical utility cannot be claimed at this stage, as validation with actual frozen shoulder patient cohorts is required. Full article

(This article belongs to the Special Issue Innovative Research in Human–Computer Interactions)

► Show Figures

Graphical abstract

31 pages, 3368 KB

Open AccessArticle

nSim-RV: A Reproducible RISC-V Framework for Scheduler-Aware Timing Scalability Under Increasing Task Concurrency

by Nicolai Iuga, Nicoleta Cristina Gaitan, Ionel Zagan and Vasile Gheorghita Gaitan

Computers 2026, 15(7), 447; https://doi.org/10.3390/computers15070447 - 15 Jul 2026

Viewed by 687

Abstract

As RISC-V processors are increasingly considered for embedded real-time and control-oriented systems, evaluating how timing behavior changes under increasing task concurrency becomes essential. Adding runnable tasks can amplify preemptions, context-switch activity, response-time variability, execution jitter, and deadline pressure. Existing RISC-V simulation and virtual-platform [...] Read more.

As RISC-V processors are increasingly considered for embedded real-time and control-oriented systems, evaluating how timing behavior changes under increasing task concurrency becomes essential. Adding runnable tasks can amplify preemptions, context-switch activity, response-time variability, execution jitter, and deadline pressure. Existing RISC-V simulation and virtual-platform environments mainly target architectural exploration, functional validation, or full-system execution, and do not directly provide a controlled workflow for isolating scheduler-induced timing degradation across large configuration spaces. This paper presents nSim-RV, a configurable and reproducible RISC-V simulation and orchestration framework for scheduler-aware timing scalability evaluation. The framework combines automated campaign generation, deterministic workload configuration, structured dataset aggregation, duplicate validation, and timing-oriented metric extraction. The evaluation compares a standard shared-pipeline execution model with an nMPRA-inspired preserved-context mode under identical scheduler and workload conditions. The campaign includes CoreMark, Dhrystone, and a deterministic synthetic RT-Control workload, 2–32 concurrent tasks, 50 k–1 M cycle observation windows, cache-disabled and cache-enabled configurations, and four-stage and five-stage pipeline organizations, resulting in 864 validated configurations. Results show that increasing task concurrency amplifies timing variability and deadline pressure. Preserved-context execution reduces switching-induced disturbance and delays or reduces higher-pressure timing behavior in several trajectories. Under the five-stage cache-disabled RT-Control configuration at N = 32, it reduces the deadline miss ratio from 3.74% to 2.21%, corresponding to a 41.1% relative reduction, with the clearest benefits observed for Dhrystone and RT-Control at intermediate–high task counts. Full article

► Show Figures

Graphical abstract

62 pages, 1440 KB

Open AccessReview

From 2G to 5G: Literature Review of Identification and Location Attacks in Cellular Networks

by Daniel Asimionesei and Nirvana Alina Popescu

Computers 2026, 15(7), 446; https://doi.org/10.3390/computers15070446 - 14 Jul 2026

Viewed by 638

Abstract

Cellular networks from 2G to 5G have evolved to improve performance, reliability, and security. However, the protection of user identity and location remains a persistent challenge across generations. Although newer architectures introduce stronger authentication, temporary identifiers, and privacy-preserving mechanisms, attacks such as IMSI [...] Read more.

Cellular networks from 2G to 5G have evolved to improve performance, reliability, and security. However, the protection of user identity and location remains a persistent challenge across generations. Although newer architectures introduce stronger authentication, temporary identifiers, and privacy-preserving mechanisms, attacks such as IMSI catching, paging-based tracking, downgrade attacks, and identifier correlation continue to affect cellular systems. This study analyzes identification and location attacks in 2G, 3G, 4G, 5G Non-Standalone (5G NSA), and 5G Standalone (5G SA) networks. The analysis focuses on studies published between 2018 and 2026 and follows a structured methodology based on research questions, inclusion and exclusion criteria, quality assessment, and comparative synthesis of studies identified in scientific databases. Following the selection process, 86 studies were included in the final analysis. The review compares attack vectors, affected identifiers, exploited procedures, adversary capabilities, reported performance metrics, and proposed mitigations for each network generation. The results show that vulnerabilities inherited from previous generations, especially GSM/2G, remain relevant in modern architectures through downgrade attacks, fallback mechanisms, and LTE anchoring in 5G NSA deployments. The main contribution of this study is a cross-generational comparative synthesis of identity and location attacks in cellular networks. The findings highlight the need for rigorous standards implementation, effective detection mechanisms, reduced reliance on legacy technologies, and practical evaluation of mitigation solutions in real cellular environments. Full article

(This article belongs to the Special Issue Intrusion Detection and Trust Provisioning in Edge-of-Things Environment)

► Show Figures

Figure 1

35 pages, 2088 KB

Open AccessArticle

Machine Learning-Based Mobile Traffic Classification for QoS-Oriented Network Management

by Mohammed Aqeel Ismail and Okuthe P. Kogeda

Computers 2026, 15(7), 445; https://doi.org/10.3390/computers15070445 - 13 Jul 2026

Viewed by 661

Abstract

The increasing complexity and volume of mobile network traffic present significant challenges to maintain consistent Quality of Service (QoS) across diverse applications. Accurate traffic classification enables application-aware resource allocation by distinguishing applications with different bandwidth, latency, and reliability requirements. Traditional classification techniques, including [...] Read more.

The increasing complexity and volume of mobile network traffic present significant challenges to maintain consistent Quality of Service (QoS) across diverse applications. Accurate traffic classification enables application-aware resource allocation by distinguishing applications with different bandwidth, latency, and reliability requirements. Traditional classification techniques, including port-based identification and Deep Packet Inspection (DPI) have become inadequate and less effective due to widespread encryption, port masquerading, and growing privacy concerns. This paper presents a supervised learning-based approach for application-level network traffic classification specifically as a foundation for QoS optimization in future 5G networks. Since publicly available labeled 5G traffic datasets remain limited, this study uses the MIRAGE-2019 mobile traffic dataset as a proxy dataset to evaluate the proposed classification framework. A Random Forest classifier was implemented using flow-level statistical features extracted from the mobile application traffic. The framework further incorporates a rule-based QoS policy mapping informed by RFC 4594 DiffServ service class guidelines to assign application-specific priority levels, bandwidth requirements, latency sensitivity, and jitter tolerance. Experimental evaluation achieved an overall classification Accuracy of 71.83%, a Macro F1-score of 0.6701, and a Weighted F1-score of 0.7227 across twenty mobile applications. Although the experiments were conducted using a pre-5G mobile traffic dataset, the results demonstrate that supervised machine learning can effectively classify encrypted mobile application traffic and provide a practical foundation for application-aware QoS policy enforcement in future 5G and next-generation mobile networks. Full article

(This article belongs to the Section Cloud Continuum and Enabled Applications)

► Show Figures

Figure 1

15 pages, 1201 KB

Open AccessArticle

Hybrid Educational Ecosystem of a Metauniversity: Integrating a Web Platform and an Immersive Digital Twin in Engineering Education

by Madina Ipalakova, Dana Tsoy, Sanzhar Otkilbayev, Yevgeniya Daineko, Danil Sharipov and Umitkhan Turzhanov

Computers 2026, 15(7), 444; https://doi.org/10.3390/computers15070444 - 13 Jul 2026

Viewed by 446

Abstract

The timely integration of technology in education is a key factor in developing competitive specialists. Given this, the development of a digital educational platform for training engineering specialists is of strategic importance. The paper presents the development of a proprietary digital educational platform [...] Read more.

The timely integration of technology in education is a key factor in developing competitive specialists. Given this, the development of a digital educational platform for training engineering specialists is of strategic importance. The paper presents the development of a proprietary digital educational platform using immersive technologies for training engineering specialists while providing educators with easier access to their educational progress and ability to spend less time processing it. The results are considered data from an initial test, aimed primarily at assessing user perception, usability, and the potential of the immersive environment. The platform’s effectiveness stems from its ability to simulate complex processes and allow students to independently control the experiment. Full article

(This article belongs to the Section Human–Computer Interactions)

► Show Figures

Figure 1

21 pages, 8493 KB

Open AccessArticle

Conceptual Framework of Common Misconceptions Regarding Generative AI in Elementary School Students Using Concurrent Think-Aloud Protocols

by Marianna Thode and Ioannis Paliokas

Computers 2026, 15(7), 443; https://doi.org/10.3390/computers15070443 - 12 Jul 2026

Viewed by 564

Abstract

Education about AI, including its opportunities and limitations, is essential for responsible academic development. It helps students, teachers, parents, researchers and practitioners understand both the strengths and the limitations of generative AI and ensures that it is applied in ways that are ethical [...] Read more.

Education about AI, including its opportunities and limitations, is essential for responsible academic development. It helps students, teachers, parents, researchers and practitioners understand both the strengths and the limitations of generative AI and ensures that it is applied in ways that are ethical and socially responsible. Without this knowledge, there is a risk that generative AI will be misused or misunderstood, reducing its potential benefits. This study evaluates an educational game designed to integrate generative AI in elementary education (in art lessons), aiming to create awareness regarding the limitations of generative AI. Data was collected from 204 game sessions in two primary education schools in Greece, using serious games for art education, while the accuracy of AI outputs and the usability and level of user satisfaction were recorded. Following these tasks, researchers used the concurrent think-aloud (CTA) protocol to identify specific usability issues and responses of elementary students in biased, incomplete or inaccurate AI results. Experimental findings from this small-scale exploratory pilot study (16 students) indicated that combining RTA with serious games in art courses successfully reveal how children understand AI outputs and evaluate them based on their initial expectations. Full article

(This article belongs to the Special Issue The Digital Transformation of Education: Trends, Technologies, and Responsible Innovation)

► Show Figures

Figure 1

26 pages, 762 KB

Open AccessArticle

Adaptive Fusion of Bug Report Titles and Descriptions for Automated Bug Severity Classification in Software Maintenance

by Thananchai Khamket, Jatuphum Juanchaiyaphum, Theeraya Uttha, Manasawee Kaenampornpan and Jantima Polpinij

Computers 2026, 15(7), 442; https://doi.org/10.3390/computers15070442 - 11 Jul 2026

Viewed by 437

Abstract

Bug severity classification supports software maintenance by helping developers prioritize defect resolution. Most existing approaches combine bug report titles and descriptions into a single representation, often assuming that both textual components contribute equally to severity prediction. However, concise titles and detailed descriptions may [...] Read more.

Bug severity classification supports software maintenance by helping developers prioritize defect resolution. Most existing approaches combine bug report titles and descriptions into a single representation, often assuming that both textual components contribute equally to severity prediction. However, concise titles and detailed descriptions may provide different types of severity-related information. This study investigates their relative contribution and proposes an Adaptive Title–Description Fusion (ATDF) framework that separately encodes title and description representations and adaptively estimates their contribution during feature fusion. Experiments were conducted on Mozilla Bugzilla repositories using a multiclass classification setting with five severity categories. The proposed framework was compared with title-only, description-only, direct concatenation, and standard gated fusion baselines. ATDF achieved the highest overall performance, with a Macro-F1 score of 0.771, compared with 0.756 for standard gated fusion and 0.739 for direct concatenation. Additional analyses showed that title and description contributions varied across severity categories and repositories under the current experimental setting. Overall, the findings suggest that considering the relative contribution of title and description information can provide a practical approach to improving bug severity classification while offering additional insight into title–description fusion behavior in software repositories. Full article

(This article belongs to the Special Issue Advanced Software Engineering in the Age of Artificial Intelligence: Best Practices, Challenges, and Opportunities)

► Show Figures

Figure 1

40 pages, 30352 KB

Open AccessArticle

Elite-Guided Collaborative Stochastic Social Learning Optimization for LSTM-Based Carbon Emission Forecasting

by Fan Yang and Lixin Lyu

Computers 2026, 15(7), 441; https://doi.org/10.3390/computers15070441 - 10 Jul 2026

Viewed by 423

Abstract

To address the difficulty of accurately capturing the dynamic patterns of carbon emission time series—characterized by nonlinearity, non-stationarity, and complex fluctuations—this paper proposes a carbon emission prediction model based on an elite-guided collaborative social spider learning optimization algorithm (EGC-SSLO) integrated with a Long [...] Read more.

To address the difficulty of accurately capturing the dynamic patterns of carbon emission time series—characterized by nonlinearity, non-stationarity, and complex fluctuations—this paper proposes a carbon emission prediction model based on an elite-guided collaborative social spider learning optimization algorithm (EGC-SSLO) integrated with a Long short-term memory (LSTM) network. First, considering the limitations of the standard stochastic social learning optimization (SSLO) algorithm in complex high-dimensional optimization problems, such as insufficient elite information guidance, weak local exploitation in the later stages, and a tendency to become trapped in local optima, three complementary improvement strategies are introduced. The adaptive elite mean-guided search strategy enhances the search directionality by incorporating the cooperative information of the best individual and the elite mean. The worst-individual hybrid Cauchy–Lévy search mechanism achieves a dynamic balance between early-stage global exploration and late-stage local exploitation through long-range Lévy flights and fine-grained Cauchy perturbations. The quadratic directional exploitation strategy further refines the search trajectory of candidate solutions, thereby improving convergence accuracy. These three strategies significantly enhance the optimization performance without increasing the time complexity order of the algorithm. Experimental results on the CEC2017 (30-dimensional), CEC2020 (20-dimensional), and CEC2022 (20-dimensional) benchmark suites demonstrate that EGC-SSLO consistently outperforms classical algorithms such as PSO, GWO, and HHO, as well as their improved variants, in terms of convergence accuracy, convergence speed, and robustness. Furthermore, the Wilcoxon rank-sum test and Friedman test confirm that the observed improvements are statistically significant. Finally, an EGC-SSLO-LSTM carbon emission prediction model is constructed and applied to daily carbon emission data in China from 2019 to 2025 for empirical analysis. The experimental findings show that the EGC-SSLO-LSTM model markedly outperforms both the standard LSTM and SSLO-LSTM approaches across key evaluation metrics, including mean absolute error (MAE), root mean square error (RMSE), and the coefficient of determination (R²). In particular, the MAE is decreased by 39.9% and 4.64% compared with the two benchmark models, respectively, which highlights the strong effectiveness and practical potential of the proposed method in real-world carbon emission forecasting applications. Full article

(This article belongs to the Section AI-Driven Innovations)

► Show Figures

Figure 1

24 pages, 456 KB

Open AccessArticle

From RISC to Risk: Exception Handling as a Gateway to Exploitation

by Mina Soltani Siapoush and Jim Alves-Foss

Computers 2026, 15(7), 440; https://doi.org/10.3390/computers15070440 - 10 Jul 2026

Viewed by 452

Abstract

RISC-V’s open and extensible design improves flexibility, but it also permits vendors to customize trap behavior and reserved opcode space in ways that can affect exception handling. This paper presents a scoping review of RISC-V exception-handling security, focusing on how traps, privilege transitions, [...] Read more.

RISC-V’s open and extensible design improves flexibility, but it also permits vendors to customize trap behavior and reserved opcode space in ways that can affect exception handling. This paper presents a scoping review of RISC-V exception-handling security, focusing on how traps, privilege transitions, control registers, and handler routines can become attack surfaces when they are not correctly protected. We organize the literature by a five-stage exception lifecycle and map documented attacks and defenses to the architectural state they affect. Our review shows that exception-handling vulnerabilities arise from exposed control registers, insufficiently protected trap entry and return paths, and flawed handler execution. Recent attacks, such as GhostWrite, halt-and-catch-fire conditions, and side-channel exploits, illustrate how weaknesses in exception management can lead to privilege escalation, denial of service, or data leakage. In contrast, defenses are unevenly distributed across the lifecycle: backward-edge return-time integrity is addressed by multiple dedicated mechanisms, while dispatch-time and privilege-transition protection remain largely indirect and are often embedded within broader trusted-execution designs. Because the primary studies report heterogeneous cost metrics, the overhead values we extract are presented as reported rather than normalized across a common benchmark. The literature suggests that lightweight mechanisms, such as Physical Memory Protection and software-based control-flow enforcement, generally incur modest overhead, whereas stronger approaches, such as CHERI-RISC-V and Trusted-Execution Environments, may impose substantially higher costs on exception-heavy workloads. Overall, our findings indicate that exception handling should be treated as a security-critical boundary and that future work should emphasize stage-specific defenses, formal verification, and tighter integration between hardware and software protections. Full article

► Show Figures

Figure 1

36 pages, 3098 KB

Open AccessArticle

Benchmarking Fault-Tolerance Characteristics of Actor-Based Runtimes

by Luís Nogueira and Jorge Coelho

Computers 2026, 15(7), 439; https://doi.org/10.3390/computers15070439 - 10 Jul 2026

Viewed by 398

Abstract

Fault tolerance is a fundamental requirement of distributed systems, and actor-based runtimes provide a widely adopted approach for building resilient and highly concurrent applications. Although several actor ecosystems offer mechanisms for supervision, failure detection, and recovery, comparative studies frequently focus on performance metrics [...] Read more.

Fault tolerance is a fundamental requirement of distributed systems, and actor-based runtimes provide a widely adopted approach for building resilient and highly concurrent applications. Although several actor ecosystems offer mechanisms for supervision, failure detection, and recovery, comparative studies frequently focus on performance metrics rather than fault-tolerance behaviour. This paper presents a language-independent benchmarking framework for evaluating fault tolerance in actor-based runtimes. The framework was implemented using three representative ecosystems: Elixir/BEAM, Scala/Akka, and Go/Proto.Actor. A distributed chat-based benchmark application was used to measure throughput, reconnection latency, and failure-detection latency under recurring transient failures. All implementations followed an equivalent architecture and were executed under identical experimental conditions. The study deliberately targets a single, well-defined fault model: the supervised crash recovery of in-memory, effectively stateless actor services, in which chat actors are abruptly terminated and restarted by their supervisors while clients rediscover and reconnect to them. Stateful recovery (actor state, mailbox contents, in-flight or persistent messages), as well as multi-node network effects, are explicitly out of scope. Accordingly, the benchmark characterises supervised crash–recovery behaviour for largely stateless actor services rather than providing a comprehensive evaluation of actor-based fault tolerance. The results reveal distinct trade-offs among the evaluated ecosystems. Elixir achieved the highest throughput and the lowest throughput variability under fault conditions, while Scala/Akka consistently provided the lowest reconnection and failure-detection latencies, particularly at large scale. Go/Proto.Actor remained competitive in throughput-oriented scenarios but showed greater degradation in recovery-related metrics as concurrency increased. The results indicate that no single runtime dominates all evaluated dimensions of recovery behaviour. Beyond the runtime comparison, this work contributes a reproducible benchmarking framework that provides a foundation for future empirical studies of actor-based runtime recovery under controlled fault conditions. Full article

(This article belongs to the Special Issue Edge and Fog Computing for Internet of Things Systems (3rd Edition))

► Show Figures

Figure 1

27 pages, 2808 KB

Open AccessSystematic Review

A Scoping Review of the Literature on Swarm Intelligence Applications in Water Scheduling

by Cheslin van Wyk, Taryn Michael and Colin Chibaya

Computers 2026, 15(7), 438; https://doi.org/10.3390/computers15070438 - 10 Jul 2026

Viewed by 345

Abstract

Water scheduling is a complex optimization problem that requires efficient and adaptive solution approaches. Metaheuristic techniques, particularly swarm intelligence models, have increasingly been applied to address these challenges. This study presents a scoping review that maps and synthesizes the existing literature on the [...] Read more.

Water scheduling is a complex optimization problem that requires efficient and adaptive solution approaches. Metaheuristic techniques, particularly swarm intelligence models, have increasingly been applied to address these challenges. This study presents a scoping review that maps and synthesizes the existing literature on the application of swarm intelligence in water scheduling. Guided by the PRISMA-ScR framework and the JBI Population–Concept–Context (PCC) model, relevant studies published between 2015 and 2025 were identified across multiple databases. From an initial pool of 1357 studies, only 23 met the inclusion criteria and were subjected to detailed analysis. The findings reveal a strong concentration of research on water distribution networks, coupled with limited methodological diversity across the reviewed studies. There is an absence of explicit focus on resource-constrained or arid environments contexts where water-scheduling challenges are often most acute. Geographically, the literature is heavily skewed toward Asia, with the majority of studies conducted in China (n = 7) and Iran (n = 6). In contrast, only one study originated from Africa and one from Australia despite the disproportionate severity of water scarcity challenges across the African continent. The review exposes a critical gap in the literature and underscores the need for more context-aware, hybrid swarm intelligence models that explicitly account for the socio-economic and environmental constraints of water-stressed regions. Full article

► Show Figures

Figure 1

Journal Menu

Journal Browser

Computers, Volume 15, Issue 7 (July 2026) – 68 articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI