2026-04-01
Hussein Mohammed Barakat, Sawal Hamid Md Ali, Hayam Alyasiri
To support digital transformation and the increasing demand for reliable, scalable, and secure communication infrastructures in e-government initiatives, the Iraqi government has upgraded legacy institutional networks and expanded broadband and fiber-optic deployments. However, many public institutions continue to rely on poorly planned wireless LANs or inadequately structured fiber-optic networks, leading to limitations in performance, scalability, security, and quality of service. This paper presents a measurement-informed simulation study of two Iraqi institutional networks: The Information and Communications Technology Company (ITPC), which primarily depends on wireless LAN infrastructure, and the Ministry of Industry and Minerals, which operates a fiber-optic backbone. Real-world baseline measurements are first collected to characterize the operational state of the existing networks. Based on these observations, multiple redesign scenarios are developed and evaluated using OPNET Modeler 14.5 to analyze traffic behavior, throughput, and delay under realistic application workloads. For each network, several redesign scenarios are examined. Scenario 1 represents a structured Ethernet redesign using conventional device deployment, while Scenario 2 improves performance by replacing hubs with switches and upgrading selected links. Scenario 3, proposed as the optimized solution, employs device regrouping, reduced switch count, and higher-capacity backbone links to achieve a more scalable and efficient architecture. Simulation results show that, for ITPC, Scenario 3 increased throughput to approximately 9 Mbps while reducing average delay to 0.0000019 s, compared with 5.8 Mbps and 0.00020 s in the existing Wi-Fi network. For the Ministry of Industry and Minerals, Scenario 3 consistently outperformed other scenarios across key applications. Overall, the results demonstrate that carefully structured and optimized network architectures can significantly outperform poorly planned wireless and fiber-optic deployments. The study highlights the importance of systematic design and simulation-based validation in developing secure, scalable, and cost-effective network infrastructures for e-government, particularly in developing countries.
DOI: 10.3389/frcmn.2026.17715042026-02-09
Zhenning Chen, Zihe Xu, Yihan Ding, Youren Wang
IntroductionFederated learning (FL) enables model training on edge devices using local data while aggregating model updates at a central server without exchanging raw data, thereby preserving privacy. However, achieving satisfactory convergence accuracy with low communication energy remains challenging. This work investigates a three-tier clustered FL (CFL) architecture to improve global training performance and communication efficiency through joint device clustering and resource scheduling.MethodsWe analyze how clustering strategies influence learning convergence and communication energy consumption. Based on this analysis, we propose a clustering method that jointly accounts for gradient cosine similarity and communication distance. A simplified procedure is further developed for device association and cluster-head selection, with the goals of improving intra-cluster data balance and reducing the overall communication distance to the server.ResultsSimulations demonstrate that the proposed method consistently improves model accuracy while reducing communication energy consumption compared with random clustering and similarity-based clustering baselines.DiscussionThese results indicate that jointly considering update similarity and communication distance in CFL can effectively balance learning quality and communication cost, offering a practical approach for energy-efficient federated training in edge networks.
DOI: 10.3389/frcmn.2026.17488152025-10-30
Akbar Ali, Adnan Nadeem, Noureen Zafar, Muhammad Shiraz
The deployment of various sensors including inductive loops, radars, GPS devices, cameras and floating car data (FCD) in intelligent transportation systems generates a stream of heterogeneous data, further complicated by exogenous factors like weather conditions and temporal patterns (e.g., peak hours, weekends). For urban traffic development planning, the accurate prediction of congestion under the influence of these exogenous factors remains a major challenge. The proliferation of these diverse data sources creates a complex prediction environment, demanding advanced analytical frameworks. To address this issue, we propose a novel Fusion-based Generative Adversarial Network with Gated Recurrent Unit (F-GGRU) framework. The F-GGRU develops a generic data pipeline for integrating and preprocessing multi-source data, featuring advanced techniques for outlier removal, fuzzy logic-based automatic labeling, and Generative Adversarial Networks (GANs) for class balancing. Extensive experimentation was conducted on a novel real-time dataset from the Safe City Islamabad Pakistan (SCIP) project, integrating heterogeneous and exogenous features. The results demonstrate that our proposed F-GGRU framework achieves superior performance, with 98% accuracy, 0.99 precision, 0.98 recall, and a 0.98 F1-score. This significantly outperforms a suite of benchmark models, including Logistic Regression, Random Forest, XGBoost, and deep learning baselines like ANN, which achieved accuracies between 77% and 83% with correspondingly lower precision, recall, and F1-scores. Significantly, hyperparameter tuning and validation on a second independent dataset (CityPulse, Aarhus) confirmed the proposed framework robustness and generalizability, achieving even higher performance 99.42% accuracy and 0.99 AUC. These findings affirm that the F-GGRU framework is a robust and generalizable solution for real world traffic congestion prediction in smart cities.
DOI: 10.3389/frcmn.2025.16664872025-10-29
Hitesh Mohapatra
IntroductionThis work presents an adaptive ant colony (AdCO) framework for dynamic task management in heterogeneous Non-Terrestrial Network–Internet of Things (NTN-IoT) systems integrating Unmanned Aerial Vehicles (UAVs) and Low Earth Orbit (LEO) satellites. The framework addresses key challenges such as stochastic mobility, intermittent connectivity, and latency-sensitive operations common in large-scale IoT deployments.MethodsThe proposed approach employs adaptive pheromone learning, heuristic control, and multi-timescale scheduling. It follows a hierarchical co-optimization strategy, where UAV swarms perform edge-side task allocation while LEO satellites handle relay scheduling during orbital passes. Event-triggered pheromone resets and distributionally robust cost modeling are introduced to maintain stability and adaptability under dynamic network conditions.ResultsSimulation results demonstrate superior performance compared to classical Ant Colony Optimization (ACO) and recent meta-heuristic methods. The proposed model achieves higher task completion ratios, reduced end-to-end latency, and enhanced energy-normalized throughput across different orbital configurations, traffic patterns, and link failures.DiscussionThe findings confirm the efficiency and resilience of the proposed framework in NTN-IoT operations. Its adaptability makes it suitable for critical applications such as disaster response, precision agriculture, and maritime monitoring, where real-time coordination and reliability are essential.
DOI: 10.3389/frcmn.2025.16913462025-10-24
Israel Tommy, Taoreed Akinola, Xiangfang Li, Lijun Qian
IntroductionBeam-level traffic forecasting plays a vital role in the optimization of 5G networks by enabling proactive resource allocation and congestion control. However, the task is complicated by inherent data sparsity and the presence of multi-scale temporal dynamics, making accurate predictions difficult to achieve using conventional models.MethodsTo address these challenges, we propose a Gated Recurrent Unit (GRU)-based Multi-Task Learning (MTL) framework, enhanced by a weighted ensemble approach. We systematically evaluate the performance of six forecasting models—Linear Regression, DLinear, XGBoost, Echo State Network (ESN), Long Short-Term Memory (LSTM), and GRU-MTL—across three input sequence lengths (168-h, 24-h, and 8-h) using real-world beam-level data from the ITU AI for Good initiative.ResultsExperimental findings reveal that the GRU-MTL model significantly outperforms traditional baselines, achieving a Mean Absolute Error (MAE) of 0.2136 on 168-h sequences compared to LSTM’s 0.3223. Long sequences (168-h) reduce MAE by 56% relative to short 8-h windows, effectively mitigating the effects of sparsity. Furthermore, an ensemble of top-performing models (MTL, XGBoost, and Linear Regression) yields additional gains, reducing MAE to 0.2105—a 1.45% improvement over MTL alone. DiscussionThese results highlight the importance of long-term temporal context and model diversity for robust traffic prediction in sparse environments. The proposed framework offers practical guidelines: 168-h forecasting windows are optimal for weekly planning, and model ensembling enhances generalization across varying beam activity levels. This study contributes a scalable and accurate solution for spatio-temporal traffic forecasting in next-generation wireless networks.
DOI: 10.3389/frcmn.2025.16584612025-10-17
Yigang Shen, Lei Xie, Ming Li
In tactical communication networks, highly dynamic topologies and frequent data exchanges create complex spatiotemporal dependencies among link states. However, most existing intelligent routing algorithms rely on simplified model architectures and fail to capture these spatiotemporal correlations, resulting in limited situational awareness and poor adaptability under dynamic network conditions. To address these challenges, this study proposes an intelligent path selection method—Deep Reinforcement Learning with Spatiotemporal-aware Link State Guidance Algorithm (DRLSGA). The algorithm builds upon the Proximal Policy Optimization (PPO) framework to develop an intelligent decision-making model and integrates a link state feature extraction module that combines Gated Recurrent Units (GRU) and a Graph Attention Network (GAT). This design enables the model to learn long-term temporal dependencies and spatial structural relationships from sequential link state data, thereby enhancing perception and decision-making capability. An attention mechanism is further introduced to highlight salient features within link state sequences, while an optimal routing strategy is derived through a deep reinforcement learning-based training process. Experimental results demonstrate that, compared with the existing DRL-ST algorithm, DRLSGA reduces average end-to-end latency by at least 2.07%, lowers the packet loss rate by 1.65%, and increases average throughput by up to 2.59% under high-traffic conditions. Moreover, the proposed algorithm exhibits stronger adaptability to highly dynamic network topologies.
DOI: 10.3389/frcmn.2025.16359822025-10-16
Farhan Nisar, Muhammad Amin, Muhammad Touseef Irshad, Hassan Jalil Hadi, Naveed Ahmad, Mohamad Ladan
The pervasive growth of the Internet of Things (IoT) necessitates efficient communication technologies, among which Long Range Wide Area Network (LoRaWAN) is prominent due to its long-range, low-power characteristics. A significant challenge in dense LoRaWAN deployments is the efficient management of resources, particularly Spreading Factor (SF) allocation. In this paper, we propose a machine learning-based approach for optimal SF allocation to enhance network performance. We developed a simulation-driven framework utilizing the ns-3 simulator to generate a comprehensive dataset mapping network conditions, including RSSI, SNR, device coordinates, and distance to the gateway, to optimal SF assignments determined through an energy-aware optimization process. An XGBoost model was trained on this dataset to predict the optimal SF based on real-time network parameters. Our methodology focuses on balancing packet delivery ratio and energy consumption. The performance evaluation demonstrates that the trained XGBoost model effectively classifies optimal SFs, exhibiting strong diagonal dominance in the confusion matrix and achieving competitive accuracy with efficient computational characteristics, making it suitable for resource-constrained LoRaWAN environments.
DOI: 10.3389/frcmn.2025.1665262