The Internet of Vehicles (IoV) connects to the internet via a 5G network and covers a wide range of different network architectures. In this environment, ubiquitous network connections are formed. High mobility of vehicles requires frequent handovers in heterogeneous networks due to the presence of multiple radio access technologies. For vehicles at short range, dedicated short‐range communication (DSRC) is used, while better technology is needed for long‐range communication. 4G long‐term evolution (4G LTE) features large bandwidth and wider cell coverage but has high transmission time intervals. 5G provides higher data rates and reduced transmission latency; but because it cannot penetrate thick objects, it often suffers from blockage issues. The integration of DSRC and cellular network technology in IoV has aroused research interest in recent years due to the need to study and identify appropriate network selection in a heterogeneous IoV (HIoV) environment so that handover can be done efficiently. Previous research selects a network directly or connects to available radio access networks, resulting in frequent handovers. To address this challenge, a hybrid algorithm in heterogeneous IoV network (HAHIN) is proposed as an integrated framework that conditionally coordinates handover decision, network selection, and vehicle‐to‐vehicle (V2V) route optimization. The proposed work develops an integrated decision framework for HIoV by coordinating handover validation, network selection, and V2V routing across DSRC, 4G LTE, and 5G mmWave technologies, where handover necessity is explicitly validated before initiating target‐network selection and subsequent V2V route optimization. The necessity for a handover is determined using a dynamic Q‐learning algorithm. Network selection is performed through a fuzzy convolutional neural network (F‐CNN) algorithm. When data transfer is required between vehicles, the jellyfish search optimization (JSO) algorithm facilitates V2V chain routing. Thus, the three established techniques are applied sequentially to address distinct but interconnected HIoV decision stages. The validation results and performance evaluations demonstrate that the proposed approach outperforms existing state‐of‐the‐art methods.
Reconfigurable intelligent surfaces (RISs) have emerged as a compelling technology for next‐generation wireless networks, enabling programmable control of electromagnetic wave propagation to improve coverage, spectral efficiency, and energy efficiency. However, practical deployment remains constrained by challenges in channel estimation, real‐time phase‐shift optimization, and scalability in dynamic environments. Digital twin (DT) technology provides a promising solution by constructing high‐fidelity virtual replicas of wireless systems that enable predictive modeling, continuous monitoring, and closed‐loop RIS control. While recent works have begun exploring DT–RIS integration, existing surveys lack a unified framework that connects application‐level objectives with the underlying computational mechanisms. To address this gap, this paper introduces a dual‐perspective taxonomy that organizes DT–RIS research along two complementary dimensions: (i) functional enablers, covering latency reduction and MEC performance, resource management, and security; and (ii) methodological enablers, including reinforcement learning, federated learning, generative models, and convex optimization. This joint perspective reveals how specific algorithmic strategies underpin distinct network objectives, exposing interdependencies overlooked in prior single‐axis analyses. Building on this framework, the survey synthesizes DT–RIS applications and identifies a fundamental trade‐off between DT modeling fidelity and synchronization overhead that limits real‐time operation. Finally, the paper outlines key open challenges, including CSI acquisition overhead, DT update latency, and control signaling cost, and discusses directions toward scalable and practically deployable DT–RIS architectures for 6G networks.
Low‐light image enhancement remains a challenging problem due to poor illumination, increased noise, and color degradation. In this paper, we introduce a new enhancement framework based on quality‐guided fuzzy sigmoid selection network (QFFS‐Net). First, multiple enhancement candidates are generated by applying fuzzy‐based sigmoid transformations to the luminance channel in the Lab color space. These candidates are then evaluated by a lightweight convolutional neural network that assigns each a perceptual quality score. The predicted quality scores are normalized using a softmax‐based fusion strategy to combine the enhancement candidates into a fused intermediate image. This intermediate image is subsequently processed by a lightweight residual refinement network to correct residual errors and produce the final enhanced output. This design allows the model to exploit complementary information from multiple enhancement hypotheses instead of relying solely on a single transformation. The proposed method achieves comparable results of PSNR and SSIM against the benchmark datasets LOL and LOLv2‐Synthetic. It also provides visually consistent results on no‐reference datasets such as LIME and DICM. The proposed method provides a better trade‐off between enhancing lighting, maintaining structure, and color fidelity compared to existing methods. QFFS‐Net offers an effective and computationally efficient approach to low‐light image enhancement without relying on complex end‐to‐end architectures or large training datasets. The work code can be downloaded via the link: https://github.com/hazimghd/QFFS-Net .
Nowadays, the full bridge‐based dual active bridge (FB‐DAB) converters are used in various applications. However, these converters suffer from higher switching losses and increased harmonic distortion, thus limiting their overall performance. Keeping these drawbacks of conventional FB‐DAB converters in view, this paper presents a five‐level dual active bridge (5L‐DAB) converter as an advanced solution to enhance efficiency while reducing the total harmonic distortion (THD) during DC‐DC power conversion. The 5L‐DAB converter has 5L voltage waveforms at both ends of the converter. This research focuses on modeling and analyzing the performance of the 5L‐DAB converter. For this, firstly, the model of the 5L‐DAB converter has been developed under MATLAB/Simulink, and then its performance has been analyzed. To realize the efficacy of the presented 5L‐DAB converter, its performance has been compared with the conventional FB‐DAB converter. Detailed simulations and comparative studies with the FB‐DAB converter demonstrate that the 5L‐DAB converter achieves a marked reduction in THD while maintaining improved power conversion efficiency. Furthermore, a real‐time Typhoon Hardware‐in‐the‐Loop (HIL) emulator confirms the practical feasibility of the design. The real‐time HIL‐based analysis further validates the converter’s effectiveness across different phase‐shift operating points in both forward and backward power‐flow modes, demonstrating its applicability for battery‐based energy storage system, next‐generation electric vehicle charging, and renewable energy integration.
An improved hybrid metaheuristic algorithm that combines the genetic algorithm (GA) for global search initiation and the black hole (BH) algorithm for local search is proposed. The proposed algorithm is developed to address the accuracy and robustness challenges associated with the original hybrid genetic black hole (GABH) algorithm, particularly when applied to high‐dimensional and real‐world engineering optimization problems. Variable population‐size GA based on the saw‐tooth scheme and a local BH algorithm with its modified method of generating new stars or solutions to replace eliminated stars are used to improve the performance of the original GABH algorithm. The proposed algorithm’s efficacy is evaluated using a suite of twenty‐three CEC benchmark functions, the welded beam, and the tension/compression spring design engineering benchmark. Comparative analyses of the proposed algorithm against traditional and state‐of‐the‐art optimization algorithms, including GA, PSO, mayfly, BH, and the original hybrid GABH, demonstrate that the improved algorithm exhibits superior performance in terms of optimal value, variability, and accuracy. This was further validated by the statistical significance test, which was done using the Friedman test and the Holm–Šidák post hoc method for pairwise comparison.
