Low-latency content delivery is essential in the Internet of Vehicles (IoV) to support autonomous driving, cooperative perception, and infotainment services. However, rapidly changing vehicular mobility and demand patterns limit the effectiveness of existing content prediction and caching strategies, which often capture either short-term temporal trends or long-range dependencies, but not both. This paper proposes a hybrid deep learning architecture that integrates Long Short-Term Memory (LSTM) networks with Transformer encoders to jointly model fine-grained temporal dynamics and global correlations in content requests. The resulting popularity predictions are incorporated into a reinforcement learning (RL)-based caching policy, enabling proactive and adaptive cache placement at roadside units (RSUs) within an end-to-end optimization framework. Simulation results across representative IoV scenarios show that the proposed approach consistently improves cache hit ratio, retrieval latency, and prediction accuracy compared with LSTM-only, Transformer-only, Least Frequently Used (LFU), and Least Recently Used (LRU) baselines. Ablation studies further demonstrate the complementary strengths of the hybrid components, highlighting improved convergence behavior and robustness under varying demand distributions.

Mobile Wireless Sensor Networks (MWSNs) enhance traditional wireless sensor networks by allowing sensor nodes to move, resulting in continuously changing network topologies. Although this mobility enables advanced applications such as disaster response, intelligent transportation systems, and mission-critical monitoring, it poses major challenges for secure and scalable key management in large-scale deployments. Most existing key management and key pre-distribution schemes are tailored to static or lightly mobile networks and therefore suffer from limited scalability, excessive memory consumption, inefficient key utilization, and increased vulnerability to node capture when applied to highly mobile environments. This paper proposes a mobility-aware, zone-based key management scheme that integrates an enhanced composite key distribution mechanism with dynamic key refinement. The network is partitioned into logical zones, each maintaining an independent key pool to confine security breaches and improve scalability. To adapt to mobility-induced topology changes, sensor nodes continuously refine their key rings by preserving only the cryptographic keys associated with persistent neighbor relationships. This selective retention strategy significantly reduces storage overhead while strengthening resilience against key compromise and unauthorized access. Comprehensive analytical modeling and performance evaluations demonstrate that the proposed scheme achieves higher secure connectivity, stronger resistance to node capture attacks, and improved scalability compared to existing approaches, particularly in dense and highly mobile MWSN scenarios.

Many cloud providers offer very high precision services to exploit Optical Character Recognition (OCR). However, there is no provider offers Tifinagh Optical Character Recognition (OCR) as Web Services. Several works have been proposed to build powerful Tifinagh OCR. Unfortunately, there is no one developed as a Web Service. In this paper, we present a new architecture of Tifinagh Handwriting Recognition as a web service based on a deep learning model via Google Colab. For the implementation of our proposal, we used the new version of the TensorFlow library and a very large database of Tifinagh characters composed of 60,000 images from the Mohammed Vth University in Rabat. Experimental results show that the TensorFlow library based on a Tensor processing unit constitutes a very promising framework for developing fast and very precise Tifinagh OCR web services. The results show that our method based on convolutional neural network outperforms existing methods based on support vector machines and extreme learning machine.

Many cloud providers offer very high precision services to exploit Optical Character Recognition (OCR). However, there is no provider offers Tifinagh Optical Character Recognition (OCR) as Web Services. Several works have been proposed to build powerful Tifinagh OCR. Unfortunately, there is no one developed as a Web Service. In this paper, we present a new architecture of Tifinagh Handwriting Recognition as a web service based on a deep learning model via Google Colab. For the implementation of our proposal, we used the new version of the TensorFlow library and a very large database of Tifinagh characters composed of 60,000 images from the Mohammed Vth University in Rabat. Experimental results show that the TensorFlow library based on a Tensor processing unit constitutes a very promising framework for developing fast and very precise Tifinagh OCR web services. The results show that our method based on convolutional neural network outperforms existing methods based on support vector machines and extreme learning machine.

Reliable data transport is an essential requirement for many multimedia applications in wireless sensor networks. Actually, an efficient transport protocol for these applications must take into account not only reliability and energy consumption factors but also memory occupancy and data delivery delay. Recently, many research works have been conducted in this area, however the proposed protocols treat some of these aspects and neglect others. Contrarily, in this paper we present a novel transport solution designed to provide 100% reliability without making light of other factors. Through different mechanisms, we attempt to reach this objective with congestion avoidance and good performances in terms of energy consumption, delivery delay, and memory storage. The proposed protocol, called congestion avoidance with reliable transmission and energy efficiency (CARTEE), attains these goals through several mechanisms, namely: fixed sliding window transmission, alternative implicit/explicit acknowledgement, a new congestion detection technique, and distributed transmission rate adjustment. To evaluate the proposed protocol, we have conducted simulations using ns-3. The obtained results confirm the efficiency and scalability of CARTEE and demonstrate that it outperforms the recent proposed transport protocols in terms of reliability, congestion avoidance, data cache occupancy, and latency.

Providing quality of service (QoS) for real-time multimedia applications such as video streaming in mobile ad hoc networks (MANETs) is an important challenge. MANETs are characterized by lack of fixed infrastructure, dynamic topology, and limited resources that make more difficult multimedia applications transport and run on this networks. To overcome this challenge, video coding techniques combined to multiple routing paths (multipath) is a promising technique for supporting transmission of multiple video streams with appropriate QoS over mobile ad hoc networks. In this paper, firstly, many issues and different techniques for video streaming over MANET have been reviewed and secondly two multi paths routing protocols (M-AODV and MDSDV) have been evaluated in order to improve QoS for real-time multimedia applications. Results show that none of these two protocols is better than the other. In certain situations (throughput and load network with high mobility) is M-AODV but in others (network load and reliability for large-scale network) is MDSDV protocol which displays good performance. It is also noted that these two protocols provide between acceptable and good quality and a small jitter regardless of nodes number in medium mobility.