Background subtraction plays an important role in intelligent video surveillance since it is one of the most used tools in motion detection. If scientific progress has enabled to develop sophisticated equipment for this task, algorithms used should be improved as well. For the past decade a background subtraction technique called ViBE is gaining the field. However, the original algorithm has two main drawbacks. The first one is ghost phenomenon which appears if the initial frame contains a moving object or in the case of a sudden change in the background situations. Secondly it fails to perform well in complicated background. This paper presents an efficient background subtraction approach based on ViBE to solve these two problems. It is based on an adaptive radius to deal with complex background, on cumulative mean and pixel counting mechanism to quickly eliminate the ghost phenomenon and to adapt to sudden change in the background model.

The rapid deployment of Next Generation Networks (NGNs) such as 4G/5G, fiber-optics, and software-defined infrastructures in Cameroon, has introduced new cybersecurity challenges, particularly in the face of increasingly sophisticated cyber threats. Traditional intrusion detection and prevention systems lack the scalability, adaptability, and real-time intelligence required to defend against modern attacks such as zero-day exploits, Advanced Persistent Threats (APTs), and insider threats. This paper proposes an AI-based Cyber Threat Intelligence (CTI) and prediction framework specifically designed for the Cameroonian NGN ecosystem. The proposed system leverages machine learning, threat intelligence feeds, and predictive analytics to identify, correlate, and forecast potential cyber threats in real time. It integrates data from diverse sources including telecom infrastructures, national CSIRTs, and open-source threat databases. A prototype implementation using simulated and public NGN traffic datasets demonstrates the system’s effectiveness in early threat detection and attack prediction. The framework aims to support decision-making and proactive cyber defense strategies in Cameroon's telecommunications and public sector environments. Keywords Artificial Intelligence (AI), Cyber Threat Intelligence (CTI), Intrusion Detection, Next Generation Networks (NGN), Cameroon, Threat Prediction, Machine Learning, 5G Security, Federated Learning, Network Security, SIEM Integration, Real-Time Analytics.

This paper introduces PREFAP, an approach to solve the multi-sensor patrolling problem in unknown environment. The multi-sensor patrolling problem consists in moving a set of sensors on a pre-set territory such that each part of this territory is visited by the sensor agents as often as possible. Eachsensor has a communicational radius and a sensory radius. indeed, optimal patrol can only be achieved if the duration between two visits of the same area of the environment is as minimal as possible. This time between two visits is called idleness. Thus, an effective patrol technique must make it possible to minimize idleness in the environment.That is why after a deep analysis of the existing resolution's approaches, we propose a hybrid approach of resolution with three components: perceptionreaction, field of strength and learning. In absence of obstacles, the perception-reaction component gives to the sensors a purely reactive behavior, as a function of their local perceptions, which permit them to move easily in their environment. The strength module enables the sensors to avoid the obstacles in the environment. As regards to the learning module, it allows the sensors to get out of blocking situations encountered during obstacle avoidance. This approach, called PREFAP, must be able to minimize idleness in different areas of the environment. The simulation results obtained show that the approach developed effectively minimizes idleness in the environment. This allows on the one hand, to have a regular patrol in the environment; on the other hand, thanks to the minimization of idleness of the areas of the environment, PREFAP will allow the sensors to quickly detect the various possible events which can occur in different areas of the environment.

The deployment of Next Generation Networks (NGNs) across Africa is accelerating digital connectivity and economic growth. However, this transformation is accompanied by escalating cybersecurity risks that challenge the effectiveness of traditional perimeter-based security architectures. This paper presents an Artificial Intelligence (AI)-driven Zero Trust Security (ZTS) framework specifically designed for NGN environments within the African context. Rooted in the "never trust, always verify" principle, the proposed framework leverages machine learning and behavioral analytics to enable continuous authentication, dynamic access control, and real-time threat mitigation. The integration of AI enhances the system’s ability to detect anomalous activities, respond to zero-day attacks, and adapt security policies based on contextual awareness. Recognizing Africa’s unique challenges—including infrastructural constraints, diverse regulatory landscapes, and limited skilled cybersecurity personnel, the model emphasizes scalability, automation, and cost-efficiency. Simulations and case studies demonstrate the proposed solution’s effectiveness in mitigating insider threats, lateral movement, and advanced persistent threats (APTs) within NGN infrastructures. The findings suggest that AI-powered ZTS can serve as a foundational cybersecurity strategy for building resilient, intelligent, and secure digital ecosystems in Africa. Index Terms: Artificial Intelligence (AI), Zero Trust Security (ZTS), Next Generation Networks (NGN), Cybersecurity, Network Access Control, Behavioral Analytics, Advanced Persistent Threats (APTs), African Telecommunications, Context-Aware Security, Dynamic Threat Detection, Identity and Access Management (IAM), Digital Infrastructure Security, Policy-Based Security.

The increasing complexity of multi-vendor network infrastructures presents substantial challenges in ensuring robust cybersecurity. Conventional network defense mechanisms are often inadequate to counter the evolving sophistication and dynamism of contemporary cyber threats. Furthermore, the proliferation of 5G networks, Internet of Things (IoT) devices, and cloud computing technologies has accelerated digital transformation, driving economic growth and connectivity while simultaneously expanding the attack surface. As cyberattacks become more advanced, maintaining network security and data integrity has become increasingly difficult. This paper investigates the application of Artificial Intelligence (AI) in enhancing network security by addressing critical challenges in predictive threat detection, real-time anomaly response, and network optimization. AI techniques demonstrate exceptional effectiveness in these domains, enabling proactive identification and mitigation of threats while improving overall network performance. Experimental results indicate that AI-driven security systems achieve up to 92% accuracy in cyber threat detection, reduce average incident response time to under 1.5 minutes, and enhance bandwidth allocation efficiency by 35% during peak traffic periods. The study further introduces a conceptual framework for AI-based network security automation, integrating machine learning, predictive analytics, and natural language processing within network monitoring tools. This model enables autonomous threat detection, response, and prevention, contributing to more resilient and adaptive cybersecurity infrastructures. KEYWORDS: Artificial Intelligence, Cybersecurity, Data Networking, Predictive Analytics, Network Optimization