榴莲视频

Cyber security Center at PMU is a state-of-the-art center in the region that is working in multiple domains to enhance cybersecurity in interconnected ecosystems such as the Internet of Things (IoT), Internet of Unmanned Aerial Vehicles (IoUAV), and Internet of Medical Things (IoMT) to name few. 

A research team led by Professor Farhan Ullah is currently focusing on applications emphasizing real-time protection for IoT devices, drones, and healthcare systems, mitigating sophisticated malware and intrusion threats. Another significant focus is the development of smart surveillance systems capable of autonomously detecting threats in critical infrastructure and healthcare environments. For future applications, the research team aims to integrate blockchain technology to secure data sharing across critical infrastructures, ensuring transparency and reliability. 

Leveraging generative AI models, the current research projects at PMU Cybersecurity Center seek to simulate and predict emerging cyber threats proactively. Additionally, efforts are being made to develop adversarial resilient systems that detect and mitigate manipulations targeting AI-driven security frameworks.

The scope of the cybersecurity center research center spans diverse cybersecurity challenges, employing innovative methodologies. A key focus is on developing feature representation techniques, including network traffic analysis, API-Call Graphs, and image-based malware classification. In some of the ongoing research projects advanced Intrusion Detection Systems (IDS) are being designed for resource-constrained edge devices by utilizing deep learning models and metaheuristic optimization techniques like Deep Particle Swarm Optimization (DPSO). Another core research aspect being considered at the center is the implementation of Federated Learning (FL) frameworks that facilitate collaborative learning without sharing raw data, maintaining privacy while improving model robustness. The research also prioritizes the development of adversarial defense mechanisms, employing hybrid approaches like PGD and IGSM to counter sophisticated cyber threats. Finally, the integration of generative AI techniques enhances proactive threat mitigation through malware analysis and attack prediction.

Novelty and Uniqueness of the Works at the Cybersecurity Center

The uniqueness of the researches lies in their innovative methodologies and groundbreaking contributions. The integration of deep transfer learning with adversarial attack detection provides an unprecedented approach to enhancing cybersecurity. The development of lightweight IDS ensures optimal functionality in resource-constrained edge environments. Furthermore, the creation of domain-specific datasets tailored for IoUAVs and IoMT applications bridges a significant gap in the field, enabling focused advancements in intrusion detection and malware analysis.

Privacy-preserving FL for edge-based healthcare informatics

Additionally, a significant highlight is attaining 99.27% classification accuracy using novel feature representations combining byte-level image processing and API-Call Graphs. The introduction of robust hybrid adversarial defense mechanisms has significantly improved the detection and mitigation of advanced cyber threats. The deployment of FL frameworks enables secure collaboration across distributed systems without compromising privacy. Additionally, the development of benchmark datasets from real-world IoUAV scenarios provides a valuable resource for advancing cybersecurity research. Finally, the adoption of generative AI models facilitates proactive threat mitigation, enabling the simulation and neutralization of emerging threats effectively.