AI FOR OPERATING SYSTEM OPTIMIZATION: A REVIEW OF INTELLIGENT SCHEDULING, RESOURCE ALLOCATION, AND SECURITY

Shayan Abbas; Faryal Abbas Abdi; Sumera Mehmood

Authors

Shayan Abbas Author
Faryal Abbas Abdi Author
Sumera Mehmood Author

Keywords:

Artificial Intelligence in Operating Systems, AI-driven Task Scheduling, Predictive Memory Management, Intelligent Resource Allocation, AI-enhanced OS Security, Reinforcement Learning in OS, Deep Learning for System Optimization, Kernel-Level AI Integration, Real-time AI Inference, OS Performance Benchmarking, Adaptive Operating Systems, Interpretable Machine Learning, Federated Learning for OS, Smart Containers and AI-kernels, Unified AI-OS Architecture

Abstract

This paper presents a literature review of the incorporation of Artificial Intelligence (AI) in optimization of operating system (OS), and focused on the efforts of Artificial Intelligence in the area of task scheduling, resource allocation, memory management, and security management. Despite the fact that reinforcement learning, LSTM networks, and autoencoders are AI models that have demonstrated measurable improvements in specific OS subsystems, their application remains isolated and disjointed. The main limitations described in the research are the lack of cross-domain integration, inefficiency in real-time, non-interpretability of models, and the absence of common benchmarking processes. The paper presents the review of the recent literature with the assistance of which it is possible to focus on the need of low-latency, interpretable, and modular AI models that could be applied to the operating system components on the kernel level. It also highlights the importance of having open platform APIs and unhomogenized datasets to increase scalability and reproducibility. The findings can be applied as a guideline to the development of intelligent, adaptive and secure operating systems that are capable of self-optimization and autonomous decision making in a broad range of computing environments.