International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 12 | Dec 2024
p-ISSN: 2395-0072
www.irjet.net
A Comprehensive Survey on AI-Enhanced CPU Scheduling in Real-Time Environments: Techniques, Challenges, and Opportunities 1Mourya Teja Yalamanchili, 2Parnita Hiremath, 3Kethan Mulpuri 1 Product Specialist, Electronics and Computer Technology, Indiana State University, Terre Haute, Indiana 2 Product Specialist, Information science and engineering, B V Bhoomaraddi College of Engineering and
Technology,Hubli, Karnataka, India
3Software Development Engineer Technology Management, University of Bridgeport, Bridgepor, Connecticut
------------------------------------------------------------------------***------------------------------------------------------------------------ABSTRACT: The increasing complexity of real-time systems in critical applications such as autonomous vehicles, industrial automation, and telecommunications necessitates more sophisticated CPU scheduling strategies to ensure tasks are executed within strict timing constraints. Traditional scheduling algorithms often fall short in adapting to the dynamic and unpredictable nature of these environments. This survey provides a comprehensive examination of AI-enhanced CPU scheduling techniques in real-time systems, exploring how Artificial Intelligence (AI) and Machine Learning (ML) can optimize scheduling decisions. We analyze various AI-driven approaches, including reinforcement learning, deep neural networks, and evolutionary algorithms, focusing on their ability to predict task execution times, dynamically adjust to workload changes, and improve overall system performance. The survey also delves into the challenges of integrating AI with real-time systems, such as the need for low-latency processing and the complexity of real-time learning. Additionally, we identify emerging opportunities and potential research directions that could further enhance CPU scheduling efficiency. Our findings suggest that AI-enhanced CPU scheduling not only offers significant improvements in meeting real-time constraints but also opens new avenues for creating more adaptive and resilient systems in increasingly demanding real-time environments.
Monotonic Scheduling (RMS) and Earliest Deadline First (EDF), while foundational, often struggle to cope with the dynamic and unpredictable nature of modern real-time environments, leading to potential inefficiencies and missed deadlines (Liu & Layland, 1973; Buttazzo, 2005). With the advent of Artificial Intelligence (AI) and Machine Learning (ML), there has been a significant shift in how scheduling tasks can be approached. AI and ML offer the ability to learn from data, adapt to changing conditions, and make intelligent decisions based on real-time inputs. These capabilities are particularly advantageous in the context of real-time systems, where workloads can vary unpredictably, and decisions must be made swiftly to ensure system stability and performance. The integration of AI into CPU scheduling presents an opportunity to enhance the adaptability, efficiency, and overall performance of real-time systems, paving the way for more intelligent and responsive computing environments (Huang et al., 2019; Yang et al., 2020).
KEYWORDS: real-time CPU scheduling, machine learning (ML), supervised learning, intelligent solutions
I.
INTRODUCTION
In the rapidly evolving landscape of computing, real-time systems play a crucial role in various critical applications, ranging from industrial automation and telecommunications to autonomous vehicles and aerospace systems. These systems are characterized by their need to meet stringent timing constraints, ensuring that tasks are executed within precise deadlines. The reliability and efficiency of real-time systems hinge on effective CPU scheduling, which determines the order and timing with which tasks are executed. Traditional CPU scheduling algorithms, such as Rate
© 2024, IRJET
|
Impact Factor value: 8.315
|
Fig 1: Scheduling Types
ISO 9001:2008 Certified Journal
|
Page 785