Author: Dr. Sarla More

ISBN: 978-81-994205-6-4

DOI: https://doi.org/10.59646/492

Date of Publication: October 10, 2025

About the Book: 

In the rapidly evolving field of high-performance computing, efficient process scheduling in multiprocessor environments has become essential for achieving optimal utilization, reduced waiting times, and improved overall system performance. The book “Some New Methods for Ready Queue Processing Time Estimation in Multiprocessor Environment” presents an in-depth exploration of innovative statistical and algorithmic techniques designed to enhance the accuracy of ready queue processing time estimation. As traditional scheduling algorithms often struggle with unpredictable workloads, varying process priorities, and dynamic execution patterns, this work introduces new probabilistic and regression-based approaches that provide more reliable estimations and adaptive control in complex multiprocessor systems. The book begins with an overview of fundamental and modern CPU scheduling algorithms, including First Come First Serve, Shortest Job First, Round Robin, Priority Scheduling, Multilevel Feedback Queue, Real-Time Scheduling, Stride Scheduling, and cloud-based task scheduling, establishing the theoretical foundation for subsequent research. It then presents a series of original methods — starting with the Modified Group Lottery Scheduling (MGLS) algorithm, which refines mean time estimation through sampling and simulation under varied allocation conditions; followed by imputation-based ready queue estimation strategies that minimize mean squared error (MSE) and variance; and further extending into size-measure-informed models that improve predictive accuracy by incorporating process-specific information. The later chapters explore regression-type estimators under the risk of sudden system failures, integrating cost analysis and simulation for practical application in fault-tolerant computing environments, and culminate in a generalized framework capable of addressing multiple estimation scenarios with uncertainty and cost optimization. This unified body of work not only enhances theoretical understanding but also offers practical insights into the design of intelligent scheduling systems for cloud computing, distributed networks, and real-time embedded architectures. Intended for researchers, academicians, and postgraduate students in computer science and engineering, this book bridges the gap between operating system theory and statistical modeling, providing a robust foundation for advancing scheduling efficiency in multiprocessor computing. The authors express sincere gratitude to their mentors, peers, and the broader research community whose guidance and collaboration have shaped this contribution, and they hope this book will inspire continued exploration into intelligent, adaptive, and cost-aware scheduling methods for the next generation of computational systems.