Chapter 5: CPU Scheduling: Algorithms, Multi-Processor, and Real-Time Scheduling

CPU Scheduling: Algorithms, Multi-Processor, and Real-Time Scheduling begins by describing the objectives of scheduling—maximizing CPU utilization, throughput, and fairness while minimizing turnaround time, waiting time, and response time—and explains how scheduling criteria influence performance trade-offs. The chapter introduces the different levels of scheduling: long-term scheduling (job scheduling), medium-term scheduling (process suspension and resumption), and short-term scheduling (CPU scheduling). It then explores core CPU scheduling algorithms, including First-Come, First-Served (FCFS), Shortest Job First (SJF) with preemptive and non-preemptive variations, Priority Scheduling, and Round Robin (RR), as well as multilevel queue and multilevel feedback queue scheduling for complex environments. Concepts such as preemption, dispatcher latency, and context-switch time are discussed in depth. The chapter covers evaluation methods for scheduling algorithms, including deterministic modeling, queuing models, simulation, and implementation-based testing. Multiprocessor scheduling is also addressed, with discussions of load balancing, processor affinity, and real-time scheduling requirements for hard and soft deadlines. The chapter concludes with case studies on CPU scheduling in Windows, Linux, and real-time operating systems, illustrating how theoretical models are applied in real-world OS designs to achieve efficient and predictable process execution.