Scholarlink

<p>Many tasks compete with one another in most modern systems for shared resources such as processors, memory, network, sensors, or other devices. When several jobs are sharing a resource, they must be coordinated and the access times for each job must be specified. Scheduling is a popular synchronization technique. Context switching, as the term implies, is the process of moving resource assignments across jobs to carry out a schedule. However, during scheduling and context switching overheads are incurred. It follows that process management is one of the important tasks performed by the operating system. The performance of the system depends on the process scheduling algorithms. The main aim of the CPU scheduling algorithms is to minimize waiting time, turnaround time, response time and context switching and maximizing CPU utilization along with optimizing or enhancing other performance metrics. This work presents a modified version of the EDZL (Earliest Deadline until Zero Laxity), a scheduling algorithm for real-time multiprocessor systems, with a reduced number of context switches to improve the performance of the algorithm. The proposed scheduler, PP-EDZL (Partially Pre-emptive Earliest Deadline until Zero Laxity with Laxity Based Reordering) uses two methods to achieve its goals. Experimental results demonstrate a 30–50% reduction in context switches and a 20–35% improvement in deadline adherence compared to traditional EDZL, while maintaining real-time guarantees.&#xa0;The algorithm also reduces average waiting and turnaround times, making it suitable for hard real-time systems where overhead minimization is critical.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

PP-LR EDZL: A Novel Laxity-Based Reordering and Partially Pre-emptive Approach to Enhancing the Earliest Deadline Until Zero Laxity Scheduler

  • 期刊论文

  • 发表日期：

    2026年6月18日

查看全文
  • Naveed Suhail Madni,
  • Sheikh Amir Fayaz,
  • Razeef Mohd,
  • Majid Zaman

摘要

Many tasks compete with one another in most modern systems for shared resources such as processors, memory, network, sensors, or other devices. When several jobs are sharing a resource, they must be coordinated and the access times for each job must be specified. Scheduling is a popular synchronization technique. Context switching, as the term implies, is the process of moving resource assignments across jobs to carry out a schedule. However, during scheduling and context switching overheads are incurred. It follows that process management is one of the important tasks performed by the operating system. The performance of the system depends on the process scheduling algorithms. The main aim of the CPU scheduling algorithms is to minimize waiting time, turnaround time, response time and context switching and maximizing CPU utilization along with optimizing or enhancing other performance metrics. This work presents a modified version of the EDZL (Earliest Deadline until Zero Laxity), a scheduling algorithm for real-time multiprocessor systems, with a reduced number of context switches to improve the performance of the algorithm. The proposed scheduler, PP-EDZL (Partially Pre-emptive Earliest Deadline until Zero Laxity with Laxity Based Reordering) uses two methods to achieve its goals. Experimental results demonstrate a 30–50% reduction in context switches and a 20–35% improvement in deadline adherence compared to traditional EDZL, while maintaining real-time guarantees. The algorithm also reduces average waiting and turnaround times, making it suitable for hard real-time systems where overhead minimization is critical.