AI 中文总结
本文针对单机器公平重复调度问题，解决了前人研究中遗留的三个复杂性问题，为该领域提供了新的理论进展。
AI 中文摘要
我们重新研究单机器下公平重复调度问题的求解。该问题包含n个客户和由q个时段（天）构成的规划周期，每天每个客户提交一个作业，需由机器处理。目标是构建q个调度方案（每天一个），使每个客户获得的服务质量（QoS）达到预定义阈值；QoS可采用客户作业在整个规划周期内的总等待时间、总完成时间等任意标准调度准则。该问题已有研究，前人针对多种QoS准则给出了复杂性分类和近似算法，但仍有若干重要问题未解决。本文解决了其中三个问题，并确定了未来研究的若干方向。
英文摘要
We revisit the problem of finding fair solutions to repetitive scheduling problems with a single machine. In this problem, we are given a set of $n$ clients and a planning horizon consisting of $q$ periods (days). Each day, every client submits a single job that must be processed by the machine. The objective is to construct a set of $q$ schedules, one for each day, such that the quality of service (QoS) received by each client meets a predefined threshold. The QoS measure may be any standard scheduling criterion, such as the total waiting time or total completion time of a client's jobs over the entire planning horizon. This problem has been studied in the literature, with previous works providing complexity classifications and approximation algorithms for various QoS measures. Nevertheless, several important questions remain open. In this paper, we resolve three of these questions and identify several additional directions for future research.
Comments12 pages
