Due-Date Setting and Priority Sequencing in a Multiclass M/G/1 Queue
研究多类M/G/1排队系统中同时设定截止日期和优先级排序的问题，目标是在满足拖期比例或平均拖期时间约束下最小化加权平均截止日期提前期，提出多种参数和非参数策略，仿真表明其优于传统方法。
The problem of simultaneous due-date setting and priority sequencing is analyzed in the setting of a multiclass M/G/1 queueing system. The objective is to minimize the weighted average due-date lead time (due-date minus arrival date) of jobs subject to a constraint on either the fraction of tardy jobs or the average job tardiness. Several parametric and nonparametric due-date setting policies are proposed that depend on the class of arriving job, the state of the queueing system at the time of the job's arrival, and the sequencing policy (the weighted shortest expected processing time rule) that is used. In a simulation experiment performed on a two-class M/M/1 system, these policies outperformed traditional due-date setting policies, and due-date setting had a larger impact on performance than priority sequencing.