{"title":"质量相关随机网络:先进先出总是比后进先出好吗?","authors":"Y. Liberman, U. Yechiali","doi":"10.1145/3388831.3388838","DOIUrl":null,"url":null,"abstract":"In various service and production systems, such as metal processing, fresh food industry, polymer forming, etc., the quality of the final product is a function of its total sojourn time in the system. We consider (i) an M/G/1-type single-server queue, as well as (ii) open tandem Jackson networks, and analyze the quality of products traversing through each system under two service disciplines -- FIFO and LIFO. Although the mean sojourn times in an M/G/1 queue are equal under both service disciplines (that is, E [WLIFO] = E [WFIFO]), the variance of WLIFO is larger than the variance of WFIFO. We show that, when quality is the measure of performance, FIFO is not necessarily better than LIFO. We consider several service-time distributions and show under what values of the parameters one discipline is better (or worse) than the other. In particular, the mean quality under LIFO is better than the mean quality under FIFO for all values of the traffic intensity. Moreover, the mean quality under the FIFO service regime drops sharply to 0 when the traffic intensity approaches 1, while the mean quality under LIFO is bounded below. Numerical results as functions of the system parameters are presented and discussed for both the M/G/1 queue and the tandem Jackson network.","PeriodicalId":419829,"journal":{"name":"Proceedings of the 13th EAI International Conference on Performance Evaluation Methodologies and Tools","volume":"48 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-05-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Quality-Dependent Stochastic Networks: Is FIFO Always Better Than LIFO?\",\"authors\":\"Y. Liberman, U. Yechiali\",\"doi\":\"10.1145/3388831.3388838\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"In various service and production systems, such as metal processing, fresh food industry, polymer forming, etc., the quality of the final product is a function of its total sojourn time in the system. We consider (i) an M/G/1-type single-server queue, as well as (ii) open tandem Jackson networks, and analyze the quality of products traversing through each system under two service disciplines -- FIFO and LIFO. Although the mean sojourn times in an M/G/1 queue are equal under both service disciplines (that is, E [WLIFO] = E [WFIFO]), the variance of WLIFO is larger than the variance of WFIFO. We show that, when quality is the measure of performance, FIFO is not necessarily better than LIFO. We consider several service-time distributions and show under what values of the parameters one discipline is better (or worse) than the other. In particular, the mean quality under LIFO is better than the mean quality under FIFO for all values of the traffic intensity. Moreover, the mean quality under the FIFO service regime drops sharply to 0 when the traffic intensity approaches 1, while the mean quality under LIFO is bounded below. Numerical results as functions of the system parameters are presented and discussed for both the M/G/1 queue and the tandem Jackson network.\",\"PeriodicalId\":419829,\"journal\":{\"name\":\"Proceedings of the 13th EAI International Conference on Performance Evaluation Methodologies and Tools\",\"volume\":\"48 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-05-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 13th EAI International Conference on Performance Evaluation Methodologies and Tools\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3388831.3388838\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 13th EAI International Conference on Performance Evaluation Methodologies and Tools","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3388831.3388838","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
在各种服务和生产系统中,如金属加工、生鲜食品工业、聚合物成型等,最终产品的质量是其在系统中总停留时间的函数。我们考虑了(i) M/G/1型单服务器队列,以及(ii)开放式串联Jackson网络,并分析了在两种服务原则(FIFO和LIFO)下穿越每个系统的产品质量。虽然M/G/1队列在两种服务纪律下的平均停留时间相等(即E [WLIFO] = E [WFIFO]),但WLIFO的方差大于WFIFO的方差。我们表明,当质量是性能的衡量标准时,先进先出不一定比后进先出更好。我们考虑几种服务时间分布,并显示在参数的什么值下,一个规程比另一个规程更好(或更差)。特别是,在所有交通强度值下,后进先出的平均质量都优于先进先出的平均质量。此外,当交通强度接近1时,FIFO服务模式下的平均质量急剧下降到0,而LIFO服务模式下的平均质量有界于以下。给出并讨论了M/G/1队列和串列Jackson网络的系统参数函数的数值结果。
Quality-Dependent Stochastic Networks: Is FIFO Always Better Than LIFO?
In various service and production systems, such as metal processing, fresh food industry, polymer forming, etc., the quality of the final product is a function of its total sojourn time in the system. We consider (i) an M/G/1-type single-server queue, as well as (ii) open tandem Jackson networks, and analyze the quality of products traversing through each system under two service disciplines -- FIFO and LIFO. Although the mean sojourn times in an M/G/1 queue are equal under both service disciplines (that is, E [WLIFO] = E [WFIFO]), the variance of WLIFO is larger than the variance of WFIFO. We show that, when quality is the measure of performance, FIFO is not necessarily better than LIFO. We consider several service-time distributions and show under what values of the parameters one discipline is better (or worse) than the other. In particular, the mean quality under LIFO is better than the mean quality under FIFO for all values of the traffic intensity. Moreover, the mean quality under the FIFO service regime drops sharply to 0 when the traffic intensity approaches 1, while the mean quality under LIFO is bounded below. Numerical results as functions of the system parameters are presented and discussed for both the M/G/1 queue and the tandem Jackson network.