非强占有限优先权单服务台排队模型研究

Research on a non-preemptive limited-priority single-server queue model

根据大数据信息处理系统处理优先级的信息流存在的拥塞问题的研究,建立了一种新的基于爱尔朗分布的非强占有限优先权单服务台排队模型。该模型对有限优先权进行限制,增加不同优先级信息流处理排序的公平性,防止优先权队列信息流长时间使用服务台导致大数据信息处理系统无优先权信息流拥塞崩溃问题,增强系统的鲁棒性。通过对文中模型的分析研究,推导出信息流在系统内的平均等待时间、平均逗留时间和平均队长,并通过MATLAB2010a建立仿真模型实验。验证文中研究得到的系统排队等待时间等理论结果正确,模型稳定性更强。

According to the research on the congestion problem of the priority information flow processing in big data information processing system,a new single-server queuing model based on Erlang distribution based on non-preemptive limited priority is established.This model restricts the limited priority,and increases the fairness of the processing sequence of different priority information flow which prevents the priority queue information flow from using the service desk for a long time,and since it could cause congestion and collapse due to the non-priority information flow in the big data information processing system,and enhances the system Robustness.Through the analysis and research of the model in this paper,the average waiting time,average staying time and average queue length of the information flow in the system are derived,and a simulation model is established through MATLAB2010a for experimental verification.It is verified that the theoretical results such as the waiting time of the system studied in this paper are correct,and the model is more stable and stronger.