MQSD: A Scheduling Mechanism of Mixed Queue in the Integrated Intelligent Network
Jiangshan Zheng, Li Yang, Chengsheng Pan, Huaifeng Shi · Journal of Physics Conference Series · 2021
Abstract In the integrative intelligent network, the burst and heterogeneous characteristics of the aggregation node traffic lead to increased queuing delay and frame loss rate of the business flow. It is difficult to meet the quality of service (QoS) requirements of the differentiated business flow. In response to the above problems, a mixed queue scheduling based on demand service (MQSD) is proposed for the integrated intelligent network of heaven and earth. Different scheduling algorithms are used to schedule different business flows, which distinguishes the different needs of the business. The weight of business flow is calculated using the construction judgment matrix method, and is determined to be acceptable through a consistency test. By comparing the available service volume with the demanded service volume, the available service volume to the business flow will be dynamically adjusted. Establish a simulation experiment platform to schedule different types of randomly generated data streams. Simulation results show that the MQSD mechanism reduces queuing delay and frame loss rate caused by burstiness while meeting QoS requirements.