【摘要】 针对云视频会议系统中MCU(Multi-point Control Unit,多点控制单元)负载动态变化的特性,将流量对MCU负载产生的影响考虑在内,设计并使用流量自适应因子,提出了一种综合考虑了影响MCU负载的内在和外在因素的流量自适应算法。该算法将影响MCU负载的多个参数等价转化为负载权值,同时优先将同一个视频会议下的终端分配在一个相同的MCU上,以使得具有较低MCU级联延时。通过性能测试,将该算法用于对云环境下MCU进行调度,可以使得各MCU负载更加均衡,音视频的转发时延更短。更多还原

【Abstract】 Based on the characteristics of the MCU’s(Multi-point Control Unit)dynamic change of load in cloud video conference system and considering the impact of traffic on MCU load,this paper proposes a traffic adaptive algorithm by designing and using the traffic adaptive factor.This algorithm takes the intrinsic and external factors that affect the MCU load into account.Specifically,the multiple parameters which affect the MCU load are equivalently transformed into the load weight,and the terminals which belongs to the same video conference are assigned to the same MCU in priority,so that a lower MCU latency can be ensured.Through the performance test,this algorithm can be used to schedule the MCU in the cloud environment,leading to a more balance MCU load and a shorter video transmission delay.更多还原