含噪声混合云大数据流量SDN控制方法仿真

Simulation of SDN Control Method for Large Data Flow in Noisy Hybrid Cloud

互联网的快速发展,使连入的网络负载增多,导致数据传输过程中吞吐量较低。为了更好的对大数据流量开展实时控制,提出基于SDN技术的混合云大数据流量实时控制方法。利用ICEEMD算法对数据流量中的噪声实施分离去除,在得到较为准确的数据流量后,在尽量缩小SDN所需时间的同时学习并优化流量控制器的部署条件。在保证时延性能的同时兼顾链路的可靠性,部署流量控制器,达到流量实时控制的目的。经实验验证,所提方法对于流量的控制较为稳定、数据吞吐量多且丢失率低于0.02%,表明所提方法能够有效控制大数据流量。

With the rapid development of the Internet,the network load increases,leading to low throughput during data transmission.In order to control big data traffic in real-time,this paper presented a real-time control method of hybrid cloud big data traffic based on SDN technology.At first,the ICEEMD algorithm was adopted to separate and remove the noise from the data flow.After obtaining more accurate data flow,the time required for SDN was reduced as much as possible while the deployment conditions of the flow controller were learned and optimized.Moreover,the traffic controller was deployed while ensuring the delay performance and considering the reliability of the link.Finally,the real-time traffic control was achieved.Experimental results prove that the proposed method is more stable in traffic control and has more data throughput.And the loss rate is less than 0.02%,indicating that the method can effectively control big data traffic.