Traffic Flow Prediction Model In Data Center Networks By Deep Learning Approach

Real-time and high-priority applications have grown enormously during the last few years.Such applications require guaranteed bandwidth in packet-switched networks,minimum network delay,and reliable delivery of packets needed in the service provider network.Multiprotocol Label Switching(MPLS)provides these requirements without compromising scalability.MPLS used labels to speed up the traffic flow in which switching of data packets can be done based on labels rather than routed based on destination address as in Internet Protocol(IP)network.It saves the time needed for a router to look up the next node's address to forward the packet.The Internet Protocol is a selection of open protocols.Since it is not operated by any one institution,it can be used by anybody or any entity.While MPLS does not have encryption,it is a virtual private network that is isolated from the rest of the Internet.MPLS is thus regarded as a safe mode of transport.The aim of this study is to demonstrate that MPLS networks are superior to IP networks.This system deals with the performance of the MPLS-based network is compared with the IP-based network.There are fifteen routers to construct the MPLS and IP network.In this system,to monitor the packet performance of MPLS network and IP network.Network Simulator(NS2)is used to simulate different performances based on the number of packets dropped.The number of packets received,packet delay,packet jitter,packet throughput,packet delivery ratio,end to end delay MPLS,and IP network.