Research And Improvement Of NDN Routing And Forwarding Strategy Based On Link Congestion Status

With the rapid development of Internet,it is feasible to connect everything to Internet.From social networking services to propagation of a lot of pictures,videos and documents,the ability of Internet that users required now is no longer basic connectivity.Today,the Internet accommodates a large number of contents which are produced and consumed by user.People need a huge content delivery network,a faster,more efficient and more secure content entrance,but no longer concern about the location of contents.This paper focuses on Named Data Networking(NDN).NDN shifts from today's focus on where(addresses and hosts)to what(the content that users and applications care about).NDN retrieves data from network through name,data is independent of location,applications,storage and various transmission methods.In addition,in-network cache is enabled in NDN.NDN project extends the intelligence from routing plane to forwarding plane.Due to the datagram state is maintained in the data plane,NDN forwarding plane is built-in adaptive.Forwarding layer is no longer simply read the routing table information,but has the ability of both reading and writing.The unique forwarding module that NDN has makes the forwarding decision through integration of both information from the routing plane and forwarding plane.Meanwhile,because of the built-in support of loop-free multipath forwarding,NDN can make better use of network resources and recover from network problems more efficient.The network layer is not responsible for congestion control in traditional IP network.While the adaptive forwarding capability brings a new network congestion control mechanism for NDN network.Through feedback messages and rate adjustment at routing node,NDN can achieve flow balance hop by hop.Network traffic is dispersed throughout the entire network first,if the traffic is still overload,the feedback message will go back to the consumers,then it is the consumers' responsibility to slow down the Interest sending rate.Through the use of adaptive forwarding capability,this paper improves the NDN network congestion control mechanisms by improving packet forwarding process.This paper introduces a predictive congestion control algorithm.The algorithm extends the network congestion feedback message types,add detection mechanisms of different level of network congestion,and proposes several rate adjustment schemas for different congestion level.This paper aims to solve the problems that current NDN network congestion mechanisms have.The congestion control now depends on inaccuracy data calculation and a single rate adjustment method.It is still possible to cause congestion problems and fairness issues of content retrieving.Finally,a ns-3 based NDN simulator,namely ndnSIM simulator was used to evaluate the performance of considered algorithms.We compare the predictive congestion control algorithm with the current NDN congestion control mechanisms through several indicators,such as sensitivity,link utilization and network throughput,in response to different topologies environment.PCC algorithm responses to network congestion faster,and have a timelier and more accurate feedback.PCC algorithm not only reduces the frequency of occurrence of network congestion,but also reduces the losses caused by network congestion.