With the advent of the information age,the network scale and application types have been expanding rapidly.The defects and shortcomings of the current TCP/IP net-work architecture make the network face great challenges in terms of security,scala-bility and mobility,etc.While relying on the incremental improvement solutions can not thoroughly solve these problems caused by the irrationality of the Internet's own architecture design.Named Data Networking(NDN)as the most competitive one in the new generation of revolutionary network architecture,its development has been widely concerned by researchers at home and abroad.Network congestion as a crucial issue in network transmission,appropriate congestion control strategies are vital to network transmission performance and quality of service.Due to the large difference between in network architecture and numerous design concepts,the existing multiple congestion control strategies for TCP/IP network can not be directly migrated to NDN.Therefore,the research of congestion control in NDN has significant value.The main works of this dissertation are as follows.Firstly,based on the existing research achievements of congestion problems,a Re-ceiver Interest Flow Control based on Variable-Structure Congestion Control Proto-col(RIFVCP)is proposed.The intermediate nodes actively detect the current congestion state of the interfaces by estimating the load status,and send the feedback information to the receivers timely through the returned data packets.The receivers adjust the send-ing rate of the interest packet according to the feedback information,thereby controlling the transmission rate of data packet in network.In addition,combining with the inter-mediate node-based congestion control algorithm,a joint congestion control strategy is formed.The transmission rate of interest packet is jointly adjusted from the interme-diate node and the content request node to improve the network bandwidth utilization.For the data flow of the same service type,the purpose of equalizing the link resources of multiple data flows has achieved by balancing the sending rate of the receiver.Ex-perimental results have shown that RIFVCP can improve bandwidth utilization,as well as can achieve fairness in resource allocation among multiple data flows.Besides,the majority of congestion control researchs have not considered the dif-ferent service attributes of the data flows.When the data flows of multiple service types coexist in the network,network only provides best-effort services for all data flows.In the current application trend of sharp increasing multimedia services,if the priority of the data flow is not distinguished,the transmission performance and the network qual-ity of service will greatly reduced.In this regard,an Interest-based Dynamic Weighted Round Robin(IDWRR)algorithm has been proposed in this dissertation.When the in-termediate nodes receive the interest packet,the interest packet is entered into different sub-queues by classifying the service type.In combination with the congestion control algorithm based on interest packet rate shaping,the allocation of weights is dynami-cally distributed by periodically detecting the queue length of each sub-queue,thereby changing the bandwidth occupancy and transmission delay of these data flows.The ex-perimental results have shown that the proposed algorithm can achieve better bandwidth utilization and allocate bandwidth to different service flows,which improves the quality of service and transmission performance of the network. |