Research On Congestion Control Algorithm For Future Internet

With the rapid development of the Internet, there are increasing interests in exploiting new business requirements. And the increasing data and users have been beyond the capability of tradition Internet. So many new problems of network are revealed in the way of evolution. Now many countries have engaged in the research on Future Internet architecture in the world, which facilitated the generation of Information-centric networks (ICN).Information-centric networks is one of the most important future internet architecture, which uses the information-centric network model instead of the tradition address-centric network communication model. ICN is more focusing on the information acquisition, and is the host-to-network communications. The two most significant features of CCN, routing-by-name and universal caching, allow for much better network transmission efficiency. However, since ICN is the name based routing and has the features of connectionless transmission and source uncontrollability, network congestion is inevitable. The congestion control algorithm is important for Information-centric networks to ensure the efficiency and stability of the network.In order to improve the congestion control in ICN, we present the network codes-based Information-Centric Networking transmission control protocol. The main works in this paper are as follows:1. There is a multi-source problem when information was transmitted in ICN, which results the redundancy of network data and hampers transmission speed. With the analysis of the network coding, we combine the network coding and hamper transmission speed together. This method uses network coding to hide the difference of multiple sources, significantly speed up the information retrieval and improve the controllability of a network-oriented large-scale data in ICN.2. Congestion control algorithm plays a key role in ensuring the transmission performance in ICN. The key question on congestion control in Information-Centric Networking is how to judge the congestion level of networks and how to adjust the window size according to the congestion level of the network. After investigate and analysis the existing congestion control methods in ICN, we present the congestion control indicator based on the Short-term throughput and the variation of RTT. And the congestion window adjusting strategies concerning Short-term throughput is more adaptable for ICN and improves the network throughput.