Adaptation and Evaluation of Widely Used TCP Flavours in CCN
Asanga Udugama, Jinglu Cai, Carmelita Göerg · Mobile Networks and Management · 2014
Content Centric Networking (CCN) is a paradigm shift from the way how networks of today work. The focus of networking in CCN is on the content and not on the hosts that are involved in a communication. One of the key cornerstones of today’s communication model is the use of flow and congestion control to pipeline data and take appropriate action when congestion is perceived to exist in a network. TCP of the Internet protocol suite has shown us how application performance is enhanced in different communication situations. An interesting area of research is how TCP-like flow and congestion control can be adapted for CCN. The work presented here adapts the most widely used TCP flavours of NewReno, Compound and Cubic to operate in CCN. Due to the architectural differences that CCN has over IP based networks, this work identifies a number of additional algorithms to cater to the issues associated with these differences. Finally, the performance of these adapted TCP flavours and the algorithms are evaluated in an OPNET based simulator.