A Review on AI-Enabled Congestion Control Schemes for Content Centric Networks

Abstract

Content centric networks (CCN) offer more advantages over conventional TCP/IP networks in areas like content distribution. However, congestion control functionalities in CCN present challenges such as detecting congestion, over-reducing windows for non-congested paths, and addressing fairness issues. Most existing studies employed congestion control mechanisms similar to those in TCP. In addition, the existing mechanisms were based on conventional optimization rules to adjust the rate at which Interest packets are sent to request data from downstream nodes. However, such existing mechanisms do not consider the changes in network status and caching strategy due to multi-path and multi-source transmission. Moreover, they are based on assumptions about link bandwidth. In this paper, we study the problem of congestion control in CCN and discuss its challenges. In addition, we review the existing congestion control schemes in CCN based on machine learning. Finally, we highlight the open research issues to spur further investigations. © 2023 IEEE.