A Content Store-Based Module for Congestion Control Algorithms of Named Data Networking

Abstract

NDN (Named Data Networking) is an increasingly important topic with the realm of the future Internet architecture research. Naturally, a new congestion control mechanism is important for a new Internet architecture. Existing research on NDN congestion control mainly considers the one-Interest-one-Data transport mode, internally, they control the Data sending rate by adjusting the Interest sending rate, thus to realize the congestion control. However, as one of the key features in NDN, in-network storage is rarely exploited. Furthermore, the problem of bursty traffic is also not fully considered and corresponding resolution is absent. Therefore, in this paper, we propose a novel congestion control module based on interacting with Content Store (CS-based) in NDN network to reduce packet loss caused by bursty traffic by taking advantage of the characteristic of in-network storage of NDN. CS-based is an independent module which can combine with existing NDN congestion control mechanisms (non-CS-based) to further reduce the packet loss rate caused by congestion especially by bursty traffic, increase throughput and improve the network resource utilization. Simulation results based on ndnSIM indicate that, compared with non-CS-based, CS-based has better performance under the circumstances that bursty traffic occurs and can mitigate NDN congestion better.