Effective NDN congestion control based on queue size feedback
Sichen Song, Lixia Zhang · 2022
Named data networking (NDN) can improve the consumer data retrieval throughput with its built-in multicast data delivery, innetwork caching, and ability to support multi-path forwarding. However, their realization brings challenges. In this work, we first examine how multi-path forwarding and in-network caching can interfere with consumer measurements for congestion control. Based on the results, we propose a congestion control solution, NDN-QSF, that can work effectively in the presence of in-network caching. In NDN-QSF, forwarders estimate upstream bandwidth and use queue size as congestion feedback to inform downstream routers to limit interest transmission rates. We further adapt and extend NDN-QSF to enable routers to make informed multi-path forwarding decisions. We evaluated NDN-QSF through simulation experimentation and our results show that NDN-QSF can effectively control congestion by using queue size as congestion feedback and improve network throughput with multi-path forwarding.