AFSndn: A novel adaptive forwarding strategy in named data networking based on Q-learning

Named Data Networking (NDN) is a new network architecture, which employs a new content-centric communication model to replace the traditional host-centric communication model. In TCP/IP network, data packets are forwarded by routers according to routing table established previously. While in NDN, routing nodes can dynamically make forwarding decisions based on network status. By considering this forwarding feature, we proposed a novel adaptive forwarding strategy in Named Data Networking (AFSndn) based on Q-learning to minimize the delivery time. AFSndn is divided into two phases—Exploration phase and Exploitation phase. The Exploration phase aims to collect information, while the Exploitation phase aims to dynamically forward interest packets. Simulation experiment results show that AFSndn has better performance compared to others famous algorithms.