Charging on the Route: An Online Pricing Gateway Congestion Control for ICNs

The Information-Centric Networking (ICN) paradigm has emerged to shift the current host-based network model to a content-oriented one in order to cope with the dominant content-based services in the Internet. Congestion control is a fundamental design concern to support massive content delivery in ICN. While the existing flow-based and hop-by-hop congestion control mechanisms suffer from complexity and compatibility issues, we propose RevMax, a gateway-aware congestion control mechanism based on the architecture of NDN, a well-known ICN platform, to overcome the drawbacks. In the proposed mechanism, the gateway offers a price to each end-user, which urges the user to adjust the request rate according to the price. The optimal pricing policy for the gateway is shown to be formulated as a revenue maximization problem, and an efficient algorithm is proposed to derive the optimal solution. The proposed RevMax mechanism is implemented in NDN and compared to PCON, a state-of-the-art congestion control mechanism for ICNs. Extensive experiments show that RevMax achieves higher throughput, lower network latency, and better fairness in a variety of network scenarios.