Modelling the Pending Interest Table occupancy in CCN with interest timeout and retransmission

Recent studies have shown that the Pending Interest Table (PIT) in most Information-Centric Networking proposals can represent a bottleneck for the network performance. Newly arrived interest packets at the PIT are dropped if the PIT becomes full. In-network caching may fail to reduce the traffic load on the PIT if the cache dynamics is high as memory in routers is limited. How to manage the occupancy of the PIT in the presence of high traffic load remains an interesting but challenging problem. To this end, this paper presents an analytical model to estimate the PIT occupancy distribution via a continuous-time Markov chain model. In the model, we take into account caching, interest aggregation, interest timeouts and retransmission.We validate the model via simulation experiments to show the accuracy of our model. Our proposed model is useful in the design of traffic controllers for managing the PIT occupancy.