On pending interest table in named data networking

Internet has witnessed its paramount function transition from host-to-host communication to content dissemination. Named Data Net-working (NDN) and Content-Centric Networking (CCN) embrace this shift by identifying each piece of content by a name, and NDN/CCN communications are based on these names, rather than IP addresses. Pending Interest Table (PIT) in NDN/CCN keeps track of the In-terest packets that are received but yet un-responded, which PIT brings NDN/CCN significant features, such as communicating with-out the knowledge of source or destination, loop and packet loss detection, multipath routing, better security, etc. All these features enable NDN/CCN to be an information/content-centric network. Every Interest packet needs to be inserted into PIT and each Data packet removed from PIT. Intuitively, due to high-speed packet ar-rival rate, the PIT may have a huge size and demand extremely high lookup, insert and delete frequencies. Though researches on NDN are currently in full swing, this issue of exact PIT size and access frequency still remains untouched. This paper presents a thorough study of PIT for the first time. We first conduct real trace-driven measurements on PIT by mapping a backbone router trace from IP to NDN scenario at the perspective of applications. Evaluation results on a 20 Gbps gateway link show that PIT contains 1.5 M entries, and the lookup, insert and delete frequencies are 1.4 M/s, 0.9 M/s and 0.9 M/s, respectively. Facing with this challenging issue and to make PIT more scalable, we fur-ther propose an Name Component Encoding (NCE) solution, which assigns codes (integers) to the components of names in the PIT, to shrink PIT size and accelerate PIT access operations. By NCE, the memory consumption can be reduced by up to 87.44%, and the ac-cess performance significantly advanced, satisfying the access speed required by PIT. Moreover, PIT exhibits good scalability on large name sets. At last, We propose that we place PIT on the outgoing line-cards of routers, which meets the NDN design and eliminates the cumbersome synchronization problem among multiple PITs on the line-cards. 1