On-line popularity monitoring method based on bloom filters and hash tables for differentiated traffic

Abstract

Towards line speed and accurateness on-line content popularity monitoring on Content Centric Networking (CCN) routers, we propose a three-stage scheme based on Bloom filters and hash tables for differentiated traffic. At the first stage, we decide whether to deliver the content to the next stage depending on traffic types. The second stage consisting of Standard Bloom filters (SBF) and Counting Bloom filters (CBF) identifies the popular content. Meanwhile, a scalable sliding time window based monitoring scheme for different traffic types is proposed to implement frequent and real-time updates by the change of popularities. Hash tables according with sliding window are used to record the popularity at the third stage. Simulation results reveal that this method reaches a 40Gbps processing speed at lower error probability with less memory, and it is more sensitive to the change of popularity. Additionally, the architecture which can be implemented in CCN router is flexible and scalable.