An Almost-zero Latency Lightweight Mechanism for Caching Decision in ICN Content Router
Qifeng Yang, Haojiang Deng, Lingfang Wang, Yiqiang Sheng · 2019
Research has shown that in-network caching for named data in ICN is an effective way to eliminate redundant network traffic. The caching strategy is the key to improve the efficiency of content distribution in the ICN. However, current complicated caching strategies have high latency and couple with network forwarding, which results in terrible forwarding performance and poor scalability of throughput. In this paper, we propose a novel mechanism for making caching decisions in the source. In our mechanism, to improve the forwarding speed, the decision result is encoded in the cache control field of the content packet so the content packet can be quickly filtered and forwarded by on-path routers with an almost-zero latency. In order to improve the accuracy of caching decision in the mechanism, we further present a lightweight caching strategy based on popularity ranking. In the experiment, we set up a POF-based ICN soft router with a split architecture and implement our caching decision mechanism inside. The result of the experiment shows that on-path decision latency can be reduced by at most 96.7 percentage in our mechanism and a scalable line-speed of at least 10 Gbps is guaranteed. Our mechanism is also lightweight which achieves similar performance with less than 1/20 memory consumption of other popularity-based strategies.