Information-Centric Networking (ICN) is a future network architecture designed to solve problems of current end-to-end communication network. The key aspect of ICN is that in-network nodes can cache the contents, and hence the caching scheme is important for the efficient use of cache and content delivery. This means that research on caching strategy is the key to improve network performance for ICNs. Still, how to better utilize in-network caching of ICNs with low costs remains a great challenge.

Aimed to tackle this challenge, recently, ZHU Xiaodong and his colleagues from the Institute of Acoustics of the Chinese Academy of Sciences proposed a popularity-based neighborhood collaborative caching algorithm for ICNs, which will give support for the management of in-network caching.

Researchers used in-network nodes to track the popularity of contents, and en-route and one-hop neighborhood nodes to make caching decision collaboratively. The popularity estimation and comparison process were also optimized as the requirement of quick process for in-network nodes.

Contents are hierarchically located at in-network cache nodes, the more popular, the closer to clients, thus reducing the redundancy of contents, and then clients can get what they need faster(figure 1). In addition, the computational overhead of popularity comparison is significantly decreased as long as maintain the minimum popularity index.

Figure 1. Information-centric Network. (Image by ZHU Xiaodong)

On-path and off-path nodes are efficiently combined to collaboratively cache contents, with ICN’s quality of service improved.

In the simulation experiments, real-world topologies and different client placed scenes were used to test this new scheme, and it performed better in terms of latency, cache hit ratio and path stretch compared with currently known algorithms and ideal situations. Meanwhile, the overhead of the algorithm both on estimation of popularity and node interaction was low budget.

In sum, the proposed algorithm provides a practical choice for ICN caching decision strategy with good performance and acceptable overhead. It also sheds light on cache management of network in the future.

The paper entitled “Popularity-based neighborhood collaborative caching for information-centric networks” was published on 36th IEEE International Performance Computing and Communications Conference.

Reference:

ZHU Xiaodong, WANG Jinlin, WANG Lingfang, QI Weining. Popularity-Based Neighborhood Collaborative Caching for Information-Centric Networks. 36th IEEE International Performance Computing and Communications Conference, 2017. DOI: 10.1109/PCCC.2017.8280453

Contact:

WANG Rongquan 
Institute of Acoustics, Chinese Academy of Sciences, 100190 Beijing, China
E-mail: wangrongquan@mail.ioa.ac.cn