CyHy: A Centrality-aware Hybrid Caching Strategy in Information-centric Networking
Ashim Dey, Asaduzzaman Asaduzzaman, Mohammad Shamsul Arefin · 2024
Network functionality has to be moved from the central server to the nearby user due to the recent explosion in multimedia applications' volume, worldwide accessibility, and rapid development. Inversely, typical Internet communications are inefficient as their main purpose is the transfer of data between hosts. As a result, a content-oriented networking paradigm known as information-centric networking (ICN) has surfaced. Researchers are trying to resolve implementation issues as ICN continues to develop. Allocating the space among cache-enabled nodes to maximize their benefits is a significantly difficult task because ICN has limited caching capability. In this work, a centrality-aware hybrid caching (CyHy) scheme has been proposed where cache storage distribution and content caching selection are made based on node relevance rather than uniformly dividing space among nodes. As node centrality metrics have a significant impact on caching decisions, we considered degree centrality which provides a local view, and betweenness centrality which provides a global view to find the important nodes in the network. Certain adjustments are suggested to the request packet and data packet structures to accommodate the proposed caching strategy. Finally, two real-world internet topologies are used for extensive modeling, and the results are compared with other standard caching techniques. The results of the simulation demonstrate that suggested caching outperforms others in terms of caching hit ratio, latency, and link load.