Consecutive Caching and Adaptive Retrieval for In-Network Big Data Sharing

Abstract

Information-centric networking (ICN) is a promising paradigm to support in-network big data sharing. However, it suffers from the problem of the segmented cached chunks resulting in low throughput and the large Interest packet overhead (IPO). The existing work cannot address these problems well, largely due to their insufficient considerations on the interplay between caching and data transport mechanisms. Through our experiments, we observe that ICN data chunks are cached in a distributed manner. Based on these observations, we propose consecutive data chunk caching (ConCaching) and adaptive data chunk retrieval (ACUR) to bridge the gap between caching and transport, where intermediate nodes on transmission path only cache the consecutive data chunks, while users can adjust the range of requested data chunks to maximize the throughput. Through the intensive simulations, we show that the proposed mechanisms can achieve better performance, in terms of higher throughput and substantial reduction in IPO compared with the existing pipeline mechanism in ICN.