A software-defined caching scheme for the Internet of Things

Abstract

Exploiting Content-Centric Networking (CCN) caching capabilities in which contents are cached on intermediate nodes can be beneficial in IoT as it can decrease the latency, reduce required transmission hops, limit traffic load on the content producer and improves availability. This article presents a new scheme for caching the contents in IoT environments. In the proposed method, devices are grouped into clusters where the cluster heads act as the cache controller. Additionally, we consider a global SDN/Cache controller (GSCC), which is responsible for orchestrating cache decisions in the whole IoT network. Such a centrally managed caching system increases the efficiency of resource usage in the IoT network. In the proposed scheme, the decision about caching the content is made in three steps: 1) determining the value of content and make decisions about caching it, 2) determining the candidate cluster, and 3) selecting candidate nodes for caching the content. In each step, several metrics are taken into account, and Multi-Criteria Decision Making (MCDM) approaches such as Analytical Hierarchy Process (AHP) and TOPSIS are used to select the best option based on considered criteria. Simulation results show that our proposed caching method can achieve an average cache hit rate of 72% and decreases the average hop count of content retrievals by 42%. Moreover, our results confirm the superiority of our algorithm over some existing methods in terms of different evaluation metrics.