Fog-based caching mechanism for IoT data in information centric network using prioritization

Abstract

In-Network caching is one of the most prominent features of Information-Centric Networks (ICN). This feature, designed to reduce the content delivery time, can improve data availability, especially during the sleeping mode of the original data producer. However, the fact that caching mechanism is simultaneously performed with data forwarding, limited processing power, and limited memory capacity of routers has posed challenges in fully profiting from this feature. The synchronization of caching mechanism with data forwarding limits the speed of decision-making in content placement policies. These challenges as well as the limitations of ICN routers have not been taken into consideration by the existing strategies for content placement of IoT data. In this paper, the restrictions for content placement policies in ICN-based IoT are outlined, and a data prioritization-based approach that prevents ICN nodes from filling their Content Store (CS) with inappropriate and inferior data is proposed to improve cache efficiency. Besides, fog computing is used as a middle layer between IoT devices and ICN networks to resolve the speed limits of decision-making and improve cache performance. In the fog layer, prioritization is performed based on the popularity and freshness of data object, and, thus, low-priority data are removed from the caching mechanism of ICN nodes. As the result shows the total number of cached data is reduced. Freshness is an essential property of multiple IoT data that significantly affects caching performance. That's why we prioritized data based on popularity and freshness. Eventually, the simulation results performed by the NDNsim simulator show that decreasing the number of data objects could improve the cache mechanism efficiency in terms of both delay and hit ratio.