NoD: A content retrieval scheme for intermittently-connected IoT networks

Abstract

In recent years, Internet of Things (IoT) devices have been deployed even to the most isolated and hard-to-reach areas. Content retrieval from such communication-disruptive remote areas therefore becomes a challenging task, mainly due to the increased infrastructure cost. In this paper, we present a gateway-oriented scheme which facilitates content retrieval from remote IoT deployments, using mobile devices. We call our scheme NDN-over-DTN (NoD), as it combines the information-centric design of the Named Data Networking (NDN) architecture with the store-carry-and-forward approach of Delay/Disruption-Tolerant Networking (DTN). In this work, we evaluate the performance of NoD in a real IoT-content-retrieval scenario, in terms of Average Delay, Cache Hit, Interest Satisfaction and Overhead Ratio. We model the behavior of NoD analytically and introduce an improved NoD version which utilizes cross-layer interaction between NDN and DTN, called NoD-Discovery. We show, both analytically and experimentally, that NoD can exploit jointly the benefits of NDN and DTN and achieve significant performance gains in highly disruptive environments. Notably, our experiments demonstrated that NoD-Default can achieve up to a 48.77% decrease in Average Delay compared to host-centric DTN, while NoD-Discovery can improve performance further by decreasing the Overhead Ratio up to 70.34%, compared to the original NoD-Default implementation.