SDRank: An Adaptable Service Selection for IoT Based on Ranking

Abstract

The Internet of Things (IoT) is a computing paradigm merging the physical object to the internet and enabling an interaction through the existing network protocols. The development of networking technology and the device's computing capacity drives the number of connected objects that increase rapidly. While this gives many benefits, it also brings challenges. One of them is the service selection capability. Issues arises in this context are related to the availability of service information and the dynamic nature of the devices. There are many of them sharing the same functionality but in fact, it's attributes changes over time. Thus, a service selection based on similarity and provide a handling of device changes is required. This research proposed the IoT service model and IOT-WSDL as a description model to support IoT-specific characteristic. Also, we introduced the SDRank, an implementation of adaptable service selection for IoT. It utilizes ranking to provide a selection method and to enable the adaptation of service changes. The models are useful in describing and sorting services to generate ranking. We compared two ranking methods, namely Service Rating and Analytical Hierarchy Process (AHP). The result shows that AHP provide a more relevant and consistent ranking than Service Rating without significant performance degradation.