A QoS Oriented Approach for Sensing Services Selection in IoT Environment using Heterogeneous Similarity Metric with MCDA Methods

Selection of an appropriate sensing service amongst large number of competitive services is a significant issue in “sensing as a service” environment provided by Internet of Things (IoT). Selection of services is performed by indexing and ranking services according to Quality of Service (QoS) attributes. These attributes are heterogeneous in nature containing both qualitative and quantitative attributes. Most of the existing indexing and ranking approaches for IoT services consider only quantitative similarity metrics. Thus, it is a measure of concern to consider qualitative attributes as well. To address this challenge a Sensing Service Search Model called SSSM is designed. The model employs a pre-filtering approach and two indexing and ranking algorithms. The proposed algorithms integrate Heterogeneous Similarity Metric (HSM) with Multi Criteria Decision Analysis (MCDA) methods. QoS and Reputation Model called QoSRM is also proposed to describe quantitative as well as qualitative QoS attributes and help users to find reputed or trustworthy services. The effectiveness of the proposed approach is presented with suitable experiments and their results.