Network Aware Composition for Internet of Thing Services

To enhance the adoption of Internet of Things (IoT) philosophy for the internet, research into IoT service composition has gathered momentum. In a distributed IoT environment, identifying IoT service among a set of similar service offerings that meets both functional and performance requirements of an IoT application has become important. However, the performance of a service cannot be guaranteed. Therefore service’s QoS and network characteristics are required to aggregate IoT services. Most existing composition approaches only consider non-network related QoS properties at the application tier. However they do not consider the network parameters such as network latency at the application level in selection and composition of services. Therefore we propose two evolutionary algorithms for IoT service composition that consider not only QoS but also network latency at the IoT application layer. The algorithms are discussed and results of evaluation are presented. The results indicate that our algorithms are efficient in finding QoS optimal and low latency solutions.