| Abstract |
The Internet of Things (loT) is an emerging technology which aims at facilitating human tasks by globally connecting things such as home appliances, ventilation systems, industrial and agricultural machines, sensors and vehicles. The Internet of Urban Things is a kindred topic that utilizes novel communication technologies to pave the way for a Smart City, providing citizens with value-added services. Service providers in a smart city (e.g. banks, shops, parking lots, transportation system) are equipped with loT-based smart features to provide citizens with a desirable quality of experience (QoE). A decision to use several city services is, consequently, a service composition problem, whereby acquiring each service from the best nearby service provider, the optimal order of receiving the city services, and the optimal route to reach each one can improve the overall Q0E, for the composite service. In this scenario, web-based loT, and more specifically, RESTful IoT, affords information regarding available services and service providers, the number of current clients in a city service provider, and city traffic, in order to facilitate decision making. To achieve a solution to the urban service composition problem, we incorporate the information provided by the IoT and propose a Q0E-aware service composition method for the smart city based on the generalized traveling salesman problem. (GTSP). The brute-force and genetic algorithms are utilized to solve the problem. Our real testbed experiments demonstrate that our proposals, taking into account the optimal sequence of services, can outperform other service composition methods when transferred to the smart city context in terms of time and total Q0E. |