Using web services-based wireless sensor networks for estimating the free size of parking places

Abstract

The aim of this paper is to measure the impact that web services have on wireless sensor networks (WSN) when they are used for estimating the free size of parking places in smart cities. In particular, each wireless sensor mote (WSM) of these WSNs supports a fuzzy inference system (FIS)-based algorithm which predicts the aforementioned free sizes. Moreover, the inputs of this algorithm are the measurements acquired by the adjacent WSMs regarding the variation of magnetic fields caused by the chassis of vehicles. Well known web services technologies were used as the implementation vehicle for exchanging these measurements. Moreover, the well known network simulator 2 (NS-2) was used for acquiring simulation results concerning the energy consumption of the battery of the WSMs. These results have shown that Web services significantly increase the battery consumption. Furthermore, in order to model this consumption, a data fitting problem was solved and the cubic polynomial was fitted to the simulation data. This polynomial and the one obtained by modeling the error of the FIS were utilized in order to form a real-valued cost function. This function was minimized to get the optimal values for the distance between the WSMs and the sampling period based on which the output of the algorithm should be re-computed.