Abstract
In our previous work, we proposed an approach that consists in using feature extraction to reduce the magnetic-field signal data
to generate a signature, which can be used to estimate the user location in an indoor environment. Each signature contains 10
seconds of magnetic-field data. In this paper, we investigate whether by reducing the time signature length decrease or not the
accuracy of the indoor location system (ILS). In order to find out, we perform experiments in two indoor environments: an office
building and a residential home. In both environments, we collect information of the magnetic-field and we variate the fingerprint
length in 0.1, 1, 2, 5 and 10 seconds, to verify whether its length affects the accuracy of the ILS Model. The results indicate that
signature length is an important issue to be considered in the development of the ILS, since it does affect the accuracy of the system.