A WiFi RSSI ranking fingerprint positioning system and its application to indoor activities of daily living recognition

WiFi received signal strength indicator seem to be the basis of the most widely used method for indoor positioning systems driven by the growth of deployed WiFi access points, especially within urban areas. However, there are still several challenges to be tackled: its accuracy is often 2–3？m, it is prone to interference and attenuation effects, and the diversity of radio frequency receivers, for example, smartphones, affects its accuracy. Received signal strength indicator fingerprinting can be used to mitigate against interference and attenuation effects. In this article, we present a novel, more accurate, received signal strength indicator ranking–based method that consists of three parts. First, an access point selection based on a genetic algorithm is applied to reduce the positioning computational cost and increase the positioning accuracy. Second, Kendall tau correlation coefficient and a convolutional neural network are applied to extract the ranking features for estimating locations. Third, an extended Kalman filter is then used to smooth the estimated sequential locations before multi-dimensional dynamic time warping is used to match similar trajectories or paths representing activities of daily living from different or the same users that vary in time and space. In order to leverage and evaluate our indoor positioning system, we also used it to recognise activities of daily living in an office-like environment. It was able to achieve an average positioning accuracy of 1.42？m and a 79.5% recognition accuracy for nine location-driven activities