Wireless infrastructure such as WiFi gained wide interest by researchers and private companies to determine the location of a target object in a building. Hence, we observe an increase in popularity of in building WiFi solutions. However, significant fluctuations in WiFi signal strength are affecting the measurement quality and as a result the fingerprinting positioning approach. The accuracy of the positioning will be improved if the RSS measured at different distances between transmitter (Tx) to receiver (Rx) is distinctly different. Several filtering methods may be applied to reduce the data noise for estimating. Many methods demonstrate great noise reduction. The time taken to compute data becomes limited to mobile robot application since real-time is required. The requirement for real time data singles out mobile robot application as an effective solution. In this paper, the focus will be on the difference of RSS measurements with varying Tx-Rx distances, following the design of experiment (DoE). Further we include a comparison between the Kalman filtering and weight fusion, which is Weight-Average of the Top-n Populated values model (WATP). Errors in estimation and reduction in computation time are taken into consideration.
