Reliability Analysis of RSSI for Localization in Small Scale WSNs

Today, wireless sensor networks (WSNs) are finding places in a wide range of applications concerning with remote and mobile measurement, personal tracking also for medical purposes, monitoring of industrial processes, mobile agents and automated warehouses to cite a few. In most of such applications, it is also required to estimate the position of the sensor node or the distances among the nodes. Then, many localization techniques have been studied in recent years, which can be adopted in small scale WSNs and that are characterized by cheapness and simplicity. Among all, those based upon the measurement of the Received Signal Strength (RSS) have found a large deployment since, commonly, the wireless communication modules implement built-in features able to provide an estimation of RSS by means of the RSS Indicator (RSSI). As described in the related literature, RSSI depends upon many practical issues, potentially sources of bias and variability that can influence significantly the localization accuracy. At the moment a methodological approach for analyzing these important aspects and a suitable uncertainty model for qualifying the RSSI measurements are not yet available. In this framework, the paper proposes a systematic approach for reliably investigating the RSSI features in localization applications. In particular, a suitable measurement set-up is proposed for identifying the intrinsic variability of RSSI measurements and for studying in a systematic way the effects of external interferences on the RSSI values.