Single antenna sensors in the rapidly emerging Internet-of- Things (IoT) are attractive due to their simplicity and low cost. However, determining their own positions autonomously using only a single antenna is challenging. This paper presents a novel approach for autonomous downlink localization of single-antenna receivers using Reconfigurable Intelligent Surfaces (RIS) and a tracking process using the complex extended Kalman filter (EKF). Simulation results show that the considered RIS-aided wireless radio system can provide accurate localization and continuous fast tracking down to the centimeter level, especially when multiple RISs are deployed. Furthermore, various factors affecting the system performance are analyzed in detail.