Abstract:

Machine learning (ML) has emerged as a transformative approach for radio frequency (RF) signal source localization and emitter tracking, providing data-driven solutions that extend beyond traditional analytical models. Progress in this area is fundamentally linked to the availability of datasets that capture realistic propagation characteristics across diverse scenarios. Real-world data acquisition remains costly, time consuming, and subject to regulatory and technical constraints, which motivates the use of both measured and synthetic datasets. The aim of this article is to survey publicly available datasets and simulators, categorizing their scope, applications, and limitations for advancing RF localization research. Particular attention is given to their role in supporting business applications such as industrial automation, wireless infrastructure optimization, and location-based services. By consolidating existing measurement datasets and open-source simulators, this work establishes a comprehensive reference to guide reproducible experimentation and the development of robust ML solutions in RF emitter localization and tracking.