A Survey on Detection, Tracking and Identification in Radio Frequency-Based Device-Free Localization

Abstract

The requirement of active localization techniques to attach a hardware device to the targets that need to be located can be difficult or even impossible for certain applications. For this reason, there has been an increasing interest in tagless or device-free localization (DFL) approaches. In particular, the research domain of RF-based device-free localization has been steadily evolving since its inception slightly over a decade ago. Many novel techniques have been developed regarding the three core aspects of DFL: detection, tracking, and identification. The increasing use of channel state information (CSI) has contributed considerably to these developments. In particular, the progress it enabled regarding the exceptionally difficult ‘identification problem’ has been highly impressive. In this survey, we provide a comprehensive overview of this evolutionary process, describe essential DFL concepts and highlight several key techniques whose creation marked important milestones within this field of research. We do so in a structured manner in which each technique is categorized according to the DFL core aspect it emphasizes most. Additionally, we discuss current blocking issues within the state-of-the-art and suggest multiple high-level research directions which will aid in the search towards eventual solutions.

Device-Free Localization and Identification Using Sub-GHz Passive Radio Mapping

“…Finally, we would like to note that RF-based device-free localization has become an increasingly popular topic of research [4] with a variety of potential applications [6,24,25]. In particular, the rise of the Internet-of-Things has led to significant interest in the use of "signals of opportunity" from already existing network infrastructure for DFL applications [26].…”

“…Therefore, in order to aid the further development of this field, there will be a need for public datasets from a variety of different environments, experimental setups and signals. The lack of public datasets is a major problem within the research field of RF-based DFL in general and we have previously discussed this in our earlier work [4]. Additionally, in the context of large-scale device-free crowd estimation, we have published one such dataset ourselves [27].…”

“…In the years since the aforementioned seminal study was published, significant advances have been made within this field of research. Many different techniques have been developed which focus on one or more of the three main aspects [4]. While initial approaches tended to solely make use of the impact of targets on the Received Signal Strength (RSS) of RF signals, an increasing number of studies is focused on other signal descriptors-most notably Channel State Information (CSI) [5].…”

MAMPI-UWB—Multipath-Assisted Device-Free Localization with Magnitude and Phase Information with UWB Transceivers

“…In contrast to camera-based systems, device-free localization preserves privacy and may work in non-line-of-sights conditions as well as in darkness and smoke or with obstructed sensors. An overview of radio-frequency-based device-free localization systems is given in [ 3 ].…”

RSSI-Based for Device-Free Localization Using Deep Learning Technique

“…It can also be known as device-free localization (DFL) or sensor-less sensing. Most of the work utilizes radio frequency (RF) to create a wireless sensor network, where specifically the propagation of the radio is actually the sensors [9]. In particular, a received signal strength indicator (RSSI) can also be used to measure signal variations between sensor nodes.…”