Ubiquitous Localization (UbiLoc): A Survey and Taxonomy on Device Free Localization for Smart World

摘要

The "Smart World" envisioned by technology will be achieved by the penetration of intelligence into ubiquitous things, including physical objects, cyber-entities, social-elements or individuals, and human thinking. The development of Smart World is enabled by diverse applications of wireless sensor networks (WSNs) into those components identified as things. Such a smart-world will have features controlled significantly by the location information. Control and Policy information of Smart World services, often addressed as location-based services (LBSs), are governed by location data. Localization thus becomes the key enabling technology for Smart World facilities. It is generally classified as active and passive techniques in nature. Active localization is a widely adopted localization scheme where the target is detected and tracked carries a tag or attached device. The other category, Passive methods, defines targets to be localized as free of carrying a tag or device, hence also referred to as device-free localization (DFL) or sensor-less localization. The passive approach is a well suited for the development of diverse smart world applications with ubiquitous localization. DFL schemes fall into a wide range of application scenarios within the Smart World ecosystem. A few notable examples are occupancy detection, identity definition, positioning, gesture detection, activity monitoring, pedestrian and vehicle-traffic flow surveillance, security safeguarding, ambient intelligence-based systems, emergency rescue operations, smart work-spaces and patient or elderly monitoring. In this paper, the revolution of DFL technologies have been reviewed and classified comprehensively. Further, the emergence of the Smart World paradigm is analyzed in the context of DFL principles. Moreover, the inherent challenges within the application domains have been extensively discussed and improvement strategies for multi-target localization and counting approach are discussed. Finally, current trends and future research directions have been presented.