PROJECT TITLE :

Mercury: An Infrastructure-Free System for Network Localization and Navigation - 2018

ABSTRACT:

Location-awareness permits a variety of rising applications on mobile devices. For indoor applications, a fascinating way of getting real-time locations is by combining different sources of positional information, like the inertial measurements, ranging measurements, and map information with an infrastructure-free system that does not depend on any customized hardware. These sources of knowledge can be incorporated into the paradigm of network localization and navigation (NLN). However, there still lacks an infrastructure-free localization system that applies the insights of NLN to effectively fuse completely different varieties of data. During this Project, we tend to gift the Mercury system, which realizes the key concepts of NLN, including the exploitation of spatiotemporal cooperation and the use of environmental knowledge. We tend to design a real-time belief propagation algorithm to fuse inertial measurements and vary measurements among totally different users with map info. We tend to implement this algorithm within the Mercury system fashioned by a network of smartphones, and evaluate its localization accuracy through experimentation. Results show that Mercury provides reliable location data and that combining spatiotemporal cooperation with environmental information remarkably reduces the location uncertainty of users. Moreover, the performance of Mercury is additional sturdy to imperfect initial positional knowledge compared with that of existing systems.