| KEYWORD |
Machine Learning techniques for Positioning, Navigation and Timing in IoT devices
keywords ARTIFICIAL INTELLIGENCE, COMMUNICATIONS NETWORKS, GLOBAL NAVIGATION SATELLITE SYSTEM, IOT AND SENSORS, IOT, M2M COMMUNICATIONS, REAL TIME MONITORING, MACHINE LEARNING, ARTIFICIAL NEURAL NETWORKS
Reference persons DANIELE JAHIER PAGLIARI, ALEX MINETTO
Research Groups Navigation Signal Analysis and Simulation group (NavSAS)
Thesis type EXPERIMENTAL / DEVELOPMENT
Description Nowadays, Global Navigation Satellite Systems (GNSS) serve a growing number of users as GNSS receivers are embedded in a plethora of electronic devices. GNSS technology leverages Time-of-Arrival (ToA) measurements that are interpreted as receiver-to-satellite ranges, i.e. pseudoranges. These measurements must be continuously corrected from intrinsic biases (e.g., relativity, atmospheric delays) to be exploited for the estimation of Position, Velocity, and Time (PVT) through linear or Bayesian estimators. These corrections are related to the data carried by the so-called navigation message, whose demodulation is typically performed in the receiver's processing loop.
IoT electronics host "by definition" low-power-consumption network connectivity, thus enabling possible new paradigms for the PVT estimation, known as collaborative or cooperative PNT methods. By assuming the availability of a variable number of commercial GNSS receivers hosted by cars, smartphones, and wearables, and a possible background data link to share GNSS data, this thesis aims at providing an answer to the following questions:
Is it possible to avoid pseudorange correction and demodulation of the navigation message thanks to a set of networked, collaborative surrounding users? Can machine learning techniques, such as KNN, support PVT estimation in IoT devices by minimizing their computational load? Are these techniques capable of a quick and reliable PNT estimation in such devices by relying on surrounding data provided by collaborative users?
Required skills Required skills include MATLAB and Python programming. Furthermore, a basic knowledge of computer architectures and embedded systems is necessary. Desired (but not required) skills include some familiarity with basic machine/deep learning concepts and prelminary knowledge on Global Navigation Satellites Systems (GNSS)
Deadline 13/12/2022
PROPONI LA TUA CANDIDATURA