A GNSS-only Cooperative Framework for Positioning of Networked Android Devices

The massive growth of the Information and Communication Technologies (ICT) and the increasing number of pervasive electronic devices are transforming cities in Smart Cities. This technology-driven revolution has raised the interest of researches to provide benefits to residents in several fields: from mobility to energy efficiency, passing by health and environmental challenges, the smart paradigm promises a better quality of life.?? Considering such a context, an increasing number of location-based services are demanding for greater accuracy and precision. Telecommunication networks are pursuing ubiquitous connectivity among different types of devices, granting a constant exchange of data and information whitin paradigms such as Internet Of Things (IOT). The network ubiquity can be considered the enabling key for answering to most of the requirements of Position and NavigationTechnologies (PNT).?? Indeed, this pressing request of accuracy and precision in positioning systems has required the integration of sensors such as Integrated Measurement Units (IMU) to ensure improved performances in standalone satellite navigation. The fusion of sensors and GNSS has been deeply investigated, leading to robust and reliable solutions for navigation. Despite these remarkable results, sensor-less solutions are still appealing in the field of ultra low cost solutions, such as components for mass-market devices. Cooperative Positioning (CP) represents another strategy to achieve a positioning improvement. Many solutions have been explored in this area, with the greater part requiring a specific hardware to be exploited. For example, in robotics the relative positioning among devices - demanded by the cooperative algorithm - is provided by a set of exteroceptive sensors such as radars and ultrasonic units; in vehicular navigation, some cooperative strategies are achieved through short range communication networks (Vehicle to vehicle (V2V) paradigm) or through a distributed ad hoc infrastructure (Vehicle to Infrastructure (V2I) paradigm). This thesis aims to develop a framework for Collaborative Positioning relying on a relative unexplored solution: the evaluation of the unknown inter range distance from the GNSS measurements redundancy. In addition to this, the relevance of this work is the creation of a service that can be integrated in the native Application Programming Interface (API) of the mobile device a in order to improve its positioning output. The framework allows the exchange of raw GNSS measurements among ultra low cost receivers interconnected by a general purpose network, i.e. 4G/LTE or Wi-Fi, leading to the estimation the inter agent distance. These measurements are then integrated to improve the calculation of the GNSS-only PVT solution. The framework is composed by a variable number of nodes (agents) connected in star topology with a central Cooperative Positioning Services (CPS). The CPS is a set of services running on a server implementing a data distribution channel for the agents. The agents run a mobile Android application called Cooperative Positioning Application (CPA) to execute the CP algorithm. To evaluate the performances of the CP, a remarkable set of experiments has been executed, investigating various test scenarios.