Abstract

Indoor positioning, also referred to as indoor localization, shall be defined as the process of providing accurate people or objects coordinates inside a covered structure, such as an airport, an hospital,and any other building. The applications and services which are enabled by indoor localization are various, and their number is constantly growing. Industrial monitoring and control, home automation and safety, security, logistics, information services, ubiquitous computing, health care, and ambient assisted living (AAL) are just a few of the domains that indoor positioning technology can benefit. A significant example is offered by local information pushing. In this case, a positioning system sends information to a user based on her/his location. For instance, a processing plant may push workflow information to employees regarding operating and safety procedures relevant to their locations in the plant. The positioning system tracks each employee, and has knowledge of the floor plan of the facility as well as the procedures. When an employee walks inside a defined perimeter of a particular area, such as packaging department, the positioning systems displays on the user’s personal device assistant (PDA) information regarding the work expected to be done in that area This significantly increases efficiency and safety by ensuring that employees follow carefully designed guidelines. Location-enabled applications like this are becoming commonplace and will play important roles in our everyday life. The design of a positioning system for indoor applications is to be regarded as a challenging task. In fact, the global positioning system (GPS) is a great solution for outdoor uses, but its applicability is strongly limited indoors because the signals coming from GPS satellites cannot penetrate the structure of most buildings. For this reason, considerable research interest for alternative non-satellite-based indoor positioning solutions has arisen in the last years. Actually, the positioning problem is strictly related with the measurement of the distance between the object to be located and a number of landmarks with known coordinates. Then, the position of the target is commonly determined by means of appropriate statistical or geometrical algorithms. Several approaches have been proposed, and various are the fundamental technologies that have been used so far. Today, distance measurement between two objects can be easily obtained by using laser-, optical- and ultrasounds-based devices. However, evident drawbacks of these systems are their sensitivity to line-of-sight (LOS) constraint, and the strict object-to-object bearing requirement. The latter becomes an even worse downside when the topology of the system dynamically changes due to mobility of the target. On the other hand, wireless-based ranging solutions are more insensitive to obstacles and non-alignment condition of devices. In addition, they may take advantage of existing radio modules and infrastructures used for communications. Accordingly, the ever grow- ing popularity of mobile and portable embedded devices provided with wireless connectivity has encouraged the study and the development of radio-frequency (RF)-based positioning techniques. The core of such systems is the measurement of distance-related parameters of the wireless signal. The two most common approaches for wireless ranging are based on received-signal-strength (RSS) and message time-of-flight (ToF) measurements, respectively. In particular: • the RSS-based method relies on the relationship between the measured received signal power and the transmitter-receiver distance, assuming that the signal propagation model and the transmitted power are known; • the ToF-based technique leans on the measured signal propagation time and the light speed, owing to the fundamental law that relates distance to time. The work presented in this dissertation is aimed at investigating and defining novel techniques for positioning in indoor environment based on wireless distance measurements. In particular, this study is devoted to the analysis and in depth evaluation of the use of RSS and ToF measurements for different indoor positioning applications. State of the art techniques relying on RSS- and ToF-based ranging methods have already proven to be effective for the localization of objects inside buildings. Nevertheless, several limitations exist (e.g., on the accuracy of the ranging, its impact on localization algorithms, etc.). The work presented in this dissertation attempts to: 1. investigate the main sources of uncertainty affecting RSS- and ToF-based indoor distance measurement; 2. analyze the impact of ranging error on the accuracy of positioning; 3. propose, on the basis of the understanding gained from 1. and 2., novel and effective systems in order to overcome the above-mentioned limitations and improve localization performance. The novel contributions of this thesis can be summarized as follows: • In-depth analysis of both RSS- and ToF-based distance measurement techniques, in order to assess advantages and disadvantages of each of them. • Guidelines for using different ranging methods in different conditions and applications. • Implementation and field testing of a novel data fusion algorithm combining both RSS and ToF techniques to improve ranging accuracy. • Theoretical and simulation-based analysis of chirp spread spectrum (CSS) signals for low-level timestamping. • Experimental assessment of CSS-based timestamping as key enabler for high accuracy ToF-based ranging and time synchronization.