Visible Liqht Based Non-cooperative Indoor Localization

With the rapid development of economy and urbanization of China, position information service plays more and more important role in our daily life and research area. After the long time development, GPS (Global Positioning system) has become a mature outdoor position technology, it can provide excellent position service with its huge database, wonderful hardware and software support. For example, Google Map and Amap can offer you positioning service in the outdoor space all over the world. However, due to the shielding and interference of building structures, traditional GPS positioning method cannot be applied to indoor scenarios. Therefore, how to design a new indoor positioning method has become an urging and emerging research topic in recent years.In the past few years, many scholars and research institutions have used Wi-Fi, sound, magnetic signals, the RFID (Radio Frequency IDentification) and many other ways to achieve high accuracy indoor positioning. As LED (Light Emitting Diode) gradually becomes the mainstream indoor lighting equipment, localization technology based on the controllable LED becomes more and more popular. Localization technique based on LED has high precision and strong anti-interference ability. By selecting appropriate frequency, the technique can avoid a large number of environmental interference while offering a high lighting performance. This dissertation utilizes the frequency-controllable characteristic of LED, and proposes a visible light based non-cooperative indoor positioning system. Compared with the traditional positioning system based on mobile intelligent equipment, non-cooperative localization system can independently achieve the positioning function, without any help from users and shows no need to wear auxiliary equipment. This system, which is suitable for workshop, prisons and other more specific, complex indoor scene, can locate multiple targets simultaneously in indoor environment, and has the characteristics of strong expansibility, low energy consumption, high precision and low cost.To begin with, this dissertation designs all the hardware necessary for the system operation, mainly including a frequency-controllable LED module and a non-cooperative optical information acquisition node. Then an autonomous lighting system and environmental light acquisition array are established. Then an autonomous lighting system and environmental light acquisition array are established. The LED module has independent power supply and driver circuit, which is accurately controlled by a single chip microcomputer. The LED module is able to provide stable and long-lasting lighting. The design of non-cooperative optical information acquisition node is very concise and it can collect optical information precisely. The optical information acquired by the acquisition node is transferred to the host computer through wired communication for further position calculation.Next, an indoor localization algorithm in the typical indoor environments is designed based on a new shielding model. Firstly, a single target localization algorithm is proposed. Then, a multi-target localization algorithm is designed by leveraging the K-means clustering algorithm and the screening and optimization algorithm. The reasonability and feasibility of the proposed algorithm is verified through multiple sets of simulations and experiments, and comparisons of various optimization methods are discussed.In the end, with the combination of the hardware units and the algorithms, this dissertation realizes a complete LED-based non-cooperative indoor localization system, which is able to realize multi-target localization according to the number of people. The performance of the system is evaluated in the practical scenarios. Further, this dissertation also evaluates the minimum positioning unit of the system through a large number of experimental tests after the discussion of light source and sensor placement scheme. The tests include both single-target and multi-target location scenarios, and the performance of stability, precision and anti-interference are verified.