On the Interactive Behavior of Wireless Communication with the Physical World

The recent emergence of wireless communications association to the physical world is pervasive everywhere including human monitoring, position tracking, smart transportation, and wireless Internet, which underscores the reliability, accuracy, and efficiency of the wireless communication among wireless devices with their physical environments. To fulfill these requirements, the interactions of wireless communication systems with the physical world including environmental effects, wireless device behavior, and human wireless usage needs to be investigated. The main challenges of the wireless communications interacting with the physical world are: (1) wireless communication environment is complex and unpredictable, resulting in unreliable signal transmission when the wireless signal goes through such environment undergoing reflection, diffraction, and distortion. (2) The system-on-chip wireless devices have their physical behavior, which is affected by environmental variations and service time. However, for a distributed Internet of Things (IoT) system, the collective performance highly relies on the efficiency of the individual devices and their synchronized schedule. Therefore, achieving accurate time synchronization in wireless networks is challenging. (3) Bandwidth scarcity. Due to the bandwidth limitations, how to smartly assign users to use the limited resources is a challenging problem. Therefore, in this thesis, we conduct extensive research to study and analyze those challenges to pave the way for developing advanced wireless communication systems. In this dissertation, to explore the environmental behavior, an environmental robust localization system is modeled and implemented by taking advantage of the transmit power diversity in IoT system. Besides, a multiple input and multiple output (MIMO) channel estimation research using a deep learning approach is studied to further investigate the environmental behavior of the physical world effect on the wireless communication system. Moreover, an investigation about the behavior of the system clock and RF clock (a clock for generating carrier frequency) of the wireless devices is performed to make the IoT system work harmoniously and efficiently. Moreover, to solve the problem of spectrum scarcity, first, the real WiFi data is analyzed to abstract the characteristics of the WiFi users and then, a deep learning-based approach is adopted to classify the behavior of the WiFi users. Finally, a dynamic spectrum sharing model is built based on the class-specific methodology from the pricing perspective, where the efficiency of the model has been validated by experiments