Research On Key Problems Of Communications In Self-Organizing Networks

Self-Organizing Networks (SONs) consist of elements that can automatically for-m into a network. Towards successful communication, SONs have the following key problems:where the terminals are, how to transmit the messages towards the destina-tions, and how to transmit the messages in a secure manner. Focusing on such SONs as wireless sensor networks, smartphone systems, and vehicular networks, this thesis considers the problems of localization, routing algorithm optimization and privacy pro-tection.(1) We propose three localization methods. Two of them are for sensor networks and one is for smartphone. In wireless sensor networks, sensed data usually make sense only if we know where the data are collected, so localization for sensor nodes is critical. We propose a method for locating sensors with linear topology. Real-world experiments have indicated that our method has very high accuracy. We also propose a method for estimating the distance between two sensor nodes in a range-free manner, which has been shown to greatly improve the localization accuracy of existing localization algorithms. For smartphone, it is widely believed that the Global Positioning System (GPS) in the smartphone has an accuracy enough for locating people, but our real-world measurements suggest that this is not the case. Our third localization method is for smartphone. It tracks the location of a pedestrian with an accuracy much better than the built-in GPS component. (2) We propose three methods for routing algorithm optimization. Due to limited energy supply, maximizing the lifetime of sensor nodes is the objective for routing algorithm optimization in sensor networks. In the three proposed methods, two are for sensor networks, with one being an exact algorithm for data gathering tree design problem and the other being an approximation algorithm for data aggregation tree design problem. The third method is for assigning smartphones to APs for WiFi access, and the objective is to maximize the minimum throughput. (3) We propose one method for privacy protection. Specifically, we design a scheme to let two cars extract a shared secret key from the channel dynamics. The extracted key is information-theoretically secure, i.e., it is secure against an adversary with unlimited computing power.