Urban VANET performance optimization

Urban VANET (Vehicular Ad hoc NETworks) performance optimization concerns the improvement of wireless signal quality between two arbitrary selected nodes moving within along city streets. It includes three procedures: VANET architecture modeling; wireless signal simulation; and signal quality optimization techniques. The first procedure converts real-world map data into a network graph according to the requirement of the optimization algorithm. The second step analyzes a communication route between two network nodes and calculates received signal quality with the information provided by the network model. The final operation optimizes the signal quality to an expected level by choosing appropriate communication route between two wireless nodes. In this thesis, three optimization techniques are presented: EP (Evolutionary Programming), SG (Stochastic Geometry) and SW (Small World). EP is a widely applied optimization strategy based on Darwin’s natural selection and evolution theory. It is effective with an enormous number of data support, and it can provide detailed route information. However, it requires enough time to evolve to an optimal solution. SG is a statistical tool to analyze points’ distribution within a multi-dimensional space, and it was recently applied on wireless network analysis. Given the distribution characteristics of an urban area, SG can calculate average data loss rate of a communication route. However, it cannot provide detailed route information. SW is a widely accepted model to represent people’s relationship in social networks, and it can be used in VANET analysis. SW provides a simplified network architecture compared with EP an SG. However, it requests additional long-range communication equipment and consumes more energy. The thesis is divided into three parts. Chapter 1 introduces the history of VANET and its architecture (in this research, it is a combination of Ad hoc network and WSN (Wireless Sensor Network). Chapter 2 and 3 presents literature review of EP and SG. Chapter 4, 5, and 6 discusses how to implement EP, SG and SW on Boston VANET. At the end of each chapter, a conclusion is presented and a discussion on the author’s contribution is given