Opportunistic Spectrum Utilization for Vehicular Communication Networks
Recently, vehicular networks (VANETs), has become the key technology of the next-generation intelligent transportation systems (ITS). By incorporating wireless communication and networking capabilities into automobiles, information can be efficiently and reliably disseminated among vehicles, road side units, and infrastructure, which enables a number of novel applications enhancing the road safety and providing the drivers/passengers with an information-rich environment. With the development of mobile Internet, people want to enjoy the Internet access in vehicles just as anywhere else. This fact, along with the soaring number of connected vehicles and the emerging data-craving applications and services, has led to a problem of spectrum scarcity, as the current spectrum bands for VANETs are difficult to accommodate the increasing mobile data demands. In this thesis, we aim to solve this problem by utilizing extra spectrum bands, which are not originally allocated for vehicular communications. In this case, the spectrum usage is based on an opportunistic manner, where the spectrum is not available if the primary system is active, or the vehicle is outside the service coverage due to the high mobility. We will analyze the features of such opportunistic spectrum, and design efficient protocols to utilize the spectrum for VANETs. Firstly, the application of cognitive radio technologies in VANETs, termed CR-VANETs, is proposed and analyzed. In CR-VANETs, the channel availability is severely affected by the street patterns and the mobility features of vehicles. Therefore, we theoretically analyze the channel availability in urban scenario, and obtain its statistics. Based on the knowledge of channel availability, an efficient channel access scheme for CR-VANETs is then designed and evaluated. Secondly, using WiFi to deliver mobile data, named WiFi offloading, is employed to deliver the mobile data on the road, in order to relieve the burden of the cellular networks, and provide vehicular users with a cost-effective data pipe. Using queueing theory, we analyze the offloading performance with respect to the vehicle mobility model and the users' QoS preferences. Thirdly, we employ device-to-device (D2D) communications in VANETs to further improve the spectrum efficiency. In a vehicular D2D (V-D2D) underlaying cellular network, proximate vehicles can directly communicate with each other with a relatively small transmit power, rather than traversing the base station. Therefore, many current transmissions can co-exist on one spectrum resource block. By utilizing the spatial diversity, the spectrum utilization is greatly enhanced. We study the performance of the V-D2D underlaying cellular network, considering the vehicle mobility and the street pattern. We also investigate the impact of the preference of D2D/cellular mode on the interference and network throughput, and obtain the theoretical results. In summary, the analysis and schemes developed in this thesis are useful to understand the future VANETs with heterogeneous access technologies, and provide important guidelines for designing and deploying such networks.