Since the invention of the first car available to masses, the 1908 Ford Model T, technology has advanced towards making car travel safer for occupants and bystanders. In recent years, wireless communication has been introduced in the vehicular industry as a means to avoid accidents and save lives.
Wireless communication may sometimes be challenging due to obstacles in the physical world that interact with wireless signals. Such obstacles may be dynamic, e.g. other vehicles in the traffic flow, or static, e.g. nearby buildings. Two scenarios are defined to describe those cases. The obstructed line-of-sight (OLOS) scenario is described as the case where a smaller obstacle, usually a vehicle, is placed in-between a transmitter and a receiver. This obstacle usually partially blocks communication and the receiver often moves in an out of the line-of-sight. The non line-of-sight (NLOS) scenario is described as the case where a larger obstacle completely blocks communication between a transmitter and a receiver. An example would be a building at an intersection which shadows the communication between two vehicles. In this thesis the OLOS and NLOS scenarios are investigated from different points of view.
In chapter 2, a road side unit (RSU) that has been constructed and evaluated for integrating older vehicles without wireless communication with newer vehicles using wireless communication is described. Older vehicles are being detected using a universal medium-range radar and their position and speed vectors are broadcasted wirelessly to newer vehicles. Tests have been performed by using the system in parallel with wireless enabled vehicles; by comparing the content in the messages obtained from both systems, the RSU has been found to perform adequately. Accuracy tests have been performed on the system and Kalman filtering has been applied to improve the accuracy even further.
Chapter 3 focuses on the OLOS scenario. A truck as an obstacle for wireless vehicular communication is being investigated. Real life measurements have been performed to characterize and model the wireless channel around the truck. The distance dependent path loss and additional shadowing loss due to the truck is analyzed through dynamic measurements. The large scale fading, delay and Doppler spreads are characterized as a measure of the channel dispersion in the time and frequency domains. It has been found that a truck as an obstacle reduces the received power by 12 and 13 dB on average in rural and highway scenarios, respectively. Also, the dispersion in time and frequency domains is highly increased when the line-of-sight is obstructed by the truck. A model for power contributions due to diffraction around the truck has also been proposed and evaluated using the previously mentioned real life measurements. It has been found that communication may actually be possible using solely diffraction around a truck as a propagation mechanism.
Finally, in chapter 4 a wireless channel emulator that has been constructed and evaluated is described. Modem manufacturers face a challenge when designing and implementing equipment for highly dynamic environments found in vehicular communication. For testing and evaluation real-life measurements with vehicles are required, which is often an expensive and slow process. The channel emulator proposed is designed and implemented using a software defined radio (SDR). The emulator together with the proposed test methodology enables quick on-bench evaluation of wireless modems. It may also be used to evaluate modem performance in different NLOS and OLOS scenarios.