On IEEE 802.11p-WAVE-based vehicular networks : modeling, evaluation, and enhancements

Abstract

Taking advantage of the advances in the areas of mobile computing and wireless communications, Intelligent Transportation Systems will govern our future. Cars equipped with transceivers will communicate while on the road via a vehicular network in order to reduce road casualties and to improve the driving experience. The protocol stack of the Wireless Access in Vehicular Environments (WAVE) has been recently proposed to enable vehicular communication on the Dedicated Short Range Communication (DSRC) frequencies. WAVE allows for the coexistence of safety-related and non-safety related vehicular applications over single-radio equipped transceivers using multi-channel operations defined in the IEEE 1609.4 protocol. Supported by both governments and the automotive industry, WAVE equipped vehicles promise to make our trips safer and more comfortable by providing the driver with numerous safety and infotainment services. However, the current state of the protocol fails to achieve the aforementioned promises and objectives. We show in this dissertation that DSRC does not provide sufficient bandwidth to guarantee proper delivery of messages. Also, the employed synchronous channel switching scheme results in a start-of-interval synchronized collisions between queued packets. Moreover, in congested scenarios, packets are prone to collisions with neighbors' transmissions due to the lack of a proper adaptation scheme to cope with the current channel contention level. Thus, WAVE is not currently able to guarantee high probability of successful packet delivery and low average packet delivery delay, which are critical for safety applications. In this dissertation, we investigate the above mentioned challenges and provide cross-layer enhancements to the WAVE stack that result in a better and safer driving experience. First, we develop a simulation tool to evaluate the performance of the WAVE stack under different scenarios. The simulation tool accurately models the behavior of the WAVE protocol and can be used to both evaluat