Text

theses

Dedicated Short-Range Communications (DSRC) is a wireless communication technology designed to support periodic information sharing between vehicles. With the shared information, a car can enhance its situational awareness and thus improve driving safety. However, as DSRC moves rapidly towards large-scale real-world deployments, several challenges remain. One particularly challenging scenario is DSRC in heterogeneous networks where different protocols/technologies coexist. These coexistence scenarios can arise when: 1) the protocols of DSRC evolve to a newer version and thus two versions of DSRC coexist in one network during a transition period; 2) DSRC shares its licensed spectrum with other unlicensed users, e.g., Wi-Fi devices; 3) Multi-technology and multi-band vehicular communications enable other emerging technologies, e.g., mmWave, on the same car.

Since DSRC was originally designed for sole operation, coexistence could degrade its performance. However, it remains unknown to what extent the DSRC performance will be degraded and whether the performance degradation can be mitigated or controlled to an acceptable level. To fill this void, this dissertation quantifies the performance of DSRC in these coexistence scenarios via accurate simulations whose simulation models were developed and calibrated based on the data collected from a set of experiments with up to four hundred DSRC transmitters, identifies the main challenges of preserving the DSRC performance and further proposes solutions to reduce the experienced performance degradation.

Specifically, for the evolved DSRC scenario, we consider two DSRC congestion control protocols as an example, CAM-DCC as the legacy DSRC protocol and LIMERIC as the evolved DSRC protocol. We first show that the CAM-DCC vehicles can experience doubled packet latency after introducing the LIMERIC vehicles into the network and identify that the performance degradation can be controlled by adjusting LIMERIC's parameters. We then propose an adaptive algorithm to maintain the performance degradation within an acceptable level.

For the DSRC-Wi-Fi spectrum sharing scenario, we first identify delayed detection, unilateral hidden terminals and backoff countdown collisions are the main challenges of preserving the performance of the legacy DSRC when applying three recently proposed spectrum sharing mechanisms to share the DSRC spectrum with unlicensed Wi-Fi devices. Simulation results indicate that sharing the DSRC spectrum by using the three mechanisms can cause more than 30% extra packet losses of DSRC transmissions. To reduce the DSRC performance degradation, we then suggest adding an extra idle period to the Wi-Fi inter-frame period and preventing Wi-Fi transmissions upon DSRC detection.

For the mmWave scenario, we show that mmWave coordination is possible via DSRC without affecting existing DSRC traffic. The driving status shared via DSRC can be further utilized to estimate the usefulness of sensing information shared via mmWave. When scheduling mmWave transmissions, these transmissions carrying more useful information can then be scheduled with a higher priority. Compared to a simple broadcast strategy, our approach significantly improves the mmWave information sharing efficiency by up to 50%.

ETD_9485

Ph.D.

Includes bibliographical references

by Bin Cheng

doi:10.7282/t3-vdwp-j902

ETD doctoral

The author owns the copyright to this work.