NeTS: Small: SOAR: Leveraging Traffic Signatures to Improve Performance of Vehicular Networks

点击翻译按钮获取中文摘要

Vehicular networking is an enabling technology for the next-generation transportation services, such as vehicle safety communication, remote vehicle diagnosis, and network-assisted autonomous driving. A major challenge to support these applications is to achieve real-time, reliable communication using limited wireless network spectrum resource, in the presence of noise, interference, and mobility-induced wireless channel variations. This project develops a novel cross-layer framework that improves the performance of vehicular networks without incurring extra spectrum overhead. The key observation is that traffic typically exhibits high predictability due to protocol-specific header structures and the spatiotemporal correlation of application data, such as vehicle positions, on-board sensor readings, and the range to surrounding objects. This project takes a system approach to leveraging such traffic signatures to optimize vehicular network performance. The approach is based on a top-down cross-layer architecture, which consists of 1) a traffic signature extraction algorithm that achieves assured accuracy under the dynamics of network traffic; 2) traffic signature-aware PHY algorithms that improve the robustness and resilience to channel variations and interference; and 3) a comprehensive link layer that exploits traffic signatures in fine-grained optimization of power control, rate adaptation and error recovery.This project has broad implications for future vehicular network systems in multiple application domains that demand high communication reliability under spectrum resource constraints. The solution framework is transparent to vehicular network applications and can be integrated with various wireless standards, including 802.11p, WiMAX and LTE, for providing better Quality of Service (QoS) in both inter-vehicle and vehicle-to-infrastructure networks. Educational and outreach activities include introduction of vehicular network systems into new graduate courses with software, testbed, and labs developed in the course of this project, and recruitment of women and minority students for participation in the project.