Development of DSRC Radios upon the IEEE 802.11p Standard and Physical Layer Security for Automated and Connected Vehicles Applications


Weidong Xiang, PhD
Associate Professor, Electrical and Computer Engineering, U-M Dearborn


May 2014 – April 2016


  1. To study, develop and implement DSRC radios for automated and connected vehicle applications under fast-varying mobility and hostile environments.
  2. Upon the development of DSRC radios, develop lightweight, scalable and robust physical modules for secure DSRC wireless access.


The foremost task in this project is to conduct extensive field-testing on the developed DSRC OBU/RSU under diverse road scenarios, in different traffic periods and with common weather conditions. Next, several relevant advanced physical technologies/schemes will be explored in order to enhance the transmission robustness; even they have not been included in current IEEE 802.11p standard. Upon development of DSRC radios, the multiple-input multiple-output (MIMO), beam forming, and spectrum awareness based channel selection will be developed and tested. Authentication is the first step to establish secure communication between two parties. This project proposes two building blocks to encrypt/hide challenge and response messages at the physical layer: one exploits the reciprocity, randomness, and location decorrelation properties of the wireless fading channel, and the other takes advantage of the time coherence and randomness of the wireless fading channel. For securing vehicular communication, we propose a new key generation mechanism without channel estimation. In this two-generation scheme, two communication parties send random OFDM signals to each other without preamble or reference signals while a shared secret can be established by exploiting channel reciprocity.


The DSRC prototype and development kit can be used to conduct extensive field testing under diverse road scenarios, in different traffic periods and with common weather conditions. Advanced base band algorithms, including robust and fast time synchronization, Doppler shift frequency estimate and compensation, channel estimate and advanced coding schemes can be modified and tuned by users. The prototype operates in a real-time mode for both low-speed event-driven safety critical message transmissions, as well as high-speed periodic wireless access for infotainment.

Build up an advanced road information platform at Mcity. Based on the current outcome, we are able to build an exemplar road information platform with up to 10 SDR base stations and more than 30 SDR terminals in the Mcity Test Facility for autonomous driving. The easy-to-use SDR platform is able to realize emerging and future wireless systems with base-band modules including those FFT/IFFT, modulation/demodulation, encoder/decoder, spectrum/channel estimate, OFDM and MIMO function blocks defined by WiFi and LTE standards. The Road Information Platform can be  acclimatized and parameterized. The former feature indicates that the platform is environmentally adaptive and able to 1) sense ambient spectrum in sufficient sweep speed with acceptable false alarm and missed alarm; 2) predict channel utilization featuring self-evolution (An old horse knows the way – a Chinese idiom); and 3) tuned to fit the undergoing service. The latter describes the controllability and multitasking of the system.

Big Data Based Wireless Channel Models and Channel Emulators. Modern wireless communications are criticized by the fact that real-world performances do not achieve their designed goals, including the widely deployed cellular and WiFi systems. One of the main factors is the lack of accurate wireless channel models. Actually, wireless channels are too complicated and diverse to be described by either a random distribution, mathematically, or with perfect close forms; even if they are highly desired in the system design. Likewise for the electromagnetic theory model and ray-tracing simulator. The PIs will present and study big data based wireless channel models by capturing extensive channel impulse response (CIR) under categorized electromagnetic environments followed by feature based modeling. The more sample of CIR it has, the more accurate the channel model will be. The channel models to be studied cover indoors, outdoors, vehicles, homes, industrial workshops, urban, suburban, and rural areas. Moreover, big data based channel emulators will be developed and made available to related research.

Web sites that reflect results of the project: