Contemporary communication systems are based on modular architectures, where the role of the physical layer is essentially confined to error correction, whereas information security is typically dealt with at the upper layers of the protocol stack. We consider a security architecture that does exactly the opposite: information sequences are first converted to longer channel codewords which are then encrypted using a classical stream cipher. Although this approach requires longer encryption sequences, our analysis shows that the natural randomness of the noisy communication channel can be used effectively against known-plaintext attacks. We also address practical implementation issues in physical-layer encryption and discuss their impact on the system architecture and on the security performance.

This work presents a V2V video transmission system to help on a car-overtaking decision using the IEEE 802.11p/WAVE communication technology. The system consists on a ffmpeg based video encoder that encapsulates the data received from a camera placed in a front vehicle into HTTP Post packets, and forwards the packet to the rear vehicle through the On Board Units (OBUs) installed on both vehicles. The rear vehicle presents the images to the driver using a visualization device. The proposed system was evaluated through real vehicular experiments in two distinct scenarios: urban and highway. Performance studies were focused on delay measurements, re- transmission rate, signal strength and bandwidth consumption related with the traveling speeds and video quality. Results show that the communication delay is higher in the highway scenario, mainly due to the distance between vehicles and the different speeds. However, promising results regarding the maximum delay and the average number of retransmissions foresee important inputs for future services of assisted-driving, in general, and car- overtaking assistance, in particular.