Optimized Transmission of Loss Tolerant Information Streams for Real-Time Vehicle-to-Vehicle Communications

This paper addresses the real-time transmission optimization of loss tolerant information streams over Vehicular Ad Hoc Networks. The proposed FEC and Interleaving Real Time Optimization (FIRO) technique dynamically adapts the two well-known error resilience techniques to the actual channel conditions, according to the stream constraints in terms of maximum allowed packet loss rate and tolerated delay. To increase the quality of the communication, FIRO implements a dynamic feedback mechanism based on periodic RTCP Receiver Reports. Frequency of reports is dynamically adapted based on the accuracy of the estimates. The FEC and Interleaving parameters are updated based on both short- and long-term channel estimates. To evaluate the performance of FIRO in a realistic environment, in this paper the proposed technique is applied to real time vehicle-to-vehicle video streaming. However, FIRO can also be implemented in other scenarios where a loss tolerant information flow is streamed over a lossy channel with real time constraints. Results show that FIRO consistently improves the performance of the received video stream at both network and application layers, thanks to the joint effect of FEC, Interleaving, dynamic parameter update and dynamic update of the inter-arrival frequency of the receiver reports. As a side benefit of the packet-level design, FIRO can be implemented on Off-The-Shelf IEEE 802.11 devices.