Abstract
Vehicular networking applications often use multi-hop wireless broadcasting as a primary data dissemination mechanism. Therefore, protocols that efficiently and thoroughly propagate application data while adapting to a wide range of network density, vehicle distribution pattern, channel quality, and other conditions are critical for vehicular communications. Here, we design the Statistical Vehicular Broadcast (SVB) protocol to efficiently distribute data via multi-hop broadcast in vehicular networks. First, we present an automated optimization technique for the design of threshold functions in statistical broadcasting methods. Next, we compare and analyze known statistical techniques, including different fundamental methods, assessment delay algorithms, and failsafe mechanisms. All combinations of these techniques are given threshold functions optimized using the proposed automated procedure then are evaluated in a wide range of simulations. High-level statistical design principles and recommendations are established based on analysis of these results. Finally, we apply those principles to design SVB. It is evaluated in JiST/SWANS and is shown to achieve a high target reachability level while consuming less bandwidth than similar protocols across urban and highway vehicular networking scenarios.Copyright (c) 2014 John Wiley & Sons, Ltd.