Bayesian networks based reliable broadcast in vehicular networks

Reliability is a key requirement of multi-hop safety message broadcasting. DSRC/802.11p MAC layer has strict reliability requirement for ITS safety applications. In city environment, transmission in wireless medium is vulnerable to packet collisions and interferences. Cross channel interference is quite predominant in the presence of high rise buildings and concurrent transmissions. Achieving high reliability (e.g., 0.99) in the presence of all kinds of communication vulnerabilities is a major challenge in urban environments. To the best of our knowledge, none of the existing emergency message dissemination schemes, in the literature, achieves high reliability in lossy channel. In this paper, we propose a novel scheme that achieves high reliability for message dissemination while satisfying delay requirements, for various channel conditions. This paper proposes a new broadcast scheme, called Bayesian networks and unipolar orthogonal Code based Reliable multi-hop Broadcast (BCRB) to address this issue. Our objective is to guarantee strict reliability requirement (e.g., 99%) in each hop using broadcast repetitions. We propose an approach, based on using Bayesian networks, that exploits periodic beacons to accurately infer 802.11p link reception quality at each hop. Using this information, a sender determines an optimal number of repetitions, multiple forwarders and their positions. To combat interference, during broadcast repetitions, BCRB uses Uni-Polar Orthogonal Codes (UPOC). For multi-hop transmissions, multiple forwarders cooperatively communicate at each hop with the objective to achieve high broadcast reliability in next hop. Simulation results show that BCRB achieves very high reliability. Furthermore, BCRB satisfies delay requirements for time-sensitive vehicular applications with relatively low overhead. BCRB achieves 35-40% improvement in comparison to existing schemes.