ABSTRACT
In recent years, there has been a trend of sensor networks being embedded into vehicular ad hoc networks (VANETs). For VANET, vehicles are equipped with sensors, and the output of such sensors is fed to various VANET devices. There are numerous benefits of VANETs: road safety, public service, driving experience, and vehicle business/entertainment.
Message delivery in mobile networks can be in the form of unicast, multicast, or broadcast. Broadcasting is preferable as it offers a simple technique to spread messages to all vehicles available in the VANET. Broadcasting messages may, however, lead to excessive redundancy, contention, and collision of messages, which are commonly known as a broadcast storm for VANET messages. Intelligent control of broadcast mechanisms may address such problems.
This chapter deals with various subjects related to VANET communication control and routing protocols. It discusses in detail various intelligence algorithms, known in the artificial intelligence field as classifiers, for controlling the VANET communication and solving the broadcast storm problem. The applications of Bayesian decision theory, Maximum-likelihood and Bayesian estimation, Nearest-neighbor estimation, support vector machine, and artificial neural network to improve rebroadcast selection problem are investigated. The chapter proves that the intelligent control of VANET communication works very well in reducing message delays compared with flooding technique and significantly reduces the message redundancy of dense vehicular networks.
