Abstract
In vehicular ad hoc networks, vehicles need to exchange their recent mobility information at a high rate to maintain network agility and to preserve the performance of applications. Unfortunately, a high broadcasting rate affects the performance of both network reliability and information accuracy. The aim of this paper is to reduce the broadcasting rate while preserving information accuracy. A Driving-Situation-Aware Adaptive Broadcasting Rate Scheme (DSA-ABR) is proposed based on effective mobility prediction algorithm operates in between message transmissions, to reduce the communication rate. The scheme contains two algorithms which are Self-Predictor and Neighboring-Predictor based on an adaptive version of the Extended Kalman Filter. Firstly, the Self-Predictor algorithm estimates the current mobility state, with the help of the previous mobility state and knowledge about the driving situation and measurement uncertainties. Individual driving situation prediction models are obtained online through training on historical data. A vehicle decides whether to send or omit the beacon messagesbased on the accuracy of the Self-Predictor. Secondly, the Neighbouring-Predictor algorithm predicts the omitted or lost beacon messages with the help of knowledge shared by the sender vehicles. The results show the effectiveness and the efficiencyof the proposed scheme under unreliable communication conditions.