Safety Information Dissemination Path Prediction in High-Density Vehicular Wireless Networks

Abstract

Many safety applications needed safety warning information to be disseminated to all vehicles in an area quickly. A path prediction model was built from the fused evidences by evidence theory, in which the basic probability assignment function of the local load and the old path load could be achieved using fuzzy set. And the loads became the evidences of the dissemination path prediction respectively. The prediction result with the maximal confidence was regarded as the quasi-real-time dissemination path load. Simulations show that the average end-to-end delay of safety information under heavy load with path prediction is less than the half of the delay without path prediction. However, the average packets loss rate of two schemes is almost the same.