Traffic hazard prediction based on neighbor nodes for vehicle safety communications on a highway

Abstract

Vehicle safety communications is an important technology for protecting against automobile accidents. The number of neighbor nodes is important in the automobile industry, which is becoming increasingly more customer-oriented. For this, many protocols that forward a safe message have been studied to protect against chain-reaction collisions when an automobile accident occurs. In this paper, we propose a traffic hazard prediction model based on neighbor nodes for vehicle safety communications on a highway. Highways have high potential of 'chain-reaction collisions' due to the high velocity of vehicles. As a preventive method against chain-reaction collisions, native and intelligent broadcasting sends emergency messages to rear vehicles. However, these methods are ineffective when vehicles are concentrated in one area. This paper offers a broadcasting method considering stopping distance that features an improved back-off algorithm based on the number of neighbor nodes. In comparing the two different methods, our broadcasting method showed a 7 % improvement in frame reception success rate than native and intelligent broadcasting. Finally, this paper suggests empirical applications to verify the adequacy and validity of this system.