Optimisation methods for performance of communication interaction based on cooperative vehicle infrastructure system

Abstract

Vehicular ad-hoc network (VANET) is the main information distribution way in VANET. In this paper, the traditional handshake mechanism of enhanced distributed channel access (EDCA) model protocol in IEEE 802.11p and the request to send/clear to send (RTS/CTS) control frame were optimised, then a more reasonable and adaptive back off algorithm based on contention window was proposed, and it's called A-EDCA, and then the competition mechanism of multi communication mode based on adaptive neural network was proposed. After finishing the nonlinear neural network modelling and least mean square (LMS) learning algorithm, variety of traffic scenarios were built to verify and evaluate competition mechanism in OPNET Modeller. The simulation results show that A-EDCA can reduce the likelihood of wireless channel conflict effectively and has significant performance in term of transmission delay and throughput. The proposed competition mechanism of multi communication mode can mainly forecast the optimal communication mode accurately.