Title: Improved predictive contention window-based broadcast optimisation strategy in vehicular cyber-physical system

Authors: Yan Huo; Yanfei Lu; Jianmin Ren; Tao Jing

Addresses: School of Electronics and Information Engineering, Beijing Jiaotong University, Beijing 100044, China ' School of Electronics and Information Engineering, Beijing Jiaotong University, Beijing 100044, China ' School of Electronics and Information Engineering, Beijing Jiaotong University, Beijing 100044, China ' School of Electronics and Information Engineering, Beijing Jiaotong University, Beijing 100044, China

Abstract: Vehicles in the vehicular cyber-physical system should avoid intense competition by adjusting contention windows of media access control (MAC) protocol. Accordingly, we propose an improved prediction contention window adjustment mechanism to optimise the broadcast performance. The set should be updated via a feedback mechanism, so as to solve the problem of changeable vehicular communication scenarios. After that, an improved Markov model is designed to analyse the backoff procedure and to obtain the window sizes related to the attribute set, for the purpose of creating the reference table by Bayesian personalised ranking algorithm. Besides, an estimation mechanism based on hidden Markov is provided to predict the following vehicle's state and to generate a mapping table that consists of the attribute set and size of the window. Comparing with traditional adjustment schemes, extensive simulations demonstrate that the proposed strategy could optimise the broadcast performance effectively in terms of reducing collision probability and transmission delay.

Keywords: vehicular CPS; broadcast; contention window; fuzzy inference; HMM; hidden Markov model.

DOI: 10.1504/IJSNET.2017.086972

International Journal of Sensor Networks, 2017 Vol.25 No.2, pp.115 - 125

Available online: 30 Sep 2017 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article