(User behavior)-based LS prediction model and anti-LS reverse-leadership strategy in the (VLC/RF)-V2X network

Abstract

The hybrid vehicle to everything (V2X) network based on visible light communication (VLC) and radio frequency (RF) is an excellent choice for future vehicle communication. In this paper, the (user behavior)-based longitudinal-separation (LS) prediction model and the reverse-leadership strategy in the hybrid V2X network are firstly proposed. By user behavior, the LS is divided into active separation (ASP) and passive separation (PSP). By characteristics of ASP and PSP, the longitudinal-separation prediction model is constructed. By this model, ASP and PSP can be predicted respectively before they occur. By the prediction result of the model, a reverse-leadership strategy for the platoon is presented. By this strategy, communication links can be self-adaptively changed to avoid outages caused by LS. In this scheme (model＋strategy), LS can be predicted timely and communication links can be handed over in advance. Finally, the validity of the model and the effectiveness of the scheme are demonstrated by simulations based on data from NGSIM data set and PTV-Vissim simulation software. Besides, the proposed scheme not only has its comparative advantages of fewer outages and handovers, but also can decrease the path loss, average outage time and handover delay significantly.