Estimation of Packet Loss Rate at Wireless Link of VANET--RPLE

Abstract

Vehicular Ad hoc Network (VANET), as a subclass of mobile Ad hoc Networks (MANETs), is a promising approach for traffic security, mobile internet and V2V communication. While the packet loss estimation of VANET impacts routing protocol and transmission control algorithm, it turns out to be an important issue. In this paper, we propose an estimation algorithm for packet loss on VANET, RPLE (Real time packet loss estimation), which uses few probe packets and enhance the estimation accuracy than traditional methods for packet loss on wireless networks. Due to node motion and complex radio environment, packet loss estimation in VANET is a hard problem. However, packet loss estimation impacts routing protocol and transmission control algorithm of VANET, so it is an important issue. Based on mass on-road measurements, we have obtained the general statistical principals of packet loss rate (PLR) by distances on VANET, and we use Gaussian Mixture Model (GMM) to present PLR probability density. In that case, to get real-time PLR by sending a few probe packets, RPLE use maximum a posteriori probability method to estimate the current PLR based on the GMM of PLR probability density and the probing result. In real urban road scenario, the experiment results indicate that RPLE can get more accurate PLR estimation by few probe packets immediately.