A Methodology for Studying VANET Performance with Practical Vehicle Distribution in Urban Environment

TL;DR

This paper introduces a novel methodology using stochastic traffic modeling to analyze VANET performance with realistic vehicle distributions, enabling better protocol optimization in urban environments.

Contribution

The paper presents an original stochastic traffic model for VANETs that accounts for non-homogeneous vehicle distribution and derives analytical performance metrics.

Findings

Accurate analytical throughput and progress estimates for different routing strategies.

Validation of the model through extensive simulations.

Optimal transmission probabilities can be computed based on vehicle location data.

Abstract

In a Vehicular Ad-hoc Network (VANET), the amount of interference from neighboring nodes to a communication link is governed by the vehicle density dynamics in vicinity and transmission probabilities of terminals. It is obvious that vehicles are distributed non-homogeneously along a road segment due to traffic controls and speed limits at different portions of the road. The common assumption of homogeneous node distribution in the network in most of the previous work in mobile ad-hoc networks thus appears to be inappropriate in VANETs. In light of the inadequacy, we present in this paper an original methodology to study the performance of VANETs with practical vehicle distribution in urban environment. Specifically, we introduce the stochastic traffic model to characterize the general vehicular traffic flow as well as the randomness of individual vehicles, from which we can acquire the mean dynamics and the probability distribution of vehicular density. As illustrative examples, we demonstrate how the density knowledge from the stochastic traffic model can be utilized to derive the throughput and progress performance of three routing strategies in different channel access protocols. We confirm the accuracy of the analytical results through extensive simulations. Our results demonstrate the applicability of the proposed methodology on modeling protocol performance, and shed insight into the performance analysis of other transmission protocols and network configurations in vehicular networks. Furthermore, we illustrate that the optimal transmission probability for optimized network performance can be obtained as a function of the location space from our results. Such information can be computed by road-side nodes and then broadcasted to road users for optimized multi-hop packet transmission in the communication network.