Temporal complex networks modeling applied to vehicular ad-hoc networks

TL;DR

This paper introduces a temporal graph modeling approach for VANETs, demonstrating its superiority over traditional aggregated models in accurately capturing network dynamics and improving deployment strategies.

Contribution

It proposes the use of temporal graphs and measures for VANETs, showing their effectiveness through comparative analysis and simulation-based evaluation.

Findings

Temporal models outperform aggregated models in capturing network dynamics.

Using temporal measures improves RSU deployment coverage.

Temporal betweenness as a feature enhances genetic algorithm performance.

Abstract

VANETs solutions use aggregated graph representation to model the interaction among the vehicles and different aggregated complex network measures to quantify some topological characteristics. This modeling ignores the temporal interactions between the cars, causing loss of information or unrealistic behavior. This work proposes the use of both temporal graphs and temporal measures to model VANETs applications. To verify the viability of this model, we initially perform a comparative analysis between the temporal and aggregated modeling considering five different real datasets. This analysis shows that the aggregated model is inefficient in modeling the temporal aspects of networks. After that, we perform a network evaluation through a simulation by considering the impact of temporal modeling applied to the deployment of RSUs. First, we compare a solution based on our temporal modeling with a greedy algorithm based on an aggregated model to choose the positions of RSUs. In a scenario with 70 RSUs, we have 77% and 65% of coverage in the temporal and aggregated model (greedy algorithm), respectively. Second, we evaluate the use of aggregated and temporal measures applied as features in a genetic algorithm. The approach with temporal betweenness had the better result with 90% of the coverage area against 61% of aggregated one applied to the same scenario.