Application Layer Modeling in Vehicle Networks: Cooperative Maneuver Use Case

TL;DR

This paper explores application layer modeling in vehicle networks, focusing on cooperative maneuvers for autonomous merging, and compares simulation tools to support upcoming 5G-enabled vehicular services.

Contribution

It extends research on application layer modeling in vehicle networks, evaluates current simulation environments, and demonstrates modeling of cooperative maneuvers using VSimRTI.

Findings

VSimRTI effectively models cooperative vehicle maneuvers.

Comparison of simulation tools highlights strengths and limitations.

Application layer modeling supports autonomous vehicle functionalities.

Abstract

In recent years, network function virtualization and a software defined focus has allowed networks to become flexible and extensible in ways not possible previously. Although network modeling tools such as NS-2, NS-3, and OMNet++ have been extended with modules and code to support the absolute latest wireless protocols and medium access standards - there has been a growing gap in simulation of layers above medium access which recent 5G use cases are designed to support. In this thesis, I extend research into the topic of application layer modeling in wireless networks, with focus on upcoming deployments of enhanced vehicle services. In particular, I begin in discussing the evolution of vehicle network standards and use cases, along with the most recent initiative `Project 5GCar`; funded by the European Commission. I go narrower, from the macro trend of vehicular network standards evolution, to compare the state of the art in simulation stacks, designed to support application models in vehicle networks. Within this comparison, simulation environments OVNIS, Veins, iTetris, and VSimRTI are compared for their capabilities. Finally, to measure the qualitative performance of application layer modeling in vehicle networks, I take the cooperative maneuver use case, presented under Project 5GCar; to design an autonomous merge algorithm - completing the steps required to program and model the application in vehicles using the VSimRTI simulation stack.