Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Comparative Study of Simulators for Vehicular Networks

Rida Saghir, Thenuka Karunathilake, Anna Förster

TL;DR

The paper addresses the challenge of evaluating VANET simulators under realistic mobility and network conditions. It adopts SUMO for mobility and compares OMNeT++ (Veins+INET) and ns-3, including offline/online SUMO coupling for ns-3. Through Bremen city experiments, it reports memory usage, computational time, and packet delivery ratio, revealing trade-offs between memory, speed, and delivery reliability. The findings guide researchers in selecting simulators based on network scale, metrics of interest, and available tooling, emphasizing the impact of coupling and logging on performance.

Abstract

Vehicular Adhoc networks (VANETs) are composed of vehicles connected with wireless links to exchange data. VANETs have become the backbone of the Intelligent Transportation Systems (ITS) in smart cities and enable many essential services like roadside safety, traffic management, platooning, etc with vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications. In any form of research testing and evaluation plays a crucial role. However, in VANETs, real-world experiments require high investment, and heavy resources and can cause many practical difficulties. Therefore, simulations have become critical and the primary way of evaluating VANETs' applications. Furthermore, the upfront challenge is the realistic capture of the networking mechanism of VANETs, which varies from situation to situation. Several factors may contribute to the successful achievement of a random realistic networking behavior. However, the biggest dependency is a powerful tool for the implementation, which could probably take into account all the configuration parameters, loss factors, mobility schemes, and other key features of a VANET, yet give out practical performance metrics with a good trade-off between investment of resources and the results. Hence, the aim of this research is to evaluate some simulators in the scope of VANETs with respect to resource utilization, packet delivery, and computational time.

Comparative Study of Simulators for Vehicular Networks

TL;DR

The paper addresses the challenge of evaluating VANET simulators under realistic mobility and network conditions. It adopts SUMO for mobility and compares OMNeT++ (Veins+INET) and ns-3, including offline/online SUMO coupling for ns-3. Through Bremen city experiments, it reports memory usage, computational time, and packet delivery ratio, revealing trade-offs between memory, speed, and delivery reliability. The findings guide researchers in selecting simulators based on network scale, metrics of interest, and available tooling, emphasizing the impact of coupling and logging on performance.

Abstract

Vehicular Adhoc networks (VANETs) are composed of vehicles connected with wireless links to exchange data. VANETs have become the backbone of the Intelligent Transportation Systems (ITS) in smart cities and enable many essential services like roadside safety, traffic management, platooning, etc with vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications. In any form of research testing and evaluation plays a crucial role. However, in VANETs, real-world experiments require high investment, and heavy resources and can cause many practical difficulties. Therefore, simulations have become critical and the primary way of evaluating VANETs' applications. Furthermore, the upfront challenge is the realistic capture of the networking mechanism of VANETs, which varies from situation to situation. Several factors may contribute to the successful achievement of a random realistic networking behavior. However, the biggest dependency is a powerful tool for the implementation, which could probably take into account all the configuration parameters, loss factors, mobility schemes, and other key features of a VANET, yet give out practical performance metrics with a good trade-off between investment of resources and the results. Hence, the aim of this research is to evaluate some simulators in the scope of VANETs with respect to resource utilization, packet delivery, and computational time.
Paper Structure (16 sections, 5 figures, 2 tables)

This paper contains 16 sections, 5 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Selected Simulators
  4. SUMO
  5. OMNeT++
  6. OMNeT++ as a Vehicular Network Simulator
  7. ns-3
  8. ns-3 as a Vehicular Network Simulator
  9. Experiments and Performance Analysis
  10. Memory Usage
  11. Computational Time
  12. Packet Delivery Ratio
  13. Shortcomings Of The Simulators
  14. Problems with OMNET++
  15. Problems with ns-3
  16. ...and 1 more sections

Figures (5)

  • Figure 1: Available VANET Simulators with number of citations
  • Figure 2: SUMO Network used for the simulations
  • Figure 3: The amount of memory usage (%) vs the simulation duration (seconds)
  • Figure 4: The computational time (hours) vs the simulation duration (seconds)
  • Figure 5: Packet delivery ratio (%) vs the simulation duration (seconds)