[Paper Review] Observations on OMNeT++ Real-Time Behaviour

Christina Obermaier, Christian Facchi|arXiv (Cornell University)|Sep 7, 2017

Vehicular Ad Hoc Networks (VANETs)6 references3 citations

TL;DR

This paper evaluates OMNeT++'s real-time capabilities for Hardware-in-the-Loop (HIL) testing of Vehicular Ad Hoc Networks (VANETs) using the Artery framework. It proposes a HIL testbed connecting simulations to real VANET hardware via software-defined radio, finding that while most VANET algorithms tolerate millisecond-level time losses, multi-hop communication requiring precise timing (e.g., SIFS-based ACKs) is severely impacted by simulation delays, limiting current HIL applicability for such scenarios.

ABSTRACT

OMNeT++ is a widely used platform for all types of network simulations. The open source simulation framework Artery can be used to perform Vehicular Ad Hoc Network (VANET) simulations. This paper presents an approach for connecting this simulation and real-world VANET hardware to extend the test range and investigates the real-time behaviour of the simulation. As a Device Under Test (DUT) depends on real-time data to perform properly, different simulation scenarios running different hardware setups are presented. Additionally, the paper deals with the impacts of real-time losses on the test run outcomes. Most time dependant algorithms like the duplicate packet detection do not need very accurate real-time data and thus could be verified using the presented approach. Otherwise, in some cases such as testing of multi-hop communication, accurate real time is crucial.

Motivation & Objective

• To assess whether OMNeT++ can provide real-time simulation data suitable for Hardware-in-the-Loop (HIL) testing of VANET devices.
• To investigate the impact of real-time losses in OMNeT++ simulations on the correctness of VANET protocol testing.
• To determine under which conditions real-time inaccuracies invalidate test results, especially for time-critical protocols.
• To propose a solution—using an SDR as a time-aware 802.11p proxy—to enable accurate multi-hop testing in HIL environments.
• To identify the limitations of current simulation setups for functional HIL testing of VANET hardware.

Proposed method

• Integrates Artery, a VANET simulation framework for OMNeT++, with real-world 802.11p hardware via a software-defined radio (SDR) gateway.
• Uses the TraCI interface to couple SUMO-based vehicle mobility with OMNeT++ simulation, enabling dynamic vehicle movement and message exchange.
• Employs a custom real-time scheduler (Algorithm 1) to monitor and log simulation-to-real-time deviations, flagging unacceptable delays.
• Tests multiple scenarios with varying numbers of vehicles (up to four) on different hardware to measure real-time performance.
• Proposes an SDR-based proxy that emulates MAC-layer behavior (e.g., sending ACKs within SIFS) based on simulation state to correct timing inaccuracies.
• Evaluates the impact of event clustering (simultaneous packet transmissions) on real-time loss due to high event density in single-core simulations.

Experimental results

Research questions

• RQ1 To what extent does OMNeT++ maintain real-time fidelity when simulating VANETs with dynamic vehicle mobility?
• RQ2 Which VANET protocols are sensitive to real-time losses, and which can tolerate millisecond-level timing deviations?
• RQ3 Can a software-defined radio proxy compensate for simulation timing inaccuracies to enable valid multi-hop communication testing?
• RQ4 What are the performance bottlenecks in OMNeT++ simulations that lead to real-time loss, especially under high event load?
• RQ5 How many vehicles can be realistically simulated in real time on current hardware, and what is the impact of hardware performance on timing accuracy?

Key findings

• OMNeT++ simulations using a single core exhibit significant real-time losses, especially when multiple events are triggered simultaneously.
• Most VANET protocols—such as duplicate packet detection, DENM expiry, and replay attack detection—are resilient to real-time losses of a few milliseconds.
• Multi-hop communication relying on precise timing, such as SIFS-based ACKs (32 µs), is highly sensitive to simulation delays and fails when OMNeT++ runs faster than real time.
• Real-time losses exceeding 1–2 seconds can invalidate DENM message processing, particularly in scenarios requiring timely response.
• The current simulation setup limits functional HIL testing to three to four vehicles depending on hardware, due to real-time performance constraints.
• A proposed SDR proxy that emulates MAC-layer behavior based on simulation state could enable accurate multi-hop testing, but requires integration with simulation state tracking.