[Paper Review] LIMoSim: A Lightweight and Integrated Approach for Simulating Vehicular Mobility with OMNeT++

Benjamin Sliwa, Johannes Pillmann+2 more|arXiv (Cornell University)|Sep 6, 2017

Vehicular Ad Hoc Networks (VANETs)5 references3 citations

TL;DR

LIMoSim is a lightweight, integrated vehicular mobility simulation framework for OMNeT++ that eliminates the need for external tools or interprocess communication by natively embedding microscopic mobility models like IDM and MOBIL within the network simulator. It enables efficient, accurate, and scalable simulation of vehicular dynamics and communication in VANETs and LTE scenarios using real-world OSM map data, with demonstrated feasibility in an LTE handover evaluation showing realistic mobility and signal behavior.

ABSTRACT

Reliable and efficient communication is one of the key requirements for the deployment of self-driving cars. Consequently, researchers and developers require efficient and precise tools for the parallel development of vehicular mobility and communication. Although current state-of-the-art approaches allow the coupled simulation of those two components, they are making use of multiple specialized simulators that are synchronized using interprocess communication, resulting in highly complex simulation setups. Furthermore, the compatibility of those simulators requires constant attention as they are developed independently. In this paper, we present a lightweight and integrated approach for simulating vehicular mobility directly in Objective Modular Network Testbed in C++ (OMNeT++) and INET without the need for external tools or Interprocess Communication (IPC). The proposed framework Lightweight ICT-centric Mobility Simulation (LIMoSim) is available as Open Source software and can easily be combined with other third-party extension frameworks for providing vehicular mobility based on well-known microscopical models. In contrast to existing approaches, the amount of necessary preprocessing steps for simulation setups is significantly reduced. The capabilities of LIMoSim are demonstrated by a proof of concept evaluation in combination with the Long Term Evolution (LTE) simulation framework SimuLTE.

Motivation & Objective

• To address the complexity and compatibility issues of coupling separate mobility and communication simulators in vehicular network research.
• To reduce preprocessing and simulation setup overhead by integrating microscopic mobility models directly into OMNeT++.
• To enable seamless integration with existing OMNeT++ frameworks like SimuLTE and INETMANET without additional configuration.
• To provide a lightweight, open-source alternative to SUMO for vehicular mobility simulation in network-centric research.
• To support future integration of strategic mobility models and live visualization of KPIs in simulation.

Proposed method

• LIMoSim is implemented as a native extension to the INET framework within OMNeT++, using the framework's modular architecture for transparent integration.
• It embeds established microscopical mobility models such as the Intelligent Driver Model (IDM) and MOBIL for lane-changing behavior directly into the simulation kernel.
• The framework uses OpenStreetMap (OSM) data for realistic urban road topologies and supports vector graphics export for visualization.
• Mobility parameters like position, velocity, acceleration, and RSSI are computed and exposed in real time to communication modules without IPC.
• The simulation setup is fully contained within OMNeT++, avoiding the need for TraCI or TCP-based synchronization with external tools like SUMO.
• The framework supports integration with LTE simulation stacks such as SimuLTE, enabling end-to-end evaluation of mobility-communication interactions.

Experimental results

Research questions

• RQ1 Can a lightweight, integrated mobility simulation framework be developed within OMNeT++ that eliminates the need for external simulators and IPC?
• RQ2 How does the integration of IDM and MOBIL models in OMNeT++ affect the realism and performance of vehicular network simulations?
• RQ3 To what extent can LIMoSim reduce simulation setup complexity compared to traditional SUMO-OMNeT++ coupling?
• RQ4 Can LIMoSim accurately reproduce realistic LTE handover behavior and signal dynamics in urban environments?
• RQ5 How does the framework support future extension with strategic mobility models and live KPI visualization?

Key findings

• LIMoSim successfully eliminates the need for interprocess communication and external tools like SUMO, simplifying simulation setup and improving maintainability.
• The framework enables accurate simulation of urban traffic dynamics, including frequent braking and accelerating, as reflected in velocity and acceleration traces.
• RSSI measurements show realistic signal decay with distance to serving eNBs, with four handovers observed during the trip, including a 'ping-pong' effect between eNBs 2 and 3.
• The simulation results confirm plausible LTE behavior, including handover events and signal strength variations, consistent with prior analytical models.
• The integration with SimuLTE demonstrates the feasibility of using LIMoSim in end-to-end LTE network evaluations without external synchronization.
• The framework supports future extension with strategic mobility models like SWIM and enables live visualization of key performance indicators through its UI component.