2nd International ICST Workshop on OMNeT++

Research Article

Modeling obstacles in INET/Mobility framework: motivation, integration, and performance

  • @INPROCEEDINGS{10.4108/ICST.SIMUTOOLS2009.5680,
        author={Hermann S. Lichte and Jannis Weide},
        title={Modeling obstacles in INET/Mobility framework: motivation, integration, and performance},
        proceedings={2nd International ICST Workshop on OMNeT++},
        publisher={ACM},
        proceedings_a={OMNET++},
        year={2010},
        month={5},
        keywords={Design Performance},
        doi={10.4108/ICST.SIMUTOOLS2009.5680}
    }
    
  • Hermann S. Lichte
    Jannis Weide
    Year: 2010
    Modeling obstacles in INET/Mobility framework: motivation, integration, and performance
    OMNET++
    ICST
    DOI: 10.4108/ICST.SIMUTOOLS2009.5680
Hermann S. Lichte1,*, Jannis Weide1,*
  • 1: University of Paderborn, Paderborn, Germany.
*Contact email: hermann.lichte@upb.de, jannis@upb.de

Abstract

Wireless network protocols are commonly evaluated through simulations. The achieved results may vary significantly with the modeled propagation environment. Protocols that use carrier sensing for collision avoidance may not operate properly in the presence of obstacles (e.g. buildings) that shield hosts from each other. These failures may not be recognized when using a long-term stochastic model (e.g. log-normal shadowing), leading to inadequate simulation results. Studying a simple carrier-sensing protocol, we find that when shielding effects are properly modeled, they increase collisions dramatically with increasing transmission power as opposed to the stochastic model. Thus, we formally introduce a model that describes shielding effects and which is still simple enough to be efficiently implemented. We discuss how such implementation looks like, using the Mobility Framework and the INET Framework for OMNeT++ as specific examples. We identify connection-specific caches to be crucial for the run-time performance of the model’s implementation.