1st International IEEE Conference on Pervasive Services

Research Article

Intrusion Detection with Neural Networks and Watermarking Techniques for MANET

  • @INPROCEEDINGS{10.1109/PERSER.2007.4283901,
        author={Aikaterini  Mitrokotsa and Nikos  Komnino and  Christos  Douligeris},
        title={Intrusion Detection with Neural Networks and Watermarking Techniques for MANET},
        proceedings={1st International IEEE Conference on Pervasive Services},
        publisher={IEEE},
        proceedings_a={ICPS},
        year={2007},
        month={8},
        keywords={Engines  Intrusion detection  Learning systems  Machine learning  Mobile ad hoc networks  Neural networks  Protection  Traffic control  Visualization  Watermarking},
        doi={10.1109/PERSER.2007.4283901}
    }
    
  • Aikaterini Mitrokotsa
    Nikos Komnino
    Christos Douligeris
    Year: 2007
    Intrusion Detection with Neural Networks and Watermarking Techniques for MANET
    ICPS
    IEEE
    DOI: 10.1109/PERSER.2007.4283901
Aikaterini Mitrokotsa1, Nikos Komnino2,*, Christos Douligeris3,*
  • 1: mitrokat@unipi.gr
  • 2: Athens Information Technology, 19002 Peania Attiki, Greece
  • 3: Department of Informatics, University of Piraeus, 80 Karaoli & Dimitriou Str. Piraeus, 18534, Greece
*Contact email: nkom@ait.edu.gr, cdoulig@unipi.gr

Abstract

In mobile ad hoc networks (MANET) specific Intrusion Detection Systems (IDSs) are needed to safeguard them since traditional intrusion prevention techniques are not sufficient in the protection of MANET. In this paper we present an intrusion detection engine based on neural networks combined with a protection method, which is based on watermarking techniques. We exploit the advantages of information visualization and machine learning techniques in order to achieve intrusion detection. Then, we authenticate the maps produced by the application of the intelligent techniques using a novel combined watermarking embedded method. The performance of the proposed model is evaluated under different traffic conditions, mobility patterns and visualization metrics, showing its high efficiency.