Distributed Intrusion Detection System for Computational Grids

Applying intrusion detection to the fast growing computational Grid environments improves the security which is considered to be the heart of this new field. Flexible cooperative distributed intrusion detection architecture is introduced that suits and benefits from the underlying computational Grid environment. The proposed architecture was tested using homogeneous distributed intrusion detection servers that use learning vector quantization neural network to detect the intrusion if occurred. The paper discusses the different parameters that may affect the proposed intrusion detection system showing and explaining their effects on the overall system performance.