Detecção de intrusão em redes de computadores com estatísticas PDF e classificadores neurais

• Autor(a) principal: Jorge, Leandro Silva
• Data de Publicação: 2006
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações do Mackenzie
• Texto Completo: http://dspace.mackenzie.br/handle/10899/24349
• topic: detecção de intrusão; redes de computadores; função densidade de probabilidade; redes neurais; intrusion detection; computer networks; density probability function; neural networks; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA

This work is a study about architectures, methods and results on computer networks intrusion detection to avoid Denial-of-Services attacks. The main techniques used were: packet-oriented window observation, structured hierarchical processes, anomaly based network intrusion detection by monitoring several network traffic parameters simultaneously. The Probability Density Function (PDF) has been used and statistically compared it to normal behavior referenced functions using similarity metrics. The results have been combined into an anomaly status vector classified by a neural network classifier. Two data sets have been tested to measure the performance of the proposed approach: the OPNET simulations data and the DARPA98 intrusion detection evaluation data. The results showed reliable detection in DoS attacks with high accuracy and very low false alarm rates on all data sets.