Although cloud network flows are similar to conventional network flows in many ways, there are some major differences in their statistical characteristics. However, due to the lack of adequate public datasets, the proponents of many existing cloud intrusion detection systems (IDS) have relied on the DARPA dataset which was obtained by simulating a conventional network environment. In the current thesis, we show empirically that the DARPA dataset by failing to meet important statistical characteristics of real-world cloud traffic data centers is inadequate for evaluating cloud IDS. We analyze, as an alternative, a new public dataset collected through cooperation between our lab and a non-profit cloud service provider, which contains benign data and a wide variety of attack data. Furthermore, we present a new hypervisor-based cloud IDS using an instance-oriented feature model and supervised machine learning techniques. We investigate 3 different classifiers: Logistic Regression (LR), Random Forest (RF), and Support Vector Machine (SVM) algorithms. Experimental evaluation on a diversified dataset yields a detection rate of 92.08% and a false-positive rate of 1.49% for the random forest, the best performing of the three classifiers. / Graduate

Made available in DSpace on 2014-07-29T15:08:23Z (GMT). No. of bitstreams: 1 Dissertacao - Helio Junio Gomes.pdf: 1556447 bytes, checksum: 1316eb94a42a76a12c1062c722417d10 (MD5) Previous issue date: 2010-02-26 / Secondary networks are essential for the process of distributing electrical energy to the final consumer, and their performance is related to their design, the quality of materials and equipment employed, environmental conditions, damage caused by third parties, and maintenance services. In this context, the objective of the present work is to carry out feasibility studies of the compact secondary networks in the distribution system of CELG D, an electric power distribution concessionaire and a crucial element of the electrical system of the state of Goiás, Brazil. The development of this research took into consideration the initial investment, the operational cost and the performance of these networks. It can therefore be concluded that, although the cost of implementation of the compact network is approximately 14% higher than the conventional cost, its installation is extremely viable because, in addition to reducing the preventive maintenance services required for this network, there is an improvement in the continuity of electrical power supply, two factors that are fundamental in an increasingly competitive market. / As redes secundárias são fundamentais no processo de distribuir a energia elétrica ao consumidor final e o seu desempenho está relacionado à sua concepção, às qualidades dos materiais e dos equipamentos empregados, às condições ambientais, aos danos causados por terceiros e aos serviços de manutenção. Este trabalho tem como objetivo realizar estudos de viabilidade das redes secundárias compactas no sistema de distribuição da CELG D, por ser uma concessionária de energia elétrica de grande importância para o sistema elétrico do Estado de Goiás. No desenvolvimento desta pesquisa, consideraram-se o investimento inicial, o custo operacional e o desempenho dessas redes. Conclui-se que, apesar de o custo de implantação da rede compacta ser, aproximadamente, 14% maior do que o custo da convencional, sua instalação torna-se bastante viável, pois além da redução dos serviços de manutenção preventiva necessários para essa rede, há melhoria da continuidade no fornecimento de energia elétrica, fatores esses considerados fundamentais, num mercado cada vez mais competitivo.