Rede definida por software para a detecção de anomalias e contramedidas de segurança em smart grid
Carregando...
Data
2018-03-01
Autores
Ferrari, Ricardo Cesar Câmara
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O trabalho propõe uma aplicação com o uso de desvio padrão para definir limites máximos e mínimos de pacotes e bytes para detecção de anomalias nos fluxos de comunicação entre mestre e escravos com o uso do protocolo DNP3 (Distributed Network Protocol v3.0) em uma Smart Grid, além de detecção e bloqueio de ataques originados de máquinas intrusas ou conhecidas. Atualmente, diversas pesquisas vêm sendo desenvolvidas sobre uso de sistemas Smart Grid, no entanto, sua implantação possui alguns fatores de risco. Esses fatores estão associados às redes de transmissão de dados, às tecnologias de aquisição e controle das informações, e às vulnerabilidades intrínsecas da união dessas tecnologias. A principal motivação dessa proposta origina-se da necessidade de se garantir segurança dos sistemas Smart Grid e o potencial apresentado pelas Redes Definidas por Software (Software Defined Networking – SDN) em analisar os fluxos de dados em um switch. Assim, a investigação dessas vulnerabilidades, bem como, a identificação de situações de ataques são relevantes a fim de propor soluções de defesa. Para isto, a tecnologia de SDN apresentou-se como uma solução viável e otimizada para a proteção de sistemas Smart Grid, permitindo que seja realizado um monitoramento dos fluxos entre mestre e escravos, e a coleta de informações para análise, abrindo oportunidades para aplicações de segurança em Smart Grid. Dessa forma, foram realizados três experimentos, o primeiro com o objetivo de mostrar a vulnerabilidade de uma Smart Grid, o segundo com o intuito de analisar uma aplicação SDN em uma Smart Grid e o terceiro com dois ataques DDoS (Distributed Denial of Service) em uma Smart Grid. O primeiro ataque a partir de máquinas intrusas e o segundo ataque, de escravas, permitindo analisar e monitorar o fluxo de dados e o bloqueio das portas em um Open vSwitch (OVS). Nesse contexto, um componente de um controlador SDN foi modificado para adicionar segurança e monitoramento da rede, tendo um comportamento satisfatório, identificando anomalias e conseguindo realizar bloqueios de portas das máquinas atacantes.
The work proposes an application with the use of standard deviation to define limits of maximum and minimum of packets and bytes for detection of anomalies in the communication flows between master and slave using the Distributed Network Protocol v3.0 (DNP3), besides the detection and blocking of attacks originated from intrusive or known machines. Currently several researches have been developed on the use of Smart Grid systems, however, its implementation has some risk factors. These factors are associated with data transmission networks, information acquisition and control technologies and intrinsic vulnerabilities of the union of these technologies. The main motivation of this proposal comes from the need to guarantee security of Smart Grid systems and the potential presented by Software Defined Networking (SDN). Thus, the investigation of these vulnerabilities, as well as, identification of situations of attacks are relevant in order to propose defense solutions. For this, the SDN technology has presented a viable and optimized solution for the protection of Smart Grid systems, allowing the monitoring of masterslave flows and the collection of information for analysis, opening opportunities for security applications in Smart Grid. In this way, three experiments were carried out, the first to show the vulnerability of an insecure Smart Grid, the second to analyze a SDN application in a Smart Grid and the third with two Distributed Denial of Service (DDoS) attack on a Smart Grid, the first from intrusive machines and the second from slaves, allowing to analyze and monitor the data flow and the lock of the doors in an Open vSwitch (OVS). In this context, a component of an SDN controller has been modified to add security and monitoring of the network, having a satisfactory behavior, identifying anomalies and being able to perform port blocking of the attacking machines.
The work proposes an application with the use of standard deviation to define limits of maximum and minimum of packets and bytes for detection of anomalies in the communication flows between master and slave using the Distributed Network Protocol v3.0 (DNP3), besides the detection and blocking of attacks originated from intrusive or known machines. Currently several researches have been developed on the use of Smart Grid systems, however, its implementation has some risk factors. These factors are associated with data transmission networks, information acquisition and control technologies and intrinsic vulnerabilities of the union of these technologies. The main motivation of this proposal comes from the need to guarantee security of Smart Grid systems and the potential presented by Software Defined Networking (SDN). Thus, the investigation of these vulnerabilities, as well as, identification of situations of attacks are relevant in order to propose defense solutions. For this, the SDN technology has presented a viable and optimized solution for the protection of Smart Grid systems, allowing the monitoring of masterslave flows and the collection of information for analysis, opening opportunities for security applications in Smart Grid. In this way, three experiments were carried out, the first to show the vulnerability of an insecure Smart Grid, the second to analyze a SDN application in a Smart Grid and the third with two Distributed Denial of Service (DDoS) attack on a Smart Grid, the first from intrusive machines and the second from slaves, allowing to analyze and monitor the data flow and the lock of the doors in an Open vSwitch (OVS). In this context, a component of an SDN controller has been modified to add security and monitoring of the network, having a satisfactory behavior, identifying anomalies and being able to perform port blocking of the attacking machines.
Descrição
Palavras-chave
Smart Grid, SDN, Segurança, POX, DDoS