Algoritmos de agrupación para flujos de datos en entornos centralizados y distribuidos

Los avances en el hardware permiten disponer de grandes volúmenes de datos, surgiendo aplicaciones que deben suministrar información en tiempo cuasi-real, la monitorización de pacientes, ej., el seguimiento sanitario de las conducciones de agua, etc. Las necesidades de estas aplicaciones hacen emerger el modelo de flujo de datos (data streaming) frente al modelo almacenar-para-despuésprocesar (store-then-process). Mientras que en el modelo store-then-process, los datos son almacenados para ser posteriormente consultados; en los sistemas de streaming, los datos son procesados a su llegada al sistema, produciendo respuestas continuas sin llegar a almacenarse. Esta nueva visión impone desafíos para el procesamiento de datos al vuelo: 1) las respuestas deben producirse de manera continua cada vez que nuevos datos llegan al sistema; 2) los datos son accedidos solo una vez y, generalmente, no son almacenados en su totalidad; y 3) el tiempo de procesamiento por dato para producir una respuesta debe ser bajo. Aunque existen dos modelos para el cómputo de respuestas continuas, el modelo evolutivo y el de ventana deslizante; éste segundo se ajusta mejor en ciertas aplicaciones al considerar únicamente los datos recibidos más recientemente, en lugar de todo el histórico de datos. En los últimos años, la minería de datos en streaming se ha centrado en el modelo evolutivo. Mientras que, en el modelo de ventana deslizante, el trabajo presentado es más reducido ya que estos algoritmos no sólo deben de ser incrementales si no que deben borrar la información que caduca por el deslizamiento de la ventana manteniendo los anteriores tres desafíos. Una de las tareas fundamentales en minería de datos es la búsqueda de agrupaciones donde, dado un conjunto de datos, el objetivo es encontrar grupos representativos, de manera que se tenga una descripción sintética del conjunto. Estas agrupaciones son fundamentales en aplicaciones como la detección de intrusos en la red o la segmentación de clientes en el marketing y la publicidad. Debido a las cantidades masivas de datos que deben procesarse en este tipo de aplicaciones (millones de eventos por segundo), las soluciones centralizadas puede ser incapaz de hacer frente a las restricciones de tiempo de procesamiento, por lo que deben recurrir a descartar datos durante los picos de carga. Para evitar esta perdida de datos, se impone el procesamiento distribuido de streams, en concreto, los algoritmos de agrupamiento deben ser adaptados para este tipo de entornos, en los que los datos están distribuidos. En streaming, la investigación no solo se centra en el diseño para tareas generales, como la agrupación, sino también en la búsqueda de nuevos enfoques que se adapten mejor a escenarios particulares. Como ejemplo, un mecanismo de agrupación ad-hoc resulta ser más adecuado para la defensa contra la denegación de servicio distribuida (Distributed Denial of Services, DDoS) que el problema tradicional de k-medias. En esta tesis se pretende contribuir en el problema agrupamiento en streaming tanto en entornos centralizados y distribuidos. Hemos diseñado un algoritmo centralizado de clustering mostrando las capacidades para descubrir agrupaciones de alta calidad en bajo tiempo frente a otras soluciones del estado del arte, en una amplia evaluación. Además, se ha trabajado sobre una estructura que reduce notablemente el espacio de memoria necesario, controlando, en todo momento, el error de los cómputos. Nuestro trabajo también proporciona dos protocolos de distribución del cómputo de agrupaciones. Se han analizado dos características fundamentales: el impacto sobre la calidad del clustering al realizar el cómputo distribuido y las condiciones necesarias para la reducción del tiempo de procesamiento frente a la solución centralizada. Finalmente, hemos desarrollado un entorno para la detección de ataques DDoS basado en agrupaciones. En este último caso, se ha caracterizado el tipo de ataques detectados y se ha desarrollado una evaluación sobre la eficiencia y eficacia de la mitigación del impacto del ataque. ABSTRACT Advances in hardware allow to collect huge volumes of data emerging applications that must provide information in near-real time, e.g., patient monitoring, health monitoring of water pipes, etc. The data streaming model emerges to comply with these applications overcoming the traditional store-then-process model. With the store-then-process model, data is stored before being consulted; while, in streaming, data are processed on the fly producing continuous responses. The challenges of streaming for processing data on the fly are the following: 1) responses must be produced continuously whenever new data arrives in the system; 2) data is accessed only once and is generally not maintained in its entirety, and 3) data processing time to produce a response should be low. Two models exist to compute continuous responses: the evolving model and the sliding window model; the latter fits best with applications must be computed over the most recently data rather than all the previous data. In recent years, research in the context of data stream mining has focused mainly on the evolving model. In the sliding window model, the work presented is smaller since these algorithms must be incremental and they must delete the information which expires when the window slides. Clustering is one of the fundamental techniques of data mining and is used to analyze data sets in order to find representative groups that provide a concise description of the data being processed. Clustering is critical in applications such as network intrusion detection or customer segmentation in marketing and advertising. Due to the huge amount of data that must be processed by such applications (up to millions of events per second), centralized solutions are usually unable to cope with timing restrictions and recur to shedding techniques where data is discarded during load peaks. To avoid discarding of data, processing of streams (such as clustering) must be distributed and adapted to environments where information is distributed. In streaming, research does not only focus on designing for general tasks, such as clustering, but also in finding new approaches that fit bests with particular scenarios. As an example, an ad-hoc grouping mechanism turns out to be more adequate than k-means for defense against Distributed Denial of Service (DDoS). This thesis contributes to the data stream mining clustering technique both for centralized and distributed environments. We present a centralized clustering algorithm showing capabilities to discover clusters of high quality in low time and we provide a comparison with existing state of the art solutions. We have worked on a data structure that significantly reduces memory requirements while controlling the error of the clusters statistics. We also provide two distributed clustering protocols. We focus on the analysis of two key features: the impact on the clustering quality when computation is distributed and the requirements for reducing the processing time compared to the centralized solution. Finally, with respect to ad-hoc grouping techniques, we have developed a DDoS detection framework based on clustering.We have characterized the attacks detected and we have evaluated the efficiency and effectiveness of mitigating the attack impact.

Defending against cybercrime: advances in the detection of malicious servers and the analysis of client-side vulnerabilities

Esta tesis se centra en el análisis de dos aspectos complementarios de la ciberdelincuencia (es decir, el crimen perpetrado a través de la red para ganar dinero). Estos dos aspectos son las máquinas infectadas utilizadas para obtener beneficios económicos de la delincuencia a través de diferentes acciones (como por ejemplo, clickfraud, DDoS, correo no deseado) y la infraestructura de servidores utilizados para gestionar estas máquinas (por ejemplo, C & C, servidores explotadores, servidores de monetización, redirectores). En la primera parte se investiga la exposición a las amenazas de los ordenadores victimas. Para realizar este análisis hemos utilizado los metadatos contenidos en WINE-BR conjunto de datos de Symantec. Este conjunto de datos contiene metadatos de instalación de ficheros ejecutables (por ejemplo, hash del fichero, su editor, fecha de instalación, nombre del fichero, la versión del fichero) proveniente de 8,4 millones de usuarios de Windows. Hemos asociado estos metadatos con las vulnerabilidades en el National Vulnerability Database (NVD) y en el Opens Sourced Vulnerability Database (OSVDB) con el fin de realizar un seguimiento de la decadencia de la vulnerabilidad en el tiempo y observar la rapidez de los usuarios a remiendar sus sistemas y, por tanto, su exposición a posibles ataques. Hemos identificado 3 factores que pueden influir en la actividad de parches de ordenadores victimas: código compartido, el tipo de usuario, exploits. Presentamos 2 nuevos ataques contra el código compartido y un análisis de cómo el conocimiento usuarios y la disponibilidad de exploit influyen en la actividad de aplicación de parches. Para las 80 vulnerabilidades en nuestra base de datos que afectan código compartido entre dos aplicaciones, el tiempo entre el parche libera en las diferentes aplicaciones es hasta 118 das (con una mediana de 11 das) En la segunda parte se proponen nuevas técnicas de sondeo activos para detectar y analizar las infraestructuras de servidores maliciosos. Aprovechamos técnicas de sondaje activo, para detectar servidores maliciosos en el internet. Empezamos con el análisis y la detección de operaciones de servidores explotadores. Como una operación identificamos los servidores que son controlados por las mismas personas y, posiblemente, participan en la misma campaña de infección. Hemos analizado un total de 500 servidores explotadores durante un período de 1 año, donde 2/3 de las operaciones tenían un único servidor y 1/2 por varios servidores. Hemos desarrollado la técnica para detectar servidores explotadores a diferentes tipologías de servidores, (por ejemplo, C & C, servidores de monetización, redirectores) y hemos logrado escala de Internet de sondeo para las distintas categorías de servidores maliciosos. Estas nuevas técnicas se han incorporado en una nueva herramienta llamada CyberProbe. Para detectar estos servidores hemos desarrollado una novedosa técnica llamada Adversarial Fingerprint Generation, que es una metodología para generar un modelo único de solicitud-respuesta para identificar la familia de servidores (es decir, el tipo y la operación que el servidor apartenece). A partir de una fichero de malware y un servidor activo de una determinada familia, CyberProbe puede generar un fingerprint válido para detectar todos los servidores vivos de esa familia. Hemos realizado 11 exploraciones en todo el Internet detectando 151 servidores maliciosos, de estos 151 servidores 75% son desconocidos a bases de datos publicas de servidores maliciosos. Otra cuestión que se plantea mientras se hace la detección de servidores maliciosos es que algunos de estos servidores podrán estar ocultos detrás de un proxy inverso silente. Para identificar la prevalencia de esta configuración de red y mejorar el capacidades de CyberProbe hemos desarrollado RevProbe una nueva herramienta a través del aprovechamiento de leakages en la configuración de la Web proxies inversa puede detectar proxies inversos. RevProbe identifica que el 16% de direcciones IP maliciosas activas analizadas corresponden a proxies inversos, que el 92% de ellos son silenciosos en comparación con 55% para los proxies inversos benignos, y que son utilizado principalmente para equilibrio de carga a través de múltiples servidores. ABSTRACT In this dissertation we investigate two fundamental aspects of cybercrime: the infection of machines used to monetize the crime and the malicious server infrastructures that are used to manage the infected machines. In the first part of this dissertation, we analyze how fast software vendors apply patches to secure client applications, identifying shared code as an important factor in patch deployment. Shared code is code present in multiple programs. When a vulnerability affects shared code the usual linear vulnerability life cycle is not anymore effective to describe how the patch deployment takes place. In this work we show which are the consequences of shared code vulnerabilities and we demonstrate two novel attacks that can be used to exploit this condition. In the second part of this dissertation we analyze malicious server infrastructures, our contributions are: a technique to cluster exploit server operations, a tool named CyberProbe to perform large scale detection of different malicious servers categories, and RevProbe a tool that detects silent reverse proxies. We start by identifying exploit server operations, that are, exploit servers managed by the same people. We investigate a total of 500 exploit servers over a period of more 13 months. We have collected malware from these servers and all the metadata related to the communication with the servers. Thanks to this metadata we have extracted different features to group together servers managed by the same entity (i.e., exploit server operation), we have discovered that 2/3 of the operations have a single server while 1/3 have multiple servers. Next, we present CyberProbe a tool that detects different malicious server types through a novel technique called adversarial fingerprint generation (AFG). The idea behind CyberProbe’s AFG is to run some piece of malware and observe its network communication towards malicious servers. Then it replays this communication to the malicious server and outputs a fingerprint (i.e. a port selection function, a probe generation function and a signature generation function). Once the fingerprint is generated CyberProbe scans the Internet with the fingerprint and finds all the servers of a given family. We have performed a total of 11 Internet wide scans finding 151 new servers starting with 15 seed servers. This gives to CyberProbe a 10 times amplification factor. Moreover we have compared CyberProbe with existing blacklists on the internet finding that only 40% of the server detected by CyberProbe were listed. To enhance the capabilities of CyberProbe we have developed RevProbe, a reverse proxy detection tool that can be integrated with CyberProbe to allow precise detection of silent reverse proxies used to hide malicious servers. RevProbe leverages leakage based detection techniques to detect if a malicious server is hidden behind a silent reverse proxy and the infrastructure of servers behind it. At the core of RevProbe is the analysis of differences in the traffic by interacting with a remote server.

Impact of denial of service solutions on network quality of service

The Internet has become a universal communication network tool. It has evolved from a platform that supports best-effort traffic to one that now carries different traffic types including those involving continuous media with quality of service (QoS) requirements. As more services are delivered over the Internet, we face increasing risk to their availability given that malicious attacks on those Internet services continue to increase. Several networks have witnessed denial of service (DoS) and distributed denial of service (DDoS) attacks over the past few years which have disrupted QoS of network services, thereby violating the Service Level Agreement (SLA) between the client and the Internet Service Provider (ISP). Hence DoS or DDoS attacks are major threats to network QoS. In this paper we survey techniques and solutions that have been deployed to thwart DoS and DDoS attacks and we evaluate them in terms of their impact on network QoS for Internet services. We also present vulnerabilities that can be exploited for QoS protocols and also affect QoS if exploited. In addition, we also highlight challenges that still need to be addressed to achieve end-to-end QoS with recently proposed DoS/DDoS solutions. © 2010 John Wiley & Sons, Ltd.

Threat Analysis of BlackEnergy Malware for Synchrophasor based Real-time Control and Monitoring in Smart Grid

The BlackEnergy malware targeting critical infrastructures has a long history. It evolved over time from a simple DDoS platform to a quite sophisticated plug-in based malware. The plug-in architecture has a persistent malware core with easily installable attack specific modules for DDoS, spamming, info-stealing, remote access, boot-sector formatting etc. BlackEnergy has been involved in several high profile cyber physical attacks including the recent Ukraine power grid attack in December 2015. This paper investigates the evolution of BlackEnergy and its cyber attack capabilities. It presents a basic cyber attack model used by BlackEnergy for targeting industrial control systems. In particular, the paper analyzes cyber threats of BlackEnergy for synchrophasor based systems which are used for real-time control and monitoring functionalities in smart grid. Several BlackEnergy based attack scenarios have been investigated by exploiting the vulnerabilities in two widely used synchrophasor communication standards: (i) IEEE C37.118 and (ii) IEC 61850-90-5. Specifically, the paper addresses reconnaissance, DDoS, man-in-the-middle and replay/reflection attacks on IEEE C37.118 and IEC 61850-90-5. Further, the paper also investigates protection strategies for detection and prevention of BlackEnergy based cyber physical attacks.

Threat analysis of IoT networks using artificial neural network intrusion detection system

The Internet of things (IoT) is still in its infancy and has attracted much interest in many industrial sectors including medical fields, logistics tracking, smart cities and automobiles. However, as a paradigm, it is susceptible to a range of significant intrusion threats. This paper presents a threat analysis of the IoT and uses an Artificial Neural Network (ANN) to combat these threats. A multi-level perceptron, a type of supervised ANN, is trained using internet packet traces, then is assessed on its ability to thwart Distributed Denial of Service (DDoS/DoS) attacks. This paper focuses on the classification of normal and threat patterns on an IoT Network. The ANN procedure is validated against a simulated IoT network. The experimental results demonstrate 99.4% accuracy and can successfully detect various DDoS/DoS attacks.

Active Data Collection Techniques to Understand Online Scammers and Cybercriminals

Nigerian scam, also known as advance fee fraud or 419 scam, is a prevalent form of online fraudulent activity that causes financial loss to individuals and businesses. Nigerian scam has evolved from simple non-targeted email messages to more sophisticated scams targeted at users of classifieds, dating and other websites. Even though such scams are observed and reported by users frequently, the community’s understanding of Nigerian scams is limited since the scammers operate “underground”. To better understand the underground Nigerian scam ecosystem and seek effective methods to deter Nigerian scam and cybercrime in general, we conduct a series of active and passive measurement studies. Relying upon the analysis and insight gained from the measurement studies, we make four contributions: (1) we analyze the taxonomy of Nigerian scam and derive long-term trends in scams; (2) we provide an insight on Nigerian scam and cybercrime ecosystems and their underground operation; (3) we propose a payment intervention as a potential deterrent to cybercrime operation in general and evaluate its effectiveness; and (4) we offer active and passive measurement tools and techniques that enable in-depth analysis of cybercrime ecosystems and deterrence on them. We first created and analyze a repository of more than two hundred thousand user-reported scam emails, stretching from 2006 to 2014, from four major scam reporting websites. We select ten most commonly observed scam categories and tag 2,000 scam emails randomly selected from our repository. Based upon the manually tagged dataset, we train a machine learning classifier and cluster all scam emails in the repository. From the clustering result, we find a strong and sustained upward trend for targeted scams and downward trend for non-targeted scams. We then focus on two types of targeted scams: sales scams and rental scams targeted users on Craigslist. We built an automated scam data collection system and gathered large-scale sales scam emails. Using the system we posted honeypot ads on Craigslist and conversed automatically with the scammers. Through the email conversation, the system obtained additional confirmation of likely scam activities and collected additional information such as IP addresses and shipping addresses. Our analysis revealed that around 10 groups were responsible for nearly half of the over 13,000 total scam attempts we received. These groups used IP addresses and shipping addresses in both Nigeria and the U.S. We also crawled rental ads on Craigslist, identified rental scam ads amongst the large number of benign ads and conversed with the potential scammers. Through in-depth analysis of the rental scams, we found seven major scam campaigns employing various operations and monetization methods. We also found that unlike sales scammers, most rental scammers were in the U.S. The large-scale scam data and in-depth analysis provide useful insights on how to design effective deterrence techniques against cybercrime in general. We study underground DDoS-for-hire services, also known as booters, and measure the effectiveness of undermining a payment system of DDoS Services. Our analysis shows that the payment intervention can have the desired effect of limiting cybercriminals’ ability and increasing the risk of accepting payments.