Biblio
Plenary Talk Our everyday life is more and more dependent on electronic communication and network connectivity. However, the threats of attacks and different types of misuse increase exponentially with the expansion of computer networks. In order to alleviate the problem and to identify malicious activities as early as possible Network Intrusion Detection Systems (NIDSs) have been developed and intensively investigated. Several approaches have been proposed and applied so far for these systems. It is a common challenge in this field that often there are no crisp boundaries between normal and abnormal network traffic, there are noisy or inaccurate data and therefore the investigated traffic could represent both attack and normal communication. Fuzzy logic based solutions could be advantageous owing to their capability to define membership levels in different classes and to do different operations with results ensuring reduced false positive and false negative classification compared to other approaches. In this presentation, after a short introduction of NIDSs a survey will be done on typical fuzzy logic based solutions followed by a detailed description of a fuzzy rule interpolation based IDS. The whole development process, i.e. data preprocessing, feature extraction, rule base generation steps are covered as well.
Internet of Things (IoT) is a revolutionary expandable network which has brought many advantages, improving the Quality of Life (QoL) of individuals. However, IoT carries dangers, due to the fact that hackers have the ability to find security gaps in users' IoT devices, which are not still secure enough and hence, intrude into them for malicious activities. As a result, they can control many connected devices in an IoT network, turning IoT into Botnet of Things (BoT). In a botnet, hackers can launch several types of attacks, such as the well known attacks of Distributed Denial of Service (DDoS) and Man in the Middle (MitM), and/or spread various types of malicious software (malware) to the compromised devices of the IoT network. In this paper, we propose a novel hybrid Artificial Intelligence (AI)-powered honeynet for enhanced IoT botnet detection rate with the use of Cloud Computing (CC). This upcoming security mechanism makes use of Machine Learning (ML) techniques like the Logistic Regression (LR) in order to predict potential botnet existence. It can also be adopted by other conventional security architectures in order to intercept hackers the creation of large botnets for malicious actions.
Botnets are one of the major threats on the Internet. They are used for malicious activities to compromise the basic network security goals, namely Confidentiality, Integrity, and Availability. For reliable botnet detection and defense, deep learning-based approaches were recently proposed. In this paper, four different deep learning models, namely Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), hybrid CNN-LSTM, and Multi-layer Perception (MLP) are applied for botnet detection and simulation studies are carried out using the CTU-13 botnet traffic dataset. We use several performance metrics such as accuracy, sensitivity, specificity, precision, and F1 score to evaluate the performance of each model on classifying both known and unknown (zero-day) botnet traffic patterns. The results show that our deep learning models can accurately and reliably detect both known and unknown botnet traffic, and show better performance than other deep learning models.
Cloud Storage Brokers (CSB) provide seamless and concurrent access to multiple Cloud Storage Services (CSS) while abstracting cloud complexities from end-users. However, this multi-cloud strategy faces several security challenges including enlarged attack surfaces, malicious insider threats, security complexities due to integration of disparate components and API interoperability issues. Novel security approaches are imperative to tackle these security issues. Therefore, this paper proposes CS-BAuditor, a novel cloud security system that continuously audits CSB resources, to detect malicious activities and unauthorized changes e.g. bucket policy misconfigurations, and remediates these anomalies. The cloud state is maintained via a continuous snapshotting mechanism thereby ensuring fault tolerance. We adopt the principles of chaos engineering by integrating BrokerMonkey, a component that continuously injects failure into our reference CSB system, CloudRAID. Hence, CSBAuditor is continuously tested for efficiency i.e. its ability to detect the changes injected by BrokerMonkey. CSBAuditor employs security metrics for risk analysis by computing severity scores for detected vulnerabilities using the Common Configuration Scoring System, thereby overcoming the limitation of insufficient security metrics in existing cloud auditing schemes. CSBAuditor has been tested using various strategies including chaos engineering failure injection strategies. Our experimental evaluation validates the efficiency of our approach against the aforementioned security issues with a detection and recovery rate of over 96 %.
Modern Browsers have become sophisticated applications, providing a portal to the web. Browsers host a complex mix of interpreters such as HTML and JavaScript, allowing not only useful functionality but also malicious activities, known as browser-hijacking. These attacks can be particularly difficult to detect, as they usually operate within the scope of normal browser behaviour. CryptoJacking is a form of browser-hijacking that has emerged as a result of the increased popularity and profitability of cryptocurrencies, and the introduction of new cryptocurrencies that promote CPU-based mining. This paper proposes MANiC (Multi-step AssessmeNt for Crypto-miners), a system to detect CryptoJacking websites. It uses regular expressions that are compiled in accordance with the API structure of different miner families. This allows the detection of crypto-mining scripts and the extraction of parameters that could be used to detect suspicious behaviour associated with CryptoJacking. When MANiC was used to analyse the Alexa top 1m websites, it detected 887 malicious URLs containing miners from 11 different families and demonstrated favourable results when compared to related CryptoJacking research. We demonstrate that MANiC can be used to provide insights into this new threat, to identify new potential features of interest and to establish a ground-truth dataset, assisting future research.
One of the effective ways of detecting malicious traffic in computer networks is intrusion detection systems (IDS). Though IDS identify malicious activities in a network, it might be difficult to detect distributed or coordinated attacks because they only have single vantage point. To combat this problem, cooperative intrusion detection system was proposed. In this detection system, nodes exchange attack features or signatures with a view of detecting an attack that has previously been detected by one of the other nodes in the system. Exchanging of attack features is necessary because a zero-day attacks (attacks without known signature) experienced in different locations are not the same. Although this solution enhanced the ability of a single IDS to respond to attacks that have been previously identified by cooperating nodes, malicious activities such as fake data injection, data manipulation or deletion and data consistency are problems threatening this approach. In this paper, we propose a solution that leverages blockchain's distributive technology, tamper-proof ability and data immutability to detect and prevent malicious activities and solve data consistency problems facing cooperative intrusion detection. Focusing on extraction, storage and distribution stages of cooperative intrusion detection, we develop a blockchain-based solution that securely extracts features or signatures, adds extra verification step, makes storage of these signatures and features distributive and data sharing secured. Performance evaluation of the system with respect to its response time and resistance to the features/signatures injection is presented. The result shows that the proposed solution prevents stored attack features or signature against malicious data injection, manipulation or deletion and has low latency.
The greatest threat towards securing the organization and its assets are no longer the attackers attacking beyond the network walls of the organization but the insiders present within the organization with malicious intent. Existing approaches helps to monitor, detect and prevent any malicious activities within an organization's network while ignoring the human behavior impact on security. In this paper we have focused on user behavior profiling approach to monitor and analyze user behavior action sequence to detect insider threats. We present an ensemble hybrid machine learning approach using Multi State Long Short Term Memory (MSLSTM) and Convolution Neural Networks (CNN) based time series anomaly detection to detect the additive outliers in the behavior patterns based on their spatial-temporal behavior features. We find that using Multistate LSTM is better than basic single state LSTM. The proposed method with Multistate LSTM can successfully detect the insider threats providing the AUC of 0.9042 on train data and AUC of 0.9047 on test data when trained with publically available dataset for insider threats.
The analysis of security-related event logs is an important step for the investigation of cyber-attacks. It allows tracing malicious activities and lets a security operator find out what has happened. However, since IT landscapes are growing in size and diversity, the amount of events and their highly different representations are becoming a Big Data challenge. Unfortunately, current solutions for the analysis of security-related events, so called Security Information and Event Management (SIEM) systems, are not able to keep up with the load. In this work, we propose a distributed SIEM platform that makes use of highly efficient distributed normalization and persists event data into an in-memory database. We implement the normalization on common distribution frameworks, i.e. Spark, Storm, Trident and Heron, and compare their performance with our custom-built distribution solution. Additionally, different tuning options are introduced and their speed advantage is presented. In the end, we show how the writing into an in-memory database can be tuned to achieve optimal persistence speed. Using the proposed approach, we are able to not only fully normalize, but also persist more than 20 billion events per day with relatively small client hardware. Therefore, we are confident that our approach can handle the load of events in even very large IT landscapes.
Detecting fake accounts (sybils) in online social networks (OSNs) is vital to protect OSN operators and their users from various malicious activities. Typical graph-based sybil detection (a mainstream methodology) assumes that sybils can make friends with only a limited (or small) number of honest users. However, recent evidences showed that this assumption does not hold in real-world OSNs, leading to low detection accuracy. To address this challenge, we explore users' activities to assist sybil detection. The intuition is that honest users are much more selective in choosing who to interact with than to befriend with. We first develop the social and activity network (SAN), a two-layer hyper-graph that unifies users' friendships and their activities, to fully utilize users' activities. We also propose a more practical sybil attack model, where sybils can launch both friendship attacks and activity attacks. We then design Sybil SAN to detect sybils via coupling three random walk-based algorithms on the SAN, and prove the convergence of Sybil SAN. We develop an efficient iterative algorithm to compute the detection metric for Sybil SAN, and derive the number of rounds needed to guarantee the convergence. We use "matrix perturbation theory" to bound the detection error when sybils launch many friendship attacks and activity attacks. Extensive experiments on both synthetic and real-world datasets show that Sybil SAN is highly robust against sybil attacks, and can detect sybils accurately under practical scenarios, where current state-of-art sybil defenses have low accuracy.
In recent years, cyber security threats have become increasingly dangerous. Hackers have fabricated fake emails to spoof specific users into clicking on malicious attachments or URL links in them. This kind of threat is called a spear-phishing attack. Because spear-phishing attacks use unknown exploits to trigger malicious activities, it is difficult to effectively defend against them. Thus, this study focuses on the challenges faced, and we develop a Cloud-threat Inspection Appliance (CIA) system to defend against spear-phishing threats. With the advantages of hardware-assisted virtualization technology, we use the CIA to develop a transparent hypervisor monitor that conceals the presence of the detection engine in the hypervisor kernel. In addition, the CIA also designs a document pre-filtering algorithm to enhance system performance. By inspecting PDF format structures, the proposed CIA was able to filter 77% of PDF attachments and prevent them from all being sent into the hypervisor monitor for deeper analysis. Finally, we tested CIA in real-world scenarios. The hypervisor monitor was shown to be a better anti-evasion sandbox than commercial ones. During 2014, CIA inspected 780,000 mails in a company with 200 user accounts, and found 65 unknown samples that were not detected by commercial anti-virus software.
The modern malware poses serious security threats because of its evolved capability of using staged and persistent attack while remaining undetected over a long period of time to perform a number of malicious activities. The challenge for malicious actors is to gain initial control of the victim's machine by bypassing all the security controls. The most favored bait often used by attackers is to deceive users through a trusting or interesting email containing a malicious attachment or a malicious link. To make the email credible and interesting the cybercriminals often perform reconnaissance activities to find background information on the potential target. To this end, the value of information found on the discarded or stolen storage devices is often underestimated or ignored. In this paper, we present the partial results of analysis of one such hard disk that was purchased from the open market. The data found on the disk contained highly sensitive personal and organizational data. The results from the case study will be useful in not only understanding the involved risk but also creating awareness of related threats.
The significant dependence on cyberspace has indeed brought new risks that often compromise, exploit and damage invaluable data and systems. Thus, the capability to proactively infer malicious activities is of paramount importance. In this context, inferring probing events, which are commonly the first stage of any cyber attack, render a promising tactic to achieve that task. We have been receiving for the past three years 12 GB of daily malicious real darknet data (i.e., Internet traffic destined to half a million routable yet unallocated IP addresses) from more than 12 countries. This paper exploits such data to propose a novel approach that aims at capturing the behavior of the probing sources in an attempt to infer their orchestration (i.e., coordination) pattern. The latter defines a recently discovered characteristic of a new phenomenon of probing events that could be ominously leveraged to cause drastic Internet-wide and enterprise impacts as precursors of various cyber attacks. To accomplish its goals, the proposed approach leverages various signal and statistical techniques, information theoretical metrics, fuzzy approaches with real malware traffic and data mining methods. The approach is validated through one use case that arguably proves that a previously analyzed orchestrated probing event from last year is indeed still active, yet operating in a stealthy, very low rate mode. We envision that the proposed approach that is tailored towards darknet data, which is frequently, abundantly and effectively used to generate cyber threat intelligence, could be used by network security analysts, emergency response teams and/or observers of cyber events to infer large-scale orchestrated probing events for early cyber attack warning and notification.
The Domain Name System (DNS) is widely seen as a vital protocol of the modern Internet. For example, popular services like load balancers and Content Delivery Networks heavily rely on DNS. Because of its important role, DNS is also a desirable target for malicious activities such as spamming, phishing, and botnets. To protect networks against these attacks, a number of DNS-based security approaches have been proposed. The key insight of our study is to measure the effectiveness of security approaches that rely on DNS in large-scale networks. For this purpose, we answer the following questions, How often is DNS used? Are most of the Internet flows established after contacting DNS? In this study, we collected data from the University of Auckland campus network with more than 33,000 Internet users and processed it to find out how DNS is being used. Moreover, we studied the flows that were established with and without contacting DNS. Our results show that less than 5 percent of the observed flows use DNS. Therefore, we argue that those security approaches that solely depend on DNS are not sufficient to protect large-scale networks.
Machine learning (ML) plays a central role in the solution of many security problems, for example enabling malicious and innocent activities to be rapidly and accurately distinguished and appropriate actions to be taken. Unfortunately, a standard assumption in ML - that the training and test data are identically distributed - is typically violated in security applications, leading to degraded algorithm performance and reduced security. Previous research has attempted to address this challenge by developing ML algorithms which are either robust to differences between training and test data or are able to predict and account for these differences. This paper adopts a different approach, developing a class of moving target (MT) defenses that are difficult for adversaries to reverse-engineer, which in turn decreases the adversaries' ability to generate training/test data differences that benefit them. We leverage the coevolutionary relationship between attackers and defenders to derive a simple, flexible MT defense strategy which is optimal or nearly optimal for a broad range of security problems. Case studies involving two distinct cyber defense applications demonstrate that the proposed MT algorithm outperforms standard static methods, offering effective defense against intelligent, adaptive adversaries.