Workshop "Netz- und Service-Infrastrukturen" Dieser Beitrag zum Workshop "Netz- und Service-Infrastrukturen" behandelt den Aufbau eines Network Intrusion Detection System im Campusnetz.

The intrusion detection system has been widely studied and deployed by researchers for providing better security to computer networks. The increasing of the attack volume and the dramatic advancement of the machine learning make the cooperation between the intrusion detection system and machine learning a hot topic and a promising solution for the cybersecurity. Machine learning usually involves the training process using huge amount of sample data. Since the huge input data may cause a negative effect on the training and detection performance of the machine learning model. Feature selection becomes a crucial technique to rule out the irrelevant and redundant features from the dataset. This study applied a feature selection approach that combines the advanced feature selection algorithms and attacks characteristic features to produce the optimal feature subset for the machine learning model in network intrusion detection. The optimal feature subset was created using the CSE-CIC-IDS2018 dataset, which is the most up-to-date benchmark dataset with comprehensive attack diversity and features. The result of the experiment was produced using machine learning models with decision tree classifier and analyzed with respect to the accuracy, precision, recall, and f1 score.

Network attacks remain a constant threat to organizations around the globe. Intrusion detection systems provide a vital piece of the protection needed in order to fend off these attacks. Machine learning has become a popular method for developing new anomaly-based intrusion detection systems, and in recent years, deep learning has followed suit. Additionally, data fusion is often applied to intrusion detection systems in research, most often in the form of feature reduction, which can improve the accuracy and training times of classifiers. Another less common form of data fusion is decision fusion, where the outputs of multipe classifiers are fused into a more reliable result. Recent research has produced some contradictory results regarding the efficiency of traditional machine learning algorithms compared to deep learning algorithms. This study aims to investigate this problemand provide some clarity about the relative performance of a selection of classifier algorithms, namely artificial neural network, long short-term memory and random forest. Furthermore, two feature selection methods, namely correlation coefficient method and principal component analysis, as well as one decision fusion method in D-S evidence theory are tested. The majority of the feature selection methods fail to increase the accuracy of the implemented models, although the accuracy is not drastically reduced. Among the individual classifiers, random forest shows the best performance, obtaining an accuracy of 87,87%. Fusing the results with D-S evidence theory further improves this result, obtaining an accuracy of 88,56%, and proves particularly useful for reducing the number of false positives.

In the recent years, connected surveillance systems have been witnessing an unprecedented evolution owing to the advancements in internet of things and deep learning technologies. However, vulnerabilities to various kinds of attacks both at the cyber network-level and at the physical worldlevel are also rising. This poses danger not only to the devices but also to human life and property. The goal of this thesis is to enhance the security of an internet of things, focusing on connected video-based surveillance systems, by proposing multiple novel solutions to address security issues at the cyber network-level and to defend such systems at the physical world-level. In order to enhance security at the cyber network-level, this thesis designs and develops solutions to detect network intrusions in an internet of things such as surveillance cameras. The first solution is a novel method for network flow features transformation, named TempoCode. It introduces a temporal codebook-based encoding of flow features based on capturing the key patterns of benign traffic in a learnt temporal codebook. The second solution takes an unsupervised learning-based approach and proposes four methods to build efficient and adaptive ensembles of neural networks-based autoencoders for intrusion detection in internet of things such as surveillance cameras. To address the physical world-level attacks, this thesis studies, for the first time to the best of our knowledge, adversarial patches-based attacks against a convolutional neural network (CNN)- based surveillance system designed for vehicle make and model recognition (VMMR). The connected video-based surveillance systems that are based on deep learning models such as CNNs are highly vulnerable to adversarial machine learning-based attacks that could trick and fool the surveillance systems. In addition, this thesis proposes and evaluates a lightweight defense solution called SIHFR to mitigate the impact of such adversarial-patches on CNN-based VMMR systems, leveraging the symmetry in vehicles’ face images. The experimental evaluations on recent realistic intrusion detection datasets prove the effectiveness of the developed solutions, in comparison to state-of-the-art, in detecting intrusions of various types and for different devices. Moreover, using a real-world surveillance dataset, we demonstrate the effectiveness of the SIHFR defense method which does not require re-training of the target VMMR model and adds only a minimal overhead. The solutions designed and developed in this thesis shall pave the way forward for future studies to develop efficient intrusion detection systems and adversarial attacks mitigation methods for connected surveillance systems such as VMMR.

This dissertation presents our work on network intrusion detection and intrusion sim- ulation. The work in intrusion detection consists of two different network anomaly-based approaches. The work in intrusion simulation introduces a model using explicit traffic gen- eration for the packet level traffic simulation. The process of anomaly detection is to first build profiles for the normal network activity and then mark any events or activities that deviate from the normal profiles as suspicious. Based on the different schemes of creating the normal activity profiles, we introduce two approaches for intrusion detection. The first one is a frequency-based approach which creates a normal frequency profile based on the periodical patterns existed in the time-series formed by the traffic. It aims at those attacks that are conducted by running pre-written scripts, which automate the process of attempting connections to various ports or sending packets with fabricated payloads, etc. The second approach builds the normal profile based on variations of connection-based behavior of each single computer. The deviations resulted from each individual computer are carried out by a weight assignment scheme and further used to build a weighted link graph representing the overall traffic abnormalities. The functionality of this system is of a distributed personal IDS system that also provides a centralized traffic analysis by graphical visualization. It provides a finer control over the internal network by focusing on connection-based behavior of each single computer. For network intrusion simulation, we explore an alternative method for network traffic simulation using explicit traffic generation. In particular, we build a model to replay the standard DARPA traffic data or the traffic data captured from a real environment. The replayed traffic data is mixed with the attacks, such as DOS and Probe attack, which can create apparent abnormal traffic flow patterns. With the explicit traffic generation, every packet that has ever been sent by the victim and attacker is formed in the simulation model and travels around strictly following the criteria of time and path that extracted from the real scenario. Thus, the model provides a promising aid in the study of intrusion detection techniques.

This dissertation addresses Internet background traffic generation and network intrusion detection. It is organized in two parts. Part one introduces a method to model realistic Internet background traffic and demonstrates how the models are used both in a simulation environment and in a lab environment. Part two introduces two different NID (Network Intrusion Detection) techniques and evaluates them using the modeled background traffic. To demonstrate the approach we modeled five major application layer protocols: HTTP, FTP, SSH, SMTP and POP3. The model of each protocol includes an empirical probability distribution plus estimates of application-specific parameters. Due to the complexity of the traffic, hybrid distributions (called mixture distributions) were sometimes required. The traffic models are demonstrated in two environments: NS-2 (a simulator) and HONEST (a lab environment). The simulation results are compared against the original captured data sets. Users of HONEST have the option of adding network attacks to the background. The dissertation also introduces two new template-based techniques for network intrusion detection. One is based on a template of autocorrelations of the investigated traffic, while the other uses a template of correlation integrals. Detection experiments have been performed on real traffic and attacks; the results show that the two techniques can achieve high detection probability and low false alarm in certain instances.

The Web, being the most popular component of the Internet, has been transformed from a static information-serving medium into a fully interactive platform. This platform has been used by developers to create web applications rivaling traditional desktop systems. Designing, developing and evaluating these applications require new or modified methodologies, techniques and tools because of the different characteristics they exhibit. This dissertation discusses two important areas for developing and evaluating these applications: security and data mining. In the security area, a survey using a process similar to the Goal Question Metric approach examines the properties of web application vulnerabilities. Using results from the survey, a white-box approach to identify web applications vulnerabilities is proposed. Although the approach eliminates vulnerabilities during the development process, it does not protect existing web applications that have not utilized the approach. Hence, an Anomaly-based Network Intrusion Detection System, called AIWAS, is introduced. AIWAS protects web applications through the analysis of interactions between the users and the web applications. These interactions are classified as either benign or malicious; malicious interactions are prevented from reaching the web applications under protection. In the data mining area, the method of reliability estimation from server logs is examined in detail. This examination reveals the fact that the session workload is currently obtained using a constant Session Timeout Threshold (STT) value. However, each website is unique and should have its own STT value. Hence, an initial model for estimating the STT is introduced to encourage future research on sessions to use a customized STT value per website. This research on the STT leads to a deeper investigation of the actual session workload unit. More specifically, the distributional properties of the session workload are re-examined to determine whether the session workload can be described as a heavy-tailed distribution. / Software Engineering and Intelligent Systems

No description available.

The use of Intrusion Detection Systems is a normal thing today in bigger companies, butthe solutions that are to be found in market is often too expensive for the smallercompany. Therefore, we saw the need in investigating if there is a more affordablesolution. In this report, we will show that it is possible to use low cost single boardcomputers as part of a bigger centralized Intrusion Detection System. To investigate this,we set up a test system including 2 Raspberry Pi 3 Model B, a cloud server and the use oftwo home networks, one with port mirroring implemented in firmware and the other withdedicated span port. The report will show how we set up the environment and the testingwe have done to prove that this is a working solution.

The tremendous increase in usage and complexity of modern communication and network systems connected to the Internet, places demands upon security management to protect organisations' sensitive data and resources from malicious intrusion. Malicious attacks by intruders and hackers exploit flaws and weakness points in deployed systems through several sophisticated techniques that cannot be prevented by traditional measures, such as user authentication, access controls and firewalls. Consequently, automated detection and timely response systems are urgently needed to detect abnormal activities by monitoring network traffic and system events. Network Intrusion Detection Systems (NIDS) and Network Intrusion Prevention Systems (NIPS) are technologies that inspect traffic and diagnose system behaviour to provide improved attack protection. The current implementation of intrusion detection systems (commercial and open-source) lacks the scalability to support the massive increase in network speed, the emergence of new protocols and services. Multi-giga networks have become a standard installation posing the NIDS to be susceptible to resource exhaustion attacks. The research focuses on two distinct problems for the NIDS: missing alerts due to packet loss as a result of NIDS performance limitations; and the huge volumes of generated alerts by the NIDS overwhelming the security analyst which makes event observation tedious. A methodology for analysing alerts using a proposed framework for alert correlation has been presented to provide the security operator with a global view of the security perspective. Missed alerts are recovered implicitly using a contextual technique to detect multi-stage attack scenarios. This is based on the assumption that the most serious intrusions consist of relevant steps that temporally ordered. The pre- and post- condition approach is used to identify the logical relations among low level alerts. The alerts are aggregated, verified using vulnerability modelling, and correlated to construct multi-stage attacks. A number of algorithms have been proposed in this research to support the functionality of our framework including: alert correlation, alert aggregation and graph reduction. These algorithms have been implemented in a tool called Multi-stage Attack Recognition System (MARS) consisting of a collection of integrated components. The system has been evaluated using a series of experiments and using different data sets i.e. publicly available datasets and data sets collected using real-life experiments. The results show that our approach can effectively detect multi-stage attacks. The false positive rates are reduced due to implementation of the vulnerability and target host information.