Improving Network Intrusion Detection Classifiers by Non-payload-Based Exploit-Independent Obfuscations: An Adversarial Approach

Abstract

Machine-learning based intrusion detection classifiers are able to detect unknown attacks, but at the same time they may be susceptible to evasion by obfuscation techniques. An adversary intruder which possesses a crucial knowledge about a protection system can easily bypass the detection module. The main objective of our work is to improve the performance capabilities of intrusion detection classifiers against such adversaries. To this end, we firstly propose several obfuscation techniques of remote attacks that are based on the modification of various properties of network connections; then we conduct a set of comprehensive experiments to evaluate the effectiveness of intrusion detection classifiers against obfuscated attacks. We instantiate our approach by means of a tool, based on NetEm and Metasploit, which implements our obfuscation operators on any TCP communication. This allows us to generate modified network traffic for machine learning experiments employing features for assessing network statistics and behavior of TCP connections. We perform evaluation on five classifiers: Gaussian Naïve Bayes, Gaussian Naïve Bayes with kernel density estimation, Logistic Regression, Decision Tree, and Support Vector Machines. Our experiments confirm the assumption that it is possible to evade the intrusion detection capability of all classifiers trained without prior knowledge about obfuscated attacks, causing an exacerbation of the TPR ranging from 7.8% to 66.8%. Further, when widening the training knowledge of the classifiers by a subset of obfuscated attacks, we achieve a significant improvement of the TPR by 4.21% -73.3%, while the FPR is deteriorated only slightly (0.1% -1.48%). Finally, we test the capability of an obfuscations-aware classifier to detect unknown obfuscated attacks, where we achieve over 90% detection rate on average for most of the obfuscations. intrusion detection suffers from undetected attacks such as zero-day attacks or polymorphism, enabling an exploit-code to avoid positive signature matching of the packet payload data. Therefore, researchers and developers are motivated to design new methods to detect various versions of the modified network attacks including the zero-day ones. These goals motivate the popularity of Anomaly Detection Systems (ADS) and also the classification approaches in the context 1 EAI Endorsed Transactions Preprint

ASNM Datasets: A Collection of Network Attacks for Testing of Adversarial Classifiers and Intrusion Detectors

“…A data sample of the dataset D tr refers to the vector of the network connection features, defined in Section II-B. Then, referring to [22] and [23], let X = V × Y be the space of labeled samples, where V represents the space of unlabeled samples and Y represents the space of possible labels. Let D tr = {x 1 , x 2 , .…”

“…For more experiments with this dataset, including trinominal and multi-nominal labels, detection of unknown obfuscations by a custom leave-one-out validation, and individual feature analysis, we refer the reader to [23] and [14].…”

“…) whose values are sensitive to the obfuscations include: t, size, ip of f , ip sum , tcp sum , tcp seq , tcp ack , tcp of f , tcp f lags , tcp win , tcp urp and data 12. The modifications of P c and P s of the connection c m can cause alteration of values of the original network connection features F m to new ones (see Section II-D).Then we built an obfuscation tool[23] that morphs network characteristics of a TCP connection at network and transport layers of the TCP/IP stack by applying one or a combination of several non-payload-based obfuscation techniques. Execution of direct communications (non-obfuscated…”

Evaluating Adversarial Learning on Different Types of Deep Learning-based Intrusion Detection Systems using min-max optimization

“…is the urgency to filter and reduce false alarms [7] developed to make better learning and processing for big data and a variety of attacks for future prediction. We overview deep neural networks in the following section.…”

“…[3] [7] [8] confirm that there are different growing types and methods of attacks that donates a severe challenge on vulnerabilities of DNN architecture design. The fact that the training of DNNs is based on data, classification tasks can be carefully manipulated by crafted perturbations called adversarial samples.…”