Biblio

Machine learning (ML) models are increasingly being used in the development of Malware Detection Systems. Existing research in this area primarily focuses on developing new architectures and feature representation techniques to improve the accuracy of the model. However, recent studies have shown that existing state-of-the art techniques are vulnerable to adversarial machine learning (AML) attacks. Among those, data poisoning attacks have been identified as a top concern for ML practitioners. A recent study on clean-label poisoning attacks in which an adversary intentionally crafts training samples in order for the model to learn a backdoor watermark was shown to degrade the performance of state-of-the-art classifiers. Defenses against such poisoning attacks have been largely under-explored. We investigate a recently proposed clean-label poisoning attack and leverage an ensemble-based Nested Training technique to remove most of the poisoned samples from a poisoned training dataset. Our technique leverages the relatively large sensitivity of poisoned samples to feature noise that disproportionately affects the accuracy of a backdoored model. In particular, we show that for two state-of-the art architectures trained on the EMBER dataset affected by the clean-label attack, the Nested Training approach improves the accuracy of backdoor malware samples from 3.42% to 93.2%. We also show that samples produced by the clean-label attack often successfully evade malware classification even when the classifier is not poisoned during training. However, even in such scenarios, our Nested Training technique can mitigate the effect of such clean-label-based evasion attacks by recovering the model's accuracy of malware detection from 3.57% to 93.2%.

Among many application domains of machine learning in real-world settings, cyber security can benefit from more automated techniques to combat sophisticated adversaries. Modern network intrusion detection systems leverage machine learning models on network logs to proactively detect cyber attacks. However, the risk of adversarial attacks against machine learning used in these cyber settings is not fully explored. In this paper, we investigate poisoning attacks at training time against machine learning models in constrained cyber environments such as network intrusion detection; we also explore mitigations of such attacks based on training data sanitization. We consider the setting of poisoning availability attacks, in which an attacker can insert a set of poisoned samples at training time with the goal of degrading the accuracy of the deployed model. We design a white-box, realizable poisoning attack that reduced the original model accuracy from 95% to less than 50 % by generating mislabeled samples in close vicinity of a selected subset of training points. We also propose a novel Nested Training method as a defense against these attacks. Our defense includes a diversified ensemble of classifiers, each trained on a different subset of the training set. We use the disagreement of the classifiers' predictions as a data sanitization method, and show that an ensemble of 10 SVM classifiers is resilient to a large fraction of poisoning samples, up to 30% of the training data.

Cyber reconnaissance is the process of gathering information about a target network for the purpose of compromising systems within that network. Network-based deception has emerged as a promising approach to disrupt attackers' reconnaissance efforts. However, limited work has been done so far on measuring the effectiveness of network-based deception. Furthermore, given that Software-Defined Networking (SDN) facilitates cyber deception by allowing network traffic to be modified and injected on-the-fly, understanding the effectiveness of employing different cyber deception strategies is critical. In this paper, we present a model to study the reconnaissance surface of a network and model the process of gathering information by attackers as interactions with a cyber defensive system that may use deception. To capture the evolution of the attackers' knowledge during reconnaissance, we design a belief system that is updated by using a Bayesian inference method. For the proposed model, we present two metrics based on KL-divergence to quantify the effectiveness of network deception. We tested the model and the two metrics by conducting experiments with a simulated attacker in an SDN-based deception system. The results of the experiments match our expectations, providing support for the model and proposed metrics.

Advanced targeted cyber attacks often rely on reconnaissance missions to gather information about potential targets and their location in a networked environment to identify vulnerabilities which can be exploited for further attack maneuvers. Advanced network scanning techniques are often used for this purpose and are automatically executed by malware infected hosts. In this paper we formally define network deception to defend reconnaissance and develop RDS (Reconnaissance Deception System), which is based on SDN (Software Defined Networking), to achieve deception by simulating virtual network topologies. Our system thwarts network reconnaissance by delaying the scanning techniques of adversaries and invalidating their collected information, while minimizing the performance impact on benign network traffic. We introduce approaches to defend malicious network discovery and reconnaissance in computer networks, which are required for targeted cyber attacks such as Advanced Persistent Threats (APT). We show, that our system is able to invalidate an attackers information, delay the process of finding vulnerable hosts and identify the source of adversarial reconnaissance within a network, while only causing a minuscule performance overhead of 0.2 milliseconds per packet flow on average.