Lorenzo Cavallaro

On the Limits of Information Flow Techniques for Malware Analysis and Containment

Taint-tracking is emerging as a general technique in software security to complement virtualization and static analysis. It has been applied for accurate detection of a wide range of attacks on benign software, as well as in malware defense. Although it is quite robust for tackling the former problem, application of taint analysis to untrusted (and potentially malicious) software is riddled with several difficulties that lead to gaping holes in defense. These holes arise not only due to the limitations of information flow analysis techniques, but also the nature of todays software architectures and distribution models. This paper highlights these problems using an array of simple but powerful evasion techniques that can easily defeat taint-tracking defenses. Given todays binary-based software distribution and deployment models, our results suggest that information flow techniques will be of limited use against future malware that has been designed with the intent of evading these defenses.

Phoenix: DGA-Based Botnet Tracking and Intelligence

Modern botnets rely on domain-generation algorithms (DGAs) to build resilient command-and-control infrastructures. Given the prevalence of this mechanism, recent work has focused on the analysis of DNS traffic to recognize botnets based on their DGAs. While previous work has concentrated on detection, we focus on supporting intelligence operations. We propose Phoenix, a mechanism that, in addition to telling DGA- and non-DGA-generated domains apart using a combination of string and IP-based features, characterizes the DGAs behind them, and, most importantly, finds groups of DGA-generated domains that are representative of the respective botnets. As a result, Phoenix can associate previously unknown DGA-generated domains to these groups, and produce novel knowledge about the evolving behavior of each tracked botnet. We evaluated Phoenix on 1,153,516 domains, including DGA-generated domains from modern, well-known botnets: without supervision, it correctly distinguished DGA- vs. non-DGA-generated domains in 94.8 percent of the cases, characterized families of domains that belonged to distinct DGAs, and helped researchers “on the field” in gathering intelligence on suspicious domains to identify the correct botnet.

Sandnet: network traffic analysis of malicious software

Dynamic analysis of malware is widely used to obtain a better understanding of unknown software. While existing systems mainly focus on host-level activities of malware and limit the analysis period to a few minutes, we concentrate on the network behavior of malware over longer periods. We provide a comprehensive overview of typical malware network behavior by discussing the results that we obtained during the analysis of more than 100,000 malware samples. The resulting network behavior was dissected in our new analysis environment called Sandnet that complements existing systems by focusing on network traffic analysis. Our in-depth analysis of the two protocols that are most popular among malware authors, DNS and HTTP, helps to understand and characterize the usage of these prevalent protocols.

PAriCheck: an efficient pointer arithmetic checker for C programs

Buffer overflows are still a significant problem in programs written in C and C++. In this paper we present a bounds checker, called PAriCheck, that inserts dynamic runtime checks to ensure that attackers are not able to abuse buffer overflow vulnerabilities. The main approach is based on checking pointer arithmetic rather than pointer dereferences when performing bounds checks. The checks are performed by assigning a unique label to each object and ensuring that the label is associated with each memory location that the object inhabits. Whenever pointer arithmetic occurs, the label of the base location is compared to the label of the resulting arithmetic. If the labels differ, an out-of-bounds calculation has occurred. Benchmarks show that PAriCheck has a very low performance overhead compared to similar bounds checkers. This paper demonstrates that using bounds checkers for programs or parts of programs running on high-security production systems is a realistic possibility.

LISABETH: automated content-based signature generator for zero-day polymorphic worms

Modern worms can spread so quickly that any countermeasure based on human reaction might not be fast enough. Recent research has focused on devising algorithms to automatically produce signature for polymorphic worms, required by Intrusion Detection Systems. However, polymorphic worms are more complex than non-mutating ones as they also require the identification of mutated instances. To this end, we propose Lisabeth, our improved version of Hamsa, an automated content-based signature generation system for polymorphic worms that uses invariant bytes analysis of network traffic content. We show an unknown attack to Hamsas signature generator that is contrasted by Lisabeth. Moreover, we show that our approach is able to generally improve the resilience to poisoning attacks as supported by our experiments with synthetic polymorphic worms.

Live and trustworthy forensic analysis of commodity production systems

We present HyperSleuth, a framework that leverages the virtualization extensions provided by commodity hardware to securely perform live forensic analysis of potentially compromised production systems. HyperSleuth provides a trusted execution environment that guarantees four fundamental properties. First, an attacker controlling the system cannot interfere with the analysis and cannot tamper the results. Second, the framework can be installed as the system runs, without a reboot and without loosing any volatile data. Third, the analysis performed is completely transparent to the OS and to an attacker. Finally, the analysis can be periodically and safely interrupted to resume normal execution of the system. On top of HyperSleuth we implemented three forensic analysis applications: a lazy physical memory dumper, a lie detector, and a system call tracer. The experimental evaluation we conducted demonstrated that even time consuming analysis, such as the dump of the content of the physical memory, can be securely performed without interrupting the services offered by the system.

Replay attack in TCG specification and solution

We prove the existence of a flaw which we individuated in the design of the object-independent authorization protocol (OIAP), which represents one of the building blocks of the trusted platform module (TPM), the core of the trusted computing platforms (TPs) as devised by the trusted computing group (TCG) standards. In particular, we prove, also with the support of a model checker, that the protocol is exposed to replay attacks, which could be used for compromising the correct behavior of a TP We also propose a countermeasure to undertake in order to avoid such an attack as well as any replay attacks to the aforementioned protocol

The Evolution of Android Malware and Android Analysis Techniques

With the integration of mobile devices into daily life, smartphones are privy to increasing amounts of sensitive information. Sophisticated mobile malware, particularly Android malware, acquire or utilize such data without user consent. It is therefore essential to devise effective techniques to analyze and detect these threats. This article presents a comprehensive survey on leading Android malware analysis and detection techniques, and their effectiveness against evolving malware. This article categorizes systems by methodology and date to evaluate progression and weaknesses. This article also discusses evaluations of industry solutions, malware statistics, and malware evasion techniques and concludes by supporting future research paths.

DroidScribe: Classifying Android Malware Based on Runtime Behavior

The Android ecosystem has witnessed a surge in malware, which not only puts mobile devices at risk but also increases the burden on malware analysts assessing and categorizing threats. In this paper, we show how to use machine learning to automatically classify Android malware samples into families with high accuracy, while observing only their runtime behavior. We focus exclusively on dynamic analysis of runtime behavior to provide a clean point of comparison that is dual to static approaches. Specific challenges in the use of dynamic analysis on Android are the limited information gained from tracking low-level events and the imperfect coverage when testing apps, e.g., due to inactive command and control servers. We observe that on Android, pure system calls do not carry enough semantic content for classification and instead rely on lightweight virtual machine introspection to also reconstruct Android-level inter-process communication. To address the sparsity of data resulting from low coverage, we introduce a novel classification method that fuses Support Vector Machines with Conformal Prediction to generate high-accuracy prediction sets where the information is insufficient to pinpoint a single family.

Understanding Android App Piggybacking: A Systematic Study of Malicious Code Grafting

The Android packaging model offers ample opportunities for malware writers to piggyback malicious code in popular apps, which can then be easily spread to a large user base. Although recent research has produced approaches and tools to identify piggybacked apps, the literature lacks a comprehensive investigation into such phenomenon. We fill this gap by: 1) systematically building a large set of piggybacked and benign apps pairs, which we release to the community; 2) empirically studying the characteristics of malicious piggybacked apps in comparison with their benign counterparts; and 3) providing insights on piggybacking processes. Among several findings providing insights analysis techniques should build upon to improve the overall detection and classification accuracy of piggybacked apps, we show that piggybacking operations not only concern app code, but also extensively manipulates app resource files, largely contradicting common beliefs. We also find that piggybacking is done with little sophistication, in many cases automatically, and often via library code.