Tracking user activities inside an enterprise network has been a fundamental building block for today's security infrastructure, as it provides accurate user profiling and helps security auditors to make informed decisions based on the derived insights from the abundant log data. Towards more accurate user tracking, we propose a novel paradigm named UTrack by leveraging rich system-level audit logs. From a holistic perspective, we bridge the semantic gap between user accounts and real users, tracking a real user's activities across different user accounts and different network hosts based on causal relationship among processes. To achieve better scalability and a more salient view, we apply a variety of data reduction and compression techniques to process the large amount of data. %and significantly reduce the data volume. We implement UTrack in a real enterprise environment consisting of 111 hosts, which generate more than 4 billion events in total during the experiment time of one month. Through our evaluation, we demonstrate that UTrack is able to accurately identify the events that are relevant to user activities. Our data reduction and compression modules largely reduce the output data size, producing a both accurate and salient overview on a user session profile.

Analyzing the DNS traffic of Internet hosts has been a successful technique to counter cyberattacks and identify connections to malicious domains. However, recent stealthy attacks hide malicious activities within seemingly legitimate connections to popular web services made by benign programs. Traditional DNS monitoring and signature-based detection techniques are ineffective against such attacks. To tackle this challenge, we present a new program-level approach that can effectively detect such stealthy attacks. Our method builds a fine-grained Program-DNS profile for each benign program that characterizes what should be the “expected” DNS behavior. We find that malware-injected processes have DNS activities which significantly deviate from the Program-DNS profile of the benign program. We then develop six novel features based on the Program-DNS profile, and evaluate the features on a dataset of over 130 million DNS requests collected from a real-world enterprise and 8 million requests from malware-samples executed in a sandbox environment. We compare our detection results with that of previously-proposed features and demonstrate that our new features successfully detect 190 malware-injected processes which fail to be detected by previously-proposed features. Overall, our study demonstrates that fine-grained Program-DNS profiles can provide meaningful and effective features in building detectors for attack campaigns that bypass existing detection systems.

Setuid system calls enable critical functions such as user authentications and modular privileged components. Such operations must only be executed after careful validation. However, current systems do not perform rigorous checks, allowing exploitation of privileges through memory corruption vulnerabilities in privileged programs. As a solution, understanding which setuid system calls can be invoked in what context of a process allows precise enforcement of least privileges. We propose a novel comprehensive method to systematically extract and enforce least privilege of setuid system calls to prevent misuse. Our approach learns the required process contexts of setuid system calls along multiple dimensions: process hierarchy, call stack, and parameter in a process-aware way. Every setuid system call is then restricted to the per-process context by our kernel-level context enforcer. Previous approaches without process-awareness are too coarse-grained to control setuid system calls, resulting in over-privilege. Our method reduces available privileges even for identical code depending on whether it is run by a parent or a child process. We present our prototype called PoLPer which systematically discovers only required setuid system calls and effectively prevents real-world exploits targeting vulnerabilities of the setuid family of system calls in popular desktop and server software at near zero overhead.

Today's enterprises are exposed to sophisticated attacks, such as Advanced Persistent Threats~(APT) attacks, which usually consist of stealthy multiple steps. To counter these attacks, enterprises often rely on causality analysis on the system activity data collected from a ubiquitous system monitoring to discover the initial penetration point, and from there identify previously unknown attack steps. However, one major challenge for causality analysis is that the ubiquitous system monitoring generates a colossal amount of data and hosting such a huge amount of data is prohibitively expensive. Thus, there is a strong demand for techniques that reduce the storage of data for causality analysis and yet preserve the quality of the causality analysis. To address this problem, in this paper, we propose NodeMerge, a template based data reduction system for online system event storage. Specifically, our approach can directly work on the stream of system dependency data and achieve data reduction on the read-only file events based on their access patterns. It can either reduce the storage cost or improve the performance of causality analysis under the same budget. Only with a reasonable amount of resource for online data reduction, it nearly completely preserves the accuracy for causality analysis. The reduced form of data can be used directly with little overhead. To evaluate our approach, we conducted a set of comprehensive evaluations, which show that for different categories of workloads, our system can reduce the storage capacity of raw system dependency data by as high as 75.7 times, and the storage capacity of the state-of-the-art approach by as high as 32.6 times. Furthermore, the results also demonstrate that our approach keeps all the causality analysis information and has a reasonably small overhead in memory and hard disk.

Intrusive multi-step attacks, such as Advanced Persistent Threat (APT) attacks, have plagued enterprises with significant financial losses and are the top reason for enterprises to increase their security budgets. Since these attacks are sophisticated and stealthy, they can remain undetected for years if individual steps are buried in background "noise." Thus, enterprises are seeking solutions to "connect the suspicious dots" across multiple activities. This requires ubiquitous system auditing for long periods of time, which in turn causes overwhelmingly large amount of system audit events. Given a limited system budget, how to efficiently handle ever-increasing system audit logs is a great challenge. This paper proposes a new approach that exploits the dependency among system events to reduce the number of log entries while still supporting high-quality forensic analysis. In particular, we first propose an aggregation algorithm that preserves the dependency of events during data reduction to ensure the high quality of forensic analysis. Then we propose an aggressive reduction algorithm and exploit domain knowledge for further data reduction. To validate the efficacy of our proposed approach, we conduct a comprehensive evaluation on real-world auditing systems using log traces of more than one month. Our evaluation results demonstrate that our approach can significantly reduce the size of system logs and improve the efficiency of forensic analysis without losing accuracy.

Account compromization is a serious threat to users of online social networks (OSNs). While relentless spammers exploit the established trust relationships between account owners and their friends to efficiently spread malicious spam, timely detection of compromised accounts is quite challenging due to the well established trust relationship between the service providers, account owners, and their friends. In this paper, we study the social behaviors of OSN users, i.e., their usage of OSN services, and the application of which in detecting the compromised accounts. In particular, we propose a set of social behavioral features that can effectively characterize the user social activities on OSNs. We validate the efficacy of these behavioral features by collecting and analyzing real user clickstreams to an OSN website. Based on our measurement study, we devise individual user's social behavioral profile by combining its respective behavioral feature metrics. A social behavioral profile accurately reflects a user's OSN activity patterns. While an authentic owner conforms to its account's social behavioral profile involuntarily, it is hard and costly for impostors to feign. We evaluate the capability of the social behavioral profiles in distinguishing different OSN users, and our experimental results show the social behavioral profiles can accurately differentiate individual OSN users and detect compromised accounts.

Reducing attack surface is an effective preventive measure to strengthen security in large systems. However, it is challenging to apply this idea in an enterprise environment where systems are complex and evolving over time. In this paper, we empirically analyze and measure a real enterprise to identify unused services that expose attack surface. Interestingly, such unused services are known to exist and summarized by security best practices, yet such solutions require significant manual effort.

Computer system monitoring generates huge amounts of logs that record the interaction of system entities. How to query such data to better understand system behaviors and identify potential system risks and malicious behaviors becomes a challenging task for system administrators due to the dynamics and heterogeneity of the data. System monitoring data are essentially heterogeneous temporal graphs with nodes being system entities and edges being their interactions over time. Given the complexity of such graphs, it becomes time-consuming for system administrators to manually formulate useful queries in order to examine abnormal activities, attacks, and vulnerabilities in computer systems. In this work, we investigate how to query temporal graphs and treat query formulation as a discriminative temporal graph pattern mining problem. We introduce TGMiner to mine discriminative patterns from system logs, and these patterns can be taken as templates for building more complex queries. TGMiner leverages temporal information in graphs to prune graph patterns that share similar growth trend without compromising pattern quality. Experimental results on real system data show that TGMiner is 6-32 times faster than baseline methods. The discovered patterns were verified by system experts; they achieved high precision (97%) and recall (91%).

Privacy and information security in general are major concerns that impede enterprise adaptation of shared or public cloud computing. Specifically, the concern of virtual machine (VM) physical co-residency stems from the threat that hostile tenants can leverage various forms of side channels (such as cache covert channels) to exfiltrate sensitive information of victims on the same physical system. However, on virtualized x86 systems, covert channel attacks have not yet proven to be practical, and thus the threat is widely considered a “potential risk.” In this paper, we present a novel covert channel attack that is capable of high-bandwidth and reliable data transmission in the cloud. We first study the application of existing cache channel techniques in a virtualized environment and uncover their major insufficiency and difficulties. We then overcome these obstacles by: (1) redesigning a pure timing-based data transmission scheme, and (2) exploiting the memory bus as a high-bandwidth covert channel medium. We further design and implement a robust communication protocol and demonstrate realistic covert channel attacks on various virtualized x86 systems. Our experimental results show that covert channels do pose serious threats to information security in the cloud. Finally, we discuss our insights on covert channel mitigation in virtualized environments.

With the ever-increasing number and complexity of applications deployed in data centers, the underlying network infrastructure can no longer sustain such a trend and exhibits several problems, such as resource fragmentation and low bisection bandwidth. In pursuit of a real-world applicable cloud network (CN) optimization approach that continuously maintains balanced network performance with high cost effectiveness, we design a topology independent resource allocation and optimization approach, NetDEO. Based on a swarm intelligence optimization model, NetDEO improves the scalability of the CN by relocating virtual machines (VMs) and matching resource demand and availability. NetDEO is capable of (1) incrementally optimizing an existing VM placement in a data center; (2) deriving optimal deployment plans for newly added VMs; and (3) providing hardware upgrade suggestions, and allowing the CN to evolve as the workload changes over time. We evaluate the performance of NetDEO using realistic workload traces and simulated large-scale CN under various topologies.

An enormous number of apps have been developed for Android in recent years, making it one of the most popular mobile operating systems. However, the quality of the booming apps can be a concern [4]. Poorly engineered apps may contain security vulnerabilities that can severally undermine users' security and privacy. In this paper, we study a general category of vulnerabilities found in Android apps, namely the component hijacking vulnerabilities. Several types of previously reported app vulnerabilities, such as permission leakage, unauthorized data access, intent spoofing, and etc., belong to this category.

With the ever-increasing number and complexity of applications deployed in data centers, the underlying network infrastructure can no longer sustain such a trend and exhibits several problems, such as resource fragmentation and low bisection bandwidth. In pursuit of a real-world applicable data center network (DCN) optimization approach that continuously maintains balanced network performance with high cost effectiveness, we design a topology independent resource allocation and optimization approach, NetDEO. Based on a swarm intelligence optimization model, NetDEO improves the scalability of the DCN by relocating virtual machines (VMs) and matching resource demand and availability. NetDEO is capable of (1) incrementally optimizing an existing VM placement in a data center; (2) deriving optimal deployment plans for newly added VMs; and (3) providing hardware upgrade suggestions and allowing the DCN to evolve as the workload changes over time. We evaluate the performance of NetDEO using realistic workload traces and simulated large-scale DCN under various topologies.

Power management has become increasingly important for server systems. Numerous techniques have been proposed and developed to optimize server power consumption and achieve energy proportional computing. However, the security perspective of server power management has not yet been studied. In this paper, we investigate energy attacks, a new type of malicious exploits on server systems. Targeted solely at abusing server power consumption, energy attacks exhibit very different attacking behaviors and cause very different victim symptoms from conventional cyberspace attacks. First, we unveil that today's server systems with improved power saving technologies are more vulnerable to energy attacks. Then, we demonstrate a realistic energy attack on a stand-alone server system in three steps: 1) by profiling energy cost of an open web service under different operation conditions, we identify the vulnerabilities that subject a server to energy attacks; 2) exploiting the discovered attack vectors, we design an energy attack that can be launched anonymously from remote; and 3) we execute the attack and measure the extent of its damage in a systematic manner. Finally, we highlight the challenges in defending against energy attacks, and we propose an effective defense scheme to meet the challenges and evaluate its effectiveness.

The abuse of chat services by automated programs, known as chat bots, poses a serious threat to Internet users. Chat bots target popular chat networks to distribute spam and malware. In this paper, we first conduct a series of measurements on a large commercial chat network. Our measurements capture a total of 16 different types of chat bots ranging from simple to advanced. Moreover, we observe that human behavior is more complex than bot behavior. Based on the measurement study, we propose a classification system to accurately distinguish chat bots from human users. The proposed classification system consists of two components: 1) an entropy-based classifier; and 2) a Bayesian-based classifier. The two classifiers complement each other in chat bot detection. The entropy-based classifier is more accurate to detect unknown chat bots, whereas the Bayesian-based classifier is faster to detect known chat bots. Our experimental evaluation shows that the proposed classification system is highly effective in differentiating bots from humans.

This paper presents Swift, a packet filter for high-performance packet capture on commercial off-the-shelf hardware. The key features of the Swift include: 1) extremely low filter update latency for dynamic packet filtering, and 2) gigabits-per-second high-speed packet processing. Based on complex instruction set computer (CISC) instruction set architecture (ISA), Swift achieves the former with an instruction set design that avoids the need for compilation and security checking, and the latter by mainly utilizing single instruction, multiple data (SIMD). We implement Swift in the Linux 2.6 kernel for both i386 and x86_64 architectures and extensively evaluate its dynamic and static filtering performance on multiple machines with different hardware setups. We compare Swift to BPF (the BSD packet filter)-the de facto standard for packet filtering in modern operating systems-and hand-coded optimized C filters that are used for demonstrating possible performance gains. For dynamic filtering tasks, Swift is at least three orders of magnitude faster than BPF in terms of filter update latency. For static filtering tasks, Swift outperforms BPF up to three times in terms of packet processing speed and achieves much closer performance to the optimized C filters. We also show that Swift can harness the processing power of hardware SIMD instructions by virtue of its SIMD-capable instruction set.

Binary obfuscation plays an essential role in evading malware static analysis and detection. The widely used code obfuscation techniques, such as polymorphism and metamorphism, focus on evading syntax based detection. However, statistic test and semantic analysis techniques have been developed to thwart their evasion attempts. More recent binary obfuscation techniques are divided in their purposes of attacking either statistical or semantic approach, but not both. In this paper, we introduce mimimorphism, a novel binary obfuscation technique with the potential of evading both statistical and semantic detections. Mimimorphic malware uses instruction-syntax-aware high-order mimic functions to transform its binary into mimicry executables that exhibit high similarity to benign programs in terms of statistical properties and semantic characteristics. We implement a prototype of the mimimorphic engine on the Intel x86 platform, and evaluate its capability of evading statistical anomaly detection and semantic analysis detection techniques. Our experimental results demonstrate that the mimicry executables are indistinguishable from benign programs in terms of byte frequency distribution and entropy, as well as control flow fingerprint.

The abuse of online games by automated programs, known as game bots, for gaining unfair advantages has plagued millions of participating players with escalating severity in recent years. The current methods for distinguishing bots and humans are based on human interactive proofs (HIPs), such as CAPTCHAs. However, HIP-based approaches have inherent drawbacks. In particular, they are too obtrusive to be tolerated by human players in a gaming context. In this paper, we propose a non-interactive approach based on human observational proofs (HOPs) for continuous game bot detection. HOPs differentiate bots from human players by passively monitoring input actions that are difficult for current bots to perform in a human-like manner. We collect a series of user-input traces in one of the most popular online games, World of Warcraft. Based on the traces, we characterize the game playing behaviors of bots and humans. Then, we develop a HOP-based game bot defense system that analyzes user-input actions with a cascade-correlation neural network to distinguish bots from humans. The HOP system is effective in capturing current game bots, which raises the bar against game exploits and forces a determined adversary to build more complicated game bots for detection evasion in the future.

Instant messaging (IM) has been one of most frequently used malware attack vectors due to its popularity. Distinct from other malware, it is straightforward for IM malware to find and hit the next victim by exploiting the current victim's contact list and playing social engineering tricks. Thus, the spread of IM malware is much harder to detect and suppress through conventional approaches. The previous solutions are ineffective to defend against IM malware in an enterprise-like network environment, mainly because of high false positive rate and the requirement of the IM server being inside the protected network. In this paper, we propose a novel IM malware detection and suppression mechanism, HoneyIM, which guarantees almost zero false positive on detecting and blocking IM malware in an enterprise-like network. The detection of HoneyIM is based on the concept of honeypot. HoneyIM uses decoy accounts to trap IM malware by leveraging malware spreading characteristics. Fed with accurate detection results, the suppression of HoneyIM can conduct a network-wide blocking. In addition, HoneyIM delivers attack information to network administrators in real-time so that system quarantine and recovery can be quickly performed. The core design of HoneyIM is generic, and can be applied to the scenarios that either enterprise IM services or public IM services are used in the protected network. Based on open-source IM client Pidgin and client honeypot Capture, we build a prototype of HoneyIM and validate its efficacy through both simulations and real experiments. Our results show that HoneyIM provides effective protection against IM malware in enterprise-like networks.