Biblio

Hashing has been a widely-adopted technique for nearest neighbor search in large-scale image retrieval tasks. Recent research has shown that leveraging supervised information can lead to high quality hashing. However, the cost of annotating data is often an obstacle when applying supervised hashing to a new domain. Moreover, the results can suffer from the robustness problem as the data at training and test stage may come from different distributions. This paper studies the exploration of generating synthetic data through semi-supervised generative adversarial networks (GANs), which leverages largely unlabeled and limited labeled training data to produce highly compelling data with intrinsic invariance and global coherence, for better understanding statistical structures of natural data. We demonstrate that the above two limitations can be well mitigated by applying the synthetic data for hashing. Specifically, a novel deep semantic hashing with GANs (DSH-GANs) is presented, which mainly consists of four components: a deep convolution neural networks (CNN) for learning image representations, an adversary stream to distinguish synthetic images from real ones, a hash stream for encoding image representations to hash codes and a classification stream. The whole architecture is trained end-to-end by jointly optimizing three losses, i.e., adversarial loss to correct label of synthetic or real for each sample, triplet ranking loss to preserve the relative similarity ordering in the input real-synthetic triplets and classification loss to classify each sample accurately. Extensive experiments conducted on both CIFAR-10 and NUS-WIDE image benchmarks validate the capability of exploiting synthetic images for hashing. Our framework also achieves superior results when compared to state-of-the-art deep hash models.

Deep web, a hidden and encrypted network that crawls beneath the surface web today has become a social hub for various criminals who carry out their crime through the cyber space and all the crime is being conducted and hosted on the Deep Web. This research paper is an effort to bring forth various techniques and ways in which an internet user can be safe online and protect his privacy through anonymity. Understanding how user's data and private information is phished and what are the risks of sharing personal information on social media.

Towards advancing the use of big keys as a practical defense against key exfiltration, this paper provides efficiency improvements for cryptographic schemes in the bounded retrieval model (BRM). We identify probe complexity (the number of scheme accesses to the slow storage medium storing the big key) as the dominant cost. Our main technical contribution is what we call the large-alphabet subkey prediction lemma. It gives good bounds on the predictability under leakage of a random sequence of blocks of the big key, as a function of the block size. We use it to significantly reduce the probe complexity required to attain a given level of security. Together with other techniques, this yields security-preserving performance improvements for BRM symmetric encryption schemes and BRM public-key identification schemes.

The Internet is becoming more and more content-oriented. CDN (Content Distribution Networks) has been a popular architecture compatible with the current Internet, and a new revolutionary paradigm such as ICN (Information Centric Networking) has studied. One of the main components in both CDN and ICN is considering cache on network. Despite a surge of extensive use of cache in the current and future Internet architectures, analysis on the performance of general cache networks are still quite limited due to complex inter-plays among various components and thus analytical intractability. Due to mathematical tractability, we consider 'static' cache policies and study asymptotic delay performance of those policies in cache networks, in particular, focusing on the impact of heterogeneous content popularities and nodes' geographical 'importances' in caching policies. Furthermore, our simulation results suggest that they perform quite similarly as popular 'dynamic' policies such as LFU (Least-Frequently-Used) and LRU (Least-Recently-Used). We believe that our theoretical findings provide useful engineering implications such as when and how various factors have impact on caching performance.

The deployment of Software Defined Networking (SDN) and Network Functions Virtualization (NFV) technologies is increasing, with security as a recognized application driving adoption. However, despite the potential with SDN/NFV for automated and adaptive network security services, the controller interaction presents both a performance and scalability challenge, and a threat vector. To overcome the performance issue, stateful data-plane designs have been proposed. However, these solutions do not offer protection from SDN-specific attacks linked to necessary control functions such as link reconfiguration and switch identification. In this work, we leverage the OpenState framework to introduce state-based SDN security protection mechanisms. The extensions required for this design are presented with respect to an SDN configuration-based attack. The demonstration shows the ability of the SDN Configuration (CFG) security protection mechanism to support legitimate relocation requests and to protect against malicious connection attempts.

Advanced Persistent Threats are increasingly becoming one of the major concerns to many industries and organizations. Currently, there exists numerous articles and industrial reports describing various case studies of recent notable Advanced Persistent Threat attacks. However, these documents are expressed in natural language. This limits the efficient reusability of the threat intelligence information due to ambiguous nature of the natural language. In this article, we propose a model to formally represent Advanced Persistent Threats as multi-agent systems. Our model is inspired by the concepts of agent-oriented social modelling approaches, generally used for software security requirement analysis.

As a plethora of wearable devices are being introduced, significant concerns exist on the privacy and security of personal data stored on these devices. Expanding on recent works of using electrocardiogram (ECG) as a modality for biometric authentication, in this work, we investigate the possibility of using personal ECG signals as the individually unique source for physical unclonable function (PUF), which eventually can be used as the key for encryption and decryption engines. We present new signal processing and machine learning algorithms that learn and extract maximally different ECG features for different individuals and minimally different ECG features for the same individual over time. Experimental results with a large 741-subject in-house ECG database show that the distributions of the intra-subject (same person) Hamming distance of extracted ECG features and the inter-subject Hamming distance have minimal overlap. 256-b random numbers generated from the ECG features of 648 (out of 741) subjects pass the NIST randomness tests.

Fuzzing is a software testing technique that finds bugs by repeatedly injecting mutated inputs to a target program. Known to be a highly practical approach, fuzzing is gaining more popularity than ever before. Current research on fuzzing has focused on producing an input that is more likely to trigger a vulnerability. In this paper, we tackle another way to improve the performance of fuzzing, which is to shorten the execution time of each iteration. We observe that AFL, a state-of-the-art fuzzer, slows down by 24x because of file system contention and the scalability of fork() system call when it runs on 120 cores in parallel. Other fuzzers are expected to suffer from the same scalability bottlenecks in that they follow a similar design pattern. To improve the fuzzing performance, we design and implement three new operating primitives specialized for fuzzing that solve these performance bottlenecks and achieve scalable performance on multi-core machines. Our experiment shows that the proposed primitives speed up AFL and LibFuzzer by 6.1 to 28.9x and 1.1 to 735.7x, respectively, on the overall number of executions per second when targeting Google's fuzzer test suite with 120 cores. In addition, the primitives improve AFL's throughput up to 7.7x with 30 cores, which is a more common setting in data centers. Our fuzzer-agnostic primitives can be easily applied to any fuzzer with fundamental performance improvement and directly benefit large-scale fuzzing and cloud-based fuzzing services.

Robotic vehicles and especially autonomous robotic vehicles can be attractive targets for attacks that cross the cyber-physical divide, that is cyber attacks or sensory channel attacks affecting the ability to navigate or complete a mission. Detection of such threats is typically limited to knowledge-based and vehicle-specific methods, which are applicable to only specific known attacks, or methods that require computation power that is prohibitive for resource-constrained vehicles. Here, we present a method based on Bayesian Networks that can not only tell whether an autonomous vehicle is under attack, but also whether the attack has originated from the cyber or the physical domain. We demonstrate the feasibility of the approach on an autonomous robotic vehicle built in accordance with the Generic Vehicle Architecture specification and equipped with a variety of popular communication and sensing technologies. The results of experiments involving command injection, rogue node and magnetic interference attacks show that the approach is promising.

Botnets have been a serious threat to the Internet security. With the constant sophistication and the resilience of them, a new trend has emerged, shifting botnets from the traditional desktop to the mobile environment. As in the desktop domain, detecting mobile botnets is essential to minimize the threat that they impose. Along the diverse set of strategies applied to detect these botnets, the ones that show the best and most generalized results involve discovering patterns in their anomalous behavior. In the mobile botnet field, one way to detect these patterns is by analyzing the operation parameters of this kind of applications. In this paper, we present an anomaly-based and host-based approach to detect mobile botnets. The proposed approach uses machine learning algorithms to identify anomalous behaviors in statistical features extracted from system calls. Using a self-generated dataset containing 13 families of mobile botnets and legitimate applications, we were able to test the performance of our approach in a close-to-reality scenario. The proposed approach achieved great results, including low false positive rates and high true detection rates.

In this paper, we propose a technique to detect phishing attacks based on behavior of human when exposed to fake website. Some online users submit fake credentials to the login page before submitting their actual credentials. He/She observes the login status of the resulting page to check whether the website is fake or legitimate. We automate the same behavior with our application (FeedPhish) which feeds fake values into login page. If the web page logs in successfully, it is classified as phishing otherwise it undergoes further heuristic filtering. If the suspicious site passes through all heuristic filters then the website is classified as a legitimate site. As per the experimentation results, our application has achieved a true positive rate of 97.61%, true negative rate of 94.37% and overall accuracy of 96.38%. Our application neither demands third party services nor prior knowledge like web history, whitelist or blacklist of URLS. It is able to detect not only zero-day phishing attacks but also detects phishing sites which are hosted on compromised domains.

Crypto-ransomware is a challenging threat that ciphers a user's files while hiding the decryption key until a ransom is paid by the victim. This type of malware is a lucrative business for cybercriminals, generating millions of dollars annually. The spread of ransomware is increasing as traditional detection-based protection, such as antivirus and anti-malware, has proven ineffective at preventing attacks. Additionally, this form of malware is incorporating advanced encryption algorithms and expanding the number of file types it targets. Cybercriminals have found a lucrative market and no one is safe from being the next victim. Encrypting ransomware targets business small and large as well as the regular home user. This paper discusses ransomware methods of infection, technology behind it and what can be done to help prevent becoming the next victim. The paper investigates the most common types of crypto-ransomware, various payload methods of infection, typical behavior of crypto ransomware, its tactics, how an attack is ordinarily carried out, what files are most commonly targeted on a victim's computer, and recommendations for prevention and safeguards are listed as well.

Deep learning model has been widely studied and proven to achieve high accuracy in various pattern recognition tasks, especially in image recognition. However, due to its non-linear architecture and high-dimensional inputs, its ill-posedness [1] towards adversarial perturbations-small deliberately crafted perturbations on the input will lead to completely different outputs, has also attracted researchers' attention. This work takes the traffic sign recognition system on the self-driving car as an example, and aims at designing an additional mechanism to improve the robustness of the recognition system. It uses a machine learning model which learns the results of the deep learning model's predictions, with human feedback as labels and provides the credibility of current prediction. The mechanism makes use of both the input image and the recognition result as sample space, querying a human user the True/False of current classification result the least number of times, and completing the task of detecting adversarial attacks.

This paper proposes a method to detect two primary means of using the Domain Name System (DNS) for malicious purposes. We develop machine learning models to detect information exfiltration from compromised machines and the establishment of command & control (C&C) servers via tunneling. We validate our approach by experiments where we successfully detect a malware used in several recent Advanced Persistent Threat (APT) attacks [1]. The novelty of our method is its robustness, simplicity, scalability, and ease of deployment in a production environment.

With growing popularity of Android, it's attack surface has also increased. Prevalence of third party android marketplaces gives attackers an opportunity to plant their malicious apps in the mobile eco-system. To evade signature based detection, attackers often transform their malware, for instance, by introducing code level changes. In this paper we propose a lightweight static Permission Flow Graph (PFG) based approach to detect malware even when they have been transformed (obfuscated). A number of techniques based on behavioral analysis have also been proposed in the past; how-ever our interest lies in leveraging the permission framework alone to detect malware variants and transformations without considering behavioral aspects of a malware. Our proposed approach constructs Permission Flow Graph (PFG) for an Android App. Transformations performed at code level, often result in changing control flow, however, most of the time, the permission flow remains invariant. As a consequences, PFGs of transformed malware and non-transformed malware remain structurally similar as shown in this paper using state-of-the-art graph similarity algorithm. Furthermore, we propose graph based similarity metrics at both edge level and vertex level in order to bring forth the structural similarity of the two PFGs being compared. We validate our proposed methodology through machine learning algorithms. Results prove that our approach is successfully able to group together Android malware and its variants (transformations) together in the same cluster. Further, we demonstrate that our proposed approach is able to detect transformed malware with a detection accuracy of 98.26%, thereby ensuring that malicious Apps can be detected even after transformations.

A technique and algorithms for early detection of the started attack and subsequent blocking of malicious traffic are proposed. The primary separation of mixed traffic into trustworthy and malicious traffic was carried out using cluster analysis. Classification of newly arrived requests was done using different classifiers with the help of received training samples and developed success criteria.

Decoy Routing, the use of routers (rather than end hosts) as proxies, is a new direction in anti-censorship research. Decoy Routers (DRs), placed in Autonomous Systems, proxy traffic from users; so the adversary, e.g. a censorious government, attempts to avoid them. It is quite difficult to place DRs so the adversary cannot route around them – for example, we need the cooperation of 850 ASes to contain China alone [1]. In this paper, we consider a different approach. We begin by noting that DRs need not intercept all the network paths from a country, just those leading to Overt Destinations, i.e. unfiltered websites hosted outside the country (usually popular ones, so that client traffic to the OD does not make the censor suspicious). Our first question is – How many ASes are required for installing DRs to intercept a large fraction of paths from e.g. China to the top-n websites (as per Alexa)? How does this number grow with n ? To our surprise, the same few ($\approx$ 30) ASes intercept over 90% of paths to the top n sites worldwide, for n = 10, 20...200 and also to other destinations. Investigating further, we find that this result fits perfectly with the hierarchical model of the Internet [2]; our first contribution is to demonstrate with real paths that the number of ASes required for a world-wide DR framework is small ($\approx$ 30). Further, censor nations' attempts to filter traffic along the paths transiting these 30 ASes will not only block their own citizens, but others residing in foreign ASes. Our second contribution in this paper is to consider the details of DR placement: not just in which ASes DRs should be placed to intercept traffic, but exactly where in each AS. We find that even with our small number of ASes, we still need a total of about 11, 700 DRs. We conclude that, even though a DR system involves far fewer ASes than previously thought, it is still a major undertaking. For example, the current routers cost over 10.3 billion USD, so if Decoy Routing at line speed requires all-new hardware, the cost alone would make such a project unfeasible for most actors (but not for major nation states).

To provide a comprehensive security analysis of modern networked systems, we need to take into account the combined effects of existing vulnerabilities and zero-day vulnerabilities. In addition to them, it is important to incorporate new vulnerabilities emerging from threats such as BYOD, USB file sharing. Consequently, there may be new dependencies between system components that could also create new attack paths, but previous work did not take into account those new attack paths in their security analysis (i.e., not all attack paths are taken into account). Thus, countermeasures may not be effective, especially against attacks exploiting the new attack paths. In this paper, we propose a Unified Vulnerability Risk Analysis Module (UV-RAM) to address the aforementioned problems by taking into account the combined effects of those vulnerabilities and capturing the new attack paths. The three main functionalities of UV-RAM are: (i) to discover new dependencies and new attack paths, (ii) to incorporate new vulnerabilities introduced and zero-day vulnerabilities into security analysis, and (iii) to formulate mitigation strategies for hardening the networked system. Our experimental results demonstrate and validate the effectiveness of UV-RAM.

We use model-based testing techniques to detect logical vulnerabilities in implementations of the Wi-Fi handshake. This reveals new fingerprinting techniques, multiple downgrade attacks, and Denial of Service (DoS) vulnerabilities. Stations use the Wi-Fi handshake to securely connect with wireless networks. In this handshake, mutually supported capabilities are determined, and fresh pairwise keys are negotiated. As a result, a proper implementation of the Wi-Fi handshake is essential in protecting all subsequent traffic. To detect the presence of erroneous behaviour, we propose a model-based technique that generates a set of representative test cases. These tests cover all states of the Wi-Fi handshake, and explore various edge cases in each state. We then treat the implementation under test as a black box, and execute all generated tests. Determining whether a failed test introduces a security weakness is done manually. We tested 12 implementations using this approach, and discovered irregularities in all of them. Our findings include fingerprinting mechanisms, DoS attacks, and downgrade attacks where an adversary can force usage of the insecure WPA-TKIP cipher. Finally, we explain how one of our downgrade attacks highlights incorrect claims made in the 802.11 standard.

Distractor generation is a crucial step for fill-in-the-blank question generation. We propose a generative model learned from training generative adversarial nets (GANs) to create useful distractors. Our method utilizes only context information and does not use the correct answer, which is completely different from previous Ontology-based or similarity-based approaches. Trained on the Wikipedia corpus, the proposed model is able to predict Wiki entities as distractors. Our method is evaluated on two biology question datasets collected from Wikipedia and actual college-level exams. Experimental results show that our context-based method achieves comparable performance to a frequently used word2vec-based method for the Wiki dataset. In addition, we propose a second-stage learner to combine the strengths of the two methods, which further improves the performance on both datasets, with 51.7% and 48.4% of generated distractors being acceptable.

With the advancement of unmanned aerial vehicles (UAV), 3D wireless mesh networks will play a crucial role in next generation mission critical wireless networks. Along with providing coverage over difficult terrain, it provides better spectral utilization through 3D spatial reuse. However, being a wireless network, 3D meshes are vulnerable to jamming/disruptive attacks. A jammer can disrupt the communication, as well as control of the network by intelligently causing interference to a set of nodes. This paper presents a distributed mechanism of avoiding jamming attacks by means of 3D spatial filtering where adaptive beam nulling is used to keep the jammer in null region in order to bypass jamming. Kalman filter based tracking mechanism is used to estimate the most likely trajectory of the jammer from noisy observation of the jammer's position. A beam null border is determined by calculating confidence region of jammer's current and next position estimates. An optimization goal is presented to calculate optimal beam null that minimizes the number of deactivated links while maximizing the higher value of confidence for keeping the jammer inside the null. The survivability of a 3D mesh network with a mobile jammer is studied through simulation that validates an 96.65% reduction in the number of jammed nodes.

Recently, the increase of interconnectivity has led to a rising amount of IoT enabled devices in botnets. Such botnets are currently used for large scale DDoS attacks. To keep track with these malicious activities, Honeypots have proven to be a vital tool. We developed and set up a distributed and highly-scalable WAN Honeypot with an attached backend infrastructure for sophisticated processing of the gathered data. For the processed data to be understandable we designed a graphical frontend that displays all relevant information that has been obtained from the data. We group attacks originating in a short period of time in one source as sessions. This enriches the data and enables a more in-depth analysis. We produced common statistics like usernames, passwords, username/password combinations, password lengths, originating country and more. From the information gathered, we were able to identify common dictionaries used for brute-force login attacks and other more sophisticated statistics like login attempts per session and attack efficiency.

As DDOS attacks interrupt internet services, DDOS tools confirm the effectiveness of the current attack. DDOS attack and countermeasures continue to increase in number and complexity. In this paper, we explore the scope of the DDoS flooding attack problem and attempts to combat it. A contemporary escalation of application layer distributed denial of service attacks on the web services has quickly transferred the focus of the research community from conventional network based denial of service. As a result, new genres of attacks were explored like HTTP GET Flood, HTTP POST Flood, Slowloris, R-U-Dead-Yet (RUDY), DNS etc. Also after a brief introduction to DDOS attacks, we discuss the characteristics of newly proposed application layer distributed denial of service attacks and embellish their impact on modern web services.

This paper proposes a new DNA cryptographic technique based on dynamic DNA encoding and asymmetric cryptosystem to increase the level of secrecy of data. The key idea is: to split the plaintext into fixed sized chunks, to encrypt each chunk using asymmetric cryptosystem and finally to merge the ciphertext of each chunk using dynamic DNA encoding. To generate chunks, characters of the plaintext are transformed into their equivalent ASCII values and split it into finite values. Now to encrypt each chunk, asymmetric cryptosystem is applied and the ciphertext is transformed into its equivalent binary value. Then this binary value is converted into DNA bases. Finally to merge each chunk, sufficient random strings are generated. Here to settle the required number of random strings, dynamic DNA encoding is exploited which is generated using Fibonacci series. Thus the use of finite chunks, asymmetric cryptosystem, random strings and dynamic DNA encoding increases the level of security of data. To evaluate the encryption-decryption time requirement, an empirical analysis is performed employing RSA, ElGamal and Paillier cryptosystems. The proposed technique is suitable for any use of cryptography.

As a vital component of variety cyber attacks, malicious domain detection becomes a hot topic for cyber security. Several recent techniques are proposed to identify malicious domains through analysis of DNS data because much of global information in DNS data which cannot be affected by the attackers. The attackers always recycle resources, so they frequently change the domain - IP resolutions and create new domains to avoid detection. Therefore, multiple malicious domains are hosted by the same IPs and multiple IPs also host same malicious domains in simultaneously, which create intrinsic association among them. Hence, using the labeled domains which can be traced back from queries history of all domains to verify and figure out the association of them all. Graphs seem the best candidate to represent for this relationship and there are many algorithms developed on graph with high performance. A graph-based interface can be developed and transformed to the graph mining task of inferring graph node's reputation scores using improvements of the belief propagation algorithm. Then higher reputation scores the nodes reveal, the more malicious probabilities they infer. For demonstration, this paper proposes a malicious domain detection technique and evaluates on a real-world dataset. The dataset is collected from DNS data servers which will be used for building a DNS graph. The proposed technique achieves high performance in accuracy rates over 98.3%, precision and recall rates as: 99.1%, 98.6%. Especially, with a small set of labeled domains (legitimate and malicious domains), the technique can discover a large set of potential malicious domains. The results indicate that the method is strongly effective in detecting malicious domains.