Search

Videoconferencing platforms have received a large influx of traffic due to COVID-19 and the global pandemic it has caused. Professionals, students, politicians, and high-risk security individuals utilize these platforms to communicate with their colleagues. They blindly trust these platforms without understanding how secure they are, and their sensitive data may be vulnerable as it is transmitted through these third party platforms. Zoom, Google Meet, and Microsoft Teams are the most popular among the platforms, and they are examined in this paper. These platforms’ security aspects were to be evaluated by simulating calls within a virtual environment, capturing the network traffic sent to/from the network, and analyzing the captured traffic to determine what information outside listeners had access to. It was discovered that these platforms utilized state-of-the-art encryption methods, but they did not provide end-to-end encryption since vendors as third parties have access to the end-user communication data. It was also discovered that some strange TCP connections were being established as well as unknown DNS communications. It also found TCP communications running in the background. This suggests that the platforms are not being transparent with users, and this could generate potential security concerns with users.

The need for a new authentication method such as biometrics becomes apparent as the data breaches on password-based authentication increase. However, current biometric forms of authentication become unusable once compromised. Additional limits are realized when an attacker coerces an authorized user into a forced authentication. To resolve both issues, we propose creating an authentication mechanism that depends on the user’s neurophysiological responses to chosen pieces of music (non-lyrical) measured using electroencephalographic (EEG) signals. This research paper provides a guide for creating and presenting a system that incorporates such idea for person classification and authentication. In a group study, the aim is that participant listen to individually selected music and music selected by other participants during an EEG reading. The change in the Alpha and Beta band frequencies across eight electrode EEG sensors serves as the input feature vector for a supervised machine learning algorithm that trains on the user and attacker EEG readings. Ultimately, the goal of the algorithm is to create a user-specific model to uniquely identify the respective user based on the corresponding EEG response to music and grant authentication. Our study lays a solid foundation for creating a promising EEG-based authentication system by solving the drawbacks of current biometric authentication methods.

As attacks against password-based authentication increase, the need for robust biometrics becomes apparent. Currently, two of the most popular biometric authentication systems are fingerprint and facial recognition. However, both of these biometrics become unusable once compromised. Also, an attacker might coerce the user to force authentication. Therefore, we propose an authentication mechanism that depends on the participant’s neurological responses to chosen pieces of music measured using electroencephalographic (EEG) signals. The current study proposes an authentication system that uses neurological responses to music for classification. Participants listened to individually selected music and music selected by other participants during an EEG reading. The change in the Alpha and Beta band frequencies across seven electrodes served as the input to a user specific K-Nearest Neighbors (KNN). The classifier attempts to determine if we can identify a user based on their EEG response to music. Our pilot data collection and analysis has shown promise of this authentication system with an accuracy rate between 76.4%-92.3%.

In this paper, we discuss the design and development of EEG Workbench: a prototyping framework for building electroencephalography (EEG) data analytics framework for the Android platform. EEG data describe the activity of neuron populations in the brain and are acquired via electrodes placed on specific positions in the skull. EEG signals have been shown to be sufficiently different between individuals and therefore suitable for use in the authentication process. The relatively recent introduction of consumer-grade EEG acquisition devices has made the utilization of EEG research more accessible to ubiquitous computing devices such as smartphones; however, the software developed in EEG research lacks specific implementations for Android. EEG Workbench aims to provide developers with a tool to build electroencephalography (EEG) data analytics framework for the Android platform.

This paper presents a storage enforcing remote verification scheme, PGV (Pretty Good Verification) as a bidirectional data integrity checking mechanism for cloud storage. At its core, PGV relies on the well-known polynomial hash; we show that the polynomial hash provably possesses the storage enforcement property and is also efficient in terms of performance. In addition to the traditional application of a client verifying the storage content at a remote server, PGV can also be applied to de-duplication scenarios where the server wants to verify whether the client possesses a significant amount of information about a file (and not just a partial knowledge/fingerprint of the file) before granting access to an existing file.While existing schemes are often developed to handle a malicious adversarial model, we argue that such a model is often too strong of an assumption, resulting in over-engineered, resource-intensive mechanisms. Instead, the storage enforcement property of PGV aims at removing a practical incentive for a storage server to cheat in order to save on storage space in a covert adversarial model.We theoretically prove the power of PGV by combining Kolmogorov complexity and list decoding and experimentally show the simplicity and low overhead of PGV by comparing it with existing schemes. Altogether, PGV provides a good, practical way to perform storage enforcing remote verification.

Network management is a critical component in both wired and wireless network. In wireless networks, the network dynamics changes rapidly and the network management information needs to be updated frequently. Due to its structured form and ease of usage in communication, eXtensible Markup Language (XML) is preferred as a configuration and logging tool in network management. However, it is convenient to use relational databases such as SQL to store and process data of wireless network management where frequent updates are necessary. In this paper, the authors show an automated method of converting XML documents to store in relational. We have implemented a proof of concept and compared performance to existing approach.

Drop box is a cloud based file storage service used by more than 200 million users. Its ability to seamlessly provide cloud storage with minimal user complexity is the key for its wide spread popularity. Despite of its high usability, Drop box has been recently criticized for loose ends in security. Security and usability is not always mutually exclusive, and we believe there is still a lot of room to improve Drop box's security without affecting the unique user experience. In this paper, we present a RAM analysis based method to extract the key security token for account access. In addition, we describe a new technique to bypass authentication and gain unauthorized access to Drop box accounts by using the new tray login feature on the most current Drop box client (v2.4.x). Through these exploits, we demonstrate that most of these security issues are at the level of implementation, rather than design. Finally, we describe potential resolutions that can improve Drop box's security without affecting its high usability.

Online social media such as Facebook has started transforming our lives in the past few years. One of the main user concerns in social media is the privacy of personal information. Although social media provides options to manage privacy, those user settings are often complicated, vague, and not user friendly. In this paper, we propose an automated and simple way to manage Facebook privacy by changing the privacy to three predefined levels: basic, medium, and advanced. For platform independence and wide usability, we discuss the development of the proposed application in two different flavors: a PHP application and Firefox add-on.

Phishing attacks are a form of social engineering technique used for stealing private information from users through emails. A general approach for phishing susceptibility analysis is to profile the user's personality using personality models such as the Five Factor Model (FFM) and find out the susceptibility for a set of phishing attempts. The FFM is a personality profiling system that scores participants on five separate personality traits: openness to experience (O), conscientiousness (C), extraversion (E), agreeableness (A), and neuroticism (N). However, existing approaches don't take into account the fact that based on the content, for example, a phishing email offering an enticing free prize might be very effective on a dominant O-personality (curious, open to new experience), but not to an N-personality (tendency of experiencing negative emotion). Therefore, it is necessary to consider the personality bias of the phishing email contents during the susceptibility analysis. In this paper, we have proposed a method to construct a dictionary based on the semantic similarity of prospective words describing the FFM. Words generated through this dictionary can be used to label the phishing emails according to the personality bias and serve as the key component of a personality bias classification system of phishing emails. We have validated our dictionary construction using a large public corpus of phishing email data which shows the potential of the proposed system in anti-phishing research.

Symmetric encryption provides lightweight security solution to maintain data confidentiality on devices in a resource constrained scenario such as in a tactical network. However, lightweight encryption schemes are traditionally vulnerable to linear and differential cryptanalysis as well as power analysis attack when the encryption structure is known to the attacker. For tactical network devices, this is a critical concern since they often operate in hostile scenarios and lack in physical security in most cases. Moving Target Defense (MTD) is one of the key components of cyber maneuver that reshapes friendly networks and associated assets to be resilient to cyber-attacks. In this paper, we propose a lightweight reconfigurable symmetric encryption architecture, REA, which is capable of implementing a user-defined symmetric encryption scheme as an MTD mechanism. The encryption structure can be customized from device to device based on their available resource and performance requirements. Due to the reconfigurable nature of the proposed architecture, it is not possible for an attacker to directly launch the cryptanalysis or power analysis attack before committing significant resources to retrieve the encryption structure first. We implemented a reference encryption scheme on our proposed architecture in programmable logic (FPGA) and compared it to two representative symmetric encryption methods: AES and Present. Our results show that the reference encryption consumes less resources and performs faster compared to AES. Performance of the REA reference encryption is comparable with Present, which is optimized only for low resource devices and doesn't support reconfigurability.

The Single System Image (SSI) clustering technology hides the distributed nature of the participating resources, and makes them appear as a single homogeneous computing resource to the user. An SSI cluster can utilize all the available processing power and memory from its participating resources. However, using the current implementations, an SSI can only be formed using nodes which are within one-hop distance of each other. This implies that nodes have to be within the same broadcast domain / Local Area Network (LAN) in order to participate in an SSI cluster. This limits the full potential of SSIs. In this research, we propose enhancements to the existing SSI technology to overcome the one-hop limitation, thus enabling nodes over a WAN to form SSI clusters. GENI provides a perfect platform for such experimentation. In this paper, we report our initial success in enabling Transparent Interprocess Communication Protocol (TIPC) over wide area nodes in GENI and progress in backporting the TIPC 2.0 protocol, with support for communication over WANs, to Kerrighed, an open-source software for creating SSIs.

This paper presents a storage enforcing remote verification scheme, PGV (Pretty Good Verification). While existing schemes are often developed to handle a malicious adversarial model, we argue that such a model is often too strong of an assumption, resulting in over-engineered, resource-intensive mechanisms. Instead, the storage enforcement property of PGV aims at removing a practical incentive for a storage server to cheat in order to save on storage space in a covert adversarial model. At its core, PGV relies on the well-known polynomial hash, we show that the polynomial hash provably possesses the storage enforcement property and is also efficient in terms of performance. In addition to the traditional application of a client verifying the storage content at a remote server, PGV can also be applied to de-duplication scenarios where the server wants to verify whether the client possesses a significant amount of information about a file (and not just a partial knowledge/fingerprint of the file) before granting access to an existing file. We theoretically prove the power of PGV by combining Kolmogorov complexity and list decoding, and experimentally show the simplicity and low overhead of PGV by comparing it with existing schemes. Altogether, PGV provides a good, practical way to perform storage enforcing remote verification.

A common approach for introducing security at the physical layer is to rely on the channel variations of the wireless environment. This type of approach is not always suitable for wireless networks where the channel remains static for most of the network lifetime. For these scenarios, a channel independent physical layer security measure is more appropriate which will rely on a secret known to the sender and the receiver but not to the eavesdropper. In this paper, we propose CD-PHY, a physical layer security technique that exploits the constellation diversity of wireless networks which is independent of the channel variations. The sender and the receiver use a custom bit sequence to constellation symbol mapping to secure the physical layer communication which is not known a priori to the eavesdropper. Through theoretical modeling and experimental simulation, we show that this information theoretic construct can achieve Shannon secrecy and any brute force attack from the eavesdropper incurs high overhead and minuscule probability of success. Our results also show that the high bit error rate also makes decoding practically infeasible for the eavesdropper, thus securing the communication between the sender and receiver., Comment: 9 pages, 15 figures

Apple iPhone has made significant impact on the society both as a handheld computing device and as a cellular phone. Due to the unique hardware system as well as storage structure, iPhone has already attracted the forensic community in digital investigation of the device. Currently available commercial products and methodologies for iPhone forensics are somewhat expensive, complex and often require additional hardware for analysis. Some products are not robust and often fail to extract optimal evidence without modifying the iPhone firmware which makes the analysis questionable in legal platforms. In this paper, we present a simple and inexpensive framework (iFF) for iPhone forensic analysis. Through experimental results using real device, we have shown the effectiveness of this framework in extracting digital evidence from an iPhone.

Smart phones with Internet capability are growing in popularity, due to many of their useful capabilities. Among other handy features of smart phones, Instant Messaging (IM) is very popular due to the level of convenience it provides in interpersonal communications. As the usage of IM on smart phone is increasing rapidly, it is important to take measures in advance from forensic standpoint forecasting the potential use of it in cyber crimes such as the cyber stalking and cyber bullying. Although, current IM applications for smart phones are in most cases a downsized version of the one used on traditional computers, diverse structure of file systems and storage device on different smart phones pose unique challenges to forensic examiners for recovering digital evidences of a conversation under investigation. In this work, we study and report the forensic analysis of three different IMs: AIM, Yahoo! Messenger and Google Talk, (both client based and web based version) on Apple iPhone. Our results show that the forensic analysis of IMs on smart phones has significant value and needs further attention.