May 29, 2018
Dan Gopstein , Hongwei Zhou , Phyllis Frankl and Justin Cappos
Prior work has shown that extremely small code patterns, such as the conditional operator and implicit type conversion, can cause considerable misunderstanding in programmers. Until now, the real world impact of these patterns ś known as ‘atoms of confusion’ ś was only speculative. This work uses a corpus of 14 of the most popular and inluential open source C and C++ projects to measure the prevalence and signiicance of these small confusing patterns. Our results show that the 15 known types of confusing micro patterns occur millions of times in programs like the Linux kernel and GCC, appearing on average once every 23 lines.
April 10, 2018
Rui Zhang and Quanyan Zhu
With the recent growing number of cyber-attacks and the constant lack of effective and state-of-art defense methods, cyber risks become ubiquitous in enterprise networks, manufacturing plants, and government computer systems. Cyber-insurance has become one of the major ways to mitigate the risks as it can transfer the cyber-risks to insurance companies and improve the security status of the insured. The designation of effective cyber-insurance policies requires the considerations from both the insurance market and the dynamic properties of the cyber-risks.
March 22, 2018
Minhui Xue, Alexandru Grigoras, Heather Lee and Keith Ross
Many countries today have “country-centric mobile apps” which are mobile apps that are primarily used by residents of a specific country. Many of these country-centric apps also include a location-based service which takes advantage of the smartphone’s API access to the smartphone’s current GPS location. In this paper, we investigate how such country-centric apps with location-based services can be employed to study the diaspora associated with ethnic and cultural groups. Our methodology combines GPS hacking, automated task tools for mobile phones, and OCR to generate migration statistics for diaspora.
March 21, 2018
Muhammad Junaid Farooq and Quanyan Zhu
Spectrum reservation is emerging as one of the potential solutions to cater for the communication needs of massive number of wireless Internet of Things (IoT) devices with reliability constraints particularly in mission-critical scenarios. In most mission-critical systems, the true utility of a reservation may not be completely known ahead of time as the unforseen events might not be completely predictable. In this paper, we present a dynamic contract approach where an advance payment is made at the time of reservation based on partial information about spectrum reservation utility.
March 20, 2018
Satwik Patnaik , Mohammed Ashraf , Johann Knechtel , and Ozgur Sinanoglu
Ensuring the trustworthiness and security of electronics has become an urgent challenge in recent years. Among various concerns, the protection of design intellectual property (IP) is to be addressed, due to outsourcing trends for the manufacturing supply chain and malicious end-user. In other words, adversaries either residing in the off-shore fab or in the field may want to obtain and pirate your design IP. As classical design tools do not consider such threats, there is clearly a need for security-aware EDA techniques.
March 20, 2018
Abhrajit Sengupta, Muhammad Yasin, Mohammed Nabeel, Mohammed Ashraf, Jeyavijayan Rajendran and Ozgur Sinanoglu
With the globalization of integrated circuit (IC) supply chain, the semi-conductor industry is facing a number of security threats, such as Intellectual Property (IP) piracy, hardware Trojans, and counterfeiting. To defend against such threats at the hardware level, logic locking was proposed as a promising countermeasure. Yet, several recent attacks have completely undermined its security by successfully retrieving the secret key. Here, we present stripped-functionality logic locking (SFLL), which resists all existing attacks by hiding a part of the functionality in the form of a secret key.
March 12, 2018
Quanyan Zhu and Stefan Rass
Advanced persistent threats (APT) are considered as a significant security threat today. Despite their diversity in nature and details, a common skeleton and sequence of phases can be identified that these attacks follow (in similar ways), which admits a game-theoretic description and analysis. This paper describes a general framework that divides a general APT into three major temporal phases, and fits an individual game model to each phase, connecting the games at the transition points between the phases (similarly to “milestones” accomplished during the launch of an APT).
Adaptive and Resilient Revenue Maximizing Dynamic Resource Allocation and Pricing for Cloud-Enabled IoT Systems
March 3, 2018
Muhammad Junaid Farooq and Quanyan Zhu
Cloud computing is becoming an essential component in the emerging Internet of Things (IoT) paradigm. The available resources at the cloud such as computing nodes, storage, databases, etc. are often packaged in the form of virtual machines (VMs) to be used by remotely located IoT client applications for computational tasks. However, the cloud has a limited number of VMs available and hence, for massive IoT systems, the available resources must be efficiently utilized to increase productivity and subsequently maximize revenue of the cloud service provider (CSP).
February 27, 2018
Hamza Anwar and Quanyan Zhu
Owing to the recent advances in “Big Data” modeling and prediction tasks, variational Bayesian estimation has gained popularity due to their ability to provide exact solutions to approximate posteriors. One key technique for approximate inference is stochastic variational inference (SVI) . SVI poses variational inference as a stochastic optimization problem and solves it iteratively using noisy gradient estimates. It aims to handle massive data for predictive and classification tasks by applying complex Bayesian models that have observed as well as latent variables. This paper aims to decentralize it allowing parallel computation, secure learning and robustness benefits.
February 22, 2018
Satwik Patnaik, Johann Knechtel, Mohammed Ashraf and Ozgur Sinanoglu
Here we advance the protection of split manufacturing (SM)-based layouts through the judicious and well-controlled handling of interconnects. Initially, we explore the cost-security trade-offs of SM, which are limiting its adoption. Aiming to resolve this issue, we propose effective and efficient strategies to lift nets to the BEOL. Towards this end, we design custom “elevating cells” which we also provide to the community. Further, we define and promote a new metric, Percentage of Netlist Recovery (PNR), which can quantify the resilience against gate-level theft of intellectual property (IP) in a manner more meaningful than established metrics.
Evolutionary Methods for Generating Synthetic MasterPrint Templates: Dictionary Attack in Fingerprint Recognition
February 21, 2018
Aditi Roy, Nasir Memon, Julian Togelius and Arun Ross
Recent research has demonstrated the possibility of generating “Masterprints” that can be used by an adversary to launch a dictionary attack against a fingerprint recognition system. Masterprints are fingerprint images that fortuitously match with a large number of other fingerprints thereby compromising the security of a fingerprint-based biometric system, especially those equipped with smallsized fingerprint sensors. This work presents new methods for creating a synthetic MasterPrint dictionary that sequentially maximizes the probability of matching a large number of target fingerprints.
Analyzing and Mitigating the Impact of Permanent Faults on a Systolic Array Based Neural Network Accelerator
February 17, 2018
Jeff (Jun) Zhang, Tianyu Gu, Kanad Basu and Siddharth Garg
Due to their growing popularity and computational cost, deep neural networks (DNNs) are being targeted for hardware acceleration. A popular architecture for DNN acceleration, adopted by the Google Tensor Processing Unit (TPU), utilizes a systolic array based matrix multiplication unit at its core. This paper deals with the design of faulttolerant, systolic array based DNN accelerators for high defect rate technologies. To this end, we empirically show that the classification accuracy of a baseline TPU drops significantly even at extremely low fault rates (as low as 0.006%).
February 13, 2018
Yanyan Zhuang,Albert Rafetseder, Yu Hu, Yuan Tian and Justin Cappos
Due to their omnipresence, mobile devices such as smartphones could be tremendously valuable to researchers. However, since research projects can extract data about device owners that could be personal or sensitive, there are substantial privacy concerns. Currently, the only regulation to protect user privacy for research projects is through Institutional Review Boards (IRBs) from researchers’ institutions. However, there is no guarantee that researchers will follow the IRB protocol. Even worse, researchers without security expertise might build apps that are vulnerable to attacks.
February 12, 2018
Yiming Zhang , Quan Ho Vuong , Kenny Song , Xiao-Yue Gong and Keith W. Ross
This paper considers entropy bonus, which is used to encourage exploration in policy gradient. In the case of high-dimensional action spaces, calculating the entropy and its gradient requires enumerating all the actions in the action space and running forward and back-propagation for each action, which may be computationally infeasible. We develop several novel unbiased estimators for the entropy bonus and its gradient. We apply these estimators to several models for the parameterized policies, including Independent Sampling, CommNet, Autoregressive with Modified MDP, and Autoregressive with LSTM.
ThUnderVolt: Enabling Aggressive Voltage Underscaling and Timing Error Resilience for Energy Efficient Deep Neural Network Accelerators
February 11, 2018
Jeff Zhang, Kartheek Rangineni, Zahra Ghodsi, and Siddharth Garg
Hardware accelerators are being increasingly deployed to boost the performance and energy efficiency of deep neural network (DNN) inference. In this paper we propose Thundervolt, a new framework that enables aggressive voltage underscaling of high-performance DNN accelerators without compromising classification accuracy even in the presence of high timing error rates. Using post-synthesis timing simulations of a DNN accelerator modeled on the Google TPU, we show that Thundervolt enables between 34%-57% energy savings on state-of-the-art speech and image recognition benchmarks with less than 1%loss in classification accuracy and no performance loss.
February 7, 2018
Rui Zhang and Quanyan Zhu
Distributed Support Vector Machines (DSVM) have been developed to solve large-scale classification problems in networked systems with a large number of sensors and control units. However, the systems become more vulnerable as detection and defense are increasingly difficult and expensive. This work aims to develop secure and resilient DSVM algorithms under adversarial environments in which an attacker can manipulate the training data to achieve his objective. We establish a game-theoretic framework to capture the conflicting interests between an adversary and a set of distributed data processing units.
February 3, 2018
Farhad Shirani, Siddharth Garg and Elza Erkip
In this paper, the problem of matching pairs of correlated random graphs with multi-valued edge attributes is considered. Graph matching problems of this nature arise in several settings of practical interest including social network deanonymization, study of biological data, web graphs, etc. An achievable region for successful matching is derived by analyzing a new matching algorithm that we refer to as typicality matching. The algorithm operates by investigating the joint typicality of the adjacency matrices of the two correlated graphs.
February 2, 2018
Tao Zhang and Quanyan Zhu
Vehicular ad hoc network (VANET) is an enabling technology in modern transportation systems for providing safety and valuable information, and yet vulnerable to a number of attacks from passive eavesdropping to active interfering. Intrusion detection systems (IDSs) are important devices that can mitigate the threats by detecting malicious behaviors. Furthermore, the collaborations among vehicles in VANETs can improve the detection accuracy by communicating their experiences between nodes.