This paper presents FUNDED (Flow-sensitive vUl-Nerability coDE Detection), a novel learning framework for building vulnerability detection models. Funded leverages the advances in graph neural networks (GNNs) to develop a novel graph-based learning method to capture and reason about the program's control, data, and call dependencies. Unlike prior work that treats the program as a sequential sequence or an untyped graph, Funded learns and operates on a graph representation of the program source code, in which individual statements are connected to other statements through relational edges. By capturing the program syntax, semantics and flows, Funded finds better code representation for the downstream software vulnerability detection task. To provide sufficient training data to build an effective deep learning model, we combine probabilistic learning and statistical assessments to automatically gather high-quality training samples from open-source projects. This provides many real-life vulnerable code training samples to complement the limited vulnerable code samples available in standard vulnerability databases. We apply Funded to identify software vulnerabilities at the function level from program source code. We evaluate Funded on large real-world datasets with programs written in C, Java, Swift and Php, and compare it against six state-of-the-art code vulnerability detection models. Experimental results show that Funded significantly outperforms alternative approaches across evaluation settings.

Mobile sensing has become a new style of applications and most of the smart devices are equipped with varieties of sensors or functionalities to enhance sensing capabilities. Current sensing systems concentrate on how to enhance sensing capabilities; however, the sensors or functionalities may lead to the leakage of users’ privacy. In this paper, we present WiPass, a way to leverage the wireless hotspot functionality on the smart devices to snoop the unlock passwords/patterns without the support of additional hardware. The attacker can “see” your unlock passwords/patterns even one meter away. WiPass leverages the impacts of finger motions on the wireless signals during the unlocking period to analyze the passwords/patterns. To practically implement WiPass, we are facing the difficult feature extraction and complex unlock passwords matching, making the analysis of the finger motions challenging. To conquer the challenges, we use DCASW to extract feature and hierarchical DTW to do unlock passwords matching. Besides, the combination of amplitude and phase information is used to accurately recognize the passwords/patterns. We implement a prototype of WiPass and evaluate its performance under various environments. The experimental results show that WiPass achieves the detection accuracy of 85.6% and 74.7% for passwords/patterns detection in LOS and in NLOS scenarios, respectively.

Evil-Twin is becoming a common attack in smart home environments where an attacker can set up a fake AP to compromise the security of the connected devices. To identify the fake APs, The current approaches of detecting Evil-Twin attacks all rely on information such as SSIDs, the MAC address of the genuine AP, or network traffic patterns. However, such information can be faked by the attacker, often leading to low detection rates and weak protection. This paper presents a novel Evil-Twin attack detection method based on the received signal strength indicator (RSSI). Our approach considers the RSSI as a fingerprint of APs and uses the fingerprint of the genuine AP to identify fake ones. We provide two schemes to detect a fake AP in two different scenarios where the genuine AP can be located at either a single or multiple locations in the property, by exploiting the multipath effect of the Wi-Fi signal. As a departure from prior work, our approach does not rely on any professional measurement devices. Experimental results show that our approach can successfully detect 90% of the fake APs, at the cost of a one-off, modest connection delay.

Since defects in software may cause product fault and financial loss, it is essential to conduct software defect prediction (SDP) to identify the potentially defective modules, especially in the early stage of the software development lifecycle. Recently, cross-version defect prediction (CVDP) began to draw increasing research interests, employing the labeled defect data of the prior version within the same project to predict defects in the current version. As software development is a dynamic process, the data distribution (such as defects) during version change may get changed. Recent studies utilize machine learning (ML) techniques to detect software defects. However, due to the close dependencies between the updated and unchanged code, prior ML-based methods fail to model the long and deep dependencies, causing a high false positive. Furthermore, traditional defect detection is performed on the entire project, and the detection efficiency is relatively low, especially on large-scale software projects. To this end, we proposeBugPre, a CVDP approach to address these two issues.BugPreis a novel framework that only conducts efficient defect prediction on changed modules in the current version.BugPreutilizes variable propagation tree-based associated analysis method to obtain the changed modules in the current version. Besides,BugPreconstructs graph leveraging code context dependences and uses a graph convolutional neural network to learn representative characteristics of code, thereby improving defect prediction capability when version changes occur. Through extensive experiments on open-source Apache projects, the experimental results indicate that ourBugPreoutperforms three state-of-the-art defect detection approaches, and the F1-score has increased by higher than 16%.

Deep learning (DL) has emerged as a viable means for identifying software bugs and vulnerabilities. The success of DL relies on having a suitable representation of the problem domain. However, existing DL-based solutions for learning program representations have limitations - they either cannot capture the deep, precise program semantics or suffer from poor scalability. We present Concoction, the first DL system to learn program presentations by combining static source code information and dynamic program execution traces. Concoction employs unsupervised active learning techniques to determine a subset of important paths to collect dynamic symbolic execution traces. By implementing a focused symbolic execution solution, Concoction brings the benefits of static and dynamic code features while reducing the expensive symbolic execution overhead. We integrate Concoction with fuzzing techniques to detect function-level code vulnerabilities in C programs from 20 open-source projects. In 200 hours of automated concurrent test runs, Concoction has successfully uncovered vulnerabilities in all tested projects, identifying 54 unique vulnerabilities and yielding 37 new, unique CVE IDs. Concoction also significantly outperforms 16 prior methods by providing higher accuracy and lower false positive rates.