Smartphones are closely coupled with their users and smartphone sensors can perceive users' private information. The existing studies in this area focus on user activity recognition and short-term context detection. In this paper, we show that smartphone sensors are able to profile users' long-term privacy and more sensitive information. We present the techniques of discovering users' spending level by merely using smartphone sensors. We do not access users' contacts, calendar, or call log, so that the profiling is performed in a non-intrusive manner. This paper is an alert towards the public that the privacy leakage could be far worse than imagination by just carrying smartphones.

Video source identification is very important in validating video evidence, tracking down video piracy crimes, and regulating individual video sources. With the prevalence of wireless communication, wireless video cameras continue to replace their wired counterparts in security/surveillance systems and tactical networks. However, wirelessly streamed videos usually suffer from blocking and blurring due to inevitable packet loss in wireless transmissions. The existing source identification methods experience significant performance degradation or even fail to work when identifying videos with blocking and blurring. In this paper, we propose a method that is effective and efficient in identifying such wirelessly streamed videos. In addition, we also propose to incorporate wireless channel signatures and selective frame processing into source identification, which significantly improve the identification speed. We conduct extensive real-world experiments to validate our method. The results show that the source identification accuracy of the proposed scheme largely outperforms the existing methods in the presence of video blocking and blurring. Moreover, our method is able to identify the video source in a near-real-time fashion, which can be used to detect the wireless camera spoofing attack.

Facial recognition is a popular biometric authentica-tion technique, but it is rarely used in practice for de-vice unlock or website / app login in smartphones, alt-hough most of them are equipped with a front-facing camera. Security issues (e.g. 2D media attack and vir-tual camera attack) and ease of use are two important factors that impede the prevalence of facial authentica-tion in mobile devices. In this paper, we propose a new sensor-assisted facial authentication method to over-come these limitations. Our system uses motion and light sensors to defend against 2D media attacks and virtual camera attacks without the penalty of authenti-cation speed. We conduct experiments to validate our method. Results show 95-97% detection rate and 2-3% false alarm rate over 450 trials in real-settings, indicat-ing high security obtained by the scheme ten times faster than existing 3D facial authentications (3 sec-onds compared to 30 seconds).

Digital camera identification can be accomplished based on sensor pattern noise, which is unique to a device, and serves as a distinct identification fingerprint. Camera identification and authentication have formed the basis of image/video forensics in legal proceedings. Unfortunately, real-time video source identification is a computationally heavy task, and does not scale well to conventional software implementations on typical embedded devices. In this paper, we propose a hardware architecture for source identification in networked cameras. The underlying algorithms, an orthogonal forward and inverse discrete wavelet transform and minimum mean square error-based estimation, have been optimized for 2-D frame sequences in terms of area and throughput performance. We exploit parallelism, pipelining, and hardware reuse techniques to minimize hardware resource utilization and increase the achievable throughput of the design. A prototype implementation on a Xilinx Virtex-6 FPGA device was optimized with a resulting throughput of 167 MB/s, processing 30 640 × 480 video frames in 0.17 s.

Network forensics is an emerging interdiscipline used to track down cyber crimes and detect network anomalies for a multitude of applications. Efficient capture of data is the basis of network forensics. Compared to traditional networks, data capture faces significant challenges in cognitive radio networks. In traditional wireless networks, usually one monitor is assigned to one channel for traffic capture. This approach will incur very high cost in cognitive radio networks because it typically has a large number of channels. Furthermore, due to the uncertainty of the primary user's behavior, cognitive radio devices change their operating channels dynamically, which makes data capturing more difficult. In this paper, we propose a systematic method to capture data in cognitive radio networks with a small number of monitors. We utilize incremental support vector regression to predict packet arrival time and intelligently switch monitors between channels. We also propose a protocol that schedules multiple monitors to perform channel scanning and packet capturing in an efficient manner. Monitors are reused in the time domain, and geographic coverage is taken into account. The real-world experiments and simulations show that our method is able to achieve the packet capture rate above 70% using a small number of monitors, which outperforms the random scheme by 200%-300%.

Video source identification is very important in validating video evidence, tracking down video piracy crimes and regulating individual video sources. With the prevalence of wireless communication, wireless video cameras continue to replace their wired counterparts in security/surveillance systems and tactical networks. However, wirelessly streamed videos usually suffer from blocking and blurring due to inevitable packet loss in wireless transmissions. The existing source identification methods experience significant performance degradation or even fail to work when identifying videos with blocking and blurring. In this paper, we propose a method which is effective and efficient in identifying such wirelessly streamed videos. In addition, we also propose to incorporate wireless channel signatures and selective frame processing into source identification, which significantly improve the identification speed.

In cognitive radio networks, an attacker transmits signals mimicking the characteristics of primary signals, in order to prevent secondary users from transmitting. Such an attack is called primary user emulation (PUE) attack. TV towers and wireless microphones are two main types of primary users in white space. Existing work on PUE attack detection only focused on the first category. For the latter category, primary users are mobile and their transmission power is low. These properties introduce great challenges on PUE detection and existing methods are not applicable. In this paper, we propose a novel method to detect the emulation attack of wireless microphones. We exploit the relationship between RF signals and acoustic information to verify the existence of wireless microphones. The effectiveness of our approach is validated through real-world implementation. Extensive experiments show that our method achieves both false positive rate and false negative rate lower than 0.1 even in a noisy environment.

Transmit power estimation is widely used in network monitoring, power-aware design of MANETs, primary user detection in cognitive radio networks and many other areas. Traditional methods for transmit power estimation are trilateration-based, which require an underlying infrastructure with at least three monitors. In this paper, we propose a novel transmit power estimation method which utilizes the nuance of the received signal strength at different frequencies. Our method only needs one monitor, thus has less hardware requirement and is much easier to carry around. We use a support vector machine to facilitate the estimation, and conduct real-world experiments to validate our method. The experimental results demonstrate that our method is able to achieve the accuracy as high as 90%, which in practice outperforms the trilateration method using multiple monitors.

Network forensics is widely used in tracking down criminals and detecting network anomalies, and data capture is the basis of network forensics. Compared to traditional networks, data capture faces significant challenges in cognitive radio networks. In traditional wireless networks, one monitor is usually assigned to one channel to capture traffic, which incurs very high cost in a cognitive radio network because the latter typically has a large number of channels. Furthermore, due to the uncertainty of the primary user's activity, cognitive radio devices change their operating channels randomly, which makes data capturing more difficult. In this paper, we propose a systematic method to capture data in cognitive radio networks with a small number of monitors. We utilize incremental support vector regression to predict packet arrival time and intelligently switch monitors between channels. In addition, a protocol is proposed to schedule multiple monitors to perform channel scan and packet capturing in an efficient manner. The real-world experiments and simulations show that our method is able to achieve the packet capture rate above 70% using a small number of monitors, which outperforms the random scheme by 200%-300%.

In cognitive radio networks, an adversary transmits signals whose characteristics emulate those of primary users, in order to prevent secondary users from transmitting. Such an attack is called primary user emulation (PUE) attack. There are two main types of primary users in white space: TV towers and wireless microphones. Existing work on PUE attack detection focused on the first category. However, for the latter category, primary users are mobile and their transmission power is low. These unique properties of wireless microphones introduce great challenges and existing methods are not applicable. In this paper, we propose a novel method to detect the PUE attack of mobile primary users. We exploit the correlations between RF signals and acoustic information to verify the existence of wireless microphones. The effectiveness of our approach is validated through extensive real-world experiments. It shows that our method achieves both false positive rate and false negative rate lower than 0.1.

Channel surfing is an effective method to prevent jamming attacks in wireless communications. In traditional channel surfing schemes, two parties have to negotiate beforehand, in order to agree on the channel switching sequence. However, the negotiation process itself is vulnerable to jamming attacks. In this paper, we propose a novel channel surfing method without relying on such negotiation. Taking advantage of the reciprocity of the wireless fading channel, our method switches channels according to the random channel states observed by the two parties during their communication. Therefore, it does not introduce any extra communication overhead and can achieve strong security. To evaluate our method, we carry out extensive experiments using off-the-shelf 802.11 devices in a real indoor environment. Experimental results validate the efficiency and security of our method.

Building large-scale smart environments is one of the long-term goals of ubiquitous computing. The widespread of context information in such environments necessitates an effective context retrieval mechanism. This paper proposes a novel context retrieval method based on applications' query patterns. We propose high dimensional vector to model contexts from applications' perspective, and apply the normalized inner product of high dimensional vectors to measure context distance. Contexts with similar query patterns are clustered into the same group. To improve the performance of context retrieval, we build distributed indices on each node to speed up a local search, and create shortcuts based on clustering results to facilitate query routing. We show how our proposed methods can be applied to existing context retrieval mechanisms. Our experimental results show that our method can significantly reduce retrieval cost.

Pervasive computing is to enhance the environment by embedding many computers that are gracefully integrated with human users. To achieve this, the key research thrust is to create a smart context-awareness environment which should enclose various users and satisfy different needs of the users. Building such smart environments is still difficult and complex due to lacking a uniform infrastructure that can adapt to diverse smart domains. To address this problem, we propose a context-aware computing infrastructure, called FollowMe. Our infrastructure integrates an ontology based context model and a workflow based application model with the OSGi framework. By plugging different domain contexts and applications, FollowMe can be customized to various domains.