Search for: All records

We prepare quasi-1D films of Sb2Se3 on GaAs by molecular beam epitaxy. The aligned grains and anisotropic bonding hierarchy of the Sb2Se3 unit cell together produce giant birefringence in the near-infrared. 

PbSe is a narrow bandgap IV–VI compound semiconductor with application in mid-wave infrared optoelectronics, thermoelectrics, and quantum devices. Alkaline-earth or rare-earth elements such as Sr and Eu can substitute Pb to widen the bandgap of PbSe in heterostructure devices, but they come with challenges such as deteriorating optical and electronic properties, even in dilute concentrations due to their dissimilar atomic nature. We substitute Pb instead with column IV Ge and assess the potential of rocksalt phase PbGeSe as a wider bandgap semiconductor in thin films grown by molecular beam epitaxy on GaAs substrates. Low sticking of GeSe adatoms requires synthesis temperatures below 260 °C to incorporate Ge, but this yields poor structural and compositional uniformity as determined by x-ray diffraction. Consequently, as-grown films in the range Pb0.94Ge0.06Se–Pb0.83Ge0.17Se (6%–17% Ge) show much less bandgap widening in photoluminescence than prior work on bulk crystals using absorption. We observe that post-growth rapid thermal annealing at temperatures of 375–450 °C improves the crystal quality and recovers bandgap widening. Rapid interdiffusion of Ge during annealing, however, remains a challenge in harnessing such PbGeSe materials for compositionally sharp heterostructures. Annealed 17% Ge films emit light at 3–3.1 μm with a minimal shift in wavelength vs temperature. These samples are wider in bandgap than PbSe films by 55 meV at room temperature, and the widening increases to 160 meV at 80 K, thanks to sharply different dependence of bandgap on temperature in PbSe vs PbGeSe. 

We study the relationship between the dynamics of the action $$\alpha$$ of a discrete group $$G$$ on a von Neumann algebra $$M$$, and structural properties of the associated crossed product inclusion $$L(G) \subseteq M \rtimes_\alpha G$$, and its intermediate subalgebras. This continues a thread of research originating in classical structural results for ergodic actions of discrete, abelian groups on probability spaces. A key tool in the setting of a noncommutative dynamical system is the set of quasinormalizers for an inclusion of von Neumann algebras. We show that the von Neumann algebra generated by the quasinormalizers captures analytical properties of the inclusion $$L(G) \subseteq M \rtimes_\alpha G$$ such as the Haagerup Approximation Property, and is essential to capturing ``almost periodic" behavior in the underlying dynamical system. Our von Neumann algebraic point of view yields a new description of the Furstenberg-Zimmer distal tower for an ergodic action on a probability space, and we establish new versions of the Furstenberg-Zimmer structure theorem for general, tracial $W^*$-dynamical systems. We present a number of examples contrasting the noncommutative and classical settings which also build on previous work concerning singular inclusions of finite von Neumann algebras. 

We investigate the beneficial effects of rapid thermal annealing on structure and photoluminescence of PbSe thin films on GaAs (001) grown below 150 °C, with a goal of low temperature integration for infrared optoelectronics. Thin films of PbSe deposited on GaAs by molecular beam epitaxy are epitaxial at these reduced growth temperatures, yet the films are highly defective with a mosaic grain structure with low angle and dendritic boundaries following coalescence. Remarkably, we find that rapid thermal annealing for as short as 180 s at temperatures between 300 and 425 °C in nitrogen ambient leads to extensive re-crystallization and transformation of these grain boundaries. The annealing at the same time dramatically improves the band edge luminescence at 3.7 μm from previously undetectable levels to nearly half as intense as our best conventionally grown PbSe films at 300 °C. We show using an analysis of laser pump-power dependent photoluminescence measurements that this dramatic improvement in the photoluminescence intensity is due to a reduction in the trap-assisted recombination. However, we find it much less correlated with improved structural parameters determined by x-ray diffraction rocking curves, thereby pointing to the importance of eliminating point defects over extended defects. Overall, the success of rapid thermal annealing in improving the luminescent properties of low growth temperature PbSe is a step toward the integration of PbSe infrared optoelectronics in low thermal budget, back end of line compatible fabrication processes. 

Internet of Things (IoT) devices have increased drastically in complexity and prevalence within the last decade. Alongside the proliferation of IoT devices and applications, attacks targeting them have gained popularity. Recent large-scale attacks such as Mirai and VPNFilter highlight the lack of comprehensive defenses for IoT devices. Existing security solutions are inadequate against skilled adversaries with sophisticated and stealthy attacks against IoT devices. Powerful provenance-based intrusion detection systems have been successfully deployed in resource-rich servers and desktops to identify advanced stealthy attacks. However, IoT devices lack the memory, storage, and computing resources to directly apply these provenance analysis techniques on the device. This paper presents ProvIoT, a novel federated edge-cloud security framework that enables on-device syscall-level behavioral anomaly detection in IoT devices. ProvIoT applies federated learning techniques to overcome data and privacy limitations while minimizing network overhead. Infrequent on-device training of the local model requires less than 10% CPU overhead; syncing with the global models requires sending and receiving 2MB over the network. During normal offline operation, ProvIoT periodically incurs less than 10% CPU overhead and less than 65MB memory usage for data summarization and anomaly detection. Our evaluation shows that ProvIoT detects fileless malware and stealthy APT attacks with an average F1 score of 0.97 in heterogeneous real-world IoT applications. ProvIoT is a step towards extending provenance analysis to resource-constrained IoT devices, beginning with well-resourced IoT devices such as the RaspberryPi, Jetson Nano, and Google TPU. 

We present PROVNINJA, a framework designed to generate adversarial attacks that aim to elude provenance-based Machine Learning (ML) security detectors. PROVNINJA is designed to identify and craft adversarial attack vectors that statistically mimic and impersonate system programs. Leveraging the benign execution profile of system processes commonly observed across a multitude of hosts and networks, our research proposes an efficient and effective method to probe evasive alternatives and devise stealthy attack vectors that are difficult to distinguish from benign system behaviors. PROVNINJA's suggestions for evasive attacks, originally derived in the feature space, are then translated into system actions, leading to the realization of actual evasive attack sequences in the problem space. When evaluated against State-of-The-Art (SOTA) detector models using two realistic Advanced Persistent Threat (APT) scenarios and a large collection of fileless malware samples, PROVNINJA could generate and realize evasive attack variants, reducing the detection rates by up to 59%. We also assessed PROVNINJA under varying assumptions on adversaries' knowledge and capabilities. While PROVNINJA primarily considers the black-box model, we also explored two contrasting threat models that consider blind and whitebox attack scenarios.