Search

APT detection is difficult to detect due to the long-term latency, covert and slow multistage attack patterns of Advanced Persistent Threat (APT). To tackle these issues, we propose TBDetector, a transformer-based advanced persistent threat detection method for APT attack detection. Considering that provenance graphs provide rich historical information and have the powerful attacks historic correlation ability to identify anomalous activities, TBDetector employs provenance analysis for APT detection, which summarizes long-running system execution with space efficiency and utilizes transformer with self-attention based encoder-decoder to extract long-term contextual features of system states to detect slow-acting attacks. Furthermore, we further introduce anomaly scores to investigate the anomaly of different system states, where each state is calculated with an anomaly score corresponding to its similarity score and isolation score. To evaluate the effectiveness of the proposed method, we have conducted experiments on five public datasets, i.e., streamspot, cadets, shellshock, clearscope, and wget_baseline. Experimental results and comparisons with state-of-the-art methods have exhibited better performance of our proposed method., Comment: 10 pages, 7 figures

Motivated by the concept of eddy viscosity tensor in improved Boussinesq hypothesis, a transport model of high-order eddy viscosity tensor in 2D-3C turbulence structure is derived from the second-order moment model by tensorial analysis.

Cuproptosis is a novel programmed cell death that depends on copper. The role and potential mechanism of cuproptosis-related genes (CRGs) in thyroid cancer (THCA) are still unclear. In our study, we randomly divided THCA patients from the TCGA database into a training set and a testing set. A cuproptosis-related signature consisting of six genes (SLC31A1, LIAS, DLD, MTF1, CDKN2A, and GCSH) was constructed using the training set to predict the prognosis of THCA and was verified with the testing set. All patients were classified into low- and high-risk groups according to risk score. Patients in the high-risk group had a poorer overall survival (OS) than those in the low-risk group. The area under the curve (AUC) values for 5 years, 8 years, and 10 years were 0.845, 0.885, and 0.898, respectively. The tumor immune cell infiltration and immune status were significantly higher in the low-risk group, which indicated a better response to immune checkpoint inhibitors (ICIs). The expression of six cuproptosis-related genes in our prognostic signature were verified by qRT-PCR in our THCA tissues, and the results were consistent with TCGA database. In summary, our cuproptosis-related risk signature has a good predictive ability regarding the prognosis of THCA patients. Targeting cuproptosis may be a better alternative for THCA patients.

Deep learning technologies have demonstrated remarkable effectiveness in a wide range of tasks, and deep learning holds the potential to advance a multitude of applications, including in edge computing, where deep models are deployed on edge devices to enable instant data processing and response. A key challenge is that while the application of deep models often incurs substantial memory and computational costs, edge devices typically offer only very limited storage and computational capabilities that may vary substantially across devices. These characteristics make it difficult to build deep learning solutions that unleash the potential of edge devices while complying with their constraints. A promising approach to addressing this challenge is to automate the design of effective deep learning models that are lightweight, require only a little storage, and incur only low computational overheads. This survey offers comprehensive coverage of studies of design automation techniques for deep learning models targeting edge computing. It offers an overview and comparison of key metrics that are used commonly to quantify the proficiency of models in terms of effectiveness, lightness, and computational costs. The survey then proceeds to cover three categories of the state-of-the-art of deep model design automation techniques: automated neural architecture search, automated model compression, and joint automated design and compression. Finally, the survey covers open issues and directions for future research.

To date, driver genes for pancreatic cancer treatment are difficult to pursue therapeutically. Targeting mutated KRAS, the most renowned driver gene in pancreatic cancer, is an active area of study. We discovered a gene named SEMA3C was highly expressed in pancreatic cancer cell lines and patients with a G12D mutation in KRAS. High expression of SEMA3C in patients was significantly associated with the decreased survival of pancreatic cancer patients based on the TCGA database. In pancreatic cancer cells, SEMA3C knockdown or inhibition exhibited growth/colony inhibition and cell cycle arrest. In addition, SEMA3C inhibition sensitized KRAS or MEK1/2 inhibition in pancreatic cancer cells. Overexpression of SEMA3C resulted in the induction of autophagy, whereas depletion of SEMA3C compromised induction of autophagy. SEMA3C modified the PD-L1 expression in tumor and immune cells and is correlated with the M2-like macrophage marker ARG1/CD163 expression, which could reshape the tumor microenvironment. Inhibition of SEMA3C decreased tumor formation in the xenograft model in vivo. Taken together, our data suggest that SEMA3C plays a substantial role in promoting cancer cell survival by regulating the autophagy process and impacting the tumor environment immune response. SEMA3C can be used as a novel target or marker with therapeutic or diagnostic potential in pancreatic cancer especially in tumors harboring the specific KRAS G12D mutation.

Additional file 2. Supplementary materials and methods.

Additional file 3: Figure S1. FTO inhibits cell proliferation in PTC. a. Overexpression efficiency of FTO in K1 cells determined by qRT-PCR and western blotting. b. Knockdown efficiency of si-FTO-1/2 in TPC1 cells as determined by qRT-PCR and western blotting. c. Total m6A modification levels after FTO knockdown or overexpression in PTC cells as determined by m6A RNA Methylation Quantification Kit. d-e. CCK-8 assay (d) and colony formation assay (e) showing proliferation ability in K1 cells after FTO overexpression. f-g. CCK-8 assay (f) and colony formation assay (g) showing proliferation ability after FTO knockdown in TPC1 cells. h. Flow cytometry was performed to measure cell cycle distribution after FTO knockdown in K1 cells. i. Flow cytometry was performed to measure cell cycle distribution after FTO overexpression in TPC1 cells. j. Knockdown efficiency of sh-FTO in K1 cells determined by western blotting. *P

Feature extraction and classification play an important role in brain–computer interface (BCI) systems. In traditional approaches, methods in pattern recognition field are adopted to solve these problems. Nowadays, the deep learning theory has developed so fast that researchers have employed it in many areas like computer vision and speech recognition, which have achieved remarkable results. However, few people introduce the deep learning method into the study of biomedical signals, especially EEG signals. In this paper, a wavelet transform-based input, which combines the time-frequency images of C3, Cz, and C4 channels, is proposed to extract the feature of motor imagery EEG signal. Then, a 2-Layer convolutional neural network is built as the classifier and convolutional kernels of different sizes are validated. The performance obtained by the proposed approach is evaluated by accuracy and Kappa value. The accuracy on dataset III from BCI competition II reaches 90%, and the best Kappa value on dataset 2a from competition IV is greater than many of other methods. In addition, the proposed method utilizes a resized small input, which reduces calculation complexity, so the training period is relatively faster. The results show that the method using convolutional neural network can be comparable or better than the other state-of-the-art approaches, and the performance will be improved when there is sufficient data.

Predicting human intentions paves the way of responding to personal requirements accordingly and thus realizes efficient human-machine interactions. Diverse applications, such as human-robot cooperation and smart assisted living, can benefit from precisely predicting human intentions. The electroencephalogram (EEG) signals that reflect electrical dynamics from the human scalp are commonly used to read human's intentions directly. Most conventional approaches to decoding EEG signals heavily depend on engineering experience and domain knowledge and are not able to provide satisfactory performance for widely practical usage. Recently, deep learning technology has shown dominant performance in various fields, and its development for EEG decoding has also sprung up. Considering the limitations of conventional methods for predicting human intentions and the superiority of deep learning technology, in this thesis, I design a set of deep learning frameworks from three aspects of human intention prediction. First, for the subject-dependent application, I design two end-to-end deep learning models with novel spatio-temporal preserving representations of raw EEG data to precisely identify human intentions. I further develop and realize a real-world brain typing prototype with the proposed deep learning frameworks. Besides the end-to-end fashion, I also propose a neural network ensemble model by using a reinforcement learning enhancing scheme. Given that the subject-dependent application requires an adaptation before ready to use on a new subject, I further design a series of models focusing on subject-independent applications. I first develop original deep learning models by leveraging the well-developed self-attention mechanism for effectively capturing the temporal dynamics of EEG streams. Moreover, I design a series of graph representations on top of the attention-based deep learning models to encode nonstructural EEG nodes. Lastly, even though the developed deep learning frameworks demonstrate outstanding performance, they possess the risk of leaking privacy information unintentionally. In view of this gap, I propose a novel user privacy protection framework that transforms raw data into a new format that has a "style" (privacy information) of random noise and a "content" (human intentions) of the raw data. Hence, it is free of user privacy information for training and able to protect all sensitive information at once collectively.

As a promising candidate of general-purpose transport layer protocol, the Stream Control Transmission Protocol (SCTP) has its new features such as multi-homing and multi-streaming. SCTP association can make concurrent multi-path transfer an appealing candidate to satisfy the ever increasing user demands for bandwidth by using Multi-homing feature. And multiple streams provide an aggregation mechanism to accommodate heterogeneous objects, which belong to the same application but may require different QoS from the network. In this paper, the authors introduce WM2-SCTP (Wireless Multi-path Multi-flow - Stream Control Transmission Protocol), a transport layer solution for concurrent multi-path transfer with parallel sub-flows. WM2-SCTP aims at exploiting SCTP's multi-homing and multi-streaming capability by grouping SCTP streams into sub-flows based on their required QoS and selecting best paths for each sub-flow to improve data transfer rates. The results show that under different scenarios WM2-SCTP is able to support QoS among the SCTP stream, and it achieves a better throughput.

With the development of smart grid, operation and control methods of power grid are more abundant and flexible, which makes contribution to solve the uncertain risks introduced by photovoltaic (PV) power plants, and also puts forward higher requirements for the decision-making effects of power grid. This paper first analyzes the role of three measures in the regulation and control of photovoltaic-integrated power system from the perspective of security and economy, namely inverter phase modulation, grid dynamic segmentation, and distributed power flow controller (DPFC). Then, taking DPFC as an example, a quantitative risk assessment and control method for photovoltaic-integrated power system is proposed. This method takes into account the uncertainty of photovoltaic output and the regulatory costs to quantify the operational control risk and control effect of the PV-integrated power system. Finally, IEEE RTS79 test system is established to verify the rationality and effectiveness of the proposed method.

Sodium-cooled fast reactor (SFR) is one of most promising Generation IV reactor technology and has a rapid development in recent years. Experimental Breeder Reactor II (EBR-II) designed by Argonne National Laboratory (ANL) is a typical sodium-cooled fast reactor with a sodium-bonded metallic fuel core, featured with reactor negative reactivity feedback. In order to verify and validate (V&V) the inherent safety performance of sodium-cooled fast reactor, the International Atomic Energy Agency (IAEA) established the coordinated research project (CRP) in which Xi’an Jiaotong University has participated. The “Benchmark Analysis of EBR-II shutdown heat removal tests” was conducted by ANL. SHRT-45R, namely Unprotected Loss of Flow (ULOF), is one of the reactor tests among many transients. Argonne National Laboratory has prepared a detailed benchmark specification and has provided the enough benchmark data for SHRT-45R. At the beginning of SHRT-45R both primary main pumps and the intermediate loop pump were synchronously tripped to simulate the unprotected loss of flow accident. During the test, the plant protection system (PPS) was disabled to initiate a control rod scram. The SHRT-45R test demonstrated that EBR-II could keep in safe during the potentially adverse consequences of unprotected accidents. This paper introduces the models for predicting SHRT-45R in detail and presents the results of the analysis of the Unprotected Loss of Flow (ULOF) test SHRT-45R performed in the EBR-II reactor. The thermal-hydraulic calculations are performed with Modified RELAP5 in which the thermodynamic and transport properties of liquid and vapor state sodium have been supplemented, as well as the specific heat transfer correlations. The numerical results show that the loss of forced coolant flow causes the coolant temperatures in the instrumented subassemblies XX09 and XX10 to increase to a peak point but keep at an acceptable level about 930K and 850K at the early state of accidents, and then the reactor can shut down by itself due to the negative reactivity feedback including Doppler reactivity feedback and density reactivity feedback after 70s. The variation of the key thermal-hydraulic parameters including the coolant temperatures and fuel cladding temperature in the instrumented subassembly has a good agreement with the experimental data in general. The results could not only verify and validate the inhere safety performance of sodium-cooled fast reactors during potentially unprotected accidents but also demonstrate the analysis capability of the modified RELAP5 for sodium-cooled fast reactors.

A novel hot-air anti-icing structure of engine inlet vane is put forward and anti-icing experiment is conducted in a small open-circuit icing wind tunnel. The structure is combined with impinging jet, micro channels, and exhaust releasing exhausted hot air to sweep running-back water off the coating. The experiment is run at different simulated meteorological and bleed air conditions. The vane surface temperature and running-back water film distributions of two kinds of micro channel with channel spacing of 1.5 mm and 2 mm are obtained. The results indicate that impinging jet ensures the anti-icing effect of leading edge, smaller interval of micro channels brings better anti-icing effect at same bleed air condition, and temperature of bleed air has more influence than pressure. The vented out exhausted air protects the rear skin from running-back water icing effectively.

The molten salt reactor (MSR), which is one of the generation IV reactors, can meet the demand of transmutation and breeding. The thermodynamic properties of the molten salt system like LiF-NaF-BeF2 influence the design and construction of the fuel salt and coolant in the MSR for the new generation. In this paper, the equation of state of the ternary system 15%LiF-58%NaF-27%BeF2, over the temperature range from 873.15 to 1 073.15 K at one atmosphere pressure, is described using a modified Peng-Robinson (PR) equation. The densities of the ternary system and its components are estimated by this equation directly, and compared with the experimental data. Based on the equation of state, the other thermodynamic properties such as the enthalpy, entropy and heat capacity at constant pressure are estimated by the residual function method and the fugacity coefficient method respectively. The densities calculated by PR equation are highly in agreement with the experimental data, and the enthalpy, entropy and heat capacity evaluated by the two different methods are consistent with each other. It can be concluded that the modified PR equation can be applied to evaluate the density of the molten salt system, and it is recommended that it be used as the basis to estimate the enthalpy, entropy and heat capacity of the molten salt system.

Radar detection has been an important means of detection in military and civilian monitoring and protection. In view of the circumstances where traditional radar is difficult to be applied to the distributed network, we propose a distributed networked radar detection scheme. Through the introduction and use of the potocol stack of IOT(Internet of things), message oriented middleware, real time database, we design and implement distributed networked control system based on the B/S structure which makes radar detection more flexible and efficient. Practice shows that the system can work well, and can realize the linkage between the radar and other sensors.

To simulate the ice accretion on an airfoil, a boundary moving technique is proposed to deal with the distortion of the airfoil surface due to ice accretion on the leading edge. Incorporating the two-phase model of air-supercooled droplets in the Eulerian coordinate system, this technique is applied to simulate the process of the rime ice accretion (the droplets freeze at the instant impinging on the airfoil) on the NACA 0012 airfoil, and the ice profile after ice accretion is achieved successfully. A brief comparison between the results of this paper and the experiment data indicates that the current method is applicable and effective. © 2005 Wiley Periodicals, Inc. Heat Trans Asian Res, 34(4): 226–234, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/htj.20064

Impingement heat transfer from rib roughened surface within two-dimensional arrays of circular jet has been investigated experimentally. After the jet impinges on the rib roughened surface parallel to the jet plate, it is constrained to exit in a single direction along the channel formed by the jet plate and the rib roughened surface. An initial crossflow is present which approaches the arrays through an upstream extension of the channel. The configurations considered are intended to simulate the impingement cooling midchord region of the gas turbine aerofoils in case where an initial crossflow is also present. The study covered four different relative positions of the jet hole to the ribs: jet hole before the rib (−p/4), jet hole on the rib, jet hole behind the rib (+p/4) and jet hole between the ribs (midst,+p/2). The tests were performed for Reynolds number Re = 8000 and 15000, and the nondimensional jet-to-surface spacing z/d = 1.4, 2.0 and 3.0. The test results show that the impingement heat transfer from the rib roughened surface can be considerably improved by adequately arranging the relative position of the jet hole to the ribs.Copyright © 1997 by ASME

The flow behavior of a jet impinging on a ribbed surface has an obvious effect on the impingement heat transfer. It has been observed experimentally in a special smoke wind tunnel with rectangular section. This is an interesting problem in the impingement cooling midchord region of the gas turbine aerofoils with ribbed surface, and also in the electronic component cooling. The effect of the structural parameters, such as rib height (e), pitch-to-height ratio (p/e), nondimensional jet-to-surface spacing (Z/d) and the relative position of the jet hole to the ribs etc., and the crossflow rate on the flow behavior have been investigated by the flow visualization on a variety of configuration models. The experimental results reveal the mechanism of flow behavior and explain the flow and heat transfer characteristics of the impingement cooling from ribbed surface with crossflow.Copyright © 1996 by ASME

The Molten Salt Reactor (MSR), one of the `Generation IV' concepts, is a liquid-fuel reactor, which is different from the conventional reactors using solid fissile materials due to the flow effect of fuel salt. The study on its neutronics considering the fuel salt flow, which is the base of the thermal-hydraulic calculation and safety analysis, must be done. In this paper, the theoretical model on neutronics under steady condition for a single-liquid-fueled MSR is conducted and calculated by numerical method. The neutronics model consists of two group neutron diffusion equations for fast and thermal neutron fluxes, and balance equations for six-group delayed neutron precursors considering the flow effect of fuel salt. The spatial discretization of the above models is based on the finite volume method, and the discretization equations are computed by the source iteration method. The distributions of neutron fluxes and the distributions of the delayed neutron precursors in the core are obtained. The numerical calculated results show that, the fuel salt flow has little effect on the distribution of fast and thermal neutron fluxes and the effective multiplication factor; however, it affects the distribution of the delayed neutron precursors significantly, especially the long-lived one. In addition, it could be found that the delayed neutron precursors influence the neutronics slightly under the steady condition.