Search

With the advent of the network era, network security has attracted more and more attention. As the key node in the network, network key devices play an important role in protecting the internal network and ensuring the network security. For the vulnerability of network key devices, security analysis has become an important concern of security personnel. Fuzzing is an automatic and effective vulnerability mining technology. In this paper, we propose the first coverage feedback fuzzy testing framework based on KVM PT technology for network key devices, aiming to solve the feasibility of applying fuzzy tools in network critical devices. At the same time, a fuzzy test agent technology based on firmware modification is proposed to help speed up the call of testcases. We evaluated the framework on Cisco ASA firewall, and trigger CVE-2018-0101, which proves the effectiveness of the framework.

As an effective defect detection technology for low-oil power equipment (oil-immersed CT, PT, high-pressure casing, etc.), pressure detection is attracting more and more researchers' attention. However, the current research on pressure detection is mainly focused on the experimental research of pressure threshold, and there is no research on the fault diagnosis technology of low-oil equipment based on pressure. This paper proposes a kind of based on membership degree of less oil pressure equipment fault diagnosis method. In the proposed approach, the pressure variation characteristics of low-oil equipment with different defects are compared and analyzed and the pressure is selected as the characteristic quantity. Fuzzy C-means clustering algorithm is used to classify the fault degree. According to the actual situation of oil pressure detection, the fault degree can be divided into four categories: no fault, light fault, moderate fault and severe fault, and the membership calculation formula based on pressure parameters under four types of fault is established in this paper. Finally, an example is given to prove the validity of the diagnostic method. The results of this paper have a certain engineering significance for the pressure-based state evaluation of low-oil equipment.

Code cloning considerably facilitates software development but also leads to recurring bugs and other software quality problems. In this paper, we propose a fast code clone detection method based on weighted recursive autoencoders (RAE) to measure code similarity at the function level. Different from manually defining features for code clone detection, our deep learning-based method can automatically learn program features. First, we analyze program abstract syntax trees using weighted RAE, extract the program features and encode the functions to vectors. During the modeling process, we consider node weight information in abstract syntax trees to increase the proportion of information contributed by important nodes in the final vector representation of one program. Second, we report functions with similar vectors as code clone pairs. The second phase is time consuming when analyzing large software systems because it needs quadratic pairwise comparisons. To solve this problem, we transform the clone detection problem into an approximate nearest neighbors search (ANNS) in a high-dimensional vector set and use the navigating spreading-out graph to reduce the computational time complexity. Experimental results on BigCloneBench show that our method outperforms the compared algorithm based on unweighted RAE in terms of precision, recall, and AUC value and can return clone pairs in approximately 33 min, while the compared algorithm requires approximately 14 days when performing pairwise comparisons among 785,438 functions’ vectors. Our method also outperforms many prominent tools, including Oreo, in detecting Moderately Type-3 or Type-4 clones, and our false positive rate (FPR) equals 0.055, which means few false positives. More importantly, our method has no need for labeled data, and all of the source code is released to guarantee experimental reproducibility.

The intermittence and fluctuation of output are the key factors when dealing with photovoltaic (PV) access capacity in distribution network planning. After taking the random characteristics of PV output and its relevance to load into account, the effective utilization of PV power generation capacity can be realized. This paper first adopts a Gaussian mixture model (GMM) to address the uncertainty issue of PV generation. Then, with consideration of the network structure and the temporal-spatial distribution characteristics of the load, a practical model for the evaluation of PV penetration limits is established. The effectiveness of the proposed model is validated through a case study, in which several factors that affect PV penetration are examined. This model is currently applied in an area in China for the determination of the penetration limits applied to existing and new solar farms.

Since the classical hash function widely used are costly for resource constrained devices, in this paper we propose a new lightweight hash function LNHASH through Cellular Automata(CA), in which sponge construction is applied as the mode of operation. Both linear and non-linear CA rules are employed to construct the internal permutation to achieve high diffusion and confusion. LNHASH allows making trade-offs among security, speed, energy consumption and implementation costs by adjusting the corresponding parameters. The hardware-friendly Sbox and linear layer bring low area implementation. Security analysis results show the resistance of the LNHASH scheme against the known attacks.

In power system, information transmission pays more attention to security, stability and reliability. In order to meet the requirement of safety protection of smart grid automation control, a safety protection scheme of distribution automation based on FTU is proposed. Without changing the current operation of distribution terminals, the characteristics of distribution automation system are provided, such as wide range of points, wide distribution and outdoor operation. And due to industrial-level hardware design encryption chip, secure operating system , SM1, SM2, SM3, SM4 national secret algorithm and PKI system (public key infrastructure) are adopted. Public Key Infrastructure (PKI) power digital dispatch certificate uses digital certificate authentication technology and encryption technology based on domestic commercial cryptographic algorithm. Its aim is to realize bidirectional identity authentication between distribution master station and distribution terminal. In the future distribution network business, hardware definition platform, software definition products, management and business separation are the development trend, distribution terminals support plug and play, and great challenges are faced by information security The implementation of the security protection scheme is of great significance to system security, which can effectively prevent hackers from malicious attacks on distribution network equipment and ensure the confidentiality, integrity and availability of distribution network data.

3D point cloud data, which are produced by various 3D sensors such as LIDAR and stereo cameras, have been widely deployed by industry leaders such as Google, Uber, Tesla, and Mobileye, for mobile robotic applications such as autonomous driving and humanoid robots. Point cloud data, which are composed of reliable depth information, can provide accurate location and shape characteristics for scene understanding, such as object recognition and semantic segmentation. However, deep neural networks (DNNs), which directly consume point cloud data, are particularly computation-intensive because they have to not only perform multiplication-and-accumulation (MAC) operations but also search neighbors from the irregular 3D point cloud data. Such a task goes beyond the capabilities of general-purpose processors in real-time to figure out the solution as the scales of both point cloud data and DNNs increase from application to application. We present the first accelerator architecture that dynamically configures the hardware on-the-fly to match the computation of both neighbor point search and MAC computation for point-based DNNs. To facilitate the process of neighbor point search and reduce the computation costs, a grid-based algorithm is introduced to search neighbor points from a local region of grids. Evaluation results based on the scene recognition and segmentation tasks show that the proposed design harvests 16.4$\times$ higher performance and saves 99.95% of energy than an NVIDIA Tesla K40 GPU baseline in point cloud scene understanding applications.

Microservice architecture splits giant and complex enterprise applications into fine-grained microservices, promoting agile development, integration, delivery and deployment. However, monitoring tens of thousands of microservices is extremely challenging, and debugging problems among massive microservices is like looking for a needle in a haystack. We present JCallGraph, a tracing and analytics tool to capture and visualize the microservice invocation relationship of tens of thousands of microservices with millions of containers at JD.com. JCallGraph achieves three main goals for distributed tracing and debugging: online microservices invocation construction within milliseconds, minimal overhead without any significant performance impact on real-production applications, and application-agnostic with zero-intrusion to application. Our evaluation shows that JCallGraph can accurately capture the real-time invocation relationship at massive scale and help developers to efficiently understand interactions among microservices, pinpoint root-cause of problems.

Maximizing the potential value of demand-side resources in terms of network reliability and economy is a considerable challenge because of their time-varying and potentially complex operational characteristics. It is urgent to take the interactive relationship between supply and demand side into consideration to improve the efficiency and practicability of the distribution network planning and enhance the utilization of existing equipment. In this paper, the uncertainty of DG output and the load is first modelled, and the seasonal label information is added to the model to generate the planning-oriented scenarios of DG output and the load. Considering the integration of demand-side resources, a bi-level optimal planning model is established to minimize the annual total cost of the distribution network. The proposed model is tested on the IEEE 33-node distribution system to illustrate its feasibility and validity in distribution network planning with consideration of the integration of demand-side resources.

Modeling and simulation is becoming one of the most powerful techniques in the development of modern automotive systems. UPPAAL, as a toolbox for modeling, simulation and verification of real-time systems, has by now reached a state, where they are mature enough for application on a number of case-studies in real-time controllers and real-time communication protocols. However the simulation of UPPAAL is not really real-time. In this paper we explore an approach for real-time simulation of the UPPAAL model. We represent an UPPAAL model to C++ code. Based on a transformed UPPAAL model, real-time simulation can be taken. The approach can be used to real-time simulating an UPPAAL model, and many works can be done based on the transformed UPPAAL model.

Understanding meaningful information such as transportation mode (e.g., walking, bus) from Global Positioning System (GPS) data has great advantages in urban management and environmental protection. However, the urban traffic environment has evolved from “data poor” to “data rich”, resulting in the decline in the accuracy of transportation mode detection results. In this paper, an enhanced approach for effectively detecting transportation mode with a detection model and correction method from GPS data is proposed. Specifically, we make the following contributions. First, a trajectory segmentation method is proposed to detect single-mode segments. Secondly, a Random Forest (RF)-based detection model containing several new features is introduced to enhance discrimination. Finally, a correction method is designed to improve the detection performance, which is based on the mode probability of the current segment and its adjacent segments. The results of our detection model and correction method outperform state-of-the-art research in transportation mode detection.

Based on the steel corridor of the first-stage construction for Panzhihua City Government Service Centre, this paper introduces the integral lifting scheme of multi-storey long-span beam grillage steel corridor. To ensure the safety and stability in the lifting process of steel corridor construction and to prevent sudden overload from affecting the later normal use, the finite element simulation of steel corridor before lifting and real-time monitoring during the lifting process were carried out. The results of finite element simulation were compared with the stress and deflection data monitored in real time. It shows that the numerical data obtained from the finite element methods (FEM) are slightly larger than those from the actual monitoring. It follows that the results obtained from FEM for the lifting is safer, the data obtained from FEM and monitoring agree with each other, and, accordingly, the study shows that the monitoring approach is applicable. The integral lifting scheme and monitoring approach presented here can be used as an interesting reference for similar projects.

The communication efficiency of large-scale WSNs with high node density is highly related with specific forwarding strategy in complicated wireless environments. Thus, it is the key problem that how to use the characteristic of high-density to select appropriate forwarding nodes to relieve the load imbalance and improve the communication efficiency. In this paper, we first propose a Fine-grain Gradient Sinking (FGS) model and several data forwarding strategies based on this model. Then we give the criteria of communication efficiency and emphatically analyze the communication efficiency of various forwarding strategies. Simulation results show that forwarding strategies under FGS model can achieve higher communication efficiency, and the efficiency can be further elevated by introducing the statistics of packet receive rate during the forwarding process.

This paper addresses the problem of test response compaction. In order to maximize compaction ratio, a single-output compactor based on a (n, n−1, m, 3) convolutional code is presented. When the proposed theorems are satisfied, the compactor can avoid two and any odd erroneous bits cancellations, and handle one unknown bit (X bit). When the X bits in response are clustered, multiple-weight check matrix design algorithm can be used to reduce the effect of massive X bits. Some extended experimental results show that the proposed encoder has an acceptable-level X tolerant capacity and low error cancellations probability.

The fast and accurate customer classifier is the core of Web marketing. But the common customer classifier doesn't fit for the customer classification in Web marketing, because it has certain industry limitation. In this paper, a decision tree classifier based on information entropy is proposed. First of all, the classification index system for Web marketing is established by integrating several common-used classification index strategies. After that, the information gain of each classification index is calculated by using information entropy. Finally, the decision tree is established according to the information gain of classification index, and the corresponding customer classification rule is generated. To test this method, the real e-commerce site data is divided into two parts. One is used to establish the customer classification decision tree. The other is used to test. The experiment result shows that the forecast accuracy of decision tree is about 97%, which meets the requirement of the actual classification work.

At-speed testing using external tester requires an expensive equipment, thus built-in self-test (BIST) is an alternative technique due to its ability to perform on-chip at-speed self-testing. The main issue in BIST for at-speed testing is to obtain high delay fault coverage with a low hardware overhead. This paper presents an improved loop-based BIST scheme, in which a configurable MISR (multiple-input signature register) is used to generate test-pair sequences. The structure and operation modes of the BIST scheme are described. The topological properties of the state-transition-graph of the proposed BIST scheme are analyzed. Based on it, an approach to design and efficiently implement the proposed BIST scheme is developed. Experimental results on academic benchmark circuits are presented to demonstrate the effectiveness of the proposed BIST scheme as well as the design approach.

Detection of path delay faults requires two-pattern tests. BIST technique provides a low-cost test solution. This paper proposes an approach to designing a cost-effective deterministic test pattern generator (TPG) for path delay testing. Given a set of pre-generated test-pairs with pre-determined fault coverage, a deterministic TPG is synthesized to apply the given test-pair set in a limited test time. To achieve this objective, configurable linear feedback shift register (LFSR) structures are used. Techniques are developed to synthesize such a TPG, which is used to generate an unordered deterministic test-pair set. The resulting TPG is very efficient in terms of hardware size and speed performance. Simulation of academic benchmark circuits has given good results when compared to alternative solutions.

This paper considers a BIST TPG that can highly reduce the power/energy consumption during test application without losing stuck-at fault coverage. Experimental results based on the ISCAS'85 and 89 benchmark circuits are reported to demonstrate the effectiveness of our approach.

It is a new direction to introduce the function S-rough sets to information security theory and its application. In this paper, we propose two novel schemes of the hiding and authentication of function S-rough sets laws based on subliminal channel technique, which has overcame the shortcomings of the traditional scheme that the hiding and authentication implemented step by step.

According to road and traffic characteristic indexes of high-way network, this paper illustrates a new model, which can evaluate which is the best among many programming schemes, the paper firstly researches the analyzing model for high-way network optimization which takes the evaluation objectives as set of indexes and programming schemes as set of domain by quantitative analysis, and establishes the standard index decision-making matrix and positive-negative ideal scheme decision-making matrix according to the TOPSIS. Moreover, the weights of decision-making model are determined comprehensively by means of AHP and entropy which takes the experts' knowledge and experience, decision makers' intentions and preferences and information within data itself into consideration. In addition, the optimization degree was judged by calculating the closeness of scheme matrix which runs from the ideal scheme. The example shows that this method can adjust programming schemes better, and the result was much more reasonable and practical.

In this paper, a new methodology for the design of compact planar lowpass filters in microstrip technology is proposed. This is based on cascading filter stages consisting of the combination of complementary open-loop resonators (COLR), proposed by the authors for the first time, and open stubs. By this means, we obtain sharp cutoff, compactness, and low passband insertion-loss. To illustrate the potentiality of the proposed approach, a compact S-band microstrip lowpass filter with COLR was designed, fabricated and measured.

Many applications of wireless sensor networks (WSN) are based on self-positioning of sensor nodes. In this paper, an efficient robust location algorithm is proposed based on time measurement, requiring no global time synchronization in WSN and minimal extra hardware in sensor construction phase. We first utilize the TDOA measurements to locate target sensor node, then optimize the estimated values, which are educed from the algorithm as observed values for extended Kalman filtering and position estimation. It is shown the algorithm mentioned in the paper is close to CRLB (Cramer-Rao lower bound) for the static location estimator.

Basis set of path is consisted of some of the program’s paths. The automatic generation method of basis set of path is discussed in this paper. It is built by searching the control flow graph of a program by depth-first searching method. In order to avoiding that the algorithm will never stop and reducing the searching procedure, the sub-path from the multi-indegree nodes to the end node of a program and the sub-path that contains a loop is recorded during the construction of a basis path. Some new basis paths can be constructed by merging these two kinds of sub-paths.

Data sinking is one of the typical transmission patterns in WSN(wireless sensor network) . There is inherent unbalanced traffic load distribution in such funnel like transmission. A case in hop-based sinking(HBS) model is found more intricate than simple thought that inner nodes burden more forwarding tasks,showing the inverse direction within the same hop level comparing with global trend. With a simple weighted average mechanism,a continuous gradient parameter is introduced,which will be dedicated to instructing how to forward data to sink in place of hop count,namely fine-grain gradient sinking(FGS) . Through traffic analysis and detailed simulation,in FGS model the network turns out to be smoother on traffic load distribution and more efficient on data forwarding than that in HBS model.