“Soft-Man” and Data Mining based Distributed Intrusion Detection System

Jun Zheng 보안공학연구지원센터 2016 International Journal of Security and Its Applicat Vol.10 No.8

As modern computer networks are large-scale with numerous nodes, the conventional concentrated intrusion detection system fails to work effectively. To deal with the above situation, the paper proposed a “Soft-Man” and data mining based distributed intrusion detection system (SMDMDIDS, for short). Specifically, it designed an overall structure model of the detection system, expounded the system’s communication models, and designed the communication models and cooperation methods between Soft-Mans as well. The paper also defined hierarchical cooperation models for the Soft-Mans and designed corresponding data mining models. Finally, with the help of IDS Informer tools, the paper conducted a simulation experiment on network intrusion detection. The experimental results showed that the proposed intrusion detection system in the paper had good detection performance.

M-Pattern: A novel scheme for improving the security of Android Pattern unlock against smudge attacks

Jun Zheng,Sai Krishna Chigurupati 한국통신학회 2019 ICT Express Vol.5 No.3

Smudge attacks are a threat specifically targeting Android pattern unlock. Current solutions improve the security of pattern unlock by significantly reducing its usability. In this paper, we propose a novel scheme called M-Pattern to improve the security of pattern unlock against smudge attacks without significantly compromising the usability. A security analysis was performed to demonstrate that M-Pattern can have significantly stronger security than normal pattern unlock by just using two simple patterns. A pilot user study was also carried out to investigate the usability of M-Pattern.

Scanning Rayleigh Doppler Lidar for Wind Profiling Based on Non-polarized Beam Splitter Cube Optically Contacted FPI

Jun Zheng,Dongsong Sun,Tingdi Chen,Ruocan Zhao,Yuli Han,Zimu Li,Anran Zhou,Nannan Zhang 한국광학회 2018 Current Optics and Photonics Vol.2 No.2

A Scanning Rayleigh Doppler lidar for wind profiling based on a non-polarized beam splitter cube optically contacted FPI is developed for wind measurement from high troposphere to low stratosphere in 5-35 km. Non-polarized beam splitter cube optically contacted to the FPI are used for a stable optical receiver. Zero Doppler shift correction is used to correct for laser or FPI frequency jitter and drift and the timing sequence is designed. Stability of the receiver for Doppler shift discrimination is validated by measuring the transmissions of FPI in different days and analyzed the response functions. The maximal relative wind deviation due to the stability of the optical receiver is about 4.1% and the standard deviation of wind velocity is 1.6% due to the stability. Wind measurement comparison experiments were carried out in Jiuquan (39.741°N, 98.495°E), Gansu province of China in 2015, showing good agreement with radiosonde result data. Continuous wind field observation was performed from October 16th to November 12th and semi-continuous wind field of 19 nights are presented.

The use of 18F-FDG positron emission tomography to predict tumour recurrence in hepatocellular carcinoma patients after hepatectomy

Jun ZHENG,Tiffany CL WONG,Jeff WC DAI,Albert CY CHAN,Tan-To CHEUNG,Chung-Mau LO 한국간담췌외과학회 2022 Annals of hepato-biliary-pancreatic surgery Vol.26 No.-

Non-pulsed Energy Modeling Based on Energy Consumption Subunits in Wire Electrical Discharge Machining(WEDM) Process

Jun Zheng,Xuwei Lai,Xingjian Zhou,Ankai Chen,Wang Zheng 한국정밀공학회 2019 International Journal of Precision Engineering and Vol.20 No.5

The scientific and reasonable prediction of energy consumption in WEDM process is the key for energy-saving optimization of wire-cutting process in the design stage. The existing research about WEDM energy-saving studies mainly focus on the pulsed power control strategies. And, the research on non-pulsed energy consumption of WEDM is deficient. The non-pulsed energy consumption takes up a larger proportion in the whole WEDM process through theoretical analysis and case verification, and the energy consumption characteristics are equally representative. Therefore, the paper analyzes the energy characteristics based on non-pulsed energy consumption subunits of WEDM. A non-pulsed auxiliary energy consumption and feeding energy prediction model is established, and a WEDM machine tool is used for the experiment. The accuracy of the prediction model proposed in this paper can reach more than 96%. The raise and establishment of the model provides the basis for energy saving optimization of the subsequent wire cutting process.

Two-Stage Flowshop Scheduling with Outsourcing Allowed

Lin Li,Junzheng Luan,Yichen Qiu 보안공학연구지원센터 2016 International Journal of u- and e- Service, Scienc Vol.9 No.10

In this paper we study the production-scheduling problem for a two-stage flow shop with outsourcing options. The maintenance of in-house machine and job-related discount for outsourcing cost are considered in model, in which total cost and makespan are regarded as bi-objective for the scheduling problem. Based on the analysis of model, we introduce a concept-packaging to reduce the complexity of problem and then develop heuristic algorithm to solve the problem. Computational experiments using different groups of data are conducted to test the algorithm.

Vibration isolation control performance for an innovative 3-DOF parallel stabilization platform

Shuo Jiang,Junzheng Wang,Shoukun Wang,Wei Shen 대한기계학회 2022 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.36 No.7

High vibration isolation performance of stabilization platforms for challenging terrain is strongly demanded in many fields, including material transportation, marine operation and airborne filming. In this study, we propose a 3-DOF parallel stabilization platform which can offset disturbance in vertical and horizontal directions. First, the overall design idea is explained, including system prototype, mechanical optimization design, and motion modeling. Then a force-position composite control framework based on active compliance is considered for impact cushioning under high frequency disturbance. Besides, an adaptive compliance control law fused with environmental stiffness online estimation and reference trajectory correction is designed to realize vibration isolation under variable excitations. Finally, simulations and experimental demonstrations using the developed 3-DOF stabilization platform are carried out. High vibration isolation performance shows the feasibility and effectiveness of proposed adaptive active compliance control strategy, which provides further applicability in engineering.

PDBNet: Parallel Dual Branch Network for Real-time Semantic Segmentation

Yingpeng Dai,Junzheng Wang,Jiehao Li,Jing Li 제어·로봇·시스템학회 2022 International Journal of Control, Automation, and Vol.20 No.8

To make a trade-off between accuracy and inference speed in real-time applications on the unmanned mobile platform, a novel neural network, named Parallel Dual Branch Network (PDBNet), is proposed. Firstly, a multi-scale module, namely Parallel Dual Branch (PDB), is designed to extract complete information. PDB module consists of two parallel branches to remove detailed low-level information and high-level semantic information while maintaining few parameters. Then, based on the PDB module, PDBNet, a small-scale and shallow structure, is designed for semantic segmentation. A multi-scale module tends to extract abundant information and segment the object out from the image well. The small-scale and shallow structure tends to accelerate the inference speed. So PDBNet architecture is designed to be effective both in terms of accuracy and inference speed. PDBNet adopts three downsamplings to obtain feature maps with high spatial resolution and uses PDB modules with different dilation rates to extract multi-scale features and enlarge the receptive field in the last several layers. Finally, experiments on Camvid dataset and Cityscapes dataset, we respectively get 67.7% and 69.5% Mean Intersection over Union (MIoU) with only 1.82 million parameters and quicker speed on a single GTX 1070Ti card.

Neural Approximation-based Model Predictive Tracking Control of Nonholonomic Wheel-legged Robots

Jiehao Li,Junzheng Wang,Shoukun Wang,Wen Qi,Longbin Zhang,Yingbai Hu,Hang Su 제어·로봇·시스템학회 2021 International Journal of Control, Automation, and Vol.19 No.1

This paper proposes a neural approximation based model predictive control approach for tracking controlof a nonholonomic wheel-legged robot in complex environments, which features mechanical model uncertaintyand unknown disturbances. In order to guarantee the tracking performance of wheel-legged robots in an uncertainenvironment, effective approaches for reliable tracking control should be investigated with the consideration of thedisturbances, including internal-robot friction and external physical interactions in the robot’s dynamical system. In this paper, a radial basis function neural network (RBFNN) approximation based model predictive controller(NMPC) is designed and employed to improve the tracking performance for nonholonomic wheel-legged robots. Some demonstrations using a BIT-NAZA robot are performed to illustrate the performance of the proposed hybridcontrol strategy. The results indicate that the proposed methodology can achieve promising tracking performancein terms of accuracy and stability.

Shadow Lane Robust Detection by Image Signal Local Reconstruction

Du Mingfang,Wang Junzheng,Li Nan,Li Duoyang 보안공학연구지원센터 2016 International Journal of Signal Processing, Image Vol.9 No.3

In order to resolve shadow interference and slow image processing speed in the visual navigation of unmanned vehicles on city roads, a new lane detection algorithm based on the Inverse Perspective Mapping(IPM) of vertical sub-image reconstruction using local bands was proposed. The lane’s IPM aerial image was obtained using the projective transformation of three-dimensional images of city roads. The ROI portion of the IPM map was decomposed and analyzed using the sym3 wavelet, and after analyzing and comparing the experimental results the first and second levels of the vertical sub images were selected for the reconstruction, compression, and removal of shadows from the original image. The Canny algorithm was proposed and adopted in order to extract the edge features of the reconstructed images according to real-time road image quality. A modified, polar angle, constraint-based, fast Hough transform was used to locate the candidate lanes. Finally, the main control point, straight-line-fitting algorithm was used to fit the final lanes, which achieved the precise location and recognition of the lane lines. The California Polytechnic lane data set, a public data platform, which is now widely used in road visual recognition fields around the world, was selected for the testing and verification of the algorithm. The results of the experiment and actual operation indicated that, despite having a storage space of approximately one tenth the size of similar algorithms, this detection algorithm effectively solves the issue of shadow removal, meets the real-time requirements of roadway image processing systems for unmanned vehicles with a recognition time of less than 20ms, and is robust.