15 November 2017, Volume 57 Issue 11

    

• Select all
  |

COMPUTER SCIENCE AND TECHNOLOGY
• Select
  Multi-tag identification and collision avoidance with RFIDs in blood management

  TIAN Chuan, YE Xiaojun, WANG Zuliang, LI Xin

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1121-1126. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.023

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The dynamic frame slotted ALOHA tag anti-collision Q algorithm in the EPCglobal_Class1 Gen2 protocol fails to remove empty slots, which leads to a waste of the slots. This is unacceptable in blood management which requires real-time handling and a large number of tags. A multi-tag identification and collision avoidance method was developed that divides the tag identification into two slot pre-assignments and tag identification. This helps eliminate empty slots in the Class1 Gen2 protocol which improves availability. Simulations of a blood management system indicate that this method significantly improves tag identification efficiency.
  
• Select
  APT attacks and defenses

  ZHANG Yu, PAN Xiaoming, LIU Qingzhong, CAO Junkuo, LUO Ziqiang

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1127-1133. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.024

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Advanced persistent threats (APT) have gradually evolved into a complex of social engineering attacks and zero-day exploits as some of the most serious cyberspace security threats. APT attacks often attack infrastructure and steal sensitive information with strong national strategic interests, so that cyberspace security threats evolve from random attacks to purposeful, organized, premeditated attacks. In recent years, APT attacks and defenses have rapidly developed in the cyberspace security community. The origin and development of APTs are reviewed here with analyses of the mechanism and life cycle of APTs. Then, APT defenses and detection methods are described with problems and further research directions identified.
• Select
  Improved chosen-plaintext DPA on block cipher SM4

  CHEN Jiazhe, LI Hexin, WANG Beibei

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1134-1138. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.056

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Since differential power analysis (DPA) is one of most important side-channel attacks on block ciphers implemented in chips, this paper revisits the DPA attack on hardware-implemented SM4. Reasonably choosing the plaintexts minimizes the affection of the variable input bits on the output bits, of the linear transformation of SM4, which leads to effective side-channel attacks on SM4. This paper deduces 8 bit-relationship in the chosen-plaintext setting by going into the linear transformation of SM4. Incorporating the bit-relationship with the known ones, this paper improves the previous chosen-plaintext DPA attacks on SM4, by proposing an analyzing module that makes better use of the side-channel information of the round-output bits. Experimental results show that the proposed manner improves the success rate of the chosen-plaintext DPA attacks on SM4.
• Select
  Android App behavior-intent inference based on API usage analysis

  SHEN Ke, YE Xiaojun, LIU Xiaonan, LI Bin

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1139-1144. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.057

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  An application behavior intention analysis is presented which analyzes the application program interface (API) usage in the background and the graphic user interface (GUI) state transitions in the foreground of the target App with behavior pattern recognition of the multivariate time series data at runtime. An API usage analysis based behavior intent inferring prototype was developed for Android Apps with static preprocessing, dynamic monitoring and behavior intent inference. This paper examines the effectiveness of the prototype on typical mobile Apps via case studies and validates the practicability and operability of the approach through real-world App profiling.
  
• Select
  Security analysis of dynamic provable data possession based on Hash aggregation

  ZOU Jing, LI Bin, ZHANG Li, LUO Yang, SUN Yunchuan, LI Shixian

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1145-1149,1158. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.058

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Dynamic provable data possession (DPDP) schemes are important in cloud data security to prevent the data from being lost, modified or corrupted. This paper presents security analyse of some existing schemes based on Hash aggregation and points out some security problems in the known methods and their fundamental reasons. The analyse show that Hash aggregation can effectively reduce storage, computation and transmission costs of verification due for data dynamic operations. One scheme had security vulnerabilities when verifying the relation between the position and the value of a data block that can allow a substitution attack.
• Select
  Data access control model based on data's role and attributes for cloud computing

  WANG Yuding, YANG Jiahai

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1150-1158. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.059

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The key cloud computing characteristics, such as data openness, elasticity, and sharing, complicate data access control. Traditional access control models cannot provide flexible, dynamic access control to large numbers of users with massive data files. This paper presents a data access control model based on the data's role and attribute for cloud computing. An attribute element is assigned to the data to provide role-based access control so that users can be assigned roles based on their own attributes and the tenant's attributes and current status, and can access data with different attributes. The paper illustrates the design of this model and the work processes and provides a theoretical security analysis. The results show that the model can provide dynamic, safe, fine-grained access control for users accessing data in a cloud environment.
• Select
  Hybrid feature extraction from ultrasound images for a silent speech interface

  LU Wenhuan, QU Yuexin, YANG Yalong, WANG Jianrong, DANG Jianwu

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1159-1162,1169. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.060

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Principal component analysis (PCA) and discrete cosine transform (DCT) are used to extract features from ultrasound images to build an ultrasound based silent speech interface. The critical information in the image is presented by using three hybrid feature extraction methods. The first method uses PCA to extract discrete wavelet transform coefficient features. The second and third methods truncate the DCT or Walsh-Hadamard transform coefficients to the appropriate number according to the energy with the truncated coefficients then used by PCA to extract the features. Tests show that this hybrid feature extraction method outperforms standalone PCA or DCT analyses. The block DCT-PCA method gives the best result among all the methods.
ENVIRONMENTAL SCIENCE AND ENGINEERING
• Select
  Environmental impact of Beijing's industrial decentralization based on a discrete choice modeling approach

  LI Tiankui, LIU Yi, WANG Chaoran, WANG Zishu

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1163-1169. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.061

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The ‘Non-Capital Functions’ dispersion policy is significantly affecting the spatial distribution of resources and the environmental in the Beijing-Tianjin-Hebei Region. This study estimates the changes in resource use and emissions at a regional level. The study classifies industrial sectors through a screening analysis of their policies, economics and environmental effects. The Monte Carlo method is used to determine individual firm sizes. A discrete choice model is used with location information and local environmental efficiencies. Relationships between the industrial size and the resource use and emission intensities are used to estimate the environmental impacts. The results show that during "the thirteenth five-year" plan, the industrial output in Beijing will decreases by 7%, while resource use and pollutant emissions will fall by over 9%, which will effectively alleviate water shortages in Beijing. In other cities in the region, the industrial output will rise by 5% on average, far more than the resource use and emissions increases. Various policy recommendations are presented to improve local environmental regulations.
• Select
  Source identification of accidental water pollution in a tidal river network based on a water environment model and database

  CHEN Zhengxia, DING Yi, MAO Xuhui, WANG Zheng, JIA Haifeng

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1170-1178. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.034

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The source of accidental water pollution can be accurately and quickly identified using a water environment model and database. The Foshan section of the Xijiang and Beijiang Rivers was taken as an example to build a EFDC (environmental fluid dynamics code) hydrodynamic and WASP (water quality analysis simulation program) water quality model. The WASP water quality model included the WASP-EUTRO model for normal pollutants and the WASP-TOXI model for toxic pollutants. The EFDC-WASP model sorts information on the pollutant risk source at each automatic monitoring section to establish a toxic and normal pollutant source identification database. A hypothetical scenario having the Shakou monitoring station in the Chancheng district measuringa high Cr⁶⁺ concentration was taken as an example with 3 possible pollutant emission situations identified through a database query. Initial loads for each situation 1-3 were estimated according to a first-order kinetics equation, with the location and load for each situation then calculated by the model. The pollutant source was quickly and accurately identified.
MECHANICAL ENGINEERING
• Select
  Optimization of an unpowered energy-stored exoskeleton spring stiffness for spinal cord injuries

  GUAN Xinyu, JI Linhong, WANG Rencheng

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1179-1184. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.035

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The unpowered energy-stored exoskeleton can provide precise walking assistance for spinal cord injury patients for a specified body height, weight and injury level. Since the energy-stored springs are standard components of the exoskeleton, their stiffness should be constant. The aim of this study is to select the optimal stiffness of the exoskeleton springs. The hip joint moment was related to the body height, body weight and hip joint angle through an inverse dynamics model for spinal cord injury patients walking in the single support phase from mid stance to terminal stance. The optimization method minimized the integral of the absolute moment generated by the hip joint and springs. The results show that the stiffness of spring 1 (k₁) has a normal distribution while the stiffness of spring 2 (k₂) does not. The medians of the two spring stiffnesses (the median k₁=3 180 N/m and the median k₂=1 279 N/m) can apply to patients whose heights and weights are in the P1-P99 ranges.
• Select
  Milling-layer thickness model based on a trochoid trajectory

  GUAN Liwen, Zhao Xiao, WANG Liping

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1185-1189. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.062

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The milling-layer thickness model greatly affects the prediction accuracy of instantaneous milling forces. Current milling layer thickness models can accurately predict the instantaneous change in the milling layer thickness foe small feed rates per tooth, but are less accurate with large feed rates per tooth. The study analyzes the milling blade movement pattern as the tool rotates and the work-piece feeds in to the tool. The model predicts the trochoid trajectory and the milling start and end angles. The milling-layer thickness model is then based on the trochoid trajectory. Trochoid trajectory model is more accurate for both large and small feed rates per tooth than the circle model and the equivalent uniform thickness model.
ELECTRONIC ENGINEERING
• Select
  Linear deployment performance analysis of a fractional full duplex cellular network

  BI Wenping, SU Xin, XIAO Limin, ZHOU Shidong

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1190-1195. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.027

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  In-band full duplex (FD) communications are attracting more attention due to their potential to double the spectral efficiency (SE). However, severe inter-cell interference will degrade the system performance. A fractional FD (FFD) resource allocation strategy was developed to reduce the effect of severe interference in the system SE. The radio resources are first classified into FD resource blocks and half duplex resource blocks with all the blocks allocated to the users based on the channel condition of each user. The Wyner model was used to evaluate the system performance. The user positions were used to partition the resource blocks and as the user classification criterion. Simulate SE and optimal partition radii for different classification thresholds show that this method can significantly improve the system SE.
• Select
  Outlier detection based on spatio-temporal nearest neighbors and a likelihood ratio test for sensor networks

  LIU Yimin, WEN Junjie, WANG Lanjun

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1196-1201. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.029

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  A spatio-temporal nearest neighbors and likelihood ratio test method was developed to detect outliers caused by sensor failures in sensor networks. In the space dimension, a sensor's spatially nearest neighbors were selected using a maximum posterior probability criterion while in the time dimension, the temporal nearest neighbors were previous observations from the same sensor. Each sensor's reading was evaluated based on differences between its earlier measurements and those of its neighbors with a sensor failure model and likelihood ratio test used to detect whether the sensor had failed. Tests show that this approach gives a higher detection rate for the same false alarm rate than existing outlier detection approaches. For example, for a 10% false alarm rate, the detection rate was increased by 10%-30%.
CIVIL ENGINEERING
• Select
  Mass real estate appraisals based on fuzzy mathematics

  LIU Hongyu, LI Yan

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1202-1206. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.065

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Mass real estate appraisals are important for real estate tax reform. There have been many studies of mass appraisal methods, but few studies of real estate grouping before modeling. This study presents of real estate grouping method based on fuzzy mathematics. The fuzzy nearness is used to measure the similarities between different buildings with similar ones put into the same real estate group. Then, the accuracy of this method is verified through an empirical study.
• Select
  Seismic analysis of a suction caisson foundation for offshore wind turbines

  YANG Chunbao, ZHANG Jianmin, WANG Rui

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1207-1211. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.063

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The stability of suction caisson foundations during earthquakes is evaluated for a typical suction caisson foundation of an offshore wind turbine located in a sandy seabed in the North Sea. The analysis was a unified plasticity model for the large post-liquefaction shear deformation of the sand and a three-dimensional dynamic finite element analysis to predict the seismic deformation and overturning probability during an earthquake. The analysis shows that the foundation experiences horizontal oscillations that lead to large deformations. The caisson slants with the rear moving up. The overturning probability keeps increasing and the angle eventually reaches more than 0.8°, which exceeds the safety limit. The rotation center constantly changes and the seabed pore pressure increases. Negative pore pressures can develop inside the caisson which enhance the overturning probability, but then dissipate. The seabed liquefaction depth is about 3 m, so the front of the caisson pierces into the seabed. Consequently, seabed liquefaction aggravates the deformation and overturning probability which affects the system operation.
ELECTRICAL ENGINEERING
• Select
  Stray inductance effects and optimal design of reverse bocking IGBT matrix converters

  ZHANG Tieshan, JIANG Xiaohua

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1212-1219. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.030

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  Stray inductances strongly affect switching device safety and performance in matrix converters. This paper analyses the effects of stray inductances and describes the main circuit optimization principles for reverse blocking IGBT (RB-IGBT) matrix converters. First of all, the multiple stages of an RB-IGBT turn-off transient are analyzed based on the switching device characteristics. The snubber capacitor affects the output voltage error caused by the forced commutation process and distorts the waveform. Simulations indicate that the main circuit stray inductance and the snubber capacitor have the greatest influence on RB-IGBT turn-off transient duration and surge voltage. Larger snubber capacitors give more waveform distortion. Design optimization principles are used to designan RB-IGBT matrix converter prototype. Tests show that the maximum switching surge voltage is 375 V without any snubber circuits. The practicality of the busbar structure design is validated by measurements of the output current from the matrix converter prototype.
• Select
  Voltage balancing in a 1-level modular multilevel matrix converter

  LI Yao, XU Lie, LI Yongdong

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1220-1227. https://doi.org/10.16511/j.cnki.qhdxxb.2017.21.031

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  A modular multilevel matrix converter (MMMC) can convert three phase AC input to three phase AC output. The converters use four-quadrant operation with good modularity and low frequency performance for high-voltage high-power variable frequency drive systems. MMMC designs include the topology without branch inductors and the topology with branch inductor. The MMMC without branch inductors are smaller and lighter.This paper describes a 1-level MMMC with the voltage of the H-bridge branches calculated numerically. The switching times are optimized to balance the voltages in the H-bridge branches. Tests show the effectiveness of the voltage balancing strategy.
THERMAL ENGINEERING
• Select
  Liquid film model for the flow in a twin-fluid atomization nozzle

  ZHU Yule, WU Yuxin, FENG Lele, LÜ Junfu

  Journal of Tsinghua University(Science and Technology). 2017, 57(11): 1228-1232. https://doi.org/10.16511/j.cnki.qhdxxb.2017.26.064

  Abstract ( ) Download PDF ( )   Knowledge map   Save

  The flow of gas-liquid mixtures in atomization nozzles is generally believed to not be fully developed annular flow but liquid film formation cannot be described accurately by traditional gas-liquid two phase flow models. This paper compares three slip ratio models for annular flow and their predictions of the liquid film thickness and the liquid mass flow rate with experimental data. The non-slip model under predicts the film thickness while the Ishii empirical correlation based on the hypothesis of fully developed annular flow over predicts the film thickness. A slip model was then developed from experimental data based on a non-fully developed annular flow model, the gas-liquid mass flow ratio and the inlet pressure. This model more accurately predicts the flow characteristics and the liquid film thickness in the nozzle than the other models.