Current Issue

• • 2019, Vol. 42 No. 4 Published date:28 August 2019 Last issue Next issue

  Papers

  • Business Process Predictive Monitoring Based on Sequence Encoding of Frequent Activity Sets
  • HUANG Xiao-fu, CAO Jian, TAN Yu-dong
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 1-7. DOI:10.13190/j.jbupt.2018-278
  • Abstract ( 387 )     HTML     　　
  • The problem of predicting an ongoing business process based on sequence is discussed. Business process predictive monitoring is an important part of process mining. Most of the existing research focuses on forecasting based on explicit workflow models. However, the enterprise may not implement end-to-end workflow modeling and management for the process, or only own partial execution logs due to permissions. In these cases, it is difficult to make prediction based on the complete process model. This article proposes a frequent activity set based sequence encoding method to handle the low-frequency activities in the log, and performs prediction by searching historical similar data. As the log is updated, the algorithm will adapt to model changes due to the concept drift. The algorithm is validated on real data sets.
  • References | Supplementary Material | Related Articles

  • An Anti-Interference Algorithm for RFID Readers Based on Power Adjustment and Random Access Joint Optimization
  • WANG Shuai
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 8-14. DOI:10.13190/j.jbupt.2018-125
  • Abstract ( 392 )     HTML     　　
  • A joint anti-interference optimization algorithm for multiple readers is proposed, which integrates random access and power control strategies. Firstly, considering the effects of access probability, transmitting power and channel selection on the signal-to-noise ratio (SNR) of the reader, the anti-interference problem of multiple readers is modeled as a mixed integer programming problem. Secondly, the generalized Benders decomposition method is used to decompose the algorithm into an easy-to-solve primal problem and a master problem of binary programming. It is proved that the algorithm converges to the global optimal solution. Finally, the simulation results show that the algorithm outperforms the existing schemes in tag recognition rate and fairness.
  • References | Supplementary Material | Related Articles

  • Effect of Phase Center in Directional Antenna Array on DOA Estimation
  • ZHAO Di, DENG Zhong-liang, REN Xiao-fei
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 15-23. DOI:10.13190/j.jbupt.2018-276
  • Abstract ( 504 )     HTML     　　
  • Broadband directional antenna arrays have uncertain phase centers, non-uniformly distributed far-field phases and limited main lobe widths. To settle these problems, a sectorial direction of arrival (DOA) estimation model of uniform circular array composed of directional antennas is proposed. Firstly, the phase center of one directional antenna is estimated by the optimization principle, the effective aperture of the array is deduced and the directional steering vector is constructed as well. Subsequently, considering the integrated effect of phase disturbance and phase center estimation error on DOA estimation accuracy, the error vector is analyzed and deduced. Finally, a DOA estimation model of conventional ideal point source antenna array is improved and some factors affecting DOA estimation error are simulated. Simulations show that the model has the advantages of small number of antennas and high accuracy of DOA estimation under the same error factors.
  • References | Supplementary Material | Related Articles

  • Lightweight Network Services Orchestration Environment Based on NFV for End Users
  • YUE Yi, CHENG Bo
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 24-31. DOI:10.13190/j.jbupt.2018-264
  • Abstract ( 463 )     HTML     　　
  • We propose a network service orchestration environment based on network function virtualization (NFV) technology to help end users develop and deploy network services. The environment includes:a virtual network function data model for adapting heterogeneous network functions; a development environment to assist users in the design of network service processes; and a deployment environment to assist users in the deployment and execution of network services. Finally, user-oriented experiments and system performance experiments were conducted to evaluate the network service generation environment. Our evaluation shows that lightweight virtualization orchestrator can effectively reduce the tool learning time of ordinary users and improve the development accuracy of network services.
  • References | Supplementary Material | Related Articles

  • APIT Location Algorithm Based on Weighted Hybrid Filtering and Center of Gravity Method
  • YU Xiu-wu, YU Hao, LIU Yong, XIAO Ren-rong, LI Ying
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 32-37,75. DOI:10.13190/j.jbupt.2018-313
  • Abstract ( 531 )     HTML     　　
  • An improved approximate point-in-triangulation test (APIT) localization algorithm which based on weighted hybrid filtering and center of gravity method (HFG-APIT) was proposed. Hybrid filtering is used to filter abrupt signal strength values to make the data output smoothly and stably. Weighted median is then used to improve the accuracy of received signal strength indication (RSSI), the center of gravity method is used for interior point testing to reduce misjudgment and improve localization accuracy. Simulation results show that the hybrid filtering algorithm has higher ranging accuracy than other filtering methods in processing RSSI data and the localization errors of HFG-APIT are 41.7% and 23.8% of RSSI localization algorithm based on least squares method (LSM-RSSI) and APIT localization algorithm, the localization performance is also better than the other two algorithms.
  • References | Supplementary Material | Related Articles

  • An Efficient Multi-Bit Quantum Public Key Encryption Scheme
  • ZHENG Shi-hui, WEN Kai, GU Li-ze
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 38-42. DOI:10.13190/j.jbupt.2018-287
  • Abstract ( 506 )     HTML     　　
  • After the advent of quantum computers, the widely used public key cryptosystem will be broken, so it is urgent to propose new public key schemes to resistant quantum computing attacks. A single qubit-rotation transformation and classical one-way functions (Hash functions) are used to construct a multi-qubit quantum public key encryption scheme. The analysis results show that the new scheme is against the known forward search attack and a chosen ciphertext attack. Furthermore, the number of public key qubits used in the new scheme is obviously lower than that in the scheme presented by Kawachi et al.
  • References | Supplementary Material | Related Articles

  • PMIPv6 Performance Modeling and Analysis under the Scenario of Multi-Mobile Node's Simultaneous Handoff
  • ZHU De-qing, TIAN Xian-zhong, LUAN Shen-ji
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 43-49. DOI:10.13190/j.jbupt.2018-220
  • Abstract ( 452 )     HTML     　　
  • When multi-mobile nodes(MN) simultaneously initiate handoff, the handoff process of a MN will greatly influenced by other nodes. Finite-state absorbing Markov chain is adopted to model the process of multi-MN's handoff in proxy MIPv6 and the total handoff delay of multiple MN is derived. Besides, the handoff success probability and the capacity of mobile access gateway under delay limitation are deduced. Moreover, we validate our modeling by simulating the handoff process and comparing with the existing work. The results could be a reference for network administration and configuration.
  • References | Supplementary Material | Related Articles

  • High Spectrum Efficiency Wireless Resource Management Based on Network Slicing
  • LI Qiang, LU Chang-long, CAO Bin, ZHANG Qin-yu
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 50-56. DOI:10.13190/j.jbupt.2018-235
  • Abstract ( 484 )     HTML     　　
  • In order to meet the quality of service requirements of different types of users in different application scenarios and improve spectrum efficiency, the wireless resource allocation model is constructed based on network slicing. In order to obtain the optimal resource allocation scheme, the quality of service requirements of users are transformed into the wireless resources requirements, and the wireless resource optimization problem is formulated with the goal of maximizing the total system rate. Since the optimization is a mixed integer nonlinear optimization, the complexity of optimal solution is considerably high. A solution based on Lagrange duality is proposed and the solving algorithm is given. Finally, compared with the proportional fairness algorithm and the maximum system capacity algorithm, the simulation results show that the proposed algorithm improves the system capacity while sacrificing partial fairness, which verifies the effectiveness of the proposed algorithm.
  • References | Supplementary Material | Related Articles

  • Blind Parameter Estimation of Hybrid Networking Frequency-Hopping
  • FU Wei-hong, HU Zhan
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 57-63. DOI:10.13190/j.jbupt.2018-281
  • Abstract ( 496 )     HTML     　　
  • To solve low accuracy of parameter estimation and high complexity of algorithms in frequency hopping signal under low signal-to-noise ratio (SNR) and complex electromagnetic environment, a new algorithm combined short-time Fourier transform (STFT) with smoothed pseudo Winger-Ville distribution (SPWVD) is proposed. Firstly, the received signals are converted to time-frequency domain by using STFT, and then adaptive noise reduction is used in time-frequency domain. Secondly, the accurate frequency is estimated by adaptive clustering algorithm. Thirdly, the time-frequency information of the signals is extracted and all kinds of interference are eliminated. After the network is sorted, the rough estimation range of the hopping time can be calculated. Finally, SPWVD algorithm and a truncation threshold are used to deal with the rough estimation range, and then a precise frequency estimation is obtained. Simulations show that the parameter estimation of the algorithm has high accuracy and low complexity under the condition of hybrid networking and low SNR. The parameter estimation problem of existing frequency switching time in actual frequency hoppingcommunication system has been effectively solved.
  • References | Supplementary Material | Related Articles

  • A New Graph Encryption Algorithm Based on Revised Ising Model
  • WANG Yong, WANG Guo-dong, ZHANG Zhi-qiang, ZHOU Qing
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 64-69. DOI:10.13190/j.jbupt.2018-285
  • Abstract ( 393 )     HTML     　　
  • As graph structure is widely applied in science and engineering, it has good practical value to design graph encryption algorithm. Since it is difficult to directly design an encryption algorithm for a graph, two methods are introduced to solve this problem. First, the advantages of 2D Ising model, such as simplicity and locality, are introduced to the design of encryption. Second, the problem of graph encryption is transferred to several simple problems. Encryption algorithms for one-dimension data, two-dimension data, tree-structure are proposed based on the improved Ising model. Finally, the graph encryption algorithm is implemented. Analysis and simulation demonstrate that the proposed method can fulfill basic requirements of graph encryption, including reversibility, adaptability, efficiency, randomness and diffusion.
  • References | Supplementary Material | Related Articles

  • Consistency Checking Between UML Models and Code Based on Graph Theory and FSM
  • WANG Lei
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 70-75. DOI:10.13190/j.jbupt.2018-283
  • Abstract ( 518 )     HTML     　　
  • A unified modeling language (UML) model and code consistency checking method based on graph theory and finite state machine (FSM) was proposed. The basic idea of this method was given; the static consistency checking and the dynamic consistency checking algorithms between UML models and code were discussed respectively; a support tool for this method was implemented, and C++ project UMLChecker 1.0 was checked by using this tool. The experimental results show that this method can check consistency between UML models and code, and has a high checking accuracy. For consistency checking in behavioral aspects, checking accuracy rate is improved obviously.
  • References | Supplementary Material | Related Articles

  Reports

  • Adaptive Time Window Function Optimization Based Electrocardiography Feature Waveform Recognition
  • ZHANG Jin-ling, CUI Tong, LIU Wei-bin
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 82-88. DOI:10.13190/j.jbupt.2018-023
  • Abstract ( 451 )     HTML     　　
  • An adaptive windowed wavelet transform method was proposed to make precision positioning of the electrocardiography feature waveforms. The effective starting and ending points of electrocardiography wave were identified according to the trend of Q wave and S wave. The information recognition of the electrocardiography characteristic band was realized by combining the differential threshold method to detect the QRS starting. It is shown that the proposed method can effectively identify abnormal and sub-healthy electrocardiography waveforms and improve the recognition of QRS wave group, P wave and T wave for extration of electrocardiography information and diagnosis of abnormal electrocardiography such as abnomal angina pectoris, myocardial ischemia and unstable angina pectoris.
  • References | Supplementary Material | Related Articles

  • A Malicious Code Detection Method Based on Ensemble Learning of Behavior
  • XU Xiao-bo, ZHANG Wen-bo, HE Chao, LUO Yi
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 89-95. DOI:10.13190/j.jbupt.2018-318
  • Abstract ( 625 )     HTML     　　
  • In order to solve the problem of variant malicious code and behavior analysis of unknown threat, a method for malware classification based on gradient boosting decision tree (GBDT) algorithm is researched, which learns the characteristics of code behavior and instruction sequence from a large number of samples, and realizes the intelligent malicious code classification function. GBDT algorithm is introduced into the field of malicious code detection, so that the behavior sequence of the model is interpretable, and improves its ability to detect malicious code significantly. GBDT algorithm can reflect the nature of the behavior and intention of malicious code objectively, and identify malicious code accurately.
  • References | Supplementary Material | Related Articles

  • Research on Personnel Detection for Mine Accident Based on Channel State Information
  • SUN Chao-yu, GAO Shou-wan, YANG Xu, CHEN Peng-peng, NIU Qiang
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 96-101. DOI:10.13190/j.jbupt.2018-223
  • Abstract ( 380 )     HTML     　　
  • Aiming at the problems of high cost, short detection distance and high false alarm rate in traditional rescue methods, a new method for personnel detection based on channel state information is proposed. First, the state detection module determines the degree of the activity according to the Gaussian mixed distribution. Second, the breath detection module uses periodicity of the influence of human on channel state information to extract breathing, and to find the breathing frequency of the people who remain inactive. Finally, we evaluated the system in a variety of parameters. The experimental result verifies that the system has high accuracy and robustness,and the average recognition rate can reach 90%.
  • References | Supplementary Material | Related Articles

  • Rule-Based Heuristic Algorithm for Parking Spaces Allocation in Marginal Area of Urban Underground Parking Lots
  • XU Han-zhe, HUANG Yi-bin, YANG He, ZHOU Zhong-bing, LIU Xiao
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 102-108. DOI:10.13190/j.jbupt.2018-269
  • Abstract ( 505 )     HTML     　　
  • Underground parking lots have general characteristics like complex contours, large in area, various components and requiring safety in structure. To properly deal with this specific problem underground and increase the utilization of urban parking lots, three rules and a three-stage algorithm is devised. To deal with parking spaces allocation problem in marginal area during stage 1, mathematic model is constructed. In addition, a heuristic genetic algorithm is designed to solve it. Finally, the accuracy and validity of the algorithm are proved by case studies on real underground parking lots. Via visualization of the results, the algorithm acts as auxiliary to underground parking lots designers, which can improve the efficiency of underground parking lots design and raise profit.
  • References | Supplementary Material | Related Articles

  • Load Balancing Algorithm Based on Buffer Prediction Mechanism for CDNs
  • SHUAI Qian-jun, WANG Run-ze, WANG Ke-qin, JIN Li-biao
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 109-113. DOI:10.13190/j.jbupt.2018-059
  • Abstract ( 395 )     HTML     　　
  • A requests redirection algorithm for load balance based on buffer queueing length dynamic prediction (BMP) was presented. The M/M/1 model is used to describe the load balancing problem for content distribution networks. This method provides reduction of the unnecessary redirection with the queueing length prediction mechanism. And meanwhile, the overload of single point can be avoided. Evaluation shows that the proposed BMP algorithm achieved more balancing effect in server queue length compared to the so-called control law for load balancing algorithm. And the total cost of the request response time is obviously decreased due to the natively reduction of the unnecessary request redirection.
  • References | Supplementary Material | Related Articles

  • Analysis on Cloud Service System with Variable Number of Servers
  • ZHAO Guo-xi, JIN Shun-fu, MA Cong-bian, CAO Jian-yu, XU Li-jun
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 114-120. DOI:10.13190/j.jbupt.2018-297
  • Abstract ( 392 )     HTML     　　
  • In view of the migration characteristics of virtual machines in cloud service systems, an M/M/m(t) system model with variable number of servers is proposed. A two-dimensional Markovian chain based on the queuing length process of sub-tasks and the number changing process of virtual machines is established. The steady-state distribution of the process and the steady-state performance index of the system are obtained by using matrix geometric theory. Considering the service level agreement, an approximate lower bound of the average number of virtual machines that the system should open is given, and the effects of different parameters on the queuing index and the number of virtual machines in on-state are analyzed.
  • References | Supplementary Material | Related Articles

  • Privacy Protection Strategy for Implicit Typed Resources
  • DUAN Yu-cong, ZHANG You-liang, BAI Yong, CAO Chun-jie
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 121-125. DOI:10.13190/j.jbupt.2018-301
  • Abstract ( 397 )     HTML     　　
  • Aiming at modeling and expressing of implicit privacy resources, a meta-model is proposed for semantic definition of implicit resource concepts. The meta-model explains the core concepts in data, information, knowledge wisdom (DIKW) architecture. Based on the DIKW resource concept, the existing knowledge map concept is extended to data graph, information graph and knowledge graph. An automatic abstract adjustment privacy protection solution is presented based on the difference of the processing efficiency of the same content expressed in different types of resources. Protection is specified as what types of resources to be converted and how much the conversions to be performed towards an optimal solution balancing the comprehensive value of both the investment of the protection side and the cost of attackers.
  • References | Supplementary Material | Related Articles

  • The Extraction and Tracking Trajectory of Wireless Channel Tap Clusters Based on Machine Learning
  • ZHANG Jia-chi, LIU Liu, ZHOU Tao, WANG Kai, PIAO Zhe-yan
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 126-132. DOI:10.13190/j.jbupt.2018-298
  • Abstract ( 523 )     HTML     　　
  • A new method for extraction and tracking trajectory of dynamic wireless channel tap clusters is proposed. First, the channel impulse response (CIR) denoising is achieved by back propagation (BP) neural network in time delay-amplitude dimension. Then effective taps are clustered by k-means clustering algorithm. Next, density-based spatial clustering of applications with noise (DBSCAN) algorithm is applied to remove the abnormal peak taps for every cluster. Finally, the trajectory of cluster peak taps is obtained by polynomial fitting. The simulation result shows that trajectory obtained by proposed method is approximate to geometric calculation result. Moreover, the analysis result of high speed railway measured data is consistent with the actual observations.
  • References | Supplementary Material | Related Articles

  Papers

  • Combining Text Classification and Text Matching for FAQ-Based Question Answering
  • MO Qi, WANG Xiao-jie
  • Journal of Beijing University of Posts and Telecommunications. 2019, 42(4): 76-81. DOI:10.13190/j.jbupt.2018-299
  • Abstract ( 335 )     HTML     　　
  • Text classification and text matching are two ways to solve the frequently asked questions (FAQ)-based question answering. However, using the text classification method alone cannot effectively utilize the information of the standard question itself. When the text matching method is used alone, the selection of the negative sample is a very difficult problem. A series of models that combine text classification and text matching methods are proposed, which not only can select negative examples that really need to be distinguished, but also can effectively use the information of standard questions. Experiments show that the proposed models achieve the best performance on multiple FAQ-based question answering data.
  • References | Supplementary Material | Related Articles