Current Issue

• • 2017, Vol. 40 No. 4 Published date:28 August 2017 Last issue Next issue

  Review

  • Feature Fusion Methods in Pattern Classification
  • LIU Wei-bin, ZOU Zhi-yuan, XING Wei-wei
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 1-8. DOI:10.13190/j.jbupt.2017.04.001
  • Abstract ( 660 )     HTML     　　
  • Feature fusion is an important method in pattern recognition. Image recognition problem in computer vision can be known as a special pattern classification problem, and it still have many challenges. To solve the problem, feature fusion method is able to use multi-feature of image, complement each other's advantages and get more robust and accurate results. Based on information fusion theory, analyze the theory of feature fusion and introduce the development of feature fusion methods. Besides, discuss three methods of feature fusion combining with other basic theories. Bayesian decision theory based feature fusion algorithm implements fusion decision of multi-feature. Sparse representation based feature fusion algorithm can get the joint sparse representation of multi-feature. Deep learning based feature fusion algorithm intensifies feature learning process of deep neural network model.
  • References | Supplementary Material | Related Articles

  Papers

  • Design of Distributed Shared Memory Structure for Array Processor
  • SHAN Rui, SHEN Xu-bang, JIANG Lin, ZHU Yun, SONG Hui
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 9-15. DOI:10.13190/j.jbupt.2017.04.002
  • Abstract ( 440 )     HTML     　　
  • With the increasing of number of processors, the problem of memory wall was more severely. In order to alleviate this problem, two-level mixed interconnection network was proposed: fast crossbar for local data transfer and network on chip for long distance data communication. Meanwhile data transfer mechanism was designed to support unified addressing. Two memory architecture sizes were implemented on field rpogrammable gate array, and area, frequency and power consumption were evaluated. A mixed simulation testbench based on SystemC language was developed. The simulation results show that the designed architecture has higher memory access bandwidth and lower local accessing latency.
  • References | Supplementary Material | Related Articles

  • A LTE-U Coexistence Scheme Based on Dynamic Channel Switching and Muting Strategy
  • SHEN Bin, ZHAO Si-cong
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 16-22. DOI:10.13190/j.jbupt.2017.04.003
  • Abstract ( 376 )     HTML     　　
  • Based on listen before talk (LBT) channel access mechanism, a scheme allowing the LTE-U system coexistence with WIFI system in 5 GHz unlicensed band was proposed. At the same time, by introducing the dynamic channel switching technology, it is effective to avoid the interference to the WIFI system caused by the LTE-U system which occupies the same channel for a long time. On this basis, the silence length of the LBT mechanism is optimized to further enhance the coexistence efficiency with the WIFI system. Experiments show that the proposed scheme can make LTE-U and WIFI coexist well in the same frequency band and achieve a better system performance.
  • References | Supplementary Material | Related Articles

  • Enhanced User Selection Algorithm for Massive MIMO with Quantized Precoding
  • LIU Jian-fei, TAO Ying, WANG Meng-jun, ZENG Xiang-ye, LU Jia
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 23-28. DOI:10.13190/j.jbupt.2017.04.004
  • Abstract ( 433 )     HTML     　　
  • In multi-user massive (MIMO) wireless system, where the base station equipped with a large number of antennas simultaneously serves multiple users, user selection algorithms have become a rising concern. Two user selection algorithms, in which both channel quality and multi-user interference are taken in account, were proposed. The one is implemented by combining 1-bit feedback and minimizing the preceding vector correlation, it can effectively reduce the computation complexity; In the other algorithm, the list of users is got from potential users by comparing the preceding vector correlation with a threshold, then a set of users with the largest sum of the channel gain were selected, so the sum capacity is further improved. The performance and complexity of algorithms are analyzed.
  • References | Supplementary Material | Related Articles

  • Sensor-Based Cooperative Spectrum Sensing with Multiple Antennas over Composite Fading Channels
  • LI Mei-ling, HE Wen-li, DONG Zeng-shou, LU Zhao-ming
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 29-34. DOI:10.13190/j.jbupt.2017.04.005
  • Abstract ( 417 )     HTML     　　
  • The censor-based cooperative spectrum sensing with multiple antennas (C-MA-CSS) was proposed, the closed-form expressions for the detection probability, miss-detection probability, the false-alarm probability and the secondary throughput were derived using the mixture Gamma distribution under the composite fading sensing channels. Then, the relationships between them and the number of antennas were analyzed, and the optimal algorithm for maximizing the secondary throughput was also given. Simulation results demonstrated that lower miss-detection probability can be achieved in C-MA-CSS compared to the non-censoring CSS. Besides, the secondary throughput monotonically increased rapidly first and slowly after. Meanwhile, optimal secondary throughput can be achieved with smaller number of antennas by the proposed algorithm while maintaining the target detection probability.
  • References | Supplementary Material | Related Articles

  • Automatic Updating Response Surface Method for Motion Trajectory Reliability Analysis of Manipulator
  • FENG Jia-zhen, ZHANG Jian-guo, JIA Qing-xuan, CHEN Gang, SUN Jing-yi
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 35-40. DOI:10.13190/j.jbupt.2017.04.006
  • Abstract ( 404 )     HTML     　　
  • In order to solve the efficiency and accuracy problem appeared in motion trajectory reliability analysis of the manipulator, a new automatic updating response surface method is presented. Firstly, based on the equivalent extreme value approach, the motion trajectory reliability analysis is converted into the reliability analysis about the trajectory point in which the motion error reaches its maximum. Secondly, for different random sample inputs, the particle swarm optimization algorithm is employed to get the time responses corresponding to the maximum values of the motion errors. These inputs and time responses are used to construct a Kriging response surface of time, which is used to get the estimated value of the time response rapidly and improve the efficiency of the reliability analysis. Thirdly, based on the mean square predicted error, an automatic updating mechanism is constructed. Through adding new sample points, the approximate quality of the Kriging response surface is improved, and the accuracy of the reliability analysis results is assured. A case analysis is used to demonstrate the validity of the proposed method.
  • References | Supplementary Material | Related Articles

  • Evaluation and Optimization of Feed-in Tariffs under Uncertainty
  • LI Li, ZHANG Xin
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 41-47. DOI:10.13190/j.jbupt.2017.04.007
  • Abstract ( 382 )     HTML     　　
  • Based on the strategic behaviors of investor, this article presents a policy evaluation model in which a specific Feed-in tariff (FIT) level can be evaluated and optimized. The influence of uncertain generation from RES on the optimal investment time and sizing (capacity choice) of investor was evaluated by using the proposed model. Also, from the view of a policy maker, the optimal tariff level was determined in the way of minimizing the total burden on ratepayers to achieve RES goals. By dint of the current tariff of offshore wind in China, empirical studies through the least squares Monte Carlo method and the real option theory assess and optimize the current level of FIT. It is revealed that the current tariff level of 0.137 USD/kWh for offshore wind in China would not be enough to offset the risk premium resulting from the output uncertainty. Improving the tariff level to 0.146 USD/kWh could take effect to stimulate investment.
  • References | Supplementary Material | Related Articles

  • Digital Watermarking Against Data Merging Attack for Vector Geographic Data
  • WANG Ying-ying, YANG Cheng-song, ZHU Chang-qing, REN Na, FANG Hu-sheng
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 48-53. DOI:10.13190/j.jbupt.2017.04.008
  • Abstract ( 371 )     HTML     　　
  • A watermarking algorithm against geographic-feature-based data merging attack was presented. And a watermarking algorithm against data merging attack was presented based on analyzing the influence of geographic-feature-based data merging attack according to the detection-merging-detection thinking for single geographic feature. Experiments show that the proposed algorithm has good performance on the robustness against data merging attack. Additionally, the different watermarks in the merged vector geographic data can be detected by using the algorithm.
  • References | Supplementary Material | Related Articles

  • Robust Outlier Detection Algorithm Based on k-Nearest Neighbor Region Center Migration
  • ZHAO Jian-long, QU Hua, ZHAO Ji-hong
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 54-59. DOI:10.13190/j.jbupt.2017.04.009
  • Abstract ( 434 )     HTML     　　
  • Considering the distance- and density-based outlier detection algorithms are often sensitive to a nearest neighbor parameter k, termed k-center offset outlier factor (COOF), a robust outlier detection criterion for the characterization of abnormal degree of each data object was proposed. Each data object is included in a region within its k nearest neighbors, and the center of region will migrate with the change of nearest neighbor parameter k. In general, the variation of center offset of k nearest neighbor region is greater for an outlier than a normal object. According to this observation, for each data object, COOF is defined as the accumulation of this kind of offset when increasing the nearest neighbor parameter from one to k. Finally, the outlier detection algorithm based on COOF was also presented. Through artificial data and real data experimental simulations show that COOF is insensitive to parameter k, and has more stable and accurate outlier detection performance compared to k nearest neighbor, local distance-based outlier factor and local outlier factor, which are the distance-based method and density-based method respectively.
  • References | Supplementary Material | Related Articles

  • User Association Load Balancing Algorithm Based on Self-Backhaul Aware in Dense Networks
  • TANG Lun, LIANG Rong, CHEN Wan, ZHANG Yuan-bao
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 60-67. DOI:10.13190/j.jbupt.2017.04.010
  • Abstract ( 363 )     HTML     　　
  • In order to solve the problem of load imbalance caused by irrational bandwidth allocation in dense heterogeneous networks, a self-backhaul aware user access load balancing scheme was proposed. Firstly, a user association-load balancing strategy(UA-LBS) was described based on the load state of each small base station access and backhaul resource in dense heterogeneous network. Secondly, the Q-Learning algorithm was used to allocate wireless access and backhaul bandwidth in each small base station. For different allocation factors, it can ensure user to re-access according to the UA-LBS to get different system utility, and then to get optimal bandwidth allocation strategy to ensure load balancing while achieving system utility maximization. Simulation shows that the scheme improves the network load balancing in the self-backhaul scene of dense heterogeneous network, and improves the user rate experience.
  • References | Supplementary Material | Related Articles

  • Using Dynamic Programming for Detecting Document Area of Mobile Images
  • WANG Mao-sen, NIU Shao-zhang
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 68-73. DOI:10.13190/j.jbupt.2017.04.011
  • Abstract ( 371 )     HTML     　　
  • The detection algorithm of the edge of the document area in the mobile image was presented. According to the characteristics of mobile image, a new dynamic programming model was designed. A combination of four edges of document area was proposed by using the two classification of logistic regression. Although the mobile images are generally blurred, uneven illuminated and distorted, these algorithms can solve the problem of boundary detection and combination of document area of mobile image well.
  • References | Supplementary Material | Related Articles

  • Research of Dynamic Key Encryption Algorithm in Smart Grid Communication
  • MA Li-cui, LI Mei-hong, WU Qian-qian, DU Ye
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 74-79. DOI:10.13190/j.jbupt.2017.04.012
  • Abstract ( 423 )     HTML     　　
  • To solve the security problem of data interception, forgery and tampering in the process of the smart grid communication, the dynamic key-based lightweight data encryption was improved. A scheme of data integrity authentication was proposed based on the analysis of the communication characteristics of the smart grid. The initial key was embedded in the device. The non-retransmission packets were used to synchronously update the two-party key, the exclusive-or algorithm was used to complete the data symmetric encryption and decryption. In addition, the key and the message are used as the input of the Hash algorithm, and the Hash value was used as the message digest to complete data integrity authentication. Libmodbus open source library was used to simulate the communication between master and slave devices. Experiments show that the synchronization and accuracy of key update are ensured, the key self-organization and management are realized, the data transmission is completed correctly, and the resource consumption is obviously superior to the existing algorithm.
  • References | Supplementary Material | Related Articles

  • Wireless Broadcast Advantage-Based Load-Aware Multicast Algorithm for Wireless Mesh Networks
  • WANG Ji-hong, SHI Wen-xiao
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 80-85. DOI:10.13190/j.jbupt.2017.04.013
  • Abstract ( 396 )     HTML     　　
  • Wireless broadcast advantage-based load-aware multi-channel multicast (W-LMCM) algorithm is proposed to solve the multicast routing and channel assignment problem with the goal of maximizing the number of served subscribers. W-LMCM regards links originating from the same node as a whole to fully exploit wireless broadcast advantage and save bandwidth; It uses maximal number channel selection strategy to avoid the adverse effect brought by random channel selection; While satisfying the channel capacity constraint within interference range, W-LMCM assigns channels for links with the goal of minimizing network interference, thus the problem that the number of subscribers served by interference-free channel assignment is limited can be solved. Network simulator simulation results show that W-LMCM can improve multicast throughput and network service capability efficiently.
  • References | Supplementary Material | Related Articles

  • Experiment on the Field Uniformity Validation Methods of Reverberation Chamber
  • GUO Xiao-tao, HE Zhao, WANG Shao-hua, WANG Li-feng
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 86-90. DOI:10.13190/j.jbupt.2017.04.014
  • Abstract ( 423 )     HTML     　　
  • The field uniformity of the same reverberation chamber is validated by the traditional method which is based on maximal field, and its results indicate that the technical specification of reverberation chamber is much better than that required in related international standard. A new method based on the probability distribution of field is introduced here. The field during the whole period of stirrers' rotating is taken into account. The general accordance from these two methods indicates that the new method is useful for the validation from another new point of view. The partial difference proves that the new method should be adopted for better specification supporting in some special applying, such as maximum reading distance measurement for radio frequency identification and simulation of complex wireless channel.
  • References | Supplementary Material | Related Articles

  Reports

  • Modeling and Simulation for 3D Wideband MIMO Mobile-to-Mobile Radio Channels
  • LIANG Xiao-lin, ZHAO Xiong-wen, LI Shu, WANG Qi
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 91-97. DOI:10.13190/j.jbupt.2017.04.015
  • Abstract ( 439 )     HTML     　　
  • A geometry-based stochastic scattering model for wideband multi-input multi-output (MIMO) mobile-to-mobile channels based on three dimensional (3D) two-cylinders was proposed. The proposed narrowband model is extended to wideband, and the carrier frequency and bandwidth was introduced into the model. The channel realization by using channel matrix is much more straight-forward and concise to study the channel characteristics compared with the too complicated analytical solutions available so far. The channel properties and parameters were further investigated based on the simulation model and realized channels. Simulations are compared with the two dimensional two-ring model. It can be found that the normalized power of power-delay-profile and root mean square delay spread of the proposed 3D model are smaller but the channel capacity are larger.
  • References | Supplementary Material | Related Articles

  • Construction Method of Adaptive Deep Convolutional Neural Network Model
  • ZOU Guo-feng, FU Gui-xia, WANG Ke-jun, GAO Ming-liang, SHEN Jin
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 98-103. DOI:10.13190/j.jbupt.2017.04.016
  • Abstract ( 593 )     HTML     　　
  • The construction process of traditional convolutional neural network (CNN) model has many shortcomings, such as over reliance on experience knowledge, a lot of parameters and difficult to training. At the same time, in view of the important value of constructing strategy of CNN model in complex multi-class problems, a new construction method of adaptive deep CNN model was proposed. First, the convolution layer and pooling layer of the initial CNN model are set only to include one feature map; and then, the convergence rate of CNN is used as evaluation index, the global expansion of network is carried out. After global expansion, the CNN is controlled to local expansion according to the recognition rate of cross validation samples. The local network learning is stopped until the recognition rate reaches the expected value. Finally, the training process for new samples, the adaptive incremental learning of network structure is realized by expanding some new branches. The superiority of proposed algorithm in network training time and recognition effect is verified through some image recognition experiments.
  • References | Supplementary Material | Related Articles

  • A Knowledge Annotation and Reasoning Semantic Web of Things Framework for Intelligent Diagnosis
  • YANG Yu-nong, WU Zhen-yu, ZHU Xin-ning
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 104-110. DOI:10.13190/j.jbupt.2017.04.017
  • Abstract ( 366 )     HTML     　　
  • The web of things(WoT) is one of the present advancements in which devices, objects, and sensors are getting linked to the semantic web. To solve the integration problems in developing WoT application with huge device data. A semantic annotation and reasoning based framework and approach was proposed. It presents a universal ontology for WoT applications with expanding a physical process model on the basis of reusing the traditional device description ontology. The framework also shows that semantic sensor networks, semantic web technologies, as well as reasoning mechanisms can help in real-world applications to automatically derive complex models for analytics tasks such as anomaly diagnosis and automatic control. The article demonstrates approaches for building automation system with numerous connected devices in knowledge base constructing and show how the semantic framework allows us to develop a complex WoT application.
  • References | Supplementary Material | Related Articles

  • Research on Rectangular Block Diffraction on 45 GHz Millimeter Wave
  • LI Xing-rong, LI Yong-qian, LI Bao-gang
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 111-116. DOI:10.13190/j.jbupt.2017.04.018
  • Abstract ( 345 )     HTML     　　
  • The block diffraction attenuation characteristics of wood board, mirror and white iron board were studied based on 45 GHz millimeter wave indoor communication measurements. According to different block material, the 3-ray and 4-ray block diffraction models were proposed, respectively. The rectangular block diffraction measurement experiments were carried out on 12 mm thick wood board, 6mm thick mirror, and 0.35 mm thick white iron board. Experiment shows that the average transmission attenuation coefficient of 45 GHz millimeter wave blocked by 12 mm thick rectangular wood board is about -4.28 dB, and the average error between the attenuation measurement results and the 4-ray model simulation results is about -0.02 dB. The average transmission attenuation coefficient of 45 GHz millimeter wave blocked by 0.35 mm thick rectangular white iron board is about -18.2 dB. In the case of the mirror far away from the antennas, the measured average transmission attenuation by the mirror surface is 1-2 dB larger than that of the negative one. When the mirror is near any set of antennas,the measured value by the mirror surface is similar to that of the white iron board.
  • References | Supplementary Material | Related Articles

  • Double-Threshold Joint Spectrum Sensing Based on Energy and Wavelet Transform
  • NI Shui-ping, CHANG Hui-gang
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 117-121. DOI:10.13190/j.jbupt.2017.04.019
  • Abstract ( 364 )     HTML     　　
  • Considering that the traditional energy detection algorithm is sensitive to noise, the detection accuracy is easy to be affected in low signal-to-noise ratio (SNR) conditions, and that cyclostationary detection algorithm is high computing complexity, a double-threshold joint detection algorithm based on energy detection and wavelet transform (WT) sensing was proposed. When the decision statistic falls outside the double-threshold section, the energy detection is performed and the confused region is within the double-threshold for wavelet detection. Besides, the two thresholds can be adjusted adaptively according to noise uncertainty. When the channel quality is better, the distance of the two thresholds will be minished, otherwise the distance of the two thresholds will be larger, so as to control the probability of wavelet detection. Simulation shows that the proposed algorithm can improve the detection performance effectively under low SNR conditions, and reduces the complexity.
  • References | Supplementary Material | Related Articles

  • A Joint Iterative Algorithm for LDS Mmultiuser Detecting and LDPC Code Decoding
  • LEI Wei-jia, LIU Yue-cen, XIE Xian-zhong
  • Journal of Beijing University of Posts and Telecommunications. 2017, 40(4): 122-127. DOI:10.13190/j.jbupt.2017.04.020
  • Abstract ( 476 )     HTML     　　
  • The LDS multiuser detection algorithm was researched to improve the performance of LDS system. A joint scheme of multiuser detection and channel decoding in LDS non-orthogonal multiple access systems was introduced. The reliability of multiuser detection was improved by transmitting users’ soft information after LDPC decoder back to multiuser detector. The performance of channel decoding was further improved by the enhancement of multiuser detection. Simulations, comparing with multiuser detecting and LDPC code decoding separately under the same conditions, indicate that the joint iterative algorithm can significantly improve the performance of BER. It is shown that the larger the system load is, the greater the performance improvements. At the same time, the difference in error performance between different users also reduces by the joint scheme. With the joint iterative algorithm, the performance of BER is almost same as that of single user even when the system load reaches 200%. The performance of the LDS system is significantly improved without increasing computational complexity of receiver by the joint scheme.
  • References | Supplementary Material | Related Articles