Table of Content

30 November 2017, Volume 35 Issue 6

Previous Issue    Next Issue

Communication Engineering

Multi-cell Multi-user Coordinated Resource Efficiency Optimization

QIAN Ye-wang, HE Shi-wen, YANG Lü-xi

2017, 35(6):  675-684.  doi:10.3969/j.issn.0255-8297.2017.06.001

Asbtract ( 763 )   PDF (7040KB) ( 355 )  

References | Related Articles | Metrics

Spectrum efficiency (SE) and energy efficiency (EE) are key specifications for the performance of the fifth generation (5G) wireless communications. In this paper, we study optimization of resource efficiency of a coordinated multi-cell multi-user downlink system defined as a weighted sum of SE and EE. The considered optimization is a nonconvex problem due to nonconvexity of the user rate for an interference channel. The original problem is transformed to a tractable form by exploiting the fractional problem theory and the relation between the user rate and minimum mean square error. A hierarchical iterative alternating optimization algorithm is then proposed to address the latter. Furthermore, convergence of the algorithm is shown. Numerical results are provided to validate effectiveness of the proposed algorithm.

Modified Support Vector Machine for Wireless Localization

YANG Jin-sheng, LIN Zhen-jun

2017, 35(6):  685-692.  doi:10.3969/j.issn.0255-8297.2017.06.002

Asbtract ( 887 )   PDF (5076KB) ( 279 )  

References | Related Articles | Metrics

Using support vector machine (SVM) for wireless localization suffers from instability of accuracy as the parameters are generally chosen based on experience. To solve the problem, we use simulated annealing (SA) to modify support vector machine (SA-SVM) to improve positioning accuracy. We obtain the training samples from simulation of the cellular communication system model to the SVM, and find the optimal SVM parameters in an iterative search based on SA. The obtained optimal parameters are then used in the positioning. Simulations show that, compared with the original SVM positioning method, SA-SVM method effectively improves localization accuracy, and therefore has application values.

Joint Angle and Delay Estimation for OFDM Using Unitary Transform and Structured Least Squares

GUO Li-kai, WU Ying, YIN Jie-xin, WANG Cheng

2017, 35(6):  693-705.  doi:10.3969/j.issn.0255-8297.2017.06.003

Asbtract ( 950 )   PDF (7652KB) ( 313 )  

References | Related Articles | Metrics

In an orthogonal frequency division multiplexing (OFDM) system, traditional subspace-based joint angle and delay estimation algorithms show significant performance degradation at low signal-to-noise ratio (SNR). To solve the problem, a new algorithm using unitary transform and structured least squares (SLS) is proposed. With unitary transform, data are transformed to the real number domain. Two-dimensional SLS is then used to estimate two real-valued diagonal matrices that contain information of angles and delays. A complex matrix is constructed with the two real-valued matrices, and eigenvalues of the complex matrix are calculated. The real and imaginary parts of the eigenvalues correspond to angles and delays, respectively. Since SLS takes into account the coupling relationship between noise terms and restores the estimated signal subspace matrix, its estimation performance is closer to optimum than those of the others. Simulation results show that the proposed USLS-JADE algorithm is superior to traditional subspace-based algorithms in terms of accuracy and success rate.

Music Comparison Algorithm Based on Intelligent Wearable Device

YE Yang, ZHANG Xue-fan, LIU Yuan, WANG Chen, HUANG Qing

2017, 35(6):  706-716.  doi:10.3969/j.issn.0255-8297.2017.06.004

Asbtract ( 900 )   PDF (1973KB) ( 523 )  

References | Related Articles | Metrics

We introduce a novel algorithm for music comparison based on intelligent wearable devices. The algorithm includes three steps. First, the note initial point is identified with an average amplitude method, and frequency information obtained with short time Fourier transform according to the characteristics of music. Second, the basic frequency is extracted with subharmonic summation and minimum variance methods to locate music note accurately. Third, the real-time music and standard music are compared both in the frequency and time domains to detect errors in terms of Euclidean distance. Experiments show that the algorithm has low computational complexity, good real-time performance and high accuracy. It is suitable for music comparison based on intelligent wearable device.

CI Distributed Algorithm under Non-stationary and Imperfect Communication Conditions

KUANG Hong-mei, LI Wei

2017, 35(6):  717-725.  doi:10.3969/j.issn.0255-8297.2017.06.005

Asbtract ( 826 )   PDF (2645KB) ( 289 )  

References | Related Articles | Metrics

In wireless sensor networks, distributed estimation is an important issue, in which sensor nodes estimate parameters of interest from the physical world. This paper develops a consensus-plus-innovations (CI) distributed algorithm to deal with distributed estimation. The focus is on the mean convergence performance of a CI distributed algorithm under non-stationary and imperfect communication conditions. Theoretical analysis shows that the algorithm is mean convergent and has an asymptotic normality property. Also, non-stationarity and imperfect communication conditions have no effect on the mean convergence performance and asymptotic normality. However, these conditions have an impact on the asymptotic variance of the algorithm. Validity of the CI distributed algorithm is shown by simulation.

SVM Spectrum Sensing Based on Data Pre-processing with Log Function

ZHAI Xu-ping, MENG Tian, WANG Tao

2017, 35(6):  726-734.  doi:10.3969/j.issn.0255-8297.2017.06.006

Asbtract ( 1077 )   PDF (6788KB) ( 497 )  

References | Related Articles | Metrics

To improve probability of detection and reduce training time, this paper proposes a method of support vector machine (SVM) spectrum sensing based on data preprocessing with a log function. A minimum size of training set is selected, which is applicable with good performance in spectrum sensing. The sample sets are generated with laboratory instruments. The obtained sample sets are pre-processed with a log function to increase the mean difference between sample sets with and without primary users (PU). Experimental results show that, after pre-processing, performance of spectrum sensing is significantly improved under low SNR conditions with detection accuracy 90% or better.

Signal and Information Processing

Fragile Watermarking for Fingerprint Images Based on QR Decomposition in NSCT-Domain

SHI Jun-peng, WU Yi-quan

2017, 35(6):  735-744.  doi:10.3969/j.issn.0255-8297.2017.06.007

Asbtract ( 941 )   PDF (7060KB) ( 407 )  

References | Related Articles | Metrics

To ensure security of fingerprint information in a network environment, a fragile watermarking algorithm for fingerprint images based on QR decomposition in the nonsubsampled contourlet transform (NSCT) domain is proposed. The fingerprint watermark is prepared by double encryption using Arnold transform and logistic scrambling. Twolevel NSCT is performed on a host image, and the low frequency component is divided into blocks with QR decomposition. The encrypted fingerprint watermark is embedded into the upper triangular matrix obtained with QR decomposition. The algorithm has good fidelity of the information because entries in the first row of the upper triangular matrix contain most information in the image. Experimental results show that, keeping invisibility, the proposed algorithm is vulnerable to both conventional image processing and geometric attacks, and therefore can accurately locate the tampered area.

Contrast Enhancement Forensics Based on Modified Convolutional Neural Network

DONG Wei, WANG Jian-jun

2017, 35(6):  745-753.  doi:10.3969/j.issn.0255-8297.2017.06.008

Asbtract ( 1083 )   PDF (5364KB) ( 503 )  

References | Related Articles | Metrics

Contrast enhancement forensics has drawn much attention in image forensics recently. This paper proposes a modified convolutional neural network (MCNN). The first layer in the MCNN is a preprocessing layer, which converts an input image to a binary gray-level co-occurrence matrix (BGLCM). The other layers are the same with conventional CNN, which learn features from BGLCM for classification. Compared with previous methods, feature extraction and classification can be optimized simultaneously, making the extracted features more suitable for contrast enhancement detection. Experimental results show that the proposed method can detect contrast enhanced images produced by conventional contrast enhancement and two anti-forensic techniques. It can also distinguish parameters used in the contrast enhancement.

Splicing Detection for Color Images Based on QDCT Markov method

WANG Jin-wei, LIU Ren-feng

2017, 35(6):  754-762.  doi:10.3969/j.issn.0255-8297.2017.06.009

Asbtract ( 1013 )   PDF (8797KB) ( 278 )  

References | Related Articles | Metrics

We propose a splicing detection scheme based on quaternion discrete cosine transform (QDCT). The scheme uses correlation among three channels of a color image to reduce loss of the image's inherent color information. QDCT is first applied to the images, and features are extracted with the proposed scheme in the QDCT domain. SVM is used to detect image splicing. Using the image bases CASIA1 and CASIA2, accuracy of the proposed scheme reaches 98.75% and 96.78% respectively, which is better than most of existing methods.

Adaptive Fusion Algorithm for Medical Images Using Non-subsampled Contourlet Transform

LOU Jian-qiang, DAI Wen-zhan, LI Jun-feng

2017, 35(6):  763-774.  doi:10.3969/j.issn.0255-8297.2017.06.010

Asbtract ( 931 )   PDF (7973KB) ( 424 )  

References | Related Articles | Metrics

An adaptive image fusion algorithm based on non-subsampled contourlet transform(NSCT)is proposed for medical images. Source images are first registered and then decomposed to low and high frequency sub-bands using NSCT. The NSCT coefficients in each sub-band are fused. For coefficients in the low frequency bands, a fusion rule based on regional energy, mutual information and information entropy is used. In high frequency bands, sum of modified Laplacian is used. The final image is obtained from the fused sub-images in the low and high frequency bands using inverse NSCT. Experiments are conducted for gray and color images to compare the propose method with previous algorithms. The results show that fused images using the proposed method contains more texture information, and is visually better.

Parallel Algorithm of UAV Image Match Considering Spatial Contiguity

ZHANG Chun-sen, QIU Zhen-guo, GUO Bing-xuan, XIAO Xiong-wu, ZHU Shi-huan

2017, 35(6):  775-785.  doi:10.3969/j.issn.0255-8297.2017.06.011

Asbtract ( 1038 )   PDF (13858KB) ( 407 )  

References | Related Articles | Metrics

To deal with the problem of load balancing and feature data transmission in parallel algorithms of unmaned aerial vehicle (UAV) image match, aparallel algorithm for image match considering spatial contiguity is proposed. In the feature extraction phase, initial task partition is carried out according to spatial contiguity between images. The final feature extraction task is determined by performing a fine-grain second partition based on the initial partition. In task scheduling, corresponding tasks are assigned according to the computational node state, after which other tasks are assigned. In the image matching phase, matching tasks are assigned first according to the feature extraction tasks. Matching task unit of each node is confirmed by the node number of the feature extraction task. The same method is applied to image match in the whole measured area. Experiments on a typical data set including 1 463 UAV images show that the algorithm can realize load balancing of the parallel system and reduce the amount of feature data transmission,thus significantly improving efficiency of image match.

Control and System

Prescribed Performance Nonlinear PID Control for Uncertain Continuous Systems

CHEN Long-sheng, WANG Qi

2017, 35(6):  786-796.  doi:10.3969/j.issn.0255-8297.2017.06.012

Asbtract ( 930 )   PDF (9098KB) ( 303 )  

References | Related Articles | Metrics

The nonlinear PID controller is studied for uncertain continuous systems only with the systems' input and output (I/O) data. Under the condition of continuously differentiable reference input and output signal, the prescribed performance control theory is considered. Based on the performance function and the error transformation theory, a simple nonlinear function is then designed to construct a nonlinear PID controller with feasibility proved by theoretical analysis. The proposed controller inherits the advantages of traditional PID with better adaptability robustness and anti-disturbance performance, and better flexibility in parameter adjustment as well. Effectiveness of the results is shown with numerical examples.

Design and Study of Laser Ultrasound Visual Detection System for Carbon Fiber Composite Materials

LIU Xia, SHAN Ning, MA Xiao-feng, WANG Shao-hua, LIU Tuan-jie

2017, 35(6):  797-804.  doi:10.3969/j.issn.0255-8297.2017.06.013

Asbtract ( 858 )   PDF (14497KB) ( 364 )  

References | Related Articles | Metrics

To realize fast visual detection of carbon fiber composite materials, the principle of laser ultrasound visual detection system is studied. Using a scan laser mode, the laser ultrasound visual detection system has advantages of simple structure and easy operation. It is designed based on the reversibility principle of ultrasound propagation, and can be used in fast and visual detection of carbon fiber composite materials. Experiments are carried out. The results show that the system can effectively be used to detect defects in carbon fiber composite materials. It has high accuracy and fast detection speed, only a few microseconds for each detection task. The detection results are straightforward and can display internal defects of the structure dynamically in real-time. The information of ultrasound transmission can be replayed and selected according to time frames. The system can be used to determine the location and sizes of defects. With the visual detection system, it is easy to achieve full non-contact detection. It is suitable for on-line real-time monitoring of structures made of carbon fiber composite materials.