Journal of Applied Sciences

Identification of Radio Communication Source with Multi-object Optimization

TANG Zhi-ling1,2, YANG Xiao-niu3, LI Jian-dong1

2012, 30(6): 559-565. doi:10.3969/j.issn.0255-8297.2012.06.001

Asbtract ( 2508 )

PDF (983KB) ( 1419 )

References | Related Articles | Metrics

To identify personality of a radio signal, a new method of multi-object optimization is proposed to solve a system of high order cumulants equations. The power amplifier model is first shown to be equivalent to a multi-input single-output system. A system of equations is derived from the cumulant relation between input and output. The system is solved with multi-object genetic optimization to obtain the features. The system of equations is verified by simulation, and the results of estimation are compared with computed values, showing that the proposed method can extract features from the received signal only with minor errors.

Uncertain Sensor Data Prediction for Wireless Sensor Networks

YAN Xiao-zhen, XIE Hong, WANG Tong

2012, 30(6): 566-572. doi:10.3969/j.issn.0255-8297.2012.06.002

Asbtract ( 2636 )

PDF (6624KB) ( 1558 )

References | Related Articles | Metrics

In wireless sensor networks, affected by sensor noise, sensor error, transmission error and other factors, data collected from different sensor nodes in the same sensor field are different, leading to uncertainty of the sensor data. This is a challenge for further data processing methods such as sensor data mining and query. To deal with the data uncertainty problem, an uncertain sensor data prediction method based on multiple variable principle component analysis (MVPCA) is proposed. The uncertain sensor data are first pretreated with MVPCA feature extraction to obtain the intrinsic featrure from the uncertain data. A multiple regression prediction method based on correlation analysis is then applied to the feature data for modeling and prediction. Sensor data of a real wireless sensor network are used to estimate the method. The results show that the proposed method can efficiently predict uncertian sensor data with high accuracy.

Decision Directed Channel Equalization for OFDM Underwater Acoustic Communications

MA Xue-fei1, QIAO Gang1, ZHAO Chun-hui2, MA Lu1

2012, 30(6): 573-580. doi:10.3969/j.issn.0255-8297.2012.06.003

Asbtract ( 2604 )

PDF (12412KB) ( 1615 )

References | Related Articles | Metrics

Complication of ocean channels causes sub-carrier amplitude fading and phase distortion in underwater communication. To reduce the effects of channel estimation error due to time variation and interpolation error, a decision-directed channel equalization algorithm for OFDM-based communication is proposed. Error control codes are used for decision making so that the data decisions can be used as pilots, with which channel estimation is performed. Taking into account different environment conditions, equalization is done using a decision-directed scheme with block piloting and a decision twice-equalization scheme with comb piloting. Results of simulation and lake experiments are promising, showing that the proposed scheme can effectively
improve system performance without loss of bandwidth efficiency.

Estimation of Fast Fading Channel for OFDM Systems Using Compressed Channel Expression

FANG Yong1,2, ZHAO Wei-jie1,2, WANG Min1,2

2012, 30(6): 581-587. doi:10.3969/j.issn.0255-8297.2012.06.004

Asbtract ( 2414 )

PDF (1195KB) ( 1621 )

References | Related Articles | Metrics

An algorithm using channel compressed expression for estimation of fast fading channel in orthogonal frequency division multiplexing (OFDM) systems is proposed. Specifically, a compressed channel expression based on compact channel impulse response (CIR) matrix and channel kernel vector is introduced, and an equivalent channel model is derived. Both least square (LS) and linear minimum mean square error (LMMSE) estimators are formulated to estimate the channel kernel vector. The CIR matrix is reconstructed from the channel kernel vector. Simulation results show that the proposed algorithm has better estimation accuracy and lower BER as compared to some existing estimation techniques.

Confidence Measure Based on Time and Boundary Features for Speech Keyword Spotting System

LI Wen-xin, QU Dan, LI Bi-cheng, WANG Bing-xi

2012, 30(6): 588-594. doi:10.3969/j.issn.0255-8297.2012.06.005

Asbtract ( 2392 )

PDF (1740KB) ( 1648 )

References | Related Articles | Metrics

As the posterior probability confidence measure cannot take full advantage of the variations in voice pronunciation, we propose an improved confidence measure based on time and boundary feature. A relaxation rate is introduced to have a flexible selection of the segmental arcs with the same words for the
calculation of confidence by which detections are rejected. On this basis, a keyword spotting system with large keyword table based on lattice is designed. An improved dynamic time warping (DTW) algorithm is used for keyword matching through the lattice to generate maximum keyword hypotheses. We have applied the proposed confidence measure in keyword verification. The results show that the equal error rate (EER) achieves 7% relative improvement compared to the mainstream method of calculation.

Sparse Representation of Signals Based on Wavelet Domain Wiener Filtering

ZHAO Zhi-peng1,2, CEN Yi-gang1,2, CHEN Xiao-fang3

2012, 30(6): 595-600. doi:10.3969/j.issn.0255-8297.2012.06.006

Asbtract ( 2546 )

PDF (6199KB) ( 1878 )

References | Related Articles | Metrics

A wavelet-based Wiener filter is proposed for signal sparse representation since the classical wavelet transform can not posses good sparse results for real signals. The proposed method can adaptively decrease the magnitude of each wavelet coefficient so that sparsity and compressibility of the wavelet coefficients is improved. This results in improvement of recovered signal quality of the compressed sensing algorithm. Simulation results show that, compared to the original sparse representation based on wavelet transform, the proposed algorithm can significantly improve quality of recovered signals for both signals and images.

Improved LMS Adaptive Filter with Convex Combined Variable Fractional Tap-Length

RUI Guo-sheng1, MIAO Jun2, ZHANG Yang2, XU Bin2, ZHANG Song2

2012, 30(6): 601-606. doi:10.3969/j.issn.0255-8297.2012.06.007

Asbtract ( 2183 )

PDF (1208KB) ( 1347 )

References | Related Articles | Metrics

In digital filters, the parameter setting and SNR affect performance of the variable order least mean square (LMS) algorithm, especially in a variable noise environment. Stable performance of variable order LMS is an important factor in the evaluation of filters. This paper analyzes stability of convex combined variable tap-length LMS. According to the variable error width, convex combination of adaptive filters for variable fractional tap-length LMS algorithm based on variable width is proposed. Theoretical analysis and simulation results show that, in a variable noise environment, the proposed algorithm can better suit the environment and converge faster, and is more stable in tap-length, as compared to other techniques including convex combination of adaptive filters for variable fractional tap-length LMS algorithm and adaptive filters for a variable tap-length LMS algorithm based on variable error width.

Improved Quasi-Monte-Carlo Particle Filtering and Its Application to Radar Target Tracking

CHEN Zhi-min, BO Yu-ming, WU Pan-long, LIU Zheng-fan

2012, 30(6): 607-612. doi:10.3969/j.issn.0255-8297.2012.06.008

Asbtract ( 2643 )

PDF (1726KB) ( 1563 )

References | Related Articles | Metrics

To address the difficulties in meeting the needs of precise and real-time radar maneuvering target tracking due to low precision and high computation complexity of quasi-Monte-Carlo particle filter (QMCPF), a new quasi-Monte-Carlo particle filter algorithm base on BP neural network (NQMC-PF) is proposed.
Through QMC fission sampling, this algorithm generates low-discrepancy progeny particles to replace the low-weight particles to guarantee validity and diversity of the samples. Meanwhile, the algorithm uses BP neural network to calculate the weight of offspring of particles. With different models, the algorithm is tested. Experimental results show that, compared to QMC-PF, the proposed algorithm can enhance precision and increase calculation speed, and thus is applicable to radar for tracking maneuvering targets.

Spectral Unmixing of Remote Sensing Images Using Interpolation of Wavelet Coefficients

LI Xi1, CHEN Feng-rui2, YU Zhi-feng1,3

2012, 30(6): 613-618. doi:10.3969/j.issn.0255-8297.2012.06.009

Asbtract ( 2351 )

PDF (8267KB) ( 1162 )

References | Related Articles | Metrics

This paper proposes a wavelet coefficient interpolation method, which uses the neighboring information in the spatial domain for spectral unmixing of remote sensing images. A super-resolution image is first generated using bilinear interpolation of wavelet coefficients. The new image is then classified to derive
a super-resolution classification map. Finally, an abundance map at the original spatial resolution is obtained using a counting window on the super-resolution classification map. This way, the original image is unmixed. A simulated TM image of Guangzhou City is used to verify the proposed method. It is found that the method performs best among three methods as it can make use of neighboring information in the space to improve unmixing accuracy.

Automatic Measurement of Central Cornea Thickness of Eye Anterior Segment Optical Coherence Tomography Image

SHU Peng1, SUN Yan-kui1, TIAN Xiao-lin2

2012, 30(6): 619-623. doi:10.3969/j.issn.0255-8297.2012.06.010

Asbtract ( 2425 )

PDF (3844KB) ( 1360 )

References | Related Articles | Metrics

To obtain quantitative parameters automatically and help medical diagnosis, automatic measurement of central cornea thickness based on edge detection and random sample consensus (RANSAC) is employed. The initial edge in the eye anterior segment optical coherence tomography (OCT) image is obtained with an edge detector. Upper and lower edges of the central cornea are extraxted using the RANSAC circle fitting method. The central cornea thickness is then computed based on the edges. Experiments show that the proposed method can avoid the effect of light beam crossing, and good results comparable to manual analysis can be obtained in real time, indicating that the method has potential applications in the future.

Application of Markov Random Field and Pyramid Structure in the Design of Digital Pattern Painting

JIA Qi1, Lü Xu-liang1, WU Chao2, RONG Xian-hui3

2012, 30(6): 624-628. doi:10.3969/j.issn.0255-8297.2012.06.011

Asbtract ( 2328 )

PDF (1796KB) ( 1593 )

References | Related Articles | Metrics

For the fast and automatic design of digital pattern painting, a design platform based on Markov random field and a pyramid structure is constructed. Major colors and their area percentages are derived using a clustering method. The Markov random field model is used to simulate natural texture distribution, and
the pyramid structure is used to decompose the digital pattern paintings to combat reconnaissance threats at different distances. The design platform is thus constructed. A test pattern painting is designed based on the background characteristics of a forest region. The results show that the proposed model can be used to design digital pattern painting automatically and quickly, resulting in effective improvements in efficiency and quality.

De-noising of SAR Images Based on Shearlets Transform

LIU Shuai-qi, HU Shao-hai, XIAO Yang

2012, 30(6): 629-634. doi:10.3969/j.issn.0255-8297.2012.06.012

Asbtract ( 3058 )

PDF (4180KB) ( 1609 )

References | Related Articles | Metrics

This paper proposes a de-noising algorithm for SAR images based on Shearlets transform. Shearlets transformation is multi-scale geometric analysis which possesses the advantages of Contourlet transform and Curvelet transform. For a singular curve or surface containing C2 high-dimensional signals, it is an optimal approximation. We apply Shearlets to approach SAR images, and use a bivariate threshold according to the Bayesian estimation theory to perform image de-noising. The obtained results show an increase of 2 dB in PSNR as compared to the Contourlet-based method with a bivariate threshold. Compared with the nonsubsampled Contourlet method with a bivariate threshold, the proposed method gives a higher PSNR and smoother denoised images. In addition, computation complexity is reduced.

Fractal Dimension Calculation of Spectrum Response Curve for Edge Enhancement

HU Qing-wu1, YUAN Hui1, SU Jun-ying2

2012, 30(6): 635-641. doi:10.3969/j.issn.0255-8297.2012.06.013

Asbtract ( 2720 )

PDF (10557KB) ( 1023 )

References | Related Articles | Metrics

In this paper, an algorithm for differential step measurement fractal dimension calculation of the spectrum response curve is proposed for image edge enhancement. A differential spectrum curve is first calculated. A step measurement fractal dimension algorithm is then presented for the differential spectrum curve. Finally, the differential fractal feature image is generated for the feature analysis of hyperspectral images. Experiment results show that the presented algorithm can enhance edges while preserving detailed textures and reducing noise. The proposed method can be used in the feature extraction and analysis of hyperspectral images.

Image Super-Resolution Reconstruction Based on Multi-groups of Coupled Dictionary and Alternative Learning

SUN Guang-ling, SHEN Zhou-biao

2012, 30(6): 642-648. doi:10.3969/j.issn.0255-8297.2012.06.014

Asbtract ( 2920 )

PDF (15555KB) ( 1068 )

References | Related Articles | Metrics

A super-resolution image reconstruction method based on multi-groups of coupled dictionary and alternative learning is presented. In the dictionary learning phase, the image from a training set is viewed as high resolution (HRI). A reduced and re-enlarged version of the HRI is low resolution (LRI). The difference
between HRI and LRI is the residual. The mapping between residual and LRI is obtained from the coupled dictionaries based on the joint data composed of residual patch and LRI patch features. In the reconstruction phase, an enlarged version of the input image is taken as LRI. For each LRI patch, sparse representations and corresponding errors are calculated by using low resolution components of each group of the coupled dictionary. The residual components of coupled dictionary with minimum errors is used to reconstruct the corresponding residual image patch. All reconstructed residual patches together are used to form a residual image, which is then combined with the LRI to produce an HRI. The experimental results demonstrate a satisfied superresolution reconstruction quality.

Spiking Neuron Model Based on Single-Electron Transistors

LIU Wen-peng, CHEN Xu, LU Hua-xiang

2012, 30(6): 649-654. doi:10.3969/j.issn.0255-8297.2012.06.015

Asbtract ( 2253 )

PDF (2607KB) ( 1204 )

References | Related Articles | Metrics

To implement the spiking neurons with pulse input and output, this paper studies similarity between current pulses and the Coulomb oscillations of single-electron transistor (SET). Based on the similarity, a new spiking neuron circuits model is designed and implemented with SETs, and simulated using PSPICE. The simulation results demonstrate that the model meets the needs of regular spiking of Izhikevich spiking neurons and has the capacity of spiking encoding via input and output of current pulses.

RET Antenna Scan Code Formation and Collision Resisting Algorithm

DENG Chun-jian1, WANG Xu-zhi2, LUO Ren-ze3, DENG Sheng-hua1

2012, 30(6): 655-660. doi:10.3969/j.issn.0255-8297.2012.06.016

Asbtract ( 2351 )

PDF (1655KB) ( 1436 )

Table of Content