Journal of University of Chinese Academy of Sciences

Estimation of error-in-variable varying-coefficient model with auxiliary instrument variables

LIU Zhifan, WANG Miaomiao, XIE Tianfa, SUN Zhihua

2018, 35 (1): 1-9. DOI: 10.7523/j.issn.2095-6134.2018.01.001

In this work, we consider the estimation of the variable-coefficient model when the covariates are measured with error. We do not specify any model structure of the measurement error, and do not require the knowledge of the variance of measurement error. Furthermore, repeated measurement data are not necessary. With the help of the instrument variable, we calibrate the error and obtain an estimator of the true variable. We replace the true variable by its estimator and get an estimator of the coefficient function by applying the local linear smoothing method. We prove the asymptotic normality of the proposed estimator. The simulation results show that the proposed estimator performs better than the naive estimator.

A restricted B-spline estimation of diffusion processes with applications to financial data

HOU Yangyang, XU Tiange

2018, 35 (1): 10-17. DOI: 10.7523/j.issn.2095-6134.2018.01.002

Abstract ( 257 ) PDF (KB) ( 1 )

Time-homogeneous diffusion process plays an important role in the financial market, and it is widely used for describing the stochastic dynamics of the underlying economic variables. In this work, we study nonparametric estimation of the drift and diffusion functions for the time-homogeneous diffusion process, and propose a new nonparametric regression technique based on higher-order approximations, which is called B-spline approach. The nonnegativity of the diffusion function is guaranteed by the restricted B-spline method. Our simulation results show that our method indeed outperforms the local polynomial method.

Robust optimization models for study of wireless resource scheduling problem with uncertain transmission rate

TIAN Leixia, YANG Wenguo, GAO Suixiang, JIANG Zhipeng

2018, 35 (1): 18-25. DOI: 10.7523/j.issn.2095-6134.2018.01.003

Abstract ( 256 ) PDF (KB) ( 10 )

In the long-term evolution system, the wireless resource scheduling problem with uncertain transmission rate is how to distribute resource blocks to users in each time slot to ensure user experience of time delay no matter how resource block transmission rate changes. The problem is solved by using the robust optimization method in this work. We establish the robust optimization model of uncertain wireless resource scheduling problem, and then select three kinds of special uncertain sets, i.e., box uncertain set, ellipsoid uncertain set, and uncertain set with the distribution information partly known. Based on the feature of the three sets we obtain their reasonable equivalent robust corresponding models. Finally we use a living example to verify the validity of the robust corresponding models.

An indirect time synchronization method for satellite communication

ZHANG Jie, MA Guanyi, WANG Zhaorui, HU Chao

2018, 35 (1): 26-32. DOI: 10.7523/j.issn.2095-6134.2018.01.004

Abstract ( 281 ) PDF (KB) ( 1 )

An indirect time synchronization method is proposed to be applied in forward link for time synchronization requirements of satellite spread spectrum communication system. Based on GNSS timing, an indirect clock correction loop consisting of ground-satellite round-trip link and phase lock loop is designed to correct the transmitter clock. The initial code phase of forward link spread spectrum signal is synchronized with GNSS 1 PPS at the receiving end. The spread spectrum signal timing receiving without bit synchronization is realized. A formula for time synchronization error and relationship of the error with loop bandwidth of the indirect clock correction loop are derived. The optimal loop bandwidth to minimize the synchronization error is also proposed. The satellite-ground link experiments show that variation of the synchronization error with the loop bandwidth is in agreemeent with theoretical analysis, and the minimum 1σ error of 20 ns is reached. It is concluded that the proposed method is efficient and qualified for satellite spread spectrum communication system. Under the condition of 20 ns synchronization error, the receiver has successfully finished the synchronization of the spread spectrum pseudo code and the message demodulation. The auxiliary of satellite navigation timing for communication is realized.

Characteristics of meso-micro soil fauna community under different plant configurations of urban green land in Wenjiang District of Chengdu City

HUANG Yumei, HUANG Shenglan, ZHANG Jian, LIU Pan, ZHANG Kai, WANG Ruoran, XIONG Xi

2018, 35 (1): 33-41. DOI: 10.7523/j.issn.2095-6134.2018.01.005

Abstract ( 228 ) PDF (KB) ( 2 )

An investigation on meso-micro soil fauna community structure was carried out under 4 plant configurations of urban green lands in Wenjiang District of Chengdu City. A total of 20 600 individuals of meso-micro soil fauna were obtained in different green lands, belonging to 19 orders and 45 families. Nothotylenchidae and Rhabditoidea were the dominant groups under 4 plant configurations. The number of groups and individual density of meso-micro soil fauna under lawn were higher than under the other 3 plant configurations in spring, summer, and autumn, while from autumn to winter they were higher under tree-shrub-grass land than under the other 3 plant configurations. The individual density of meso-micro soil fauna showed significant differences (F=3.23,P<0.05) among different plant configurations, and the C and J indexes of meso-micro soil fauna showed highly significant differences (F=7.48,P<0.01 for C;F=6.13,P<0.01 for J). The rank of groups and individual quantities of meso-micro soil fauna was saprozoie > predators > phytophage > omnivorous. The ratio of Acarina number to Collembola number changed with time, and it fluctuated more in winter and spring.

Effects of cloudy atmosphere on microwave signals in channels of AMSR-E

SUN Chuan, SONG Xiaoning, ZHOU Fangcheng, LI Zhaoliang

2018, 35 (1): 42-49. DOI: 10.7523/j.issn.2095-6134.2018.01.006

Abstract ( 318 ) PDF (KB) ( 13 )

Passive microwave remote sensing has the ability to obtain surface radiation information through clouds, fog, rain, and snow. Therefore, in cloudy weather, it is obviously better than thermal infrared remote sensing in land surface temperature retrieval. However, the clouds and the atmospheric molecules affect the microwave signals to some extent. The effects of cloudy atmosphere on microwave signals in the advanced microwave scanning radiometer-earth observing system (AMSR-E) channels were studied by the way of combining theoretical analysis and model simulation. Results show that the effects can be ignored in 6.925 and 10.65 GHz channels, but cannot be ignored in 18.7, 23.8, 36.5, and 89 GHz channels. The effects in the last four channels can be expressed as functions of precipitable water vapor and cloud liquid water. Based on the above study and further analysis, three channels, 18.7, 23.8, and 36.5 GHz, are selected to build a land surface temperature inversion algorithm with higher accuracy.

A background subtraction and frame subtraction combined method for moving vehicle detection in satellite video data

YUAN Yiqin, HE Guojin, WANG Guizhou, JIANG Wei, KANG Jinzhong

2018, 35 (1): 50-58. DOI: 10.7523/j.issn.2095-6134.2018.01.007

Abstract ( 326 ) PDF (KB) ( 3 )

Appearance of video satellite brings a new favorable opportunity for real-time observation of remote sensing and provides a kind of new data for dynamic monitoring and target tracking. Based on the differences in target detection between remote sensing satellite video and traditional surveillance video, this work indicates the problems about applying existing target detection algorithm directly to satellite video. A new approach which combines the background subtraction and frame subtraction technologies is proposed for better detection of the moving target in remote sensing satellite video. Moving vehicles in 4 frame images acquired from UrtheCast videos are detected by using the background subtraction method, the frame subtraction method, and the proposed method. The results indicate that the proposed method has a good ability in reducing the errors of mobile background edges and residual noises, improves the correctness and quality of detection, and has a promising application in moving target detection in satellite videos.

Analysis on time reversal imaging in presence of random media

CHANG Jingming, JIN Ming, ZENG Jiangyuan, CHEN Kunshan

2018, 35 (1): 59-65. DOI: 10.7523/j.issn.2095-6134.2018.01.008

Abstract ( 400 ) PDF (KB) ( 3 )

Time-reversal (TR) has the advantages of space-time focusing and super resolution. In this work, the effects of random media on TR focusing and imaging are analyzed. The simulation experiments are carried out to investigate the performance of TR focusing and imaging, in cases of different array sparsities, varying optical depths, different locations, and different albedos of the random medium. Results indicate that the multipath scattering within the random media enhances the super resolution capability of TR imaging. Also, the imaging performance of TR focusing shows considerable adaptability to various random media parameters and array sparsities.

Spatial structure and evolution of basic education facilities in Nanjing in China

TU Tangqi, CHEN Jianglong, WEI Yehua Dennis, LIANG Qichun, ZHANG Yinghao

2018, 35 (1): 66-74. DOI: 10.7523/j.issn.2095-6134.2018.01.009

Abstract ( 249 ) PDF (KB) ( 2 )

In this work we used GIS clustering analysis, kernel density analysis, gravitational ellipse method, and social network analysis to analyze the location and network structure of education facilities in Nanjing, beyond the traditional emphasis on quantity of education facilities and static perspectives. The spatial structure and evolution characteristics were analyzed based on the quantity, quality, and connection of basic education facilities in the eight districts of Nanjing from 2006 to 2015. The results show that the spatial concentrations of kindergartens and primary schools have increased, while the concentrations of junior high schools and high schools have decreased. The average travel distances among schools have declined except those among junior high schools. The boundaries of high-density school areas have significantly expanded except those of high school facilities. Primary schools showed a mononuclear structure expansion while middle schools were intended to form three clustering areas, and a new cluster of kindergartens has formed in Jiangning District. The development of the network of elite schools has evolved over time, and the heterogeneity of the network increased first and then declined. The boundary effect widened first and then narrowed. The center of gravity of education quality moved northward and westward. The area of higher education quality expanded, and mainly concentrated in Gulou District. This work proposes policies for future development and planning of education facilities.

Target recognition using the transfer learning-based deep convolutional neural networks for SAR images

LI Song, WEI Zhonghao, ZHANG Bingchen, HONG Wen

2018, 35 (1): 75-83. DOI: 10.7523/j.issn.2095-6134.2018.01.010

Abstract ( 806 ) PDF (KB) ( 1 )

The automatic target recognition procedure of synthetic aperture radar (SAR) generally includes two steps, feature extraction and classifier training. Based on the development of deep convolutional neural networks, we present a new method of SAR target recognition. This method automatically learns the hierarchies of features from different targets, which means it avoids the non-normalization caused by manual feature extraction. Then the transfer learning technology is applied to avert the occurrence of locally optimal solution and accelerate the training procedure. Finally we use the moving and stationary target acquisition and recognition database to verify our method.

Improved PCA method for SAR target recognition based on sparse solution

XIAO Yao, LIU Chang

2018, 35 (1): 84-88. DOI: 10.7523/j.issn.2095-6134.2018.01.011

Abstract ( 245 ) PDF (KB) ( 2 )

In this work, an improved principal component analysis method (IPCA) is proposed for SAR target recognition. Firstly, we use the sparse method to obtain the training samples which are most relevant to the test samples and their representation coefficients for the test samples. Then, using the principal component analysis (PCA) we obtain the optimal projection matrix so that different test samples after projection can be better classified by using the training sample information. The results of experiments, performed on SAR ground stationary targets based on the moving and stationary target acquisition and recognition (MSTAR) database, show that IPCA reaches higher recognition rate and better robustness to sparse aspect training samples of three true objects than PCA.

Building extraction from polarimetric SAR image based on feature selection and two-level classification

ZHANG Miaoran, LIU Chang

2018, 35 (1): 89-95. DOI: 10.7523/j.issn.2095-6134.2018.01.012

Abstract ( 295 ) PDF (KB) ( 3 )

In this work, a building extraction method, based on feature selection and two-level classification, for polarimetric SAR (POLSAR) image is proposed. Firstly, some polarimetric features and texture features are extracted from POLSAR data via refined Lee filter as the initial feature set. Then, by using the random forest as the primary classifier and evaluating features' importance meanwhile, the feature subset is obtained according to the importance rank. Secondary classification is made for the selected feature subset by the support vector machine, and the final result is achieved by combining the primary results and secondary results together using neighborhood voting. The experimental results on AIRSAR system demonstrate that the proposed method effectively improves the extraction accuracy.

Dealing with Doppler migration for passive radar based on SDCFCP

ZHANG Dan, LÜ Xiaode, LI Daojing, YANG Pengcheng, CHAI Zhihai

2018, 35 (1): 96-101. DOI: 10.7523/j.issn.2095-6134.2018.01.013

Abstract ( 252 ) PDF (KB) ( 2 )

Long-term coherent integration is a main method to improve gain in target identification for passive radar. However, tangential-speed leads to Doppler expansion based on bistatic model which decreases the gain and then degrades the detection range. In this work we discuss the contribution factor of Doppler migration and propose an algorithm to deal with Doppler migration based on sub-band double carrier frequency conjugated processing (SDCFCP). SDCFCP increases the equivalent wavelength of composite signal, makes Doppler quadratic term close to zero, and solves the Doppler migration problem. We also propose an improved method to decrease PRF at azimuth and enhance the weak target detection ability. SDCFCP can be applied conveniently with little calculation work. Moreover, the algorithm is useful on multi-target scenario. Finally, experiments based on simulated and real signals verify the proposed algorithm.

Retrieval of land surface emissivity using spectral and texture features based on neural network

XU Kaifa, LEI Bin, ZHANG Yueting

2018, 35 (1): 102-108. DOI: 10.7523/j.issn.2095-6134.2018.01.014

Abstract ( 197 ) PDF (KB) ( 8 )

Land surface emissivity is one of the most important parameters in thermal infrared remote sensing and plays a significant role in the quantitative study of thermal infrared remote sensing, the surface energy balance, and feature mapping. Retrieving the land surface emissivity from thermal infrared remote sensing data is a challenge because it presents an ill-posed problem. In this work, a method, which takes advantage of spectral and texture features of many visible channels available in the moderate resolution imaging spectroradiometer (MODIS) data and is based on back-propagation neural network to obtain land surface emissivity pixel-by-pixel, is presented. The method obtains the land surface emissivity map without the categorization of the land cover and the analysis indicates that the average error, compared to MODIS emissivity product, is within 0.002. It builds a direct relationship between reflectance data and emissivity data, and provides the possibility of obtaining precise emissivity data through the single channel thermal infrared satellite.

A lip-reading recognition approach based on long short-term memory

MA Ning, TIAN Guodong, ZHOU Xi

2018, 35 (1): 109-117. DOI: 10.7523/j.issn.2095-6134.2018.01.015

Abstract ( 410 ) PDF (KB) ( 2 )

Visual speech information is the important carrier of conversation. However, visual speech informations from different speakers are different due to various appearances of lips, various backgrounds, and various talking ways even the content of the conversation is the same. To address the problem of variety of visual speech information, we propose a new approach for lip-reading recognition based on long short-term memory (LSTM). We compute the positions of lip landmarks which describe the dynamic information of the shape as the features of the lip-reading video, and it has the characteristics of within-class consistency and between-class distinctiveness. Then we use LSTM to encode temporal information, and it learns spatio-temporal features which have the ability of discrimination and generalization. Our approach is evaluated on three public databases (GRID, MRIALC, and OuluVS) for lip-reading recognition of isolated words or phrases in speaker independent experiments. On GRID and MRIALC, the accuracy of our approach is more than 30% highter than that of the conventional approach. On OuluVS, the accuracy of our approach is comparable to state of the art. The experiment results indicate that our lip-reading recognition approach solves the problem of variety of visual speech information effectively.

The parallel methods of simultaneous equations in discontinuous deformation analysis

XIAO Yunfan, XIAO Jun, MIAO Qinghai, WANG Ying

2018, 35 (1): 118-125. DOI: 10.7523/j.issn.2095-6134.2018.01.016

Abstract ( 152 ) PDF (KB) ( 2 )

As an advanced numerical analysis method, discontinuous deformation analysis (DDA) shows relatively low efficiency. Simultaneous equations solusion is the bottleneck. This work focuses on how to choose an optimal solver in DDA. Jacobi and Jacobi preconditioned conjugate gradient methods are paralleled with OpenMP and CUDA. Three factors affecting the efficiency, computational scale, time step size, and block contact, are analyzed. Based on the tests, an empirical formula for choosing optimal solver in different computing conditions is obtained.

Joint user-association and power-control algorithm based on multiple association in ultra-dense network

ZHANG Jian, QIU Ling, CHEN Zheng

2018, 35 (1): 126-130. DOI: 10.7523/j.issn.2095-6134.2018.01.017

Abstract ( 186 ) PDF (KB) ( 2 )

Ultra-dense network is one of the key technologies to improve the capacity of the fifth generation mobile communication system. In this work, we assume that one user can connect to multiple base stations to make full use of the stations. In order to further improve the throughput of the system using more base stations, we study user-association and power-control in the multiple association condition. To consider the network throughput and the equality of users, we model this problem as a problem of maximizing the logarithm of user rate. Since this problem is difficult to solve, we divide this problem into two subproblems, which can be solved efficiently. Simulation results reveal that the proposed algorithm is better than the existing algorithms in system performance.

Deep learning model for suicidal identification of Chinese microblogs

TIAN Wei, ZHU Tingshao

2018, 35 (1): 131-136. DOI: 10.7523/j.issn.2095-6134.2018.01.018

Abstract ( 406 ) PDF (KB) ( 10 )

With the development of internet, more and more people express their emotion and feeling in social media, including suicidal ideation. There is a new opportunity for suicide prevention, if people with high suicidal risks can be identified through social media like microblog. In this work we attempt to set up the suicidal ideation recognizer with deep learning model, and address the possibility of suicidal risk assessment in social media. In order to verify the validity of the algorithm model,we also conduct statistical analysis of the key word features, and compare the incorrectly classified instances with other two algorithms. The results indicate that deep learning algorithm works more effectively for suicide risk identification.

Design and experiments of an adaptive OFDM system for visible light communication

CAI Shaoyang, CHEN Li, WANG Weidong

2018, 35 (1): 137-143. DOI: 10.7523/j.issn.2095-6134.2018.01.019

Abstract ( 299 ) PDF (KB) ( 2 )

As an emerging wireless communication technology, visible light communication (VLC) is one of the effective solutions for the shortage of wireless spectrum resource. In this work, a low-complexity design scheme of orthogonal frequency division multiplexing (OFDM) system is proposed based on the mask invariable characteristics of the received signal-to-noise rate of VLC in the electrical domain. The scheme is verified by the experiments. The experimental results show that, compared with the adaptive equal bits transmission scheme, the proposed adaptive OFDM system makes full use of the bandwidth resources of wireless optical channel, and it shows improvements in transmission rate and BER performance.

Current Issue