Table of Content

20 June 2010 Volume 37 Issue 3

  

Original Articles

Novel MAC protocol for MIMO link ad hoc networks

CHEN Dan;LI Jian-dong;LI Chang-le

J4. 2010, 37(3):  385-390+422.  doi:10.3969/j.issn.1001-2400.2010.03.001

Abstract ( 1421 )   PDF (730KB) ( 1133 )   Save

References | Related Articles | Metrics

Considering the throughput reduction caused by the packet collision of the PRP MAC, a novel MAC based on five phase reservation (FPR-PRP MAC) is proposed. The FPR-PRP MAC adopts a synchronous time structure, and it allows the nodes in the network to acquire transmission slot numbers through five phase reservation. Then in information slots, some nodes can transmit RTS packets with higher priority according to the reservation results, and other nodes contend with each other to transmit RTS packets. A receiving node determines the maximum allowable number of data streams according to all the RTS packets it has received, and put this number in the CTS packet to reply to the transmitter. The mechanism of combining reservation with contention in FPR-PRP MAC reduces the collisions of RTS packets in the network. Moreover, the synchronous time structure in FPR-PRP MAC eliminates the collisions of the control packet with the data packet. Therefore, the proposed FPR-PRP MAC can improve the performance of the network effectively. Simulation results show that, in a heavy load situation, compared with the PRP MAC, the FPR-PRP MAC can transmit many more packets in a stream-controlled way, and increase the network throughput.

SNR wall for energy detection over fading channels in cognitive radio

YAN Qi;YANG Jia-wei;ZHANG Wen;DONG Wei

J4. 2010, 37(3):  391-395+411.  doi:10.3969/j.issn.1001-2400.2010.03.002

Abstract ( 2029 )   PDF (583KB) ( 1334 )   Save

References | Related Articles | Metrics

We have obtaiend the “SNR Wall”, which is induced by noise uncertainty, of the energy detector over multipath fading channels in the cognitive radio. The robust detection model under noise uncertainty is built up. By finding the statistically averaged probability of robust detection over the fading distribution, the closed-form expression for detection probability is derived by making a proper approximation and transformation of the complex integral expression. Then, the SNR Wall is acquired by synthesizing the expressions for the detection and false alarm probability and eliminating the detection threshold. Analytical and simulation results show that the SNR Wall of energy detection over fading channels is related to the noise uncertainty, and increases dramatically with the rising of the target detection probability as well. Therefore, the SNR wall of the energy detector becomes very large when a high target probability of detection is required, and the energy detector is not applicable to the cognitive radio with a small received SNR.

Adaptive quantization parameter determination for key frames

LI Ming;CHANG Yi-lin;YANG Fu-zheng;WAN Shuai

J4. 2010, 37(3):  396-404+435.  doi:10.3969/j.issn.1001-2400.2010.03.003

Abstract ( 1591 )   PDF (754KB) ( 1229 )   Save

References | Related Articles | Metrics

Based on the rate control scheme in JVT-AA029 for the video encoders using hierarchical B frames, a quantization parameter (QP) determination method is proposed to adaptively determine the QP for key frames. This method firstly utilizes an empirical equation expressing the relation among the starting QP, the available channel bandwidth and the source characteristics to calculate the starting QP for the first key frame of the video sequence. For the other key frames, the target bit budgets and reference QPs are determined based on the actual coding properties of the previous frames. Then the QPs are calculated by employing a rate-distortion model and further adjusted according to the reference QPs. Experimental results show that the proposed method leads to an average improvement of 0.14dB over the performance of the original rate control scheme in JVT-AA029, and by integrating the proposed method, the rate control scheme outperforms the existing state-of-the-art one by 0.77dB on average.

Efficient SVC transmission system based on network coding

LU Ji;XIAO Song;WU Cheng-ke

J4. 2010, 37(3):  405-411.  doi:10.3969/j.issn.1001-2400.2010.03.004

Abstract ( 1747 )   PDF (980KB) ( 1241 )   Save

References | Related Articles | Metrics

Focusing on improving the video transmission efficiency and quality over the internet simultaneously,we put forward a video robustness transmission system, on the basis of  which the Scalable Video Coding(SVC) characteristics description framework is proposed.  The effective combination of Random Linear Network Coding (RLNC) technique and SVC technique is considered.  The SVC characteristic expressions for packetization, distortion and latency are deduced respectively. The packetization algorithm, bit stream organization algorithm and bit stream transmission algorithm derived from the characteristics are introduced to improve video reconstruction quality and decrease transmission latency. Simulation results show that the proposed algorithms can lead to better performance. The conclusions are helpful in designing a practical SVC transmission system.

A class of nonbinary quasi-cyclic LDPC codes with a low-density generator matrix

CHEN Chao;BAI Bao-ming;WANG Xin-mei

J4. 2010, 37(3):  412-416.  doi:10.3969/j.issn.1001-2400.2010.03.005

Abstract ( 1756 )   PDF (518KB) ( 1070 )   Save

References | Related Articles | Metrics

This paper presents a class of nonbinary quasi-cyclic low-density parity-check (QC-LDPC) codes, whose parity-check matrix consists of weight-2 columns. The parity-check matrix is carefully designed such that the corresponding generator matrix has some nice properties: (1) systematic, (2) quasi-cyclic, and (3) low-density, which allow a parallel encoding with low-complexity through simple shift-register circuits. Simulation results show that the proposed codes do not have performance degradation compared with randomly constructed codes.

Low rate multimode MBE vocoder with linear prediction coding

LIANG Yan-xia;YANG Jia-wei;LI Ye

J4. 2010, 37(3):  417-422.  doi:10.3969/j.issn.1001-2400.2010.03.006

Abstract ( 1682 )   PDF (1030KB) ( 1183 )   Save

References | Related Articles | Metrics

This paper presents a model of the Multimode Multi-Band Excitation (MBE) joined with the Linear Prediction vocoder. Compared with MBE coding, this vocoder has some improvements in two ways. First, the Linear Predictive Mode is combined with Spectrum Estimation, and then the linear predictive coefficients are denoted and quantized by Line Spectral Frequency parameters (LSF), which solves the problem of variable dimensions in spectrum estimation in the MBE system. Second, in Unvoiced/Voiced (U/V) Decision, the number of sub-bands in one frame is fixed, ahd then the decisions are divided into different modes. LSF vectors are quantized by different codebooks to improve the quantization result, according to the different statistic distributions of the LSF vector in different U/V determinative modes. Simulation by the C programming language demonstrates that the synthesized speech Intelligibility is high and that the speech waveforms in the time domain and spectrograms both fit well.

Multi-branch fractional multi-bit differential detection for continuous phase modulation

SUN Jin-hua;WU Xiao-jun

J4. 2010, 37(3):  423-428.  doi:10.3969/j.issn.1001-2400.2010.03.007

Abstract ( 1762 )   PDF (751KB) ( 1074 )   Save

References | Related Articles | Metrics

The multi-branch fractional multi-bit differential detection technique for improving the performance of continuous phase modulation signals is proposed. The distance between different phase states is increased by combination and decision of different multi-bit differential signals. The essence of the technique is physically explained, and a theoretical analysis of the upper bound of the fractional bit duration for maximum performance improvement is conducted. Simulation results and further analysis show that the multi-branch fractional multi-bit differential detection technique increases the distance between phase states compared with 1-bit differential detection and 2-bit differential detection, which leads to the performance improvement; when the fractional bit duration is the maximal duration in which the polarity of the differential signal is the same as the decision symbol, the maximum performance gain can be obtained.

New face recognition algorithm using tensor local and global information

WEN Hao;SUN Lei

J4. 2010, 37(3):  429-435.  doi:10.3969/j.issn.1001-2400.2010.03.008

Abstract ( 1765 )   PDF (624KB) ( 1492 )   Save

References | Related Articles | Metrics

The current algorithms based on tensor subspace manifold learning can utilize the intrinsic geometrical structure of images. But the local information and global information are not utilized sufficiently in current algorithms. A novel tensor subspace learning algorithm is proposed in this paper which is named tensor local and global projection. The local nonlinear structure of the data manifold that is the local information of the data can be preserved in the algorithm, and at the same time, the global information of data is utilized. So the discriminant between classes of data in low dimension subspace can be maximized. And the optimal tensor subspace can be obtained by iteratively computing the generalized eigenvectors and projection. The experiments on the standard face database demonstrate that the right recognition rate of the novel algorithm is higher than the recognition rate of the four algorithms named TLDA,TMFA,TLDP and TSA.

Generalized AIC based channel estimation using circularly orthogonal sequences

LI Dan-ping;ZHANG Hai-lin

J4. 2010, 37(3):  436-440+486.  doi:10.3969/j.issn.1001-2400.2010.03.009

Abstract ( 1749 )   PDF (631KB) ( 1157 )   Save

References | Related Articles | Metrics

Circularly orthogonal training sequences are designed for multiple-input multiple-output single carrier frequency domain equalization (MIMO-SCFDE) systems, which can be used in time-domain channel estimation. By using the perfect auto-correlation and cross-correlation properties of the training sequences, the proposed algorithm not only suppresses the interference between the antennas but also reduces the variance of the additive white Gaussian noise. Then the time delay of each subpath can be determined by using the generalized Akaike information criterion (GAIC), which helps to further improve the accuracy of the channel estimaton. Simulation results in sparse MIMO channels show that the GAIC based channel estimation algorithm using circularly orthogonal sequences can decrease the effect of noise on the channel estimation and reduce the mean square error greatly.

High resolution imaging method for the sparse aperture of ISAR

LI Jun;XING Meng-dao;ZHANG Lei;WU Sun-jun

J4. 2010, 37(3):  441-446+453.  doi:10.3969/j.issn.1001-2400.2010.03.010

Abstract ( 2440 )   PDF (1499KB) ( 1415 )   Save

References | Related Articles | Metrics

The conventional linear interpolation method will introduce the estimation error when the gapped data of ISAR is large. A high resolution imaging method based on compressed sensing for gapped data of ISAR is proposed in this paper. We first construct a sparse basis dictionary and linear measurement matrix. Then the convex optimization means as the base pursuit can be used to extract the scatter property and Doppler frequency information. By all the steps above, the image of ISAR is formed in the plane of range and Doppler. No interpolation is needed for the sparse aperture, and no sidelobes exit in the ISAR images. Simulation results and real sparse ISAR data validate the feasibility and superiority of the approach.

Passive localization algorithm for minimizing the measurement error

CHEN Jin-guang;LI Jie;GAO Xin-bo

J4. 2010, 37(3):  447-453.  doi:10.3969/j.issn.1001-2400.2010.03.011

Abstract ( 1893 )   PDF (589KB) ( 1118 )   Save

References | Related Articles | Metrics

In the passive localization of multiple passive sensors, there are some constraints among sensors'position, measurement error and measurements. Measurement errors will be reduced if the constraint can be used rationally. The paper derives constraint formulas for measurements of different sensors by means of every sensor's coordinate and corresponding trigonometric function. Then objective functions which regard measurement errors as variables are given. The measurement errors are minimized by the Lagrange-Newton optimization method, and then measurements can be modified. Finally, more accurate results of target localization can be obtained. The proposed algorithm does not need the initial target position and the results did not diverge, and the proposed algorithm has a better localization accuracy. Simulation results verify the effectiveness of the new algorithm.

Knowledge-aided space-time adaptive detection method

ZHOU Yu;ZHANG Lin-rang;LIU Nan;LIU Xin

J4. 2010, 37(3):  454-458+546.  doi:10.3969/j.issn.1001-2400.2010.03.012

Abstract ( 1844 )   PDF (589KB) ( 1268 )   Save

References | Related Articles | Metrics

The heterogeneous clutter results in insufficiency of available secondary data in radar Space-Time adaptive detection. To solve the problem, this paper deals with the design and the analysis of adaptive detector by exploring the Knowledge-based (KB) processing technique. Considering the theory that the outer product of independent identical Gaussian distribution vectors is the Wishart distribution, the covariance matrix in the Wald detector is assumed to be inverse Wishart distribution. The parameter matrix can be obtained by exploring the sensed environment. The proposed Knowledge-based Wald detector makes use of information from both secondary data and prior knowledge available about the sensed environment. Results show that noticeable performance improvements can be obtained especially when only a small number of secondary data is available.

Incremental principal component analysis using  the approximated covariance matrix

CAO Xiang-hai;LIU Hong-wei;WU Shun-jun

J4. 2010, 37(3):  459-463.  doi:10.3969/j.issn.1001-2400.2010.03.013

Abstract ( 1507 )   PDF (674KB) ( 1166 )   Save

References | Related Articles | Metrics

Firstly, with eigenvectors orthogonal to each other, the computation complexity of the subspace projection(SP) algorithm is reduced to 1/P of the original algorithm(where P is the number of desired eigencomponents). Then, the covariance matrix is replaced by the approximated covariance matrix which is composed of large eigenvalues and corresponding eigenvectors, the computation complexity can be reduced to 1/N of the original algorithm(where N is the input vector dimension)further. Finally, experimental results based on the ORL face database demonstrate the efficiency of the presented algorithm.

Efficient wideband digital channelized receiver design

CHANG Hong;ZHAO Guo-qing;NIU Xin-liang

J4. 2010, 37(3):  464-468.  doi:10.3969/j.issn.1001-2400.2010.03.014

Abstract ( 1887 )   PDF (684KB) ( 1453 )   Save

References | Related Articles | Metrics

The short-time Fourier transform(STFT) algorithm is an easy and efficient method for digital channelized reception with limitation in the operation speed of DSP. The method based on extend resources with parallelism pipeline architecture for high speed is supposed in this paper to realize transport data at a low speed and make efficient use of hardware.Detection and estimate parameters of signal in real-time are also realized. The method is suitable for hardware implementation at a high speed.

Study on dechirp SF ISAR imaging

LIU Ya-bo;LI Jun;LI Ya-chao;XING Meng-dao

J4. 2010, 37(3):  469-475+519.  doi:10.3969/j.issn.1001-2400.2010.03.015

Abstract ( 1571 )   PDF (1090KB) ( 1098 )   Save

References | Related Articles | Metrics

A new high-resolution ISAR imaging method including two coherency processing steps is presented for the linearly stepped-frequency (SF) modulated radar using the dechirp technique. First, the coherency preprocessing is applied to the echoes to estimate the target radial velocity and acceleration rate, and then these motion parameters are compensated. Second, a band coherence technique is used to adjust the residual phase error, and a high resolution range profile can be obtained by means of a time-domain band append technique. Finally, a well-focused ISAR image is obtained via the conventional ISAR imaging algorithm. Simulation results verify the effectiveness of the proposed method.

Design of a low input voltage startup circuit for the Boost DC-DC converter

CHEN Fu-ji;LAI Xin-quan;LI Yu-shan;YE Qiang;YUAN Bing

J4. 2010, 37(3):  476-480+506.  doi:10.3969/j.issn.1001-2400.2010.03.016

Abstract ( 1531 )   PDF (726KB) ( 1217 )   Save

References | Related Articles | Metrics

This paper presents a startup circuit for a low voltage Boost DC-DC converter. Firstly, the DC-DC converter is started in the open loop, and then the converter is powered by the output port for normal operation. Secondly, the minimum voltage of the circuit is as low as 1.5V. Thirdly, the quiescent current of the V_IN Pin is small. Finally, the circuit is designed and implemented by UMC 0.6μm BCD technology. Experimental results demonstrate that DC-DC can be started up in an input voltage greater than or equal to 1.5V. The quiescent current of the V_IN is only 18μA. With V_IN=1.5V and V_OUT=7V, the peak current of the inductor is 2A.

Novel method for the narrow band signal frequency and angle estimation

LIU Xiao-jun;LIU Cong-feng;LIAO Gui-sheng

J4. 2010, 37(3):  481-486.  doi:10.3969/j.issn.1001-2400.2010.03.017

Abstract ( 1783 )   PDF (588KB) ( 1436 )   Save

References | Related Articles | Metrics

For the problem of the frequency and direction of arrive (DOA) united estimation of the narrow band signal for a wide band receiver, a novel quick method is proposed. First, the pre-estimation is used to get the pre-estimating DOA with the reference frequency, then the space filtering method is used to make the match of the signal pre-estimating DOA with its frequency, and the corresponding frequency is used to adjust the pre-estimating DOA according to the characteristic of the antenna configuration, so that, the precise signal frequency and its DOA are obtained at the same time. Simulation indicates that the proposed method is correct and effective for estimating the frequency and DOA of a narrow signal for a wide band receiver.

Wideband digital channelized receiver design

WANG Hong-wei;ZHAO Guo-qing;WANG Yu-jun;BAO Dan

J4. 2010, 37(3):  487-491+553.  doi:10.3969/j.issn.1001-2400.2010.03.018

Abstract ( 1886 )   PDF (713KB) ( 1507 )   Save

References | Related Articles | Metrics

The digital receiver should possess the ability to process a large number of data in real-time for the wideband signal. A new kind of efficient digital channelized receiver is proposed in this paper which is based on polyphase filters and short-time FFT technology(polyphase fast Fourier transform). It can detect objects and measure parameters for the wideband signal in real-time. A digital receiver system with a 640MHz instantaneous bandwidth is built based on the structure proposed using FPGA and DSP devices. It has high accuracy of time-frequency parameter measurement and low data output ratio, which shows the structure is feasible.

Level set SAR image segmentation method combined with ratio edge information

Lü Yan;FENG Da-zheng

J4. 2010, 37(3):  492-495.  doi:10.3969/j.issn.1001-2400.2010.03.019

Abstract ( 1643 )   PDF (1072KB) ( 1106 )   Save

References | Related Articles | Metrics

A variational level set synthetic aperture radar (SAR) image segmentation method based on edge and region information is proposed. An energy functional adapted for SAR image segmentation is defined, which consists of an active contour without the edge model and the edge information of SAR image by the Ratio operator with CFAR. Partial differential equations (PDE) of curve evolution are obtained by minimization of the energy functional. To implement image segmentation, the solution of the PDEs by a variational level set approach is applied. The performance of the method is verified by both synthetic and real SAR images. It is shown that the method can accurately extract targets from the SAR image but without any despeckle step, which possesses a preferable edge accuracy.

Pattern analysis and synthesis of the conformal antennas array

ZHANG Fan;ZHANG Fu-shun;ZHAO Gang;LIN Chen

J4. 2010, 37(3):  496-501.  doi:10.3969/j.issn.1001-2400.2010.03.020

Abstract ( 2232 )   PDF (1321KB) ( 1306 )   Save

References | Related Articles | Metrics

The radiation pattern of the conformal antennas array is analyzed and synthesized by considering the effects of mutual coupling. First, we choose the tapered slot antenna as the element of the conformal array and calculate the patterns of the element in the environments of the array, and then analyze the array patterns of co-polarization and cross-polarization in terms of the combination of the element patterns. A comparison between the result of full-wave calculation by Ansoft HFSS for the whole array and that by this method is given to show the effectiveness of this method. In the meantime, geese particle swarm optimization (GPSO) is applied for the optimization of the scan patterns of co-polarization and the performance of the cross-polarization in the plane of φ=0°.

Analysis of the electrical large radiation problem with the pre-corrected multilevel fast multipole algorithm

WANG Yuan-yuan;XIE Yong-jun;WANG Peng

J4. 2010, 37(3):  502-506.  doi:10.3969/j.issn.1001-2400.2010.03.021

Abstract ( 1384 )   PDF (875KB) ( 1006 )   Save

References | Related Articles | Metrics

For increasing the solution speed of the radiation problem using the multilevel fast multipole algorithm, a novel pre-corrected method is presented. It generates a small impedance matrix in the local region near the feed point and evaluates the local surface current, which could approximately describe the characteristic near the excited point. The difference between the unknown and the local current approaches the zero vector more than the unknown current itself. So it needs fewer iterations by using the difference as the unknown vector to get the solution. With consistent precision, the analysis of the numerical results shows that this method could decrease the number of iterations and reduce the time of solution by 40%~50%.

Access control scheme in a wireless sensor network

LIU Yun;PEI Qing-qi

J4. 2010, 37(3):  507-512+528.  doi:10.3969/j.issn.1001-2400.2010.03.022

Abstract ( 1579 )   PDF (779KB) ( 1114 )   Save

References | Related Articles | Metrics

Access control is one of the most challenging security problems in sensor networks. Sensor networks provide the query service of the environmental monitoring data as the service provider to the users. The current access control scheme based on public key cryptography is high in expense, and resists node capture, query replay and DoS attacks with difficulty. This paper proposes one-way hash chain based access control schemes. To increase the scalability of users and access capability, and to resist users being compromised, several effective access control schemes based on the Merkle hash tree and revoking user access control capability are proposed. Analysis, evaluation and comparison show that these schemes have several advantages over the current access control methods of low expense in calculation, storage and communication, and resistance to node capture, query replay and DoS attacks.

Modeling and calculating of interconnect delay by considering the influence of inductance effect and process fluctuations

YANG Yang;CHAI Chang-chun;DONG Gang;YANG Yin-tang;LENG Peng

J4. 2010, 37(3):  513-519.  doi:10.3969/j.issn.1001-2400.2010.03.023

Abstract ( 1731 )   PDF (1183KB) ( 1177 )   Save

References | Related Articles | Metrics

A statistical interconnect delay model considering the influence of inductance effect and process fluctuations is presented according to the improved equivalent Elmore delay, and the expressions for mean and standard deviation of interconnect delay are obtained. Calculated results indicate that the errors of the mean and standard deviation are less than 1.27％ and 5.23％ respectively in comparison with the results calculated by the widely used Monte Carlo method, and at the same time, the time elapsed for computing by using the proposed method is less than one ten-thousandth of that for 1000 computations by using the Monte Carlo method.

Electronic transport properties of a (4， 4) carbon nanotube/silicon carbide nanotube heterojunction

LIU Hong-xia;SONG Jiu-xu;ZHANG He-ming

J4. 2010, 37(3):  520-523.  doi:10.3969/j.issn.1001-2400.2010.03.024

Abstract ( 1640 )   PDF (888KB) ( 1277 )   Save

References | Related Articles | Metrics

The transport properties of the nanotube heterojunctions are the basis for the studies of their working mechanism, which is important for the theoretical study and the practical application. First, a two-probe model for a (4, 4) carbon nanotube (CNT)/silicon carbide nanotube (SiCNT) heterojunction is established. The transport properties of the heterojunction is realized with the nonequilibrium Green's function (NEGF). From the orbitals of the molecule projected self-consistent Hamiltonian, we can see that the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) concentrate mainly on the CNT section and that the band gap of the heterojunction is about 0.48 eV. The turn-on voltages of the heterojunction under the positive and negative voltage are +2.0V and -1.6V, which can be achieved from its voltage-current characteristic.

Moving dim point target detection based on the spatial and temporal combined filter

WANG Bo;ZHANG Jian-qi

J4. 2010, 37(3):  524-528.  doi:10.3969/j.issn.1001-2400.2010.03.025

Abstract ( 1285 )   PDF (831KB) ( 1124 )   Save

References | Related Articles | Metrics

The trajectories of dim point targets are always continuous in infrared image sequences. This can be used to improve the detection ability of the temporal based detection algorithms. In this paper, we present a spatial and temporal combined detection algorithm which utilizes both spatial and temporal information on moving dim point targets. Based on the analysis of the probable trajectories of moving dim point targets, a group of filter templates are constructed. And the trajectory of the dim moving target is obtained by using the constructed template to filter the temporal detection result. Also the target occurrence time in each pixel is extracted from the temporal based algorithm to further eliminate interference of the background. Since our new algorithm uses both spatial and temporal information on dim moving targets, it can have a better detection performance than the temporal based detection algorithm. Experiments with real infrared image sequences have proved the validity of the new approach.

New method for partial discharge signals detection

LI Ya-feng;WANG Bao-bao;FENG Xiang-chu

J4. 2010, 37(3):  529-533+569.  doi:10.3969/j.issn.1001-2400.2010.03.026

Abstract ( 1822 )   PDF (758KB) ( 1225 )   Save

References | Related Articles | Metrics

In order to extract partial discharge signals(PD), this paper proposes a new method based on anisotropic diffusion. According to the characteristics of partial discharge signals, a new diffusion function is constructed. Compared with the classical diffusion functions, using the proposed diffusion function can eliminate the noise more effectively, and maintain the same computing speed. Compared with the conventional wavelet based denoising algorithms, the proposed method is not limited by the selection of the wavelet function and the optimal decomposition level of wavelet. And our method is directly implemented in the spatial domain, which makes the filter process simpler.

Key dimension filtering based search algorithm of B+Tree for image feature matching

HE Zhou-can;WANG Qing;YANG Heng

J4. 2010, 37(3):  534-540.  doi:10.3969/j.issn.1001-2400.2010.03.027

Abstract ( 2343 )   PDF (2004KB) ( 1503 )   Save

References | Related Articles | Metrics

In dealing with the issues of low efficiency and low accuracy in multiple wide-based-line image matching,this paper adopts the classical SIFT descriptor, and proposes a novel high dimensional feature search algorithm. This paper follows the distance-based similarity standard, and firstly partitions the image feature set into different classes, then establishes a B+Tree for each class, and finally gives out a key dimension filtering strategy(KDF) in the KNN search step to speed up the high dimensional feature matching. Experimental results show that the proposed algorithm, which can obtain a higher accuracy with a lower time cost than the classical KNN search algorithm such as BBF, LSH and so on, would be a help to improve the capability of multiple wide-based-line image matching.

Novel detail preserving Retinex algorithm for image enhancement

MA Shi-ping;ZHANG Ming;BI Du-yan;XU Yue-lei

J4. 2010, 37(3):  541-546.  doi:10.3969/j.issn.1001-2400.2010.03.028

Abstract ( 1877 )   PDF (1271KB) ( 1144 )   Save

References | Related Articles | Metrics

In image enhancement, due to the blurring problem of the Kimmel variation model based on the Retinex theory, a modification for the Kimmel model is presented. The new model modifies the Kimmel model by increasing an expression for preserving the detail, and then calculates the optimal illumination image by combining pyramid data structure and lifting algorithm, which reduces the computational complexity. Finally, the normal clipping stretch is deployed to process the reflectance image. Experiments show that this algorithm can effectively remove obscuration effect from color images.

Key infection and evolution for sensor networks

LIU Wen-ju;LIU Zhi-hong;PEI Qing-qi;YANG Chao

J4. 2010, 37(3):  547-553.  doi:10.3969/j.issn.1001-2400.2010.03.029

Abstract ( 1591 )   PDF (599KB) ( 984 )   Save

References | Related Articles | Metrics

To tackle key management in sensor networks, we borrow concepts from natural biological systems to handle this problem, and a lightweight key management scheme which is applicable to smart dust sensor networks, key infection and evolutionm, is proposed. We first propose a probability model to analyze the security of the key infection protocol, and then introduce a new variant of the key infection protocol, Group based Key Infection, to improve its security performance. By combining key infection with a lightweight key pre-distribution scheme, a more secure key distribution scheme, Group based Key Infection with Key Pre-distribution, is proposed. To thwart on-going cryptanalytic attacks, a key evolution scheme with key synchronization is proposed to continuously renew shared keys. Analytical results show that the proposed scheme presents a viable trade-off between security and resource consumption for smart dust sensor networks.

Quantized-feedback scheme based on capacity upperbound maximizing in the MIMO BC system

XIE Fang;YUAN Chao-wei;CHENG Tie-zheng;ZHAO Wei

J4. 2010, 37(3):  554-558.  doi:10.3969/j.issn.1001-2400.2010.03.030

Abstract ( 1608 )   PDF (561KB) ( 1207 )   Save

References | Related Articles | Metrics

The quantization and limited feedback of channel state information (CSI) for the Multiple-input Multiple-output Broadcasting Channel(MIMO BC) are considered. Through analysis, a capacity upperbound for the receiver is derived and a novel quantized-feedback scheme is constructed based on this bound maximizing. The novel scheme utilizes the benefits of the random quantization codebook and considers not only the channel direction but also the gain of the sub-channel. Simulation shows that the proposed scheme improves the system capacity in the presence of quantization errors obviously and has less complexity than the conventional method, which only includes channel direction information.

Implementation of objected-oriented mechanism in framed temporal logic language MSVL

WANG Xiao-bing;DUAN Zhen-hua

J4. 2010, 37(3):  559-564+575.  doi:10.3969/j.issn.1001-2400.2010.03.031

Abstract ( 1701 )   PDF (742KB) ( 1137 )   Save

References | Related Articles | Metrics

In order to overcome the shortages which the available temporal logic programming languages have such as existence of the frame problem, lack of object-oriented mechanisms and high formalization, the framed temporal logic language MSVL is defined, which includes some techniques such as new frame operator, await statement, nondeterministic choice statement,and object-oriented programming. The implementation plan for the interpreter for MSVL is proposed based on the normal form and normal form graph. An example for concurrent access to shared resources shows that MSVL is much more similar to high-level programming languages than other temporal logic languages, and that the implementation plan for the interpreter is workable.

Separate BPNN-predistortion method for nonlinear HPA with memory

CUI Hua;ZHAO Xiang-mo;AI Bo

J4. 2010, 37(3):  565-569.  doi:10.3969/j.issn.1001-2400.2010.03.032

Abstract ( 1774 )   PDF (812KB) ( 1136 )   Save

References | Related Articles | Metrics

To circumvent the predistortion limitation for nonlinear high power amplifiers with memory (NHPAM) in OFDM communication systems, a separate BPNN (back propagation neural network) predistortion method with its structure and algorithm and their principles is proposed by considering the distortion characteristics and structure features of NHPAM. This method separates the whole predistortion for NHPAM into memory, AM/AM and AM/PM predistortion subsystems. Then the three NN predistorters are identified by the Levenberg-Marquardt BP algorithm based on the indirect training method. Simulation results show this proposed method can reduce ACEPR (Adjacent Channel Error Power Ratio) 30dB.

Low-complexity V-BLAST detection scheme with high performance

GUO Ming-xi;JIA Chong;SHEN Yue-hong;GAO Yuan-yuan

J4. 2010, 37(3):  570-575.  doi:10.3969/j.issn.1001-2400.2010.03.033

Abstract ( 2250 )   PDF (731KB) ( 1128 )   Save

References | Related Articles | Metrics

Based on sorted QR decomposition, a new list V-BLAST detection scheme is proposed. List detection is adopted for the first layer to produce several candidate symbols. A modified sorted QR decomposition is also proposed for the case of the number of candidates equal to the constellation size. Several groups of candidate symbols can be obtained when successive interference cancellation is applied to the following layers, and the minimum Euclidean distance (MED) search is used to choose one as the output. Analysis and simulation show that the proposed scheme can achieve very good performance with low complexity. When the number of antennas is less than 5, it can achieve near ML detection performance. A threshold can be set in the process of searching MED to further reduce the complexity, and the method for determining the threshold value is also given. Simulation results show that the method can reduce more than one half of the complexity while suffering a little performance loss.

Orthogonal projection-based spatial spectrum algorithm for finding the direction of the wideband coherent source

XIE Nuo;ZHANG Run-sheng;GE Jian-hua;ZHANG Jian-li

J4. 2010, 37(3):  576-580.  doi:10.3969/j.issn.1001-2400.2010.03.034

Abstract ( 1346 )   PDF (580KB) ( 1237 )   Save

References | Related Articles | Metrics

A novel algorithm using the orthogonal projection is proposed, which avoids initially estimating the DOA and the number of sources through angle scanning and covariance matrix orthogonal projection, thereby improving the performance of direction finding by the spatial spectrum. This paper discusses the principle and realization of this method, and we compare other algorithms with the one developed here whose performance is shown to be better under the condition of a low SNR. Computer simulation illustrates that the proposed method has a better angle resolution ability and a lower RMSE.