Email alert
2010 Vol. 32, No. 3
Display Method:
2010, 32(3): 509-514.
doi: 10.3724/SP.J.1146.2009.01066
Abstract:
Affinity Propagation (AP) clustering is not fit to deal with multi-scale data cluster as well as the arbitrary shape cluster issue. Therefore, an improved affinity propagation clustering algorithm AP-VSM (Affinity Propagation based on Variable-Similarity Measure) is proposed embarking from the token of data distribution characters. First, a kind of variable-similarity measure method is devised according of characters of global and local data distribution, which has the ability of describing the characters of data clustering effectively. Then AP-VSM clustering algorithm is proposed base on the frame of traditional AP algorithm, and this method has extended data processing capacity compared with traditional AP. The simulation results show that the new method is outperforming traditional AP algorithm.
Affinity Propagation (AP) clustering is not fit to deal with multi-scale data cluster as well as the arbitrary shape cluster issue. Therefore, an improved affinity propagation clustering algorithm AP-VSM (Affinity Propagation based on Variable-Similarity Measure) is proposed embarking from the token of data distribution characters. First, a kind of variable-similarity measure method is devised according of characters of global and local data distribution, which has the ability of describing the characters of data clustering effectively. Then AP-VSM clustering algorithm is proposed base on the frame of traditional AP algorithm, and this method has extended data processing capacity compared with traditional AP. The simulation results show that the new method is outperforming traditional AP algorithm.
2010, 32(3): 515-521.
doi: 10.3724/SP.J.1146.2009.00309
Abstract:
As a new computational intelligence method, the Artificial Immune Network (AIN) is widely applied to pattern recognition and data classification. Existing artificial immune network algorithms for classifier have two major limitations: one is the scale of the networks, a large scale of networks needs high computation complexity, the other is only once presenting the antigens that can not guarantee find the optimal global classifier. A new Artificial Immune Network Classifier (AINC) algorithm is proposed in this paper. In the proposed algorithm, only one B-cell is used to denote single class in order to reduce the scale of network, and avoid the suppression operation between B-cells, moreover, a new affinity based on the correct rate is proposed to realize the evaluation strategy based on antigen priority. The proposed algorithm is extensively compared with Fuzzy C-Means (FCM), Multiple-Valued Immune Network algorithm (MVIN), and Clonal Selection Algorithm for classifier (CSA) over a test suit of several real life data sets and one SAR image. The result of experiment indicates the superiority of the AINC over FCM, MVIN and CSA on accuracy and robustness.
As a new computational intelligence method, the Artificial Immune Network (AIN) is widely applied to pattern recognition and data classification. Existing artificial immune network algorithms for classifier have two major limitations: one is the scale of the networks, a large scale of networks needs high computation complexity, the other is only once presenting the antigens that can not guarantee find the optimal global classifier. A new Artificial Immune Network Classifier (AINC) algorithm is proposed in this paper. In the proposed algorithm, only one B-cell is used to denote single class in order to reduce the scale of network, and avoid the suppression operation between B-cells, moreover, a new affinity based on the correct rate is proposed to realize the evaluation strategy based on antigen priority. The proposed algorithm is extensively compared with Fuzzy C-Means (FCM), Multiple-Valued Immune Network algorithm (MVIN), and Clonal Selection Algorithm for classifier (CSA) over a test suit of several real life data sets and one SAR image. The result of experiment indicates the superiority of the AINC over FCM, MVIN and CSA on accuracy and robustness.
2010, 32(3): 522-527.
doi: 10.3724/SP.J.1146.2009.00249
Abstract:
Aiming at information fusion problem under complex environment, an intellectualized information fusion model, NFE model, which is an organic combination of Neural network, Fuzzy reasoning and Expert system is proposed. The necessity of constructing a complete three-dimensional intelligent information fusion model is analyzed from an engineering perspective. Considering various factors including sensor performance, weather situation, electronic interfere and clutter intensity that effect the credibility of sensors, analysis for weights of sensors is provided and credibility estimator is designed. Simulation experiments prove that the NFE model could realize target recognition effectively under the circumstances of low confidence and sensor fault etc.
Aiming at information fusion problem under complex environment, an intellectualized information fusion model, NFE model, which is an organic combination of Neural network, Fuzzy reasoning and Expert system is proposed. The necessity of constructing a complete three-dimensional intelligent information fusion model is analyzed from an engineering perspective. Considering various factors including sensor performance, weather situation, electronic interfere and clutter intensity that effect the credibility of sensors, analysis for weights of sensors is provided and credibility estimator is designed. Simulation experiments prove that the NFE model could realize target recognition effectively under the circumstances of low confidence and sensor fault etc.
2010, 32(3): 528-532.
doi: 10.3724/SP.J.1146.2008.00275
Abstract:
In this paper, a modified Gaussian Mixed Model (GMM) with an embedded Auto-Associate Neural Network (AANN) is proposed. It integrates the merits of GMM and AANN. GMM and AANN as a whole are trained by means of Maximum Likelihood (ML). In the process of training, the parameters of GMM and AANN are updated alternately. AANN reshapes the distribution of the data and improves the similarity of the data in one class. Experiments show that the proposed system improves accuracy rate against baseline GMM at all SNR, maximum to 19%.
In this paper, a modified Gaussian Mixed Model (GMM) with an embedded Auto-Associate Neural Network (AANN) is proposed. It integrates the merits of GMM and AANN. GMM and AANN as a whole are trained by means of Maximum Likelihood (ML). In the process of training, the parameters of GMM and AANN are updated alternately. AANN reshapes the distribution of the data and improves the similarity of the data in one class. Experiments show that the proposed system improves accuracy rate against baseline GMM at all SNR, maximum to 19%.
2010, 32(3): 533-538.
doi: 10.3724/SP.J.1146.2009.00352
Abstract:
Due to the defects of high complexity and instability convergence performance in the array signal adaptive canceller under the correlated interference, an enhanced array signal multistage cascaded canceller is proposed by substituting the weights of the multistage Winner filter for the minimal module weights, which have the minimal module of the samples quotient between channels. The enhanced algorithm has the ability of fast convergence, less operation and performs well with non-stationary samples distorted by the correlated interference. Simulation results indicate the algorithm can reach convergent performance of the kind of Sample Matrix Inversion (SMI) algorithm using less samples.
Due to the defects of high complexity and instability convergence performance in the array signal adaptive canceller under the correlated interference, an enhanced array signal multistage cascaded canceller is proposed by substituting the weights of the multistage Winner filter for the minimal module weights, which have the minimal module of the samples quotient between channels. The enhanced algorithm has the ability of fast convergence, less operation and performs well with non-stationary samples distorted by the correlated interference. Simulation results indicate the algorithm can reach convergent performance of the kind of Sample Matrix Inversion (SMI) algorithm using less samples.
2010, 32(3): 539-543.
doi: 10.3724/SP.J.1146.2009.00235
Abstract:
A novel simplified Maximum Likelihood (ML) detection algorithm based on noise analysis is proposed for Vertical-Bell Labs Layered Space-Time (V-BLAST) system. This algorithm chooses the signal points near the reference line in a multi-dimension space to form the signal subset，finally selects the proper signal vector in the subset as final estimation of the transmitted signals based on ML algorithm. Theoretical analysis and simulation shows that the proposed algorithm not only guarantees the near optimum error rate performance, but also has lower computational complexity, and solves the noise enhancement problem.
A novel simplified Maximum Likelihood (ML) detection algorithm based on noise analysis is proposed for Vertical-Bell Labs Layered Space-Time (V-BLAST) system. This algorithm chooses the signal points near the reference line in a multi-dimension space to form the signal subset，finally selects the proper signal vector in the subset as final estimation of the transmitted signals based on ML algorithm. Theoretical analysis and simulation shows that the proposed algorithm not only guarantees the near optimum error rate performance, but also has lower computational complexity, and solves the noise enhancement problem.
2010, 32(3): 544-547.
doi: 10.3724/SP.J.1146.2008.01006
Abstract:
This paper firstly studies parameter estimation issue directly based on FFT. Thereby, the relationship between the frequency and phase estimation is exposed. Subsequently, the estimation error formula of phase difference arithmetic are deduced and validated by computer simulation. The compare of two methods shows that phase difference arithmetic possesses lesser calculation quantity. Simultaneously, it can gain highly accurate, mutually independent parameter estimation under low SNR. So phase difference arithmetic is easy to realize in engineering field much more.
This paper firstly studies parameter estimation issue directly based on FFT. Thereby, the relationship between the frequency and phase estimation is exposed. Subsequently, the estimation error formula of phase difference arithmetic are deduced and validated by computer simulation. The compare of two methods shows that phase difference arithmetic possesses lesser calculation quantity. Simultaneously, it can gain highly accurate, mutually independent parameter estimation under low SNR. So phase difference arithmetic is easy to realize in engineering field much more.
2010, 32(3): 548-553.
doi: 10.3724/SP.J.1146.2009.00211
Abstract:
The longest axis in coherence region, which gives the largest distance between any two coherences, is significant in forestry applications using coherence scattering model, because it reflects the linear variation of complex coherences with the polarization states. To improve the efficiency and accuracy of traversal search, a fast approach is proposed to estimate longest axis in this paper. It uses the line segment of the tangent points of the rectangle externally tangent to coherence region to get the initial estimation for the longest axis. Nest, it obtains the longest axis using the approximation technique. To improve forest height estimation, a forest parameter inversion technique using the longest axis is presented. Results of simulated data for pine forest show the fast approach consumes much less computation time and gives better estimations for the longest axis than traversal search method. Results also show that the introduction of longest axis in forest parameter inversion yields closer estimations to the true forest height.
The longest axis in coherence region, which gives the largest distance between any two coherences, is significant in forestry applications using coherence scattering model, because it reflects the linear variation of complex coherences with the polarization states. To improve the efficiency and accuracy of traversal search, a fast approach is proposed to estimate longest axis in this paper. It uses the line segment of the tangent points of the rectangle externally tangent to coherence region to get the initial estimation for the longest axis. Nest, it obtains the longest axis using the approximation technique. To improve forest height estimation, a forest parameter inversion technique using the longest axis is presented. Results of simulated data for pine forest show the fast approach consumes much less computation time and gives better estimations for the longest axis than traversal search method. Results also show that the introduction of longest axis in forest parameter inversion yields closer estimations to the true forest height.
2010, 32(3): 554-558.
doi: 10.3724/SP.J.1146.2009.00444
Abstract:
The issue of micro-Doppler signal separation from collected data of targets with micro-motion structures is addressed. The signal mode for target with micro-motion structures is analyzed and the results show that the Doppler frequency of main body is constant and the micro-Doppler frequency is time varying, which can be fitted by a series of piecewise linear frequency modulation signal. A method, which decomposes radar signal to AM-LFM (amplitude modulation-linear frequency modulation) components and determines each components corresponding structure by frequency modulation rate, is proposed to extract micro-motion echoes from the main body echoes. Simulated and experimental results verify the correctness of proposed method.
2010, 32(3): 559-563.
doi: 10.3724/SP.J.1146.2009.00251
Abstract:
The mathematical expression of Continuous-Chaos Frequency-Modulating Signal (CCFMS) is given. The auto-correlation function and root-mean-square bandwidth of CCFMS are derived in theory. The results show that autocorrelation function of CCFMS has excellent side lobe suppression performance. Its RMS bandwidth is proportional to frequency modulation index and RMS of chaotic signal. The performance of CCFMS is further confirmed with typical chaos. It is found that CCFMS has high time peak to side lobe ratio, flat power spectrum, almost thumbtack ambiguity function and peak to average power ratio equal to that of sinusoidal signal. The performance of CCFMS radar against noise is analyzed. Compared with LFMS radar, CCFMS radar has obvious anti-jamming advantages with signal-to-interference ratio improvement factor increased by about 5.5 dB.
The mathematical expression of Continuous-Chaos Frequency-Modulating Signal (CCFMS) is given. The auto-correlation function and root-mean-square bandwidth of CCFMS are derived in theory. The results show that autocorrelation function of CCFMS has excellent side lobe suppression performance. Its RMS bandwidth is proportional to frequency modulation index and RMS of chaotic signal. The performance of CCFMS is further confirmed with typical chaos. It is found that CCFMS has high time peak to side lobe ratio, flat power spectrum, almost thumbtack ambiguity function and peak to average power ratio equal to that of sinusoidal signal. The performance of CCFMS radar against noise is analyzed. Compared with LFMS radar, CCFMS radar has obvious anti-jamming advantages with signal-to-interference ratio improvement factor increased by about 5.5 dB.
2010, 32(3): 564-569.
doi: 10.3724/SP.J.1146.2009.00257
Abstract:
A new method for detection of multi-target in wideband signal is proposed. Firstly the adjacent correlation is employed to decrease the order of the echoes. Then the Ambiguity function is applied to the auto-term of the result of the adjacent correlation. Further a Radon transform is performed for target energy accumulation, and then the target detection can be carried out. The results of simulated data show the effectiveness of the proposed method.
A new method for detection of multi-target in wideband signal is proposed. Firstly the adjacent correlation is employed to decrease the order of the echoes. Then the Ambiguity function is applied to the auto-term of the result of the adjacent correlation. Further a Radon transform is performed for target energy accumulation, and then the target detection can be carried out. The results of simulated data show the effectiveness of the proposed method.
2010, 32(3): 570-574.
doi: 10.3724/SP.J.1146.2008.00797
Abstract:
The echo waves of radar received consist of direct wave and multipath wave which is reflected from earth (sea) surface and coherent with direct wave. Two new height finding algorithms using improved Toeplitz technique at low-angle environment are presented. The first method uses not only every row elements but also corresponding every column elements of array covariance matrix to construct a Toeplitz matrix, a new covariance matrix is obtained taking average these two matrixes, this process is equivalent to spatial smoothing and improves the accuracy of DOA estimation of elevation. The second method based on the first method, reconstructing a Toeplitz matrix to approach the subspace of above covariance matrix constraint to maximum SNR (Signal Noise Ratio), it can reduce the influence of the noise to a certain extent and improve the robust of algorithm. Theoretical analysis and simulation results demonstrate the merits of the new algorithm.
The echo waves of radar received consist of direct wave and multipath wave which is reflected from earth (sea) surface and coherent with direct wave. Two new height finding algorithms using improved Toeplitz technique at low-angle environment are presented. The first method uses not only every row elements but also corresponding every column elements of array covariance matrix to construct a Toeplitz matrix, a new covariance matrix is obtained taking average these two matrixes, this process is equivalent to spatial smoothing and improves the accuracy of DOA estimation of elevation. The second method based on the first method, reconstructing a Toeplitz matrix to approach the subspace of above covariance matrix constraint to maximum SNR (Signal Noise Ratio), it can reduce the influence of the noise to a certain extent and improve the robust of algorithm. Theoretical analysis and simulation results demonstrate the merits of the new algorithm.
2010, 32(3): 575-581.
doi: 10.3724/SP.J.1146.2008.00641
Abstract:
An ultra-wideband orthogonal Hermite pulse is an appropriate family of pulses that can represent the commonly-used waveforms in through-the-wall surveillance radar. It has advantages of flexible design，high spectrum utilization and simple design. An analytical expression of the ambiguity function for a train of Time-hopping Modulated Ultra-WideBand (TM-UWB) Hermite waveforms is derived. The relation of resolution,side-lobe suppression characteristic and unambiguous for TM-UWB impulse waveforms with the autocorrelation properties, period of time-hopping sequences, and the average pulse repetition period, order number and scaling parameter of the Hermite signal are investigated, and some application issues for through-the-wall surveillance radar,such as sidelobe suppression and clutter suppression, are also discussed.
An ultra-wideband orthogonal Hermite pulse is an appropriate family of pulses that can represent the commonly-used waveforms in through-the-wall surveillance radar. It has advantages of flexible design，high spectrum utilization and simple design. An analytical expression of the ambiguity function for a train of Time-hopping Modulated Ultra-WideBand (TM-UWB) Hermite waveforms is derived. The relation of resolution,side-lobe suppression characteristic and unambiguous for TM-UWB impulse waveforms with the autocorrelation properties, period of time-hopping sequences, and the average pulse repetition period, order number and scaling parameter of the Hermite signal are investigated, and some application issues for through-the-wall surveillance radar,such as sidelobe suppression and clutter suppression, are also discussed.
2010, 32(3): 582-586.
doi: 10.3724/SP.J.1146.2009.00259
Abstract:
When reconstructing elevation-velocity image, the observation data obtained from differential SAR tomography in baseline-time plane does not follow uniform distribution. If the elevation-velocity image of multiple scatterers is obtained using two-dimensional FFT method, the imaging result is not very good because of high sidelobes. In this paper, a new differential SAR tomography imaging algorithm is proposed based on Backus- Gilbert technique. In this algorithm, the elevation-velocity imaging is converted into an inverse problem of two- dimensional integral function, and Tikhonov regularization is used to get the regularized solution of the inverse problem. Simulation results show that the proposed algorithm can overcome the influence caused by non-uniform samples, and acquire better imaging result.
When reconstructing elevation-velocity image, the observation data obtained from differential SAR tomography in baseline-time plane does not follow uniform distribution. If the elevation-velocity image of multiple scatterers is obtained using two-dimensional FFT method, the imaging result is not very good because of high sidelobes. In this paper, a new differential SAR tomography imaging algorithm is proposed based on Backus- Gilbert technique. In this algorithm, the elevation-velocity imaging is converted into an inverse problem of two- dimensional integral function, and Tikhonov regularization is used to get the regularized solution of the inverse problem. Simulation results show that the proposed algorithm can overcome the influence caused by non-uniform samples, and acquire better imaging result.
2010, 32(3): 587-592.
doi: 10.3724/SP.J.1146.2009.00283
Abstract:
The large range migration produced by missile diving acceleration flight make SAR image difficult. This paper establishes the echo model of the missile-borne SAR by using the high order range model based on the characters of the missile movements. Considering the large scene, the change of the slant range is analyzed in details. Then, the two-dimensional point target spectrum is derived by the method of series reversion. A large scene imaging algorithm used for diving acceleration flight is presented. Finally, simulation results are presented to demonstrate the accuracy and validity of the proposed algorithms. The resolution of range and azimuth are identical with the theoretical values.
The large range migration produced by missile diving acceleration flight make SAR image difficult. This paper establishes the echo model of the missile-borne SAR by using the high order range model based on the characters of the missile movements. Considering the large scene, the change of the slant range is analyzed in details. Then, the two-dimensional point target spectrum is derived by the method of series reversion. A large scene imaging algorithm used for diving acceleration flight is presented. Finally, simulation results are presented to demonstrate the accuracy and validity of the proposed algorithms. The resolution of range and azimuth are identical with the theoretical values.
2010, 32(3): 593-598.
doi: 10.3724/SP.J.1146.2009.00472
Abstract:
SAR echo signal in high squint mode features large range walk and small range curvature. An improved CS imaging algorithm based on range-walk removal is proposed. First, range-walk is removed in time domain. Second, range curvature is corrected in frequency domain. Finally, geometry correction is used to correct geometric distortion. The coupling between azimuth and range is greatly reduced after range-walk removal. Therefore, it satisfies the imaging quality of SAR in high squint mode and the width of the scene swath is broadened as well. Simulation results illustrate that it satisfies the imaging quality of SAR in high squint mode and large scene swath.
SAR echo signal in high squint mode features large range walk and small range curvature. An improved CS imaging algorithm based on range-walk removal is proposed. First, range-walk is removed in time domain. Second, range curvature is corrected in frequency domain. Finally, geometry correction is used to correct geometric distortion. The coupling between azimuth and range is greatly reduced after range-walk removal. Therefore, it satisfies the imaging quality of SAR in high squint mode and the width of the scene swath is broadened as well. Simulation results illustrate that it satisfies the imaging quality of SAR in high squint mode and large scene swath.
2010, 32(3): 599-603.
doi: 10.3724/SP.J.1146.2009.00197
Abstract:
Spotlight SAR is an efficiency way to obtain high resolution radar image. The non-stability of the aircraft motion during the long synthetic aperture time can make the image suffer the severe defocus, and there is phase error in the range direction in the high resolution SAR system. This paper presents a novel signal processing method based on echo signal to get high definition resolution spotlight SAR image, which can make an effective estimation on the aircraft position error, the phase error in range direction and the residual envelop error. The experiment results show that the proposed approach can get the high quality SAR image from raw data.
Spotlight SAR is an efficiency way to obtain high resolution radar image. The non-stability of the aircraft motion during the long synthetic aperture time can make the image suffer the severe defocus, and there is phase error in the range direction in the high resolution SAR system. This paper presents a novel signal processing method based on echo signal to get high definition resolution spotlight SAR image, which can make an effective estimation on the aircraft position error, the phase error in range direction and the residual envelop error. The experiment results show that the proposed approach can get the high quality SAR image from raw data.
2010, 32(3): 604-608.
doi: 10.3724/SP.J.1146.2009.00321
Abstract:
A new decorrelation algorithm based on one single snapshot is proposed for Direction-Of-Arrival (DOA) estimation of coherent signals on a ULA, which reconstructs a Toeplitz matrix using the observed data vector directly. It is proved that the rank of this Toeplitz matrix is determined merely by the number of incident sources and has no relations with the coherency of the signals. Then accurate signal subspace and noise subspace can be acquired by performing eigenvalue decomposition of the matrix. Combined with the subspace kind algorithms such as MUSIC and ESPRIT, the DOA of coherent signals can be estimated. Simulation results verify that the proposed algorithm is effective.
A new decorrelation algorithm based on one single snapshot is proposed for Direction-Of-Arrival (DOA) estimation of coherent signals on a ULA, which reconstructs a Toeplitz matrix using the observed data vector directly. It is proved that the rank of this Toeplitz matrix is determined merely by the number of incident sources and has no relations with the coherency of the signals. Then accurate signal subspace and noise subspace can be acquired by performing eigenvalue decomposition of the matrix. Combined with the subspace kind algorithms such as MUSIC and ESPRIT, the DOA of coherent signals can be estimated. Simulation results verify that the proposed algorithm is effective.
2010, 32(3): 609-612.
doi: 10.3724/SP.J.1146.2009.00208
Abstract:
This paper presents a method to construct quasi-cyclic LDPC codes from prime fields. This method is a generalization of a method proposed by Lan et al to construct quasi-cyclic LDPC codes based on finite fields, and gives a much larger class of quasi-cyclic LDPC codes based on prime fields. Simulations show that quasi-cyclic LDPC codes constructed by the proposed method perform well over AWGN channels with iterative decoding.
This paper presents a method to construct quasi-cyclic LDPC codes from prime fields. This method is a generalization of a method proposed by Lan et al to construct quasi-cyclic LDPC codes based on finite fields, and gives a much larger class of quasi-cyclic LDPC codes based on prime fields. Simulations show that quasi-cyclic LDPC codes constructed by the proposed method perform well over AWGN channels with iterative decoding.
2010, 32(3): 613-616.
doi: 10.3724/SP.J.1146.2009.00160
Abstract:
This paper first presents Belief Propagation (BP) iteration algorithm in LDPC code-based public-key cryptosystems, and develops the necessary condition of private key if the probability of plaintext is equal. Then the necessary and sufficient condition of public key is deduced according to the recursion of BP iteration algorithm. Simulations show that the parameters of private key and public key are correct.
This paper first presents Belief Propagation (BP) iteration algorithm in LDPC code-based public-key cryptosystems, and develops the necessary condition of private key if the probability of plaintext is equal. Then the necessary and sufficient condition of public key is deduced according to the recursion of BP iteration algorithm. Simulations show that the parameters of private key and public key are correct.
2010, 32(3): 617-622.
doi: 10.3724/SP.J.1146.2009.00335
Abstract:
In multiuser MIMO communication networks, a new amplify-and-forward two-way relaying scheme is proposed employing OFDMA for the multiple access transmission in the first time slot and OFDM/SDMA for the broadcast transmission in the second time slot to improve the performance of the system through utilizing frequency diversity and spatial diversity. Considering the characteristic of two-way relaying, the relay beamforming matrixes are designed according to two methods respectively, i.e., signal to leakage and noise ratio (SLNR) and block diagonalization based zero-forcing (BDZF), on per subcarrier basis. The upper bound on the capacity region of this two-way relay network is also derived by using cut-set theory. Simulation results show that the proposed scheme outperforms three other two-way relaying schemes in terms of sum rate and can approach the upper bound of capacity region.
In multiuser MIMO communication networks, a new amplify-and-forward two-way relaying scheme is proposed employing OFDMA for the multiple access transmission in the first time slot and OFDM/SDMA for the broadcast transmission in the second time slot to improve the performance of the system through utilizing frequency diversity and spatial diversity. Considering the characteristic of two-way relaying, the relay beamforming matrixes are designed according to two methods respectively, i.e., signal to leakage and noise ratio (SLNR) and block diagonalization based zero-forcing (BDZF), on per subcarrier basis. The upper bound on the capacity region of this two-way relay network is also derived by using cut-set theory. Simulation results show that the proposed scheme outperforms three other two-way relaying schemes in terms of sum rate and can approach the upper bound of capacity region.
2010, 32(3): 623-628.
doi: 10.3724/SP.J.1146.2009.00296
Abstract:
More than an adaptive system, the cognitive radio system is an intelligent system. The Q-Learning of the intelligent control theory is adopted in the paper, to solve the sensing task allocation problem among cognitive users. And a Q-Learning based sensing management algorithm is proposed. The algorithm allocates sensing tasks to users through times of interaction with the environment and self-learning. The scheme of the paper works without any channel state information and estimation of primary traffic. From the simulation result, the algorithm could improve the sensing efficiency compared to the static allocation algorithm and attain to the convergence in a short time, which could be an attempt to the future intelligent cognitive radio systems.
More than an adaptive system, the cognitive radio system is an intelligent system. The Q-Learning of the intelligent control theory is adopted in the paper, to solve the sensing task allocation problem among cognitive users. And a Q-Learning based sensing management algorithm is proposed. The algorithm allocates sensing tasks to users through times of interaction with the environment and self-learning. The scheme of the paper works without any channel state information and estimation of primary traffic. From the simulation result, the algorithm could improve the sensing efficiency compared to the static allocation algorithm and attain to the convergence in a short time, which could be an attempt to the future intelligent cognitive radio systems.
2010, 32(3): 629-632.
doi: 10.3724/SP.J.1146.2009.00612
Abstract:
In this paper, one sensing-slots scheduling scheme is proposed by using cross-layer design based on IEEE802.22 draft standard. In order to improve the whole system performance, the parameters of TSS scheme in current IEEE802.22 draft are further studied and modified by using cross-layer design. A new sensing-slots scheduling scheme on the MAC layer is proposed from two aspects. First, on the basis of the second-order statistic properties of wireless fading channels on the physical layer, the scheduling frequency of fast sensing timeslot is defined. Then, in order to maximize the transmission throughput for sencondary users, the duration of fine sensing stage is optimized. It is proven that the scheme proposed in this paper makes the TSS scheme more explicit, effective and be able to adapt to varying transmission environments.
In this paper, one sensing-slots scheduling scheme is proposed by using cross-layer design based on IEEE802.22 draft standard. In order to improve the whole system performance, the parameters of TSS scheme in current IEEE802.22 draft are further studied and modified by using cross-layer design. A new sensing-slots scheduling scheme on the MAC layer is proposed from two aspects. First, on the basis of the second-order statistic properties of wireless fading channels on the physical layer, the scheduling frequency of fast sensing timeslot is defined. Then, in order to maximize the transmission throughput for sencondary users, the duration of fine sensing stage is optimized. It is proven that the scheme proposed in this paper makes the TSS scheme more explicit, effective and be able to adapt to varying transmission environments.
2010, 32(3): 633-637.
doi: 10.3724/SP.J.1146.2009.00163
Abstract:
This paper addresses the issue of joint Maximum-Likelihood (ML) time-frequency synchronization and channel estimation for Multiple-Input Multiple-Output (MIMO) systems. The resulting joint ML estimation requires solving a maximization problem with no closed-form solution. Since numerical calculation of the estimation is computationally hard, a computationally efficient closed-form ML solution is proposed using two repetitions of orthogonal training sequences. With theoretical analysis and simulations, the mean-square errors of the ML estimates versus the average SNR and the number of antennas are investigated, and then the performance of the proposed estimator is compared with the Cramer-Rao Bound (CRB). The results prove the effectiveness of proposed estimator.
This paper addresses the issue of joint Maximum-Likelihood (ML) time-frequency synchronization and channel estimation for Multiple-Input Multiple-Output (MIMO) systems. The resulting joint ML estimation requires solving a maximization problem with no closed-form solution. Since numerical calculation of the estimation is computationally hard, a computationally efficient closed-form ML solution is proposed using two repetitions of orthogonal training sequences. With theoretical analysis and simulations, the mean-square errors of the ML estimates versus the average SNR and the number of antennas are investigated, and then the performance of the proposed estimator is compared with the Cramer-Rao Bound (CRB). The results prove the effectiveness of proposed estimator.
2010, 32(3): 638-642.
doi: 10.3724/SP.J.1146.2009.00328
Abstract:
In this paper, an improved water-filling algorithm is proposed for the problem of Discrete Bit Allocation DBA in OFDM systems. The bit-water-level, defined in this paper as the water-filling level under which the power allocated to a certain sub-carrier satisfies the integer bit constraint, is used in the proposed algorithm. First, using the bit-water-levels of the sub-carrier with the maximal channel gain, bits and power are allocated to all the sub-carriers. Then, the allocation results are adapted to satisfy the total transmit power constraint. It is proved that the algorithm yields the optimal solution and its computational complexity depends only on the number of sub-carriers.
In this paper, an improved water-filling algorithm is proposed for the problem of Discrete Bit Allocation DBA in OFDM systems. The bit-water-level, defined in this paper as the water-filling level under which the power allocated to a certain sub-carrier satisfies the integer bit constraint, is used in the proposed algorithm. First, using the bit-water-levels of the sub-carrier with the maximal channel gain, bits and power are allocated to all the sub-carriers. Then, the allocation results are adapted to satisfy the total transmit power constraint. It is proved that the algorithm yields the optimal solution and its computational complexity depends only on the number of sub-carriers.
2010, 32(3): 643-648.
doi: 10.3724/SP.J.1146.2009.00332
Abstract:
Clustering is the main framework of large scale Ad Hoc networks, and one of its key technologies is spectrum allocation. In this paper, a novel three-level market-based scheme of spectrum allocation is proposed for cognitive clustered Ad Hoc networks. In this scheme, cluster headers buy spectrum from spectrum administrator of primary service according to the estimation of their demands, and a market-based algorithm which stems from microeconomic theory is adopted in the spectrum allocation within each cluster. An analysis is given on the iterative pricing algorithms of Excess Demand-Based (EDB) and Successive Over-Relaxation (SOR) for in-cluster market as well as on the process of spectrum purchase based on demand executed by cluster headers. The simulation results show that cluster profit is maximized by adopting the market-based algorithm, and that a significantly utility improvement is achieved by demand-based purchase of cluster headers compared with equal purchase. As a result, the spectrum allocation scheme proposed in this paper can improve the system performance efficiently. The convergence of EDB and SOR iterative algorithms is also proved by simulation.
Clustering is the main framework of large scale Ad Hoc networks, and one of its key technologies is spectrum allocation. In this paper, a novel three-level market-based scheme of spectrum allocation is proposed for cognitive clustered Ad Hoc networks. In this scheme, cluster headers buy spectrum from spectrum administrator of primary service according to the estimation of their demands, and a market-based algorithm which stems from microeconomic theory is adopted in the spectrum allocation within each cluster. An analysis is given on the iterative pricing algorithms of Excess Demand-Based (EDB) and Successive Over-Relaxation (SOR) for in-cluster market as well as on the process of spectrum purchase based on demand executed by cluster headers. The simulation results show that cluster profit is maximized by adopting the market-based algorithm, and that a significantly utility improvement is achieved by demand-based purchase of cluster headers compared with equal purchase. As a result, the spectrum allocation scheme proposed in this paper can improve the system performance efficiently. The convergence of EDB and SOR iterative algorithms is also proved by simulation.
2010, 32(3): 649-654.
doi: 10.3724/SP.J.1146.2009.00329
Abstract:
As the huge computational loads of panoramic video processing, most of the embedded panoramic video processing system are based on multi-cores. The problem of high-speed data communication between multi-cores must be solved. A high-speed data communication method between DSP and FPGA is proposed. In order to realize high-speed data communication, the address-bus is used to transport commands, the wave of data communication between dual-cores in DMA mode is analyzed and simulated. The experiments show that the data transmission speed is up to 588 MBps between DSP and FPGA by using those methods.
As the huge computational loads of panoramic video processing, most of the embedded panoramic video processing system are based on multi-cores. The problem of high-speed data communication between multi-cores must be solved. A high-speed data communication method between DSP and FPGA is proposed. In order to realize high-speed data communication, the address-bus is used to transport commands, the wave of data communication between dual-cores in DMA mode is analyzed and simulated. The experiments show that the data transmission speed is up to 588 MBps between DSP and FPGA by using those methods.
2010, 32(3): 655-659.
doi: 10.3724/SP.J.1146.2009.00253
Abstract:
Set out from the evolution to all-optical transparent infrastructure and to service-orient characteristic, the researches on impairment-aware routing and wavelength assignment algorithms are surveyed and analyzed. Three new dynamic impairment-aware routing and wavelength assignment algorithms (IBest, IFF, IPack)are proposed through increasing the consideration of physical impairments effect on classic routing and wavelength assignment algorithms. The results from simulation show that the proposed algorithms both can maintain its competitive edge from corresponding classic ones and are superior to the classic ones in blocking probability in realistic circumstance with impairments.
Set out from the evolution to all-optical transparent infrastructure and to service-orient characteristic, the researches on impairment-aware routing and wavelength assignment algorithms are surveyed and analyzed. Three new dynamic impairment-aware routing and wavelength assignment algorithms (IBest, IFF, IPack)are proposed through increasing the consideration of physical impairments effect on classic routing and wavelength assignment algorithms. The results from simulation show that the proposed algorithms both can maintain its competitive edge from corresponding classic ones and are superior to the classic ones in blocking probability in realistic circumstance with impairments.
2010, 32(3): 660-665.
doi: 10.3724/SP.J.1146.2009.00181
Abstract:
While current research on service context mainly focuses on service quality and seldom consider personal terminal environment, a terminal environment related context named Terminal Service Context is proposed, which could provide essential information support for ubiquitous services. A general terminal service context definition method is provided, which could draw feature parameters of terminal environment for all kinds of service scenarios, and form formalized and retrievable contexts. Furthermore, service model based on the terminal service context is given and the storage and application methods of the context are analyzed. Finally, the proposed definition method and service model are verified by a typical ubiquitous service scenario. The results show that the method can extract the context required by ubiquitous service. Based on the model, service platform can adapt to heterogeneous terminal environments dynamically, and provide individualized presentation and handoff services for users.
While current research on service context mainly focuses on service quality and seldom consider personal terminal environment, a terminal environment related context named Terminal Service Context is proposed, which could provide essential information support for ubiquitous services. A general terminal service context definition method is provided, which could draw feature parameters of terminal environment for all kinds of service scenarios, and form formalized and retrievable contexts. Furthermore, service model based on the terminal service context is given and the storage and application methods of the context are analyzed. Finally, the proposed definition method and service model are verified by a typical ubiquitous service scenario. The results show that the method can extract the context required by ubiquitous service. Based on the model, service platform can adapt to heterogeneous terminal environments dynamically, and provide individualized presentation and handoff services for users.
2010, 32(3): 666-670.
doi: 10.3724/SP.J.1146.2009.00190
Abstract:
This paper studies encapsulated packet length of circuit traffic in the multi-service transport multi- protocol label switching network, which based on ITU-T standards. And the paper compared the performance of the network according to different TDM packet length. Then, an adjustment scheme is proposed to change the packet length dynamically. With the consideration of the real-time conditions of the traffics latency and variance, this scheme calculates the suitable payload sizes in the destination node and feeds back to the source node, in order to adjust this size in the adaptation module. This scheme is implemented in a testbed based on OPNET modeler for performance evaluation. The simulation results show that, compared with traditional constant bit encapsulation scheme, this scheme works more effectively both on the transmission quality and efficiency, and meets the requirement of supporting circuit traffic.
This paper studies encapsulated packet length of circuit traffic in the multi-service transport multi- protocol label switching network, which based on ITU-T standards. And the paper compared the performance of the network according to different TDM packet length. Then, an adjustment scheme is proposed to change the packet length dynamically. With the consideration of the real-time conditions of the traffics latency and variance, this scheme calculates the suitable payload sizes in the destination node and feeds back to the source node, in order to adjust this size in the adaptation module. This scheme is implemented in a testbed based on OPNET modeler for performance evaluation. The simulation results show that, compared with traditional constant bit encapsulation scheme, this scheme works more effectively both on the transmission quality and efficiency, and meets the requirement of supporting circuit traffic.
2010, 32(3): 671-676.
doi: 10.3724/SP.J.1146.2009.00186
Abstract:
Most of network link parameters inference methods assume that link states are stationary during measurement period, and can not obtain time-varying characteristics of link parameters. In this paper, a novel nonstationary internal loss tomography method is proposed. Assume in a relatively short time window, the time-varying curves of link loss rates are described by k times continuous differentiable functions. The k-th order Taylor Serieses of these functions are estimated using network tomography approach. Then based on the estimates of each time window, the time-varying link loss rates of entire measurement period are obtained by integrating the estimates of all time windows using Inverse Distance Square Weighted algorithm. NS2 simulations show that the method is capable of tracking variation of link loss rates effectively, and superior existing stationary internal loss tomography methods.
Most of network link parameters inference methods assume that link states are stationary during measurement period, and can not obtain time-varying characteristics of link parameters. In this paper, a novel nonstationary internal loss tomography method is proposed. Assume in a relatively short time window, the time-varying curves of link loss rates are described by k times continuous differentiable functions. The k-th order Taylor Serieses of these functions are estimated using network tomography approach. Then based on the estimates of each time window, the time-varying link loss rates of entire measurement period are obtained by integrating the estimates of all time windows using Inverse Distance Square Weighted algorithm. NS2 simulations show that the method is capable of tracking variation of link loss rates effectively, and superior existing stationary internal loss tomography methods.
2010, 32(3): 677-681.
doi: 10.3724/SP.J.1146.2009.00262
Abstract:
A new data delivery scheme-PRD (Probability Replication Delivery scheme) is proposed for pervasive data gathering in DTMSN (Delay Tolerant Mobile Sensor Networks) that networks with intermittent connectivity in space. PRD consists of two key components for data transmission and queue management, respectively. The former makes decisions on when and where to transmit data messages according to the node delivery probability. The latter employs the message survival time based on probability and delivery copies to decide dropping for minimizing transmission overhead. Simulation results show that the proposed PRD data delivery scheme achieves the higher message delivery ratio with the lower transmission overhead and data delivery delay than other DTMSN data delivering approaches.
A new data delivery scheme-PRD (Probability Replication Delivery scheme) is proposed for pervasive data gathering in DTMSN (Delay Tolerant Mobile Sensor Networks) that networks with intermittent connectivity in space. PRD consists of two key components for data transmission and queue management, respectively. The former makes decisions on when and where to transmit data messages according to the node delivery probability. The latter employs the message survival time based on probability and delivery copies to decide dropping for minimizing transmission overhead. Simulation results show that the proposed PRD data delivery scheme achieves the higher message delivery ratio with the lower transmission overhead and data delivery delay than other DTMSN data delivering approaches.
2010, 32(3): 682-686.
doi: 10.3724/SP.J.1146.2009.00300
Abstract:
A frequent pattern mining algorithm FPM (Frequent Pattern Mining) is proposed. FPM not only considered the frequency but also the distribution of the frequent pattern along the time series. Based on these different types of frequent patterns, MAMC (Mixed memory Aggregation Markov Chan) is extended to FMAMC (Frequent pattern based Mixed memory Aggregation Markov Chan) model. The proposed algorithm and model are applied to a smart building project, experiment and practice both demonstrate FPM is efficient than existing algorithms and FMAMC model could more accurately predict the node behavior in WSAN than MAMC.
A frequent pattern mining algorithm FPM (Frequent Pattern Mining) is proposed. FPM not only considered the frequency but also the distribution of the frequent pattern along the time series. Based on these different types of frequent patterns, MAMC (Mixed memory Aggregation Markov Chan) is extended to FMAMC (Frequent pattern based Mixed memory Aggregation Markov Chan) model. The proposed algorithm and model are applied to a smart building project, experiment and practice both demonstrate FPM is efficient than existing algorithms and FMAMC model could more accurately predict the node behavior in WSAN than MAMC.
2010, 32(3): 687-692.
doi: 10.3724/SP.J.1146.2009.01085
Abstract:
A kind of Rekeying Algorithm based on Orbital Cluster (RAOC) is proposed in the Low Earth Orbit (LEO) satellite network. The LEO satellite network is divided into different clusters according to the characteristics of the LEO satellite network in RAOC. The rekeying process is carried out by the initiating node of the LEO satellite network and the head nodes of the clusters. Rekeying lock is introduced to ensure consistency of the rekeying. The simulation results indicate that RAOC can accomplish the rekeying automatically compared with ground-based TTC(Tracking, Telemetry and Command) algorithm and space-based TTC algorithm, and the rekeying efficiency is improved by RAOC.
A kind of Rekeying Algorithm based on Orbital Cluster (RAOC) is proposed in the Low Earth Orbit (LEO) satellite network. The LEO satellite network is divided into different clusters according to the characteristics of the LEO satellite network in RAOC. The rekeying process is carried out by the initiating node of the LEO satellite network and the head nodes of the clusters. Rekeying lock is introduced to ensure consistency of the rekeying. The simulation results indicate that RAOC can accomplish the rekeying automatically compared with ground-based TTC(Tracking, Telemetry and Command) algorithm and space-based TTC algorithm, and the rekeying efficiency is improved by RAOC.
2010, 32(3): 693-699.
doi: 10.3724/SP.J.1146.2009.00287
Abstract:
For most present threshold signature schemes, sub-sign member can not sign a message anonymously or theirs anonymity is very weak. To improve their anonymity, a strong anonymity (n, t) threshold signature scheme based on DAA (Direct Anonymous Attestation), which is adopted by Trusted Computing Group v1.2 specifications, is proposed. Compared with the others, the scheme colligates DAA, zero-knowledge proof and Feldman verifiable secret sharing technique to achieve untraceable sub-sign and insure strong anonymity of signers, even the verifier and the dealer are colluded. Besides strong anonymity, analysis shows the scheme also has the property of unforgeable share, verifiable sub-sign, and robustness etc. It can be used in the situations which desire high-level anonymity such as anonymous voting.
For most present threshold signature schemes, sub-sign member can not sign a message anonymously or theirs anonymity is very weak. To improve their anonymity, a strong anonymity (n, t) threshold signature scheme based on DAA (Direct Anonymous Attestation), which is adopted by Trusted Computing Group v1.2 specifications, is proposed. Compared with the others, the scheme colligates DAA, zero-knowledge proof and Feldman verifiable secret sharing technique to achieve untraceable sub-sign and insure strong anonymity of signers, even the verifier and the dealer are colluded. Besides strong anonymity, analysis shows the scheme also has the property of unforgeable share, verifiable sub-sign, and robustness etc. It can be used in the situations which desire high-level anonymity such as anonymous voting.
2010, 32(3): 700-704.
doi: 10.3724/SP.J.1146.2009.00214
Abstract:
For Chinese Part-Of-Speech(POS) tagging, word segmentation is a preliminary step. To reduce accumulated errors between two steps and improve the segmentation performance by utilizing POS information, segmentation and POS tagging can be performed simultaneously. In this paper, a joint segmentation and POS tagging system is proposed based on undirected graphical models which can make full use of the dependencies between the two stages. In the joint system, segmenting and tagging are viewed as the sequence labeling; moreover any connected sub-graph can be viewed as a certain dependency which can be used to find the final opinion labeling. The joint model achieves high performances with 97.19% in segmentation precision and 95.34% in POS tagging precision, which are the state-of-art performances for Chinese word segmentation and tagging on 1998-year Peoples Daily corpus.
For Chinese Part-Of-Speech(POS) tagging, word segmentation is a preliminary step. To reduce accumulated errors between two steps and improve the segmentation performance by utilizing POS information, segmentation and POS tagging can be performed simultaneously. In this paper, a joint segmentation and POS tagging system is proposed based on undirected graphical models which can make full use of the dependencies between the two stages. In the joint system, segmenting and tagging are viewed as the sequence labeling; moreover any connected sub-graph can be viewed as a certain dependency which can be used to find the final opinion labeling. The joint model achieves high performances with 97.19% in segmentation precision and 95.34% in POS tagging precision, which are the state-of-art performances for Chinese word segmentation and tagging on 1998-year Peoples Daily corpus.
Reducing Undetected Error Rate of CRC Coding Based on Lossless Compression and Implement in Circuits
2010, 32(3): 705-709.
doi: 10.3724/SP.J.1146.2009.00255
Abstract:
The Cyclic Redundancy Checking (CRC) is widely used in many applications. For a certain format of the code, the error undetected probability is nearly a constant. But more bits of the information take more chances to burst errors. This paper proposes to reduce the error burst probability with lossless compression method, i.e., Huffman coding. The probability can be less then one ten-thousandth. Consequently, a new reliability circuit for VLSI with combined Huffman and CRC coding has been designed in this paper.
The Cyclic Redundancy Checking (CRC) is widely used in many applications. For a certain format of the code, the error undetected probability is nearly a constant. But more bits of the information take more chances to burst errors. This paper proposes to reduce the error burst probability with lossless compression method, i.e., Huffman coding. The probability can be less then one ten-thousandth. Consequently, a new reliability circuit for VLSI with combined Huffman and CRC coding has been designed in this paper.
2010, 32(3): 710-713.
doi: 10.3724/SP.J.1146.2009.00149
Abstract:
The phenomena of a high order mode oscillation in a broadband klystron are described in this paper. Some methods are discussed for suppressing parasitical mode oscillation in klystron resonant cavity. One new method is used to eliminate the oscillation in the broadband klystron. The method can make the high order mode frequency change enormously but the fundamental mode frequency alter faintly through adjust the structure of a resonant cavity, thereby the high order mode oscillation is suppressed. An improved design for resonant cavity with Ansoft HFSS code and corresponding experimental approval has been performed. The testing result of the klystron with the improved resonant cavity shows that the high order mode oscillation can effectively be suppressed.
The phenomena of a high order mode oscillation in a broadband klystron are described in this paper. Some methods are discussed for suppressing parasitical mode oscillation in klystron resonant cavity. One new method is used to eliminate the oscillation in the broadband klystron. The method can make the high order mode frequency change enormously but the fundamental mode frequency alter faintly through adjust the structure of a resonant cavity, thereby the high order mode oscillation is suppressed. An improved design for resonant cavity with Ansoft HFSS code and corresponding experimental approval has been performed. The testing result of the klystron with the improved resonant cavity shows that the high order mode oscillation can effectively be suppressed.
2010, 32(3): 728-731.
doi: 10.3724/SP.J.1146.2009.00120
Abstract:
A clustering algorithm, based on ant colony optimization algorithm, is proposed, and the stationary target can be detected without prior information. This algorithm is used in practical time difference location system and its performance is demonstrated.
A clustering algorithm, based on ant colony optimization algorithm, is proposed, and the stationary target can be detected without prior information. This algorithm is used in practical time difference location system and its performance is demonstrated.
2010, 32(3): 732-736.
doi: 10.3724/SP.J.1146.2009.00129
Abstract:
Based on the analysis of Short Time Fourier Transform(STFT) and Wigner-Vill Transform(WVT) for Loran-C signal, influence of SNR at the antenna input, SGR(Sky-wave to Ground-wave Ratio) and phase difference on STFT and WVT for Loran-C signal are analyzed and simulated. The results show that peak values of STFT and WVT for Loran-C signal are steady, and noise can be restrained successfully. But they both have the issue of multi-values, moreover, the issue of multi-values of Loran-C signals WVT can be removed successfully and effectively because frequency points of its negative peak values are different when the arrival time interval of sky-wave and ground-wave is different, so a new period identification based on WVT for Loran-C is presented.
Based on the analysis of Short Time Fourier Transform(STFT) and Wigner-Vill Transform(WVT) for Loran-C signal, influence of SNR at the antenna input, SGR(Sky-wave to Ground-wave Ratio) and phase difference on STFT and WVT for Loran-C signal are analyzed and simulated. The results show that peak values of STFT and WVT for Loran-C signal are steady, and noise can be restrained successfully. But they both have the issue of multi-values, moreover, the issue of multi-values of Loran-C signals WVT can be removed successfully and effectively because frequency points of its negative peak values are different when the arrival time interval of sky-wave and ground-wave is different, so a new period identification based on WVT for Loran-C is presented.
2010, 32(3): 737-741.
doi: 10.3724/SP.J.1146.2009.00061
Abstract:
-Linear Feedback Shift Register (-LFSR) is a word-oriented feedback shift register, which has a better tradeoff between the security and efficiency. The sequence generated by -LFSR is called the -linear recurrence sequence and its characteristic polynomial is the matrix polynomial over finite field. With analysis of the algebra structure of the matrix polynomial ring over finite field, the sufficient and necessary condition for the minimal polynomial of -linear recurrence sequence to be unique is given.
-Linear Feedback Shift Register (-LFSR) is a word-oriented feedback shift register, which has a better tradeoff between the security and efficiency. The sequence generated by -LFSR is called the -linear recurrence sequence and its characteristic polynomial is the matrix polynomial over finite field. With analysis of the algebra structure of the matrix polynomial ring over finite field, the sufficient and necessary condition for the minimal polynomial of -linear recurrence sequence to be unique is given.
2010, 32(3): 742-746.
doi: 10.3724/SP.J.1146.2009.00198
Abstract:
A novel wideband omnidirectioanl planar antenna comprised two back-to-back dipoles printed on a dielectric substrate which use as unit of pattern reconfigurable antenna array is presented. The presented antenna shows a wide impedance bandwidth about 50% (VSWR2), and a stable antenna gain level of about 6.44 dBi, with small variations(2.0 dB) for omnidirectional radiation, is obtained across the used bandwidth. Details of antenna design and experimental results are proposed, and compared with excellent agreement. A constructed prototype is suitable for WiMax(802.16) application fixed and mobile solutions ,as well as base or relay stations for WLAN operation or mobile communications.
A novel wideband omnidirectioanl planar antenna comprised two back-to-back dipoles printed on a dielectric substrate which use as unit of pattern reconfigurable antenna array is presented. The presented antenna shows a wide impedance bandwidth about 50% (VSWR2), and a stable antenna gain level of about 6.44 dBi, with small variations(2.0 dB) for omnidirectional radiation, is obtained across the used bandwidth. Details of antenna design and experimental results are proposed, and compared with excellent agreement. A constructed prototype is suitable for WiMax(802.16) application fixed and mobile solutions ,as well as base or relay stations for WLAN operation or mobile communications.
2010, 32(3): 747-749.
doi: 10.3724/SP.J.1146.2009.00204
Abstract:
A new algorithm for efficient computation of the Modulated Complex Lapped Transform (MCLT) with arbitrary windowing function is presented. For the MCLT of length-2M input data sequence, the proposed method is based on computing a length-2M type-II generalized discrete Hartley transform. Comparison with existing algorithms shows that the proposed method achieves the minimal number of arithmetic operations.
A new algorithm for efficient computation of the Modulated Complex Lapped Transform (MCLT) with arbitrary windowing function is presented. For the MCLT of length-2M input data sequence, the proposed method is based on computing a length-2M type-II generalized discrete Hartley transform. Comparison with existing algorithms shows that the proposed method achieves the minimal number of arithmetic operations.
2010, 32(3): 750-754.
doi: 10.3724/SP.J.1146.2009.00119
Abstract:
Crowd object segmentation is a key issue of the object tracking and recognition in multiple cameras. Human rough models with position, scale and pose information are constructed and then get the corresponding models by using Bayesian model. Then, the foreground is segmented into blocks of similar color distribution. Then the problem of the seed blocks selection is solved thought of color and position information under human inter-occlusion, and human region is received by seed growth. For blocks with similar color, they are merged into the objects by comparing the edge energy they brought. It can be seen that the method could segment the crowd precisely, and is not sensitive to background noise from the experimental results.
Crowd object segmentation is a key issue of the object tracking and recognition in multiple cameras. Human rough models with position, scale and pose information are constructed and then get the corresponding models by using Bayesian model. Then, the foreground is segmented into blocks of similar color distribution. Then the problem of the seed blocks selection is solved thought of color and position information under human inter-occlusion, and human region is received by seed growth. For blocks with similar color, they are merged into the objects by comparing the edge energy they brought. It can be seen that the method could segment the crowd precisely, and is not sensitive to background noise from the experimental results.
2010, 32(3): 755-758.
doi: 10.3724/SP.J.1146.2009.00704
Abstract:
Considering spatial and temporal redundancy of data and demand of saving energy in Wireless Sensor Networks (WSN), a One-Dimensional Linear Regression model based Spatial and Temporal(ODLRST) data compression algorithm, is proposed. By eliminating temporal redundancy of data in single node and spatial redundancy of data among nodes respectively in WSN, ODLRST greatly compresses these data. Simulation results show that ODLRST can reduce data size sent by nodes and network traffic in WSN, and save and balance energy consumption in the network.
Considering spatial and temporal redundancy of data and demand of saving energy in Wireless Sensor Networks (WSN), a One-Dimensional Linear Regression model based Spatial and Temporal(ODLRST) data compression algorithm, is proposed. By eliminating temporal redundancy of data in single node and spatial redundancy of data among nodes respectively in WSN, ODLRST greatly compresses these data. Simulation results show that ODLRST can reduce data size sent by nodes and network traffic in WSN, and save and balance energy consumption in the network.
2010, 32(3): 759-762.
doi: 10.3724/SP.J.1146.2009.01096
Abstract:
The scheme of incremental cooperation transmission and its energy efficiency in wireless sensor networks are studied in this paper. Two important issues including when to cooperate and performance of cooperation are also solved. The expressions of energy efficiency of direct and cooperative transmission are deduced respectively and then the effects of key parameters on the performance of cooperative transmission are discussed. Simulation results demonstrate that the incremental cooperation transmission provides better energy efficiency than direct transmission when the communication distance is bigger than the threshold of the distance. Energy efficiency can be enhanced by optimizing the position of the relay node selected or the modulation level.
The scheme of incremental cooperation transmission and its energy efficiency in wireless sensor networks are studied in this paper. Two important issues including when to cooperate and performance of cooperation are also solved. The expressions of energy efficiency of direct and cooperative transmission are deduced respectively and then the effects of key parameters on the performance of cooperative transmission are discussed. Simulation results demonstrate that the incremental cooperation transmission provides better energy efficiency than direct transmission when the communication distance is bigger than the threshold of the distance. Energy efficiency can be enhanced by optimizing the position of the relay node selected or the modulation level.
2010, 32(3): 714-727.
doi: 10.3724/SP.J.1146.2009.00751
Abstract:
As a programmable logic device, Field Programmable Gate Array(FPGA)has evolved from merely a peripheral component in an electronic design to become a core processing element of digital systems over the last two decades. It finds extensive applications in many fields, such as computer hardware, communication, aviation, spaceflight and automobile-electronics, etc. The FPGA chip design research achieves a significant progress with the advance of semi-conductor technologies. This survey reviews the past history, presents status and future trend in the ever quest for high performance FPGAs.
As a programmable logic device, Field Programmable Gate Array(FPGA)has evolved from merely a peripheral component in an electronic design to become a core processing element of digital systems over the last two decades. It finds extensive applications in many fields, such as computer hardware, communication, aviation, spaceflight and automobile-electronics, etc. The FPGA chip design research achieves a significant progress with the advance of semi-conductor technologies. This survey reviews the past history, presents status and future trend in the ever quest for high performance FPGAs.
