Email alert
2008 Vol. 30, No. 8
Display Method:
2008, 30(8): 1779-1782.
doi: 10.3724/SP.J.1146.2007.01493
Abstract:
In Orthogonal Frequency Division Multiple Access (OFDMA) systems, utilizing Proportional Fair (PF) algorithm can enhance system capacity, fairness, and performance. PF algorithm is extended, and a quality of service guaranteed packet scheduling algorithm is proposed for OFDMA systems. Users with high data rate requirements but in poor instantaneous channel state may be allocated excessive subcarriers. To prevent that situation, the algorithm uses dynamic weight factor. In a scheduling process, the weight factors of remaining subcarriers update as a subcarrier is allocated. Simulation results show that the proposed algorithm has a better performance in maintaining fairness than PF algorithm for real time services.
In Orthogonal Frequency Division Multiple Access (OFDMA) systems, utilizing Proportional Fair (PF) algorithm can enhance system capacity, fairness, and performance. PF algorithm is extended, and a quality of service guaranteed packet scheduling algorithm is proposed for OFDMA systems. Users with high data rate requirements but in poor instantaneous channel state may be allocated excessive subcarriers. To prevent that situation, the algorithm uses dynamic weight factor. In a scheduling process, the weight factors of remaining subcarriers update as a subcarrier is allocated. Simulation results show that the proposed algorithm has a better performance in maintaining fairness than PF algorithm for real time services.
2008, 30(8): 1783-1786.
doi: 10.3724/SP.J.1146.2007.00197
Abstract:
In this paper, a novel QoS-guarantee efficient scheduling and resource allocation algorithm for heterogeneous traffics in the downlink of MIMO-OFDMA system is proposed, which sufficiently exploits the space，time, frequency resource to provide QoS-guarantee. The algorithm takes into consideration not only the advanced physical techniques but also the service characteristics, QoS requirements and fairness for users observed in the medium access control (MAC) layer. Simulation results show it can significantly improve the system throughput performance while guaranteeing efficient transmission for heterogeneous traffics.
In this paper, a novel QoS-guarantee efficient scheduling and resource allocation algorithm for heterogeneous traffics in the downlink of MIMO-OFDMA system is proposed, which sufficiently exploits the space，time, frequency resource to provide QoS-guarantee. The algorithm takes into consideration not only the advanced physical techniques but also the service characteristics, QoS requirements and fairness for users observed in the medium access control (MAC) layer. Simulation results show it can significantly improve the system throughput performance while guaranteeing efficient transmission for heterogeneous traffics.
2008, 30(8): 1787-1790.
doi: 10.3724/SP.J.1146.2007.00122
Abstract:
This paper proposes an iterative Maximum A Posteriori (MAP) probability channel estimation algorithm for MIMO-OFDM systems. The receiver employs the soft systematic bits and soft parity bits of MAP decoding and feeds them back to the channel estimator through nonlinear mapping. To track the time-varying channel, the estimator adopts the Recursive Least Squares (RLS) adaptive filtering algorithm so as to improve the accuracy of the estimation. Computer simulations show that the proposed algorithm can achieve much better performance than the conventional Least Squares (LS) channel estimation in both Mean Square Error (MSE) and Frame Error Rate(FER) .
This paper proposes an iterative Maximum A Posteriori (MAP) probability channel estimation algorithm for MIMO-OFDM systems. The receiver employs the soft systematic bits and soft parity bits of MAP decoding and feeds them back to the channel estimator through nonlinear mapping. To track the time-varying channel, the estimator adopts the Recursive Least Squares (RLS) adaptive filtering algorithm so as to improve the accuracy of the estimation. Computer simulations show that the proposed algorithm can achieve much better performance than the conventional Least Squares (LS) channel estimation in both Mean Square Error (MSE) and Frame Error Rate(FER) .
2008, 30(8): 1791-1795.
doi: 10.3724/SP.J.1146.2006.01527
Abstract:
It is well known that Space-Time Coded Orthogonal Frequency Division Multiplexing (STC-OFDM) is easy to be affected by the Co-Channel Interference (CCI) and it is effective to mitigate CCI by beamforming. However, the existing methods are based on the accurate estimation of source Direction-Of-Arrival (DOA). In fact, the performance of system will degrade significantly when there is error in the estimation of DOA. In this paper, a robust algorithm based on Extended Bayesian(E-Bayesian)filter is proposed to improve the performance of the beamformers with uncertainty or error in desired DOA. In this algorithm, each DOA of desired signal is regarded as a random variable composed by several discrete samples. The posterior probability of these samples will be estimated by Bayesian formula and the samples will be resampled when some posterior probability is under certain threshold. Then, the optimized weights are determined by the posterior probability of these samples. Simulation results show that the proposed algorithm can significantly improve the robustness of the beamformers to combat the co-channel interference over Rayleigh multipath fading channel.
It is well known that Space-Time Coded Orthogonal Frequency Division Multiplexing (STC-OFDM) is easy to be affected by the Co-Channel Interference (CCI) and it is effective to mitigate CCI by beamforming. However, the existing methods are based on the accurate estimation of source Direction-Of-Arrival (DOA). In fact, the performance of system will degrade significantly when there is error in the estimation of DOA. In this paper, a robust algorithm based on Extended Bayesian(E-Bayesian)filter is proposed to improve the performance of the beamformers with uncertainty or error in desired DOA. In this algorithm, each DOA of desired signal is regarded as a random variable composed by several discrete samples. The posterior probability of these samples will be estimated by Bayesian formula and the samples will be resampled when some posterior probability is under certain threshold. Then, the optimized weights are determined by the posterior probability of these samples. Simulation results show that the proposed algorithm can significantly improve the robustness of the beamformers to combat the co-channel interference over Rayleigh multipath fading channel.
2008, 30(8): 1796-1800.
doi: 10.3724/SP.J.1146.2007.00095
Abstract:
An improved Bayesian EM channel estimator is proposed for MIMO-OFDM systems. Using the candidates list of the List Sphere Decoder (LSD) and a prior information from the decoder, the approximations of the soft decision symbols in the conventional EM channel estimator are modified. The a posteriori probabilities, the first and second moments of the soft symbol decisions are calculated more accurately. Based on the prior information of the channel measured by the initial estimation, a new Maximum A Posteriori (MAP) estimation of the channel impulse response, which considers the posteriori cross-correlation of the soft symbols, is obtained. Simulation results show that the newly proposed BEM algorithm outperforms conventional EM estimation schemes both in terms of BER and mean square estimation error.
An improved Bayesian EM channel estimator is proposed for MIMO-OFDM systems. Using the candidates list of the List Sphere Decoder (LSD) and a prior information from the decoder, the approximations of the soft decision symbols in the conventional EM channel estimator are modified. The a posteriori probabilities, the first and second moments of the soft symbol decisions are calculated more accurately. Based on the prior information of the channel measured by the initial estimation, a new Maximum A Posteriori (MAP) estimation of the channel impulse response, which considers the posteriori cross-correlation of the soft symbols, is obtained. Simulation results show that the newly proposed BEM algorithm outperforms conventional EM estimation schemes both in terms of BER and mean square estimation error.
2008, 30(8): 1801-1805.
doi: 10.3724/SP.J.1146.2007.00178
Abstract:
Some Quadratic Programs (QP) have been presented for the blind channel estimation and signal detection of the time-varying OFDM systems. In this paper a method is proposed to turn the QP problem into the optimization of a Semi-Definite Program (SDP) based on the Semi-Definite Relaxation (SDR) technique. Compared to the original QPs which reach the optimal solution with exponential complexity, the SDP could be solved efficiently in polynomial time. In order to reduce the complexity involved in the transform from the solution of the SDPs to the solution of the original QPs, the correlation property of the OFDM time-frequency response between adjacent subchannels or time-slots is exploited to devise a novel randomization strategy. The simulation proves the effectiveness of the new method.
Some Quadratic Programs (QP) have been presented for the blind channel estimation and signal detection of the time-varying OFDM systems. In this paper a method is proposed to turn the QP problem into the optimization of a Semi-Definite Program (SDP) based on the Semi-Definite Relaxation (SDR) technique. Compared to the original QPs which reach the optimal solution with exponential complexity, the SDP could be solved efficiently in polynomial time. In order to reduce the complexity involved in the transform from the solution of the SDPs to the solution of the original QPs, the correlation property of the OFDM time-frequency response between adjacent subchannels or time-slots is exploited to devise a novel randomization strategy. The simulation proves the effectiveness of the new method.
2008, 30(8): 1806-1809.
doi: 10.3724/SP.J.1146.2006.02070
Abstract:
Frequency Domain Equalization for Single Carrier transmission (SC-FDE) attracted much attention due to its low complexity and performance in suppressing ISI. But SC-FDE usually can not provide soft information for Soft-In Soft-Out (SISO) decoders which Shannon limit codes relies on to achieve outstanding BER performance. In order to combine SISO decoders with SC-FDE, a method to calculate the soft information of SC-FDE output symbols is proposed in this paper, and its validity is confirmed by simulations.
Frequency Domain Equalization for Single Carrier transmission (SC-FDE) attracted much attention due to its low complexity and performance in suppressing ISI. But SC-FDE usually can not provide soft information for Soft-In Soft-Out (SISO) decoders which Shannon limit codes relies on to achieve outstanding BER performance. In order to combine SISO decoders with SC-FDE, a method to calculate the soft information of SC-FDE output symbols is proposed in this paper, and its validity is confirmed by simulations.
2008, 30(8): 1810-1814.
doi: 10.3724/SP.J.1146.2006.02034
Abstract:
A modulation identification algorithm based on blind channel identification and equalization is proposed for MPSK and MQAM signals in multi-path environments. The channel order is estimated accurately first. Then, the eigenvalue perturbation analysis is done on the basis of matrix outer-product decomposition to get the more stable performance of channel blind identification and equalization. Finally, the multistage blind subtractive clustering method is used for the equalized constellation to determine the modulation type. Compared with the existing methods, it is a simple solution with smaller number of symbols for more modulation types and is more suitable for the practical application. Simulation results prove the effectiveness of the algorithm.
A modulation identification algorithm based on blind channel identification and equalization is proposed for MPSK and MQAM signals in multi-path environments. The channel order is estimated accurately first. Then, the eigenvalue perturbation analysis is done on the basis of matrix outer-product decomposition to get the more stable performance of channel blind identification and equalization. Finally, the multistage blind subtractive clustering method is used for the equalized constellation to determine the modulation type. Compared with the existing methods, it is a simple solution with smaller number of symbols for more modulation types and is more suitable for the practical application. Simulation results prove the effectiveness of the algorithm.
2008, 30(8): 1815-1819.
doi: 10.3724/SP.J.1146.2006.02081
Abstract:
A mechanism is proposed to notify the link layer ARQ retransmission state information to TCP by the cross-layer information exchange. It can avoid the negative effect of delay variation on TCP due to the link layer ARQ retransmission. The perfectly-persistent selective repeat ARQ is adopted to offer a reliable link to TCP to avoid packet loss in satellite links. It does not require the in order delivery at the link layer, then the delay of waiting for reorder is eliminated. The mechanism is fit for satellite networks with big bandwidth delay product. Simulation results validate that it can greatly improve TCP performance in satellite networks, especially when the error frame rate is high.
A mechanism is proposed to notify the link layer ARQ retransmission state information to TCP by the cross-layer information exchange. It can avoid the negative effect of delay variation on TCP due to the link layer ARQ retransmission. The perfectly-persistent selective repeat ARQ is adopted to offer a reliable link to TCP to avoid packet loss in satellite links. It does not require the in order delivery at the link layer, then the delay of waiting for reorder is eliminated. The mechanism is fit for satellite networks with big bandwidth delay product. Simulation results validate that it can greatly improve TCP performance in satellite networks, especially when the error frame rate is high.
2008, 30(8): 1820-1823.
doi: 10.3724/SP.J.1146.2006.02085
Abstract:
In satellite communication systems, some special users which are usually fast-moving terminals may need to be supported as well as the common users sometimes. Since the two types of users exist, the communication performances for both of them should be considered. In this paper, a novel channel reservation algorithm is proposed for this case. The theoretic analysis and simulation results show that the handover performances of the special users are improved by this algorithm.
In satellite communication systems, some special users which are usually fast-moving terminals may need to be supported as well as the common users sometimes. Since the two types of users exist, the communication performances for both of them should be considered. In this paper, a novel channel reservation algorithm is proposed for this case. The theoretic analysis and simulation results show that the handover performances of the special users are improved by this algorithm.
2008, 30(8): 1824-1827.
doi: 10.3724/SP.J.1146.2006.02054
Abstract:
Chaotic Direct Sequence Spread Spectrum (CD3S) signal is more nonlinear and more complex than the conventional direct sequence spread spectrum signal. This is the merit of CD3S, but also difficulty of estimation of chaotic spread spectrum sequences. According to the difficulty, a nonlinear Resilient back PROPagation (RPROP) neural network was proposed to estimate the chaotic sequences. The proposed method takes full advantages of the neural networks nonlinearity. It does not need to search a synchronous point between symbol waveform and chaotic sequences. The coefficient of neural network is used to estimate the chaotic spread spectrum sequences. The simulation results show that the method can estimate the chaotic sequences exactly at low SNR.
Chaotic Direct Sequence Spread Spectrum (CD3S) signal is more nonlinear and more complex than the conventional direct sequence spread spectrum signal. This is the merit of CD3S, but also difficulty of estimation of chaotic spread spectrum sequences. According to the difficulty, a nonlinear Resilient back PROPagation (RPROP) neural network was proposed to estimate the chaotic sequences. The proposed method takes full advantages of the neural networks nonlinearity. It does not need to search a synchronous point between symbol waveform and chaotic sequences. The coefficient of neural network is used to estimate the chaotic spread spectrum sequences. The simulation results show that the method can estimate the chaotic sequences exactly at low SNR.
2008, 30(8): 1828-1831.
doi: 10.3724/SP.J.1146.2006.02040
Abstract:
Based on a proposed equivalent model of two concatenated modules, the conventional QRD-M algorithm for MIMO signal detection was studied and an improved QRD-M algorithm was proposed in this paper. Through complexity analysis and numerical simulations, it is proved that the proposed algorithm can provide a more flexible tradeoff between system performance and the complexity of MIMO detection.
Based on a proposed equivalent model of two concatenated modules, the conventional QRD-M algorithm for MIMO signal detection was studied and an improved QRD-M algorithm was proposed in this paper. Through complexity analysis and numerical simulations, it is proved that the proposed algorithm can provide a more flexible tradeoff between system performance and the complexity of MIMO detection.
2008, 30(8): 1832-1835.
doi: 10.3724/SP.J.1146.2006.02077
Abstract:
In this paper, a novel simplified algorithm with broken line analysis is proposed for the log-likelihood ratio calculation of 16QAM signals. In this novel algorithm several line segments are introduced to approach the LLR curves of 16QAM signals, which results that the complicated nonlinear computations in the optimum algorithm are replaced with simple linear one. The simulation show that the algorithm in this paper can be a perfect approach to the optimum LLR algorithm in calculation results; and the performance will not degrade when it is applied in BITCM systems.
In this paper, a novel simplified algorithm with broken line analysis is proposed for the log-likelihood ratio calculation of 16QAM signals. In this novel algorithm several line segments are introduced to approach the LLR curves of 16QAM signals, which results that the complicated nonlinear computations in the optimum algorithm are replaced with simple linear one. The simulation show that the algorithm in this paper can be a perfect approach to the optimum LLR algorithm in calculation results; and the performance will not degrade when it is applied in BITCM systems.
2008, 30(8): 1836-1840.
doi: 10.3724/SP.J.1146.2007.00126
Abstract:
Multiple description image coding based on quantized frame expansion performs very well over error-prone channels. In this paper, an efficient multiple description image coding based on frame expansion is proposed. Firstly, an optical uniform tight frame is designed, which can minimize the mean squared error with quantization. Then, basic unit of frame expansion is constructed from wavelet zerotrees and different frames are applied to coefficients in different sub-bands. In that way, different sub-bands gain different protections during the transmission. After quantization and entropy coding, multiple description bit streams are generated from frame expansion. The experiment results prove that compared with other methods based on frame expansion, the scheme can obtain better recovery images under the same channel condition.
Multiple description image coding based on quantized frame expansion performs very well over error-prone channels. In this paper, an efficient multiple description image coding based on frame expansion is proposed. Firstly, an optical uniform tight frame is designed, which can minimize the mean squared error with quantization. Then, basic unit of frame expansion is constructed from wavelet zerotrees and different frames are applied to coefficients in different sub-bands. In that way, different sub-bands gain different protections during the transmission. After quantization and entropy coding, multiple description bit streams are generated from frame expansion. The experiment results prove that compared with other methods based on frame expansion, the scheme can obtain better recovery images under the same channel condition.
2008, 30(8): 1841-1844.
doi: 10.3724/SP.J.1146.2007.01320
Abstract:
A high-speed data transmission system (Overlapped Time Division Multiplexing System, OvTDM) and its performace are researched in AWGN channel. Simultaneously, the channel coded OvTDM system is researched. The optimal algorithm, which is soft-in soft-out Maximum A Posteriori probability (MAP) iterative joined detection and decoding algorithm is proposed for the coded OvTDM system. Simulation results show that the BER performance of the OvTDM system is better than that of the conventional QAM system under the same spectral efficiency. For the channel coded OvTDM system, the MAP soft-in soft-out iterative algorithm fully utilizes the encoding constraints formed by the OvTDM and obtains the optimal performance.
A high-speed data transmission system (Overlapped Time Division Multiplexing System, OvTDM) and its performace are researched in AWGN channel. Simultaneously, the channel coded OvTDM system is researched. The optimal algorithm, which is soft-in soft-out Maximum A Posteriori probability (MAP) iterative joined detection and decoding algorithm is proposed for the coded OvTDM system. Simulation results show that the BER performance of the OvTDM system is better than that of the conventional QAM system under the same spectral efficiency. For the channel coded OvTDM system, the MAP soft-in soft-out iterative algorithm fully utilizes the encoding constraints formed by the OvTDM and obtains the optimal performance.
2008, 30(8): 1845-1848.
doi: 10.3724/SP.J.1146.2007.01699
Abstract:
In this paper, an improved iterative decoding algorithm based on averaged Log-Likelihood Ratio (LLR) is proposed for LDPC coded BICM systems. The extrinsic information delivered by every BP iteration is averaged and the averaged LLR values are used as a prior LLR fed back to the input of demodulator for the next iteration. This method can alleviate the LLR oscillation of certain variable nodes during BP iterations and thus make the extrinsic output of decoder is more credible. Simulation results show that, compared with the two conventional iterative decoding algorithms, this improved algorithm achieves noticeable error performance improvement with only modest increase in computation complexity.
In this paper, an improved iterative decoding algorithm based on averaged Log-Likelihood Ratio (LLR) is proposed for LDPC coded BICM systems. The extrinsic information delivered by every BP iteration is averaged and the averaged LLR values are used as a prior LLR fed back to the input of demodulator for the next iteration. This method can alleviate the LLR oscillation of certain variable nodes during BP iterations and thus make the extrinsic output of decoder is more credible. Simulation results show that, compared with the two conventional iterative decoding algorithms, this improved algorithm achieves noticeable error performance improvement with only modest increase in computation complexity.
2008, 30(8): 1849-1852.
doi: 10.3724/SP.J.1146.2007.00043
Abstract:
A novel algorithm for denoising the contaminated chaotic signals is proposed, which is based on Particle Filtering (PF), and adapted for low SNR, additive non-Gaussian noise and the chaotic dynamic system with unknown parameters. Basic idea behind the proposed algorithm is that, chaotic signal and unknown parameters in the chaotic dynamic system are considered as a high dimension state vector, and the joint posterior probability density of these state vectors can be recursively calculated by utilizing the principle of Particle Filtering, then the optimum estimation of chaotic signal can be attained. In order to overcome the degenerate phenomena caused by the rapid divergence of the chaotic orbits, an effective strategy is taken in the proposed algorithm. Kernel smoothing method and Auto Regression (AR) model are used to recursively estimate the non-time-varying and time-varying parameters, respectively. The simulation results show that, compared with the existing denoising methods, the proposed algorithm can more effectively denoise additive noise in contaminated chaotic signals.
A novel algorithm for denoising the contaminated chaotic signals is proposed, which is based on Particle Filtering (PF), and adapted for low SNR, additive non-Gaussian noise and the chaotic dynamic system with unknown parameters. Basic idea behind the proposed algorithm is that, chaotic signal and unknown parameters in the chaotic dynamic system are considered as a high dimension state vector, and the joint posterior probability density of these state vectors can be recursively calculated by utilizing the principle of Particle Filtering, then the optimum estimation of chaotic signal can be attained. In order to overcome the degenerate phenomena caused by the rapid divergence of the chaotic orbits, an effective strategy is taken in the proposed algorithm. Kernel smoothing method and Auto Regression (AR) model are used to recursively estimate the non-time-varying and time-varying parameters, respectively. The simulation results show that, compared with the existing denoising methods, the proposed algorithm can more effectively denoise additive noise in contaminated chaotic signals.
2008, 30(8): 1853-1856.
doi: 10.3724/SP.J.1146.2007.00133
Abstract:
Aiming at the problem of communication signal blind separation in the noisy circumstance, a Blind Source Separation(BSS)algorithm in the case of noise is presented which combined the robust whitening algorithm and Joint Approximate Diagonalization (JAD) of the fourth order cumulant matrix and has good separation performance. Simulation results show that the separation performance of the algorithm proposed in the paper is greatly improved upon the normal fourth order cumulant matrix JAD algorithm, and the Interference-Signal-Ratio (ISR) can be decreased 10dB without increasing much too computation quantity.
Aiming at the problem of communication signal blind separation in the noisy circumstance, a Blind Source Separation(BSS)algorithm in the case of noise is presented which combined the robust whitening algorithm and Joint Approximate Diagonalization (JAD) of the fourth order cumulant matrix and has good separation performance. Simulation results show that the separation performance of the algorithm proposed in the paper is greatly improved upon the normal fourth order cumulant matrix JAD algorithm, and the Interference-Signal-Ratio (ISR) can be decreased 10dB without increasing much too computation quantity.
2008, 30(8): 1857-1860.
doi: 10.3724/SP.J.1146.2007.00114
Abstract:
This paper discusses the influence of both quantization and noise upon the estimation results in phase estimation algorithm based on orthogonal transformation. The bias, variance and distribution property of estimated values are provided, the theoretical analysis shows that the expectation of the phase estimated is biased in most of the cases even the noise is zero-mean-valued, and the effect could not be removed by averaging. In addition the variance of the estimated result is a function of SNR (Signal-Noise Ratio), quantization and its value. This conclusion is supported by simulations.
This paper discusses the influence of both quantization and noise upon the estimation results in phase estimation algorithm based on orthogonal transformation. The bias, variance and distribution property of estimated values are provided, the theoretical analysis shows that the expectation of the phase estimated is biased in most of the cases even the noise is zero-mean-valued, and the effect could not be removed by averaging. In addition the variance of the estimated result is a function of SNR (Signal-Noise Ratio), quantization and its value. This conclusion is supported by simulations.
2008, 30(8): 1861-1864.
doi: 10.3724/SP.J.1146.2007.00066
Abstract:
The efficiency of traditional binary Huffman decoding method is very low. A new decoding method based on octonary Huffman trees is presented in this paper to improve the decoding speed. Huffman codes are represented as an octonary tree and reconstructed as a single dimensional array according to the position of each node in the tree. When decoding, three bit code elements are read from the bitstream each time, and direct numerical computation may be used to replace judge and jump operations, which improve decoding efficiency. The proposed method is also applied to VLC and RVLC decoding algorithms of MPEG-4. Experiment results demonstrate that the proposed method can greatly improve decoding efficiency without increasing memory consumption much, and it exhibits better performance than other decoding methods.
The efficiency of traditional binary Huffman decoding method is very low. A new decoding method based on octonary Huffman trees is presented in this paper to improve the decoding speed. Huffman codes are represented as an octonary tree and reconstructed as a single dimensional array according to the position of each node in the tree. When decoding, three bit code elements are read from the bitstream each time, and direct numerical computation may be used to replace judge and jump operations, which improve decoding efficiency. The proposed method is also applied to VLC and RVLC decoding algorithms of MPEG-4. Experiment results demonstrate that the proposed method can greatly improve decoding efficiency without increasing memory consumption much, and it exhibits better performance than other decoding methods.
A Generalized Fuzzy Entropy Thresholding Segmentation Method Based on the Sugeno Complement Operator
2008, 30(8): 1865-1868.
doi: 10.3724/SP.J.1146.2007.00103
Abstract:
For images with bad illumination, the traditional fuzzy entropy thresholding segmentation method can not achieve satisfactory results. In this paper a generalized fuzzy entropy thresholding method based on the Sugeno complement function is presented. Firstly, nine thresholds are obtained for an image based on the variations of the fixed point in the Sugeno complement function. Secondly, the nine thresholds are evaluated by an image segmentation quality evaluation principle. Finally, the threshold with the maximum quality evaluation value among the nine thresholds is chosen as the optimal threshold. Compared with the traditional fuzzy entropy method, new method increases the opportunity of choosing an optimal threshold and obtains better segmentation result for images with bad illumination.
For images with bad illumination, the traditional fuzzy entropy thresholding segmentation method can not achieve satisfactory results. In this paper a generalized fuzzy entropy thresholding method based on the Sugeno complement function is presented. Firstly, nine thresholds are obtained for an image based on the variations of the fixed point in the Sugeno complement function. Secondly, the nine thresholds are evaluated by an image segmentation quality evaluation principle. Finally, the threshold with the maximum quality evaluation value among the nine thresholds is chosen as the optimal threshold. Compared with the traditional fuzzy entropy method, new method increases the opportunity of choosing an optimal threshold and obtains better segmentation result for images with bad illumination.
2008, 30(8): 1869-1873.
doi: 10.3724/SP.J.1146.2006.02059
Abstract:
This paper puts forward a new algorithm for image reducing noise based on the time matrix of Pulse Coupled Neural Networks (PCNN) form the aspect of the characteristic of image impulsive noise. The time matrix is a mapping from spatial image information to time information generated from PCNN, The time matrix contains useful information related to spatial information in image processing. The results of computer simulations show that through analyzing and processing the PCNN time matrix, the image polluted by impulsive noise can be filtered efficiently and visual effects of restoration images are much batter than those obtained from the median filter, mean filter and wiener filter. This method presents higher Peak Signal-to-Noise, better capability to reduce noise, and can protect edges and details of images and be more adaptive.
This paper puts forward a new algorithm for image reducing noise based on the time matrix of Pulse Coupled Neural Networks (PCNN) form the aspect of the characteristic of image impulsive noise. The time matrix is a mapping from spatial image information to time information generated from PCNN, The time matrix contains useful information related to spatial information in image processing. The results of computer simulations show that through analyzing and processing the PCNN time matrix, the image polluted by impulsive noise can be filtered efficiently and visual effects of restoration images are much batter than those obtained from the median filter, mean filter and wiener filter. This method presents higher Peak Signal-to-Noise, better capability to reduce noise, and can protect edges and details of images and be more adaptive.
2008, 30(8): 1874-1877.
doi: 10.3724/SP.J.1146.2006.02098
Abstract:
Performance of anti-noise for the pseudo-random code phase modulation and pulse amplitude modulation combined fuze based on Doppler effect are discussed in this thesis. Firstly, the principle of them are introduced briefly. Secondly, on this basis, the overall SNR gains of the whole procedure after correlation detection are deduced in detail. Their performance of anti-noise is analyzed, which is affected by Doppler frequency, the period of pseudo-random code and the pulse width of code-element .The results show that the performance of anti-noise for the latter fuze is better than that of the pseudo-random code phase modulation fuze. And the good performance of anti-noise can be guaranteed by designing on their parameters.
Performance of anti-noise for the pseudo-random code phase modulation and pulse amplitude modulation combined fuze based on Doppler effect are discussed in this thesis. Firstly, the principle of them are introduced briefly. Secondly, on this basis, the overall SNR gains of the whole procedure after correlation detection are deduced in detail. Their performance of anti-noise is analyzed, which is affected by Doppler frequency, the period of pseudo-random code and the pulse width of code-element .The results show that the performance of anti-noise for the latter fuze is better than that of the pseudo-random code phase modulation fuze. And the good performance of anti-noise can be guaranteed by designing on their parameters.
2008, 30(8): 1878-1881.
doi: 10.3724/SP.J.1146.2007.00051
Abstract:
In this paper, a new method is proposed to estimate the DOA of wideband coherent signals. The orthogonality of projected subspaces method is used to estimate DOA. This method does not require the primary information of DOA for focusing matrix and need not the sector dividing of interpolated method. And it can deal with coherent signals and array calibration by using the Toeplitz method to modify the data covariance matrix. The simulation results illustrated the effectiveness of this method.
In this paper, a new method is proposed to estimate the DOA of wideband coherent signals. The orthogonality of projected subspaces method is used to estimate DOA. This method does not require the primary information of DOA for focusing matrix and need not the sector dividing of interpolated method. And it can deal with coherent signals and array calibration by using the Toeplitz method to modify the data covariance matrix. The simulation results illustrated the effectiveness of this method.
2008, 30(8): 1882-1885.
doi: 10.3724/SP.J.1146.2007.01475
Abstract:
The Discrete Fourier Transform (DFT) has circular shift properties. Similarly, the Discrete Chirp- Fourier Transform (DCFT) motivated by DFT also has circular shift properties. This paper introduces the circular shift of DCFT and proposes a new method to detect chirp. The new method utilities not only the chirp characteristic, but also the noise characteristic, i.e., chirp rates and initial frequencies of chirp are invariant in time, whereas chirp rates and initial frequencies of noise are random in time, therefore, it has better detection performance. Comparing with DCFT, some experimental results are provided to demonstrate the better performance of the detection method proposed by this paper in low Signal-to-Noise ratio (SNR) environments.
The Discrete Fourier Transform (DFT) has circular shift properties. Similarly, the Discrete Chirp- Fourier Transform (DCFT) motivated by DFT also has circular shift properties. This paper introduces the circular shift of DCFT and proposes a new method to detect chirp. The new method utilities not only the chirp characteristic, but also the noise characteristic, i.e., chirp rates and initial frequencies of chirp are invariant in time, whereas chirp rates and initial frequencies of noise are random in time, therefore, it has better detection performance. Comparing with DCFT, some experimental results are provided to demonstrate the better performance of the detection method proposed by this paper in low Signal-to-Noise ratio (SNR) environments.
2008, 30(8): 1886-1889.
doi: 10.3724/SP.J.1146.2007.01420
Abstract:
DOA-matrix method is one of eigenstructure-based algorithms dealing with the estimation of the 2-D Direction-Of-Arrival(DOA) for multiple narrowband sources. In this paper, the DOA-matrix subspace method is extended for DOA estimation with an array of electromagnetic vector sensors. The detailed algorithm is deduced and the algorithm computational complexity is analyzed. Compared with ESPRIT method, this method can decrease much computational complexity. Monte Carlo simulation results show that the RMS bias and the RMS standard deviation from DOA-matrix method are larger than that from ESPRIT method. The DOA-matrix method also needed more snapshots in the simulation.
DOA-matrix method is one of eigenstructure-based algorithms dealing with the estimation of the 2-D Direction-Of-Arrival(DOA) for multiple narrowband sources. In this paper, the DOA-matrix subspace method is extended for DOA estimation with an array of electromagnetic vector sensors. The detailed algorithm is deduced and the algorithm computational complexity is analyzed. Compared with ESPRIT method, this method can decrease much computational complexity. Monte Carlo simulation results show that the RMS bias and the RMS standard deviation from DOA-matrix method are larger than that from ESPRIT method. The DOA-matrix method also needed more snapshots in the simulation.
2008, 30(8): 1890-1892.
doi: 10.3724/SP.J.1146.2008.00138
Abstract:
In this paper, the twice virtual interpolations is introduced into the circle array receiving signal 2D-DOA separable estimation, twice virtual interpolations are used to get the rotational invariance factor, the factor is used to get the estimation of elevation angle, then the estimated elevation angle is substituted into the circle array manifold, thus the azimuth angle can be get by one dimensional search. This method reduced the estimation error because of the interpolation error between the virtual array and the real array efficiently, simulation results validate the efficiency of the method.
In this paper, the twice virtual interpolations is introduced into the circle array receiving signal 2D-DOA separable estimation, twice virtual interpolations are used to get the rotational invariance factor, the factor is used to get the estimation of elevation angle, then the estimated elevation angle is substituted into the circle array manifold, thus the azimuth angle can be get by one dimensional search. This method reduced the estimation error because of the interpolation error between the virtual array and the real array efficiently, simulation results validate the efficiency of the method.
2008, 30(8): 1893-1896.
doi: 10.3724/SP.J.1146.2007.00955
Abstract:
The iterated estimation of DOA and polarization for completely polarized electromagnetic waves is proposed based on an partly calibrated polarization sensitive array with different type of polarization sensitive antennas, mounted on airframe. The proposed algorithm can estimate the arrays error based on the partly calibrated polarization. Then, the polarization steering vector of single polarization sensitive antenna is estimated by this arrays error, so that the sources DOA and polarization are estimated based on the polarization steering vector. The estimating accuracy is enhanced by means of iteration. The simulations show that the proposed algorithm has convergence and robust character.
The iterated estimation of DOA and polarization for completely polarized electromagnetic waves is proposed based on an partly calibrated polarization sensitive array with different type of polarization sensitive antennas, mounted on airframe. The proposed algorithm can estimate the arrays error based on the partly calibrated polarization. Then, the polarization steering vector of single polarization sensitive antenna is estimated by this arrays error, so that the sources DOA and polarization are estimated based on the polarization steering vector. The estimating accuracy is enhanced by means of iteration. The simulations show that the proposed algorithm has convergence and robust character.
2008, 30(8): 1897-1900.
doi: 10.3724/SP.J.1146.2006.02035
Abstract:
Dempster-Shafer (D-S) algorithm is improved to fuse different features of multi-temporal SAR images to detect change detection of urban areas. Firstly, Dempster-Shafer is developed by not only considering the certainty of the evidence, but also considering the average support of the evidence to different subsets in the assignment framework so that it can give more reliable combination result. Secondly, amplitude ratio feature and Euclid distance of the probability density distribution function in Pearson graph from different temporal SAR images are extracted to present change feathers in different scales. Finally, the improved Dempster-Shafer algorithm is applied to fuse the two different features to detect change information of SAR images. An example of the Shanghai Lujiazui area using the ERS-2 SAR image well demonstrates the accuracy of the improved fusion algorithm.
Dempster-Shafer (D-S) algorithm is improved to fuse different features of multi-temporal SAR images to detect change detection of urban areas. Firstly, Dempster-Shafer is developed by not only considering the certainty of the evidence, but also considering the average support of the evidence to different subsets in the assignment framework so that it can give more reliable combination result. Secondly, amplitude ratio feature and Euclid distance of the probability density distribution function in Pearson graph from different temporal SAR images are extracted to present change feathers in different scales. Finally, the improved Dempster-Shafer algorithm is applied to fuse the two different features to detect change information of SAR images. An example of the Shanghai Lujiazui area using the ERS-2 SAR image well demonstrates the accuracy of the improved fusion algorithm.
2008, 30(8): 1901-1904.
doi: 10.3724/SP.J.1146.2007.00056
Abstract:
A scheme about multi-sensor data fusion based on adaptive filter is developed for improve tracking precision for moving power-driven target under complicated air-battle environment. At first, optimal weight for sensors are found by measuring data to optimize target point x of anytime. Secondly, the x point is put into adaptive filter as input signal. Plus matrix is adjusted according to change of state noise and observation noise of system at the same time. According to adaptive filter system state noise output and current data, weight for sensors is adjusted on line by using fuzzy logic system. Finally, the output signal is a fusion track that is gained passing through two class self-adapt signal process. The simulation result demonstrates the fusion algorithm is effective.
A scheme about multi-sensor data fusion based on adaptive filter is developed for improve tracking precision for moving power-driven target under complicated air-battle environment. At first, optimal weight for sensors are found by measuring data to optimize target point x of anytime. Secondly, the x point is put into adaptive filter as input signal. Plus matrix is adjusted according to change of state noise and observation noise of system at the same time. According to adaptive filter system state noise output and current data, weight for sensors is adjusted on line by using fuzzy logic system. Finally, the output signal is a fusion track that is gained passing through two class self-adapt signal process. The simulation result demonstrates the fusion algorithm is effective.
2008, 30(8): 1905-1908.
doi: 10.3724/SP.J.1146.2007.01784
Abstract:
In this paper, a new particle filter based on Statistical Linear Regression (SLR) is proposed for the state estimation of non-Gauss nonlinear systems. In the new algorithm, the importance density function of particle filter is generated by linearizing the nonlinear function using statistical linear regression through a set of Gauss-Hermite quadrature points estimating regression coefficient. The density function integrates the new observations into system state transition and extends the overlap fields with true posterior density. The simulation shows that the new algorithm not only has high estimation accuracy but also has better stability and less computation amount than the PF.
In this paper, a new particle filter based on Statistical Linear Regression (SLR) is proposed for the state estimation of non-Gauss nonlinear systems. In the new algorithm, the importance density function of particle filter is generated by linearizing the nonlinear function using statistical linear regression through a set of Gauss-Hermite quadrature points estimating regression coefficient. The density function integrates the new observations into system state transition and extends the overlap fields with true posterior density. The simulation shows that the new algorithm not only has high estimation accuracy but also has better stability and less computation amount than the PF.
2008, 30(8): 1909-1912.
doi: 10.3724/SP.J.1146.2007.00136
Abstract:
By analyzing imaging theories of point target, a preprocessing algorithm of point target detection based on temporal-spatial bilateral filter using adaptive neighborhoods is presented. The algorithm establishes an adaptive neighborhood for the given pixel, and a size limitation is added to the adaptive neighborhood to accelerate the algorithm. Filter template is obtained by multiplying the potential function in the domain and the potential function in the range of different neighborhoods. Compared with spatial bilateral filtering algorithm and temporal-spatial bilateral filtering algorithm using fixed neighborhoods, this preprocessing algorithm removes noise and enhances the contrast between point target and background efficiently and rapidly.
By analyzing imaging theories of point target, a preprocessing algorithm of point target detection based on temporal-spatial bilateral filter using adaptive neighborhoods is presented. The algorithm establishes an adaptive neighborhood for the given pixel, and a size limitation is added to the adaptive neighborhood to accelerate the algorithm. Filter template is obtained by multiplying the potential function in the domain and the potential function in the range of different neighborhoods. Compared with spatial bilateral filtering algorithm and temporal-spatial bilateral filtering algorithm using fixed neighborhoods, this preprocessing algorithm removes noise and enhances the contrast between point target and background efficiently and rapidly.
2008, 30(8): 1913-1917.
doi: 10.3724/SP.J.1146.2007.00123
Abstract:
The design of the rough neural network based on variable precision rough set model is studied. The condition of -approximation reduction is generalized and the criteria for selecting a -approximation reduction are introduced. In the experiment of the Brodatz texture image classification, the performance of conventional RNN(Rough Neural Network) and VPRNN(Variable Precision Rough set Neural Network) is compared. The results indicate that VPRNN not only has more simplify structure and less training time, but also, has better approximation decision-making ability and generalization ability than RNN.
The design of the rough neural network based on variable precision rough set model is studied. The condition of -approximation reduction is generalized and the criteria for selecting a -approximation reduction are introduced. In the experiment of the Brodatz texture image classification, the performance of conventional RNN(Rough Neural Network) and VPRNN(Variable Precision Rough set Neural Network) is compared. The results indicate that VPRNN not only has more simplify structure and less training time, but also, has better approximation decision-making ability and generalization ability than RNN.
2008, 30(8): 1918-1922.
doi: 10.3724/SP.J.1146.2007.00049
Abstract:
The number of Gaussian component is fixed and correlativity of class label between adjacent pixels is not considered in classical Gaussian mixture background model. As an improved version of the model, the main contribution of this paper is twofold. The first is to construct entropy image to measure the complexity of pixels intensity distribution, and further present the adaptation mechanism for automatically choosing the component number of Gaussian mixture model for each pixel according to entropy image so that the computational cost can be reduced without significantly sacrificing detection accuracy. The other is to use the membership degree to measure the degree that one pixel belongs to the background, and further fusion the local information within its adjacent region for effective pixel classification so that the classification decision becomes more reliable without significantly increasing the computation load. Experiments conducted on various real scenes demonstrate the good performance in computational speed and accuracy.
The number of Gaussian component is fixed and correlativity of class label between adjacent pixels is not considered in classical Gaussian mixture background model. As an improved version of the model, the main contribution of this paper is twofold. The first is to construct entropy image to measure the complexity of pixels intensity distribution, and further present the adaptation mechanism for automatically choosing the component number of Gaussian mixture model for each pixel according to entropy image so that the computational cost can be reduced without significantly sacrificing detection accuracy. The other is to use the membership degree to measure the degree that one pixel belongs to the background, and further fusion the local information within its adjacent region for effective pixel classification so that the classification decision becomes more reliable without significantly increasing the computation load. Experiments conducted on various real scenes demonstrate the good performance in computational speed and accuracy.
2008, 30(8): 1923-1927.
doi: 10.3724/SP.J.1146.2006.02065
Abstract:
A new robust Improved Fuzzy Partitions for K-Plane Clustering (IFP-KPC) algorithm is proposed. The proposed algorithm can reduce the sensitivity of the k-plane clustering algorithm to noises in real datasets. Also the distances to the Voronoi cell are used to give a reasonable explanation for the robustness of IFP-KPC. Experimental results demonstrate the effectiveness of IFP-KPC.
A new robust Improved Fuzzy Partitions for K-Plane Clustering (IFP-KPC) algorithm is proposed. The proposed algorithm can reduce the sensitivity of the k-plane clustering algorithm to noises in real datasets. Also the distances to the Voronoi cell are used to give a reasonable explanation for the robustness of IFP-KPC. Experimental results demonstrate the effectiveness of IFP-KPC.
2008, 30(8): 1928-1931.
doi: 10.3724/SP.J.1146.2007.00010
Abstract:
A Kernel-SOM based unsupervised nonlinear system identification algorithm is proposed. Analysis of the model running convergence of the proposed algorithm is performed, and the convergence theorem is proofed by considering both identification error and initial input error. Numerical simulation results demonstrate the effectiveness of the proposed identification algorithm and the correctness of the convergence theorem.
A Kernel-SOM based unsupervised nonlinear system identification algorithm is proposed. Analysis of the model running convergence of the proposed algorithm is performed, and the convergence theorem is proofed by considering both identification error and initial input error. Numerical simulation results demonstrate the effectiveness of the proposed identification algorithm and the correctness of the convergence theorem.
2008, 30(8): 1932-1935.
doi: 10.3724/SP.J.1146.2007.00130
Abstract:
According to the heteroclinic shilnikov theorem, a kind of piecewise linear chaotic system is presented in this paper. These systems have at least two equilibriums and at each equilibrium they have the same Jacobian. Changing the equilibriums and the separating planes, the other forms of these systems can be got. Theoretical analysis and experimental results confirm the method is effective.
According to the heteroclinic shilnikov theorem, a kind of piecewise linear chaotic system is presented in this paper. These systems have at least two equilibriums and at each equilibrium they have the same Jacobian. Changing the equilibriums and the separating planes, the other forms of these systems can be got. Theoretical analysis and experimental results confirm the method is effective.
2008, 30(8): 1936-1939.
doi: 10.3724/SP.J.1146.2007.00047
Abstract:
A previous text steganography scheme uses the mimic function and the Context-Free Grammar (CFG) to generate pseudo-natural language text to convey secret information. However, the generated text usually lacks semantic consistency and is therefore vulnerable to human evaluation. An improved mimicry technique is proposed that constructs the CFG-based grammar file from a template text that meets certain criteria so that the generated text is semantically consistent. The grammar file is constructed from a replacement text base composed of many optional words, phrases or sentences. Huffman coding is used to take into account the occurrence frequencies of various words, expressions, and syntactic structures to improve naturalness of the generated text. Thus, in addition to machine examination, the obtained stego-text is able to pass stringent human evaluation.
A previous text steganography scheme uses the mimic function and the Context-Free Grammar (CFG) to generate pseudo-natural language text to convey secret information. However, the generated text usually lacks semantic consistency and is therefore vulnerable to human evaluation. An improved mimicry technique is proposed that constructs the CFG-based grammar file from a template text that meets certain criteria so that the generated text is semantically consistent. The grammar file is constructed from a replacement text base composed of many optional words, phrases or sentences. Huffman coding is used to take into account the occurrence frequencies of various words, expressions, and syntactic structures to improve naturalness of the generated text. Thus, in addition to machine examination, the obtained stego-text is able to pass stringent human evaluation.
2008, 30(8): 1940-1943.
doi: 10.3724/SP.J.1146.2007.00054
Abstract:
To effectively suppress speckle noise and preserve structure information of SAR images, a NeighShrink despeckling based on scale space correlation is proposed in this paper. Firstly, for detail subbands of logarithmically transformed SAR images decomposed by stationary wavelet transform, the single selective scale space correlation with tunable parameter is introduced to separate the wavelet coefficients related to noise and the ones related to structure information. Subsequently, as regards the former, the NeighShrink is directly taken advantage of to obtain good smoothing effect. For the latter, the weighted NeighShrink is proposed to achieve the preservation of structure information. The experimental results verify the validity of the proposed method.
To effectively suppress speckle noise and preserve structure information of SAR images, a NeighShrink despeckling based on scale space correlation is proposed in this paper. Firstly, for detail subbands of logarithmically transformed SAR images decomposed by stationary wavelet transform, the single selective scale space correlation with tunable parameter is introduced to separate the wavelet coefficients related to noise and the ones related to structure information. Subsequently, as regards the former, the NeighShrink is directly taken advantage of to obtain good smoothing effect. For the latter, the weighted NeighShrink is proposed to achieve the preservation of structure information. The experimental results verify the validity of the proposed method.
2008, 30(8): 1944-1948.
doi: 10.3724/SP.J.1146.2007.00115
Abstract:
A new algorithm which can use lacunarity feature to discriminate vehicle target from natural clutter in SAR imagery is developed in this paper. Firstly, the variation and irregularity of back-scattered intensity for vehicle target and natural terrain are analyzed, which are resulted from their different scattering centers with different spatial arrangement and other cases. The vehicle image presents more irregularity and largeness of gaps than natural terrains image. Based on fractal theory, the lacunarity feature is estimated to measure the difference and can be used to eliminate the natural clutter. And then, the real X band SAR image data are applied to validate the above algorithm, and the performance of this algorithm is good.
A new algorithm which can use lacunarity feature to discriminate vehicle target from natural clutter in SAR imagery is developed in this paper. Firstly, the variation and irregularity of back-scattered intensity for vehicle target and natural terrain are analyzed, which are resulted from their different scattering centers with different spatial arrangement and other cases. The vehicle image presents more irregularity and largeness of gaps than natural terrains image. Based on fractal theory, the lacunarity feature is estimated to measure the difference and can be used to eliminate the natural clutter. And then, the real X band SAR image data are applied to validate the above algorithm, and the performance of this algorithm is good.
2008, 30(8): 1949-1953.
doi: 10.3724/SP.J.1146.2007.00024
Abstract:
Radar High-Resolution Range Profile (HRRP) is very sensitive to time-shift; therefore, HRRP-based Radar Automatic Target Recognition (RATR) requires efficient time-shift invariant features. A new time-shift invariant feature, i.e., amplitude spectrum difference of HRRP, is extracted from HRRP to solve the time-shift sensitivity. The result of theoretical analysis shows that, as a time-shift invariant feature, amplitude spectrum difference is more suitable for HRRP-based RATR than amplitude spectrum is. Shortest distance classifier and Support Vector Machine (SVM) classifier are designed to evaluate the recognition performance. Experimental results for measured data show that, comparing with amplitude spectrum, amplitude spectrum difference improves recognition performance remarkably.
Radar High-Resolution Range Profile (HRRP) is very sensitive to time-shift; therefore, HRRP-based Radar Automatic Target Recognition (RATR) requires efficient time-shift invariant features. A new time-shift invariant feature, i.e., amplitude spectrum difference of HRRP, is extracted from HRRP to solve the time-shift sensitivity. The result of theoretical analysis shows that, as a time-shift invariant feature, amplitude spectrum difference is more suitable for HRRP-based RATR than amplitude spectrum is. Shortest distance classifier and Support Vector Machine (SVM) classifier are designed to evaluate the recognition performance. Experimental results for measured data show that, comparing with amplitude spectrum, amplitude spectrum difference improves recognition performance remarkably.
2008, 30(8): 1954-1958.
doi: 10.3724/SP.J.1146.2006.02080
Abstract:
Interferometry with Space-borne ScanSAR is a wide swath elevation measurement technique. The properties of ScanSAR interferometric signal spectrum and Azimuth Scan Pattern Synchronization are discussed combined with the principle of ScanSAR mode. The characteristics of ENVISAT/ASAR Wide Swath SLC data are analyzed. A novel data processing method for ScanSAR interferometry is proposed based on the analysis results,. The method is validated with real data experiment.
Interferometry with Space-borne ScanSAR is a wide swath elevation measurement technique. The properties of ScanSAR interferometric signal spectrum and Azimuth Scan Pattern Synchronization are discussed combined with the principle of ScanSAR mode. The characteristics of ENVISAT/ASAR Wide Swath SLC data are analyzed. A novel data processing method for ScanSAR interferometry is proposed based on the analysis results,. The method is validated with real data experiment.
2008, 30(8): 1959-1962.
doi: 10.3724/SP.J.1146.2007.00063
Abstract:
In HF groundwave shipborne bistatic radar, the movement of radar can make the system more complex. This paper analyses the geometrical change of the mobile bistatic system, and the effect of the geometrical change to the LFMICW signal echo of target,and anylyses the relationship between phase of the intermediate frequency signal of target echo and the movement of radar in detail, and simply discusses the method to get the range and velocity of target in the complex dynamic system. Finally, the result of simulation proves the broadening of target doppler frequency when the target is near to the radar.
In HF groundwave shipborne bistatic radar, the movement of radar can make the system more complex. This paper analyses the geometrical change of the mobile bistatic system, and the effect of the geometrical change to the LFMICW signal echo of target,and anylyses the relationship between phase of the intermediate frequency signal of target echo and the movement of radar in detail, and simply discusses the method to get the range and velocity of target in the complex dynamic system. Finally, the result of simulation proves the broadening of target doppler frequency when the target is near to the radar.
2008, 30(8): 1963-1967.
doi: 10.3724/SP.J.1146.2006.02083
Abstract:
A novel method which is called polarimetric-ESPRIT (P-ESPRIT) algorithm is presented for full-polarization scattering center extraction and parameter estimation. The P-ESPRIT algorithm is a joint processing of polarization and super-resolution. It is able to estimate the number, positions, intensities and normalized scattering matrices of scattering centers instantaneously rather than the one which extracts parameters from each channel separately, and its performance is better than the latter because the fully-polarized information is used. It has computational advantage over other methods like MUSIC and ML because it neednt search. The validity is proved by the experimental results based on simulated and real data.
A novel method which is called polarimetric-ESPRIT (P-ESPRIT) algorithm is presented for full-polarization scattering center extraction and parameter estimation. The P-ESPRIT algorithm is a joint processing of polarization and super-resolution. It is able to estimate the number, positions, intensities and normalized scattering matrices of scattering centers instantaneously rather than the one which extracts parameters from each channel separately, and its performance is better than the latter because the fully-polarized information is used. It has computational advantage over other methods like MUSIC and ML because it neednt search. The validity is proved by the experimental results based on simulated and real data.
2008, 30(8): 1968-1972.
doi: 10.3724/SP.J.1146.2006.02033
Abstract:
The 3L (low data sampling rate, low measurement precision and low detection probability) features of sky-wave Over-The-Horizon Radar (OTHR) pose new challenges to the conventional track initiation methods. In this paper, a Two-Hierarchy Hough Transform track initiation method (TH-HT) is proposed. Firstly the range-time data space is used to hierarchical I Hough transform and rough candidate tracks are obtained, where Doppler is used as an accumulation constraint; then such rough candidate tracks are further refined by time-azimuth accumulation in hierarchical II Hough transform. Moreover, two theorems concerning the expectation of the number of clutters voting to each cell of each hierarchy are given, which provides a way of determining the two-hierarchy threshold value. In simulation, the proposed method is compared with the classical initiation method of OTHR-MMUPDA. Simulation results show that TH-HT has higher track detection probability and lower false track probability; meanwhile it is much less time-consuming.
The 3L (low data sampling rate, low measurement precision and low detection probability) features of sky-wave Over-The-Horizon Radar (OTHR) pose new challenges to the conventional track initiation methods. In this paper, a Two-Hierarchy Hough Transform track initiation method (TH-HT) is proposed. Firstly the range-time data space is used to hierarchical I Hough transform and rough candidate tracks are obtained, where Doppler is used as an accumulation constraint; then such rough candidate tracks are further refined by time-azimuth accumulation in hierarchical II Hough transform. Moreover, two theorems concerning the expectation of the number of clutters voting to each cell of each hierarchy are given, which provides a way of determining the two-hierarchy threshold value. In simulation, the proposed method is compared with the classical initiation method of OTHR-MMUPDA. Simulation results show that TH-HT has higher track detection probability and lower false track probability; meanwhile it is much less time-consuming.
2008, 30(8): 1973-1976.
doi: 10.3724/SP.J.1146.2007.00075
Abstract:
Interactive image segmentation methods have recently gained more and more attentions. A new interactive segmentation method is proposed based on the graph cuts. It combines several image features like texture, color and edge together through a probabilistic model. Texture and color features are modeled with histograms. Dimensionality reduction in feature space is achieved with a fisher discriminant criterion based on texton. The global optimal segmentation can be efficiently computed via graph cuts. Efficiency and accuracy of the method is demonstrated on aerial image segmentation and some other applications.
Interactive image segmentation methods have recently gained more and more attentions. A new interactive segmentation method is proposed based on the graph cuts. It combines several image features like texture, color and edge together through a probabilistic model. Texture and color features are modeled with histograms. Dimensionality reduction in feature space is achieved with a fisher discriminant criterion based on texton. The global optimal segmentation can be efficiently computed via graph cuts. Efficiency and accuracy of the method is demonstrated on aerial image segmentation and some other applications.
2008, 30(8): 1977-1980.
doi: 10.3724/SP.J.1146.2007.00527
Abstract:
A new beamforming algorithm used in GPS receiver, which can greatly depress the wideband interference, is proposed. The algorithm is based on space-time adaptive processing. Firstly, the steering vectors matrix of interference signals can be estimated by space-time 2-dimension power spectrum. Secondly, with in the null space of the steering vectors matrix of interference signals, the space-time 2-dimension weight vector can be obtained with least square method through the linear combination of orthogonal basis. And the weight vector is closest to the steering vector of the expected signal and deeply depresses the interference signals simultaneously. Finally, compared with the constrained least mean-squares algorithm, the proposed algorithm is more efficient to depress wideband interference. The experimental results show the effectiveness and stability of the method.
A new beamforming algorithm used in GPS receiver, which can greatly depress the wideband interference, is proposed. The algorithm is based on space-time adaptive processing. Firstly, the steering vectors matrix of interference signals can be estimated by space-time 2-dimension power spectrum. Secondly, with in the null space of the steering vectors matrix of interference signals, the space-time 2-dimension weight vector can be obtained with least square method through the linear combination of orthogonal basis. And the weight vector is closest to the steering vector of the expected signal and deeply depresses the interference signals simultaneously. Finally, compared with the constrained least mean-squares algorithm, the proposed algorithm is more efficient to depress wideband interference. The experimental results show the effectiveness and stability of the method.
2008, 30(8): 1981-1984.
doi: 10.3724/SP.J.1146.2007.00163
Abstract:
In Kalman filtering applications to combination satellite navigation the observation values including outliers are important effect on optimal filtering.The outliers affect the position filtering accuracy and make estimation inaccurate. In this paper,an outlier decision based on adaptive robust Kalman filtering is presented,which can keep orthogonal properties of innovation sequence.Simulation results show that the modified algorithms are effectively resistant to outliers in sampling date.
In Kalman filtering applications to combination satellite navigation the observation values including outliers are important effect on optimal filtering.The outliers affect the position filtering accuracy and make estimation inaccurate. In this paper,an outlier decision based on adaptive robust Kalman filtering is presented,which can keep orthogonal properties of innovation sequence.Simulation results show that the modified algorithms are effectively resistant to outliers in sampling date.
2008, 30(8): 1985-1988.
doi: 10.3724/SP.J.1146.2007.00059
Abstract:
A novel hierarchical mobile IP network analysis model is firstly constructed in this paper. And then a scheme is proposed to set hierarchical mobile IP optimal size of regional network. After mathematical relations between handoff latency and the regional size of hierarchical mobile IP are studied, the conclusion can be drawn that there exists optimal size of regional network. Moreover, a minimum iterative function is adopted to deduce optimal MAP regional radius when handoff latency is minimum. Research results will be benefit to hierarchical mobile IP network planning and optimization.
A novel hierarchical mobile IP network analysis model is firstly constructed in this paper. And then a scheme is proposed to set hierarchical mobile IP optimal size of regional network. After mathematical relations between handoff latency and the regional size of hierarchical mobile IP are studied, the conclusion can be drawn that there exists optimal size of regional network. Moreover, a minimum iterative function is adopted to deduce optimal MAP regional radius when handoff latency is minimum. Research results will be benefit to hierarchical mobile IP network planning and optimization.
2008, 30(8): 1989-1993.
doi: 10.3724/SP.J.1146.2006.02063
Abstract:
Analysis on Optical Burst Switching(OBS)network performance based on the Poisson model would not be exactly. In this paper, the distribution of data burst length is researched, and it is simulated that the self-similarity of data burst traffic assembled by time-based mechanism, moreover, the performance of LAUC and LAUC-VF are shown in the simulations. Simulation and theoretical results show that the assembly mechanism based on time threshold can decrease the traffics self-similarity, on the other hand, the influence of self-similarity on LAUC is small, but much greater on LAUC-VF whose burst drop probability will be decreased by 3 percent than Poisson stream.
Analysis on Optical Burst Switching(OBS)network performance based on the Poisson model would not be exactly. In this paper, the distribution of data burst length is researched, and it is simulated that the self-similarity of data burst traffic assembled by time-based mechanism, moreover, the performance of LAUC and LAUC-VF are shown in the simulations. Simulation and theoretical results show that the assembly mechanism based on time threshold can decrease the traffics self-similarity, on the other hand, the influence of self-similarity on LAUC is small, but much greater on LAUC-VF whose burst drop probability will be decreased by 3 percent than Poisson stream.
2008, 30(8): 1994-1998.
doi: 10.3724/SP.J.1146.2006.02042
Abstract:
In this paper, the problem of dynamic routing under the hose uncertain model for the full-mesh optical network architecture is considered. A novel dynamic routing algorithm-LBADF (Load Balancing with Adjustable Distribution Fraction) based on Valiant load balancing is proposed. LBADF algorithm can instantly adjust distribution fraction in Valiant load balancing according to the number of the spare wavelengths on the links to optimize the performance of the network. Computer simulation results show LBADF algorithm has the lower blocking probability for the whole network than that of VLB (Valiant Load Balancing) algorithm, which has the fixed distribution fraction. And the maximum blocking probability for all the node pairs in the network can also be reduced correspondingly in LBADF.
In this paper, the problem of dynamic routing under the hose uncertain model for the full-mesh optical network architecture is considered. A novel dynamic routing algorithm-LBADF (Load Balancing with Adjustable Distribution Fraction) based on Valiant load balancing is proposed. LBADF algorithm can instantly adjust distribution fraction in Valiant load balancing according to the number of the spare wavelengths on the links to optimize the performance of the network. Computer simulation results show LBADF algorithm has the lower blocking probability for the whole network than that of VLB (Valiant Load Balancing) algorithm, which has the fixed distribution fraction. And the maximum blocking probability for all the node pairs in the network can also be reduced correspondingly in LBADF.
2008, 30(8): 1999-2003.
doi: 10.3724/SP.J.1146.2007.00040
Abstract:
The main methods to construct P2P overlay network are index and super-node, which introduce bottleneck problem and ignore the significant influence of topology aware problem. A new P2P overlay network TPPH is established. It combines the virtue of efficient locating performance of structured P2P network and complex searching function of unstructured P2P network. Based on physical network topology, it partitioned nodes into topic domain by utilizing distributed hash table mechanism, and according to physical-proximity principle it clustered nodes within a domain. Analysis and simulation results indicate that TPPH can dramatically increase the recall and decrease the searching latency simultaneously, thus it is complex searching available and a high performance semantic overlay network architecture.
The main methods to construct P2P overlay network are index and super-node, which introduce bottleneck problem and ignore the significant influence of topology aware problem. A new P2P overlay network TPPH is established. It combines the virtue of efficient locating performance of structured P2P network and complex searching function of unstructured P2P network. Based on physical network topology, it partitioned nodes into topic domain by utilizing distributed hash table mechanism, and according to physical-proximity principle it clustered nodes within a domain. Analysis and simulation results indicate that TPPH can dramatically increase the recall and decrease the searching latency simultaneously, thus it is complex searching available and a high performance semantic overlay network architecture.
2008, 30(8): 2004-2007.
doi: 10.3724/SP.J.1146.2007.00053
Abstract:
Secure group key distribution and efficient rekeying is one of the most challenging security issues in Ad hoc networks at present. In this paper, Latin squares are used to construct orthogonal arrays, from which t-packing designs can be quickly obtained. Based on cover-free family properties, t-packing designs are adopted in key pre-distribution phase. The new scheme improves the collusion-resilience of the networks using the cover-free family properties, and enhances the key-sharing connectivity of nodes which makes key management more efficient. This paper also presents in depth theory and data analysis of the new scheme in terms of network security and efficiency.
Secure group key distribution and efficient rekeying is one of the most challenging security issues in Ad hoc networks at present. In this paper, Latin squares are used to construct orthogonal arrays, from which t-packing designs can be quickly obtained. Based on cover-free family properties, t-packing designs are adopted in key pre-distribution phase. The new scheme improves the collusion-resilience of the networks using the cover-free family properties, and enhances the key-sharing connectivity of nodes which makes key management more efficient. This paper also presents in depth theory and data analysis of the new scheme in terms of network security and efficiency.
2008, 30(8): 2008-2011.
doi: 10.3724/SP.J.1146.2007.00076
Abstract:
In this paper, a reconfigurable AHB interface component was designed. This component is for AHB slave devices in SOC. Various types of data interfaces such as register, interrupt, SRAM and FIFO are provided with high configurability. The performance and reusability are both considered. This AHB interface component was successfully applied to chips for DAB and DRM receivers. A typical application of this component in DRM receiver has an area of 0.078mm2 in 0.18 m CMOS process.
In this paper, a reconfigurable AHB interface component was designed. This component is for AHB slave devices in SOC. Various types of data interfaces such as register, interrupt, SRAM and FIFO are provided with high configurability. The performance and reusability are both considered. This AHB interface component was successfully applied to chips for DAB and DRM receivers. A typical application of this component in DRM receiver has an area of 0.078mm2 in 0.18 m CMOS process.
2008, 30(8): 2012-2016.
doi: 10.3724/SP.J.1146.2006.02095
Abstract:
A new VLSI architecture of deblocking filter is developed for H.264/AVC system. In the presented architecture, a novel filter scheduling is proposed to reduce the size of local data buffer, and an enhanced data reuse technology is adopted to reduce the number of external memory access, thus the speed of filtering process is significantly improved as well. Whats more, this architecture employs no on-chip SRAM, so there is no on-chip SRAM access. Simulation results show that the new filter can support real-time deblocking for HDTV video application when it works at 100 MHz. The synthesized logic gate count is only 16.8k with 0.18m CMOS technology.
A new VLSI architecture of deblocking filter is developed for H.264/AVC system. In the presented architecture, a novel filter scheduling is proposed to reduce the size of local data buffer, and an enhanced data reuse technology is adopted to reduce the number of external memory access, thus the speed of filtering process is significantly improved as well. Whats more, this architecture employs no on-chip SRAM, so there is no on-chip SRAM access. Simulation results show that the new filter can support real-time deblocking for HDTV video application when it works at 100 MHz. The synthesized logic gate count is only 16.8k with 0.18m CMOS technology.
2008, 30(8): 2017-2020.
doi: 10.3724/SP.J.1146.2006.02089
Abstract:
A low-power VLSI Rake receiver is proposed and realized on FPGA for wireless sensor networks used in complicated wireless environments. Low-power design strategies including reducing clock frequency, sharing of models and dynamic sleeping control are used to reduce the power consumption in order to fit the energy limitations in wireless sensor networks. Simulations and applications show that the receiver can specially reduce VLSI resource and power consumption compared to ordinary Rake receiver.
A low-power VLSI Rake receiver is proposed and realized on FPGA for wireless sensor networks used in complicated wireless environments. Low-power design strategies including reducing clock frequency, sharing of models and dynamic sleeping control are used to reduce the power consumption in order to fit the energy limitations in wireless sensor networks. Simulations and applications show that the receiver can specially reduce VLSI resource and power consumption compared to ordinary Rake receiver.
2008, 30(8): 2021-2024.
doi: 10.3724/SP.J.1146.2007.00121
Abstract:
Volterra-series behavioral model for Radio Frequency (RF) power amplifier is limited to weak nonlinearity because of high computational complexity. In order to reduce computational complexity or the number of coefficient of Volterra-series kernels, the two approaches, based on Chebyshev orthogonal polynomials function and Laguerre orthogonal polynomials function, are proposed, and the mathematical expressions of Volterra-Chebyshev and Volterra-Laguerre behavioral model is derived, and Volterra-laguerre model is simulated. Mathematical analysis and simulation results show that Volterra-Chebyshev and Volterra-Laguerre behavioral model have a simplified structure and reductive coefficients than general Volterra-series model.
Volterra-series behavioral model for Radio Frequency (RF) power amplifier is limited to weak nonlinearity because of high computational complexity. In order to reduce computational complexity or the number of coefficient of Volterra-series kernels, the two approaches, based on Chebyshev orthogonal polynomials function and Laguerre orthogonal polynomials function, are proposed, and the mathematical expressions of Volterra-Chebyshev and Volterra-Laguerre behavioral model is derived, and Volterra-laguerre model is simulated. Mathematical analysis and simulation results show that Volterra-Chebyshev and Volterra-Laguerre behavioral model have a simplified structure and reductive coefficients than general Volterra-series model.
2008, 30(8): 2025-2028.
doi: 10.3724/SP.J.1146.2008.00163
Abstract:
A new Automatic Gain Control (AGC) circuitry is proposed in this paper. The feedback control scheme of subsequent AGC closed-loop in conventional AGC frame is converted to forward control scheme in new circuitry. The power detector and the loop filter are shared by two cascaded AGC loops in new AGC circuitry, so the gain errors in foregoing AGC closed-loop can be corrected by subsequent AGC closed-loop, and total gain errors of new AGC circuitry is determined only by the gain errors in subsequent AGC closed-loop. Simulation and measurement results verify that the new AGC circuitry not only improved the gain control precision, but also decreased the response time, compared with conventional AGC circuitry.
A new Automatic Gain Control (AGC) circuitry is proposed in this paper. The feedback control scheme of subsequent AGC closed-loop in conventional AGC frame is converted to forward control scheme in new circuitry. The power detector and the loop filter are shared by two cascaded AGC loops in new AGC circuitry, so the gain errors in foregoing AGC closed-loop can be corrected by subsequent AGC closed-loop, and total gain errors of new AGC circuitry is determined only by the gain errors in subsequent AGC closed-loop. Simulation and measurement results verify that the new AGC circuitry not only improved the gain control precision, but also decreased the response time, compared with conventional AGC circuitry.
2008, 30(8): 2029-2032.
doi: 10.3724/SP.J.1146.2007.00137
Abstract:
The effect of plated helix on heat dissipation capability of the slow-wave circuit is studied in this paper. With the plated helix, the improvement of the heat dissipation capability of the slow-wave circuit is estimated and evaluated. ANSYS is used to simulate the thermal conduction in the components. The influence of helices coated with copper film, gold film, diamond film upon the heat dissipation is analyzed and verified with computer simulation. Finally, several experimental tests are performed in some slow-wave circuits with copper plated helix and gold plated helix. The tests show good agreement with the theoretical analysis.
The effect of plated helix on heat dissipation capability of the slow-wave circuit is studied in this paper. With the plated helix, the improvement of the heat dissipation capability of the slow-wave circuit is estimated and evaluated. ANSYS is used to simulate the thermal conduction in the components. The influence of helices coated with copper film, gold film, diamond film upon the heat dissipation is analyzed and verified with computer simulation. Finally, several experimental tests are performed in some slow-wave circuits with copper plated helix and gold plated helix. The tests show good agreement with the theoretical analysis.
