Email alert
2013 Vol. 35, No. 5
Display Method:
2013, 35(5): 1017-1022.
doi: 10.3724/SP.J.1146.2012.01219
Abstract:
There exist two problems with the conventional Maximal Likelihood (ML) decoding algorithms: high decoding complexity and large memory space consumption. To solve these problems, a new algorithm that is based on Viterbi and bidirectional searching algorithm is proposed. By comparing the accumulated path metrics of survived paths with the path metric of ML tail-biting path, all of which are obtained in the Viterbi searching phase, the new algorithm deletes impossible starting states and their corresponding sub-tail-biting trellises to reduce the searching space for the second phase. In the second phase, the decoding complexity can be further reduced by comparing the path metric of ML tail-biting path with the threshold used in the bidirectional searching algorithm. Combing the Viterbi algorithm and bidirectional searching algorithm, a new ML decoding algorithm for tail-biting codes, which can be performed on tail-biting trellis with high efficiency, is obtained. The results of experiments show that the new algorithm improves the decoding efficiency and reduces the memory space consumption.
There exist two problems with the conventional Maximal Likelihood (ML) decoding algorithms: high decoding complexity and large memory space consumption. To solve these problems, a new algorithm that is based on Viterbi and bidirectional searching algorithm is proposed. By comparing the accumulated path metrics of survived paths with the path metric of ML tail-biting path, all of which are obtained in the Viterbi searching phase, the new algorithm deletes impossible starting states and their corresponding sub-tail-biting trellises to reduce the searching space for the second phase. In the second phase, the decoding complexity can be further reduced by comparing the path metric of ML tail-biting path with the threshold used in the bidirectional searching algorithm. Combing the Viterbi algorithm and bidirectional searching algorithm, a new ML decoding algorithm for tail-biting codes, which can be performed on tail-biting trellis with high efficiency, is obtained. The results of experiments show that the new algorithm improves the decoding efficiency and reduces the memory space consumption.
2013, 35(5): 1023-1030.
doi: 10.3724/SP.J.1146.2012.01180
Abstract:
The existing resource allocation research for OFDMA relay networks investigate only frequency-domain allocation and power allocation with fixed time-domain allocation, which can not fit the user service changes in time-domain. An optimal energy-efficient resource allocation strategy and its simplified version are proposed for OFDMA relay network. A generalized model for resource allocation issue, which dynamically allocates time-domain, frequency-domain resources and power resources, is established. Due to the strong flexibility and adaptability, the model may apply to not only the fixed time-domain allocation system, but also the non-fixed time-domain allocation system. For non-full-buffer traffic, to ensure the users quality of service, an energy-efficiency maximization model is established for OFDMA non-cooperative relay networks by using the Lagrange multiplier algorithm to solve the issue. Taking into account the complexity of the algorithm, a simplified resource allocation strategy is proposed by using the Hungarian method. The theory and simulation results show that the optimal algorithm can get the energy-efficiency maximization, and the energy-efficiency difference between the simplified algorithm and the optimal algorithm is less than 5%, however, the complexity has been significantly reduced. Moreover, the dynamic time-domain allocation has more adaptive in uneven user distribution or link distribution than fixed time-domain allocation.
The existing resource allocation research for OFDMA relay networks investigate only frequency-domain allocation and power allocation with fixed time-domain allocation, which can not fit the user service changes in time-domain. An optimal energy-efficient resource allocation strategy and its simplified version are proposed for OFDMA relay network. A generalized model for resource allocation issue, which dynamically allocates time-domain, frequency-domain resources and power resources, is established. Due to the strong flexibility and adaptability, the model may apply to not only the fixed time-domain allocation system, but also the non-fixed time-domain allocation system. For non-full-buffer traffic, to ensure the users quality of service, an energy-efficiency maximization model is established for OFDMA non-cooperative relay networks by using the Lagrange multiplier algorithm to solve the issue. Taking into account the complexity of the algorithm, a simplified resource allocation strategy is proposed by using the Hungarian method. The theory and simulation results show that the optimal algorithm can get the energy-efficiency maximization, and the energy-efficiency difference between the simplified algorithm and the optimal algorithm is less than 5%, however, the complexity has been significantly reduced. Moreover, the dynamic time-domain allocation has more adaptive in uneven user distribution or link distribution than fixed time-domain allocation.
2013, 35(5): 1031-1036.
doi: 10.3724/SP.J.1146.2012.01120
Abstract:
This paper considers MIMO cellular downlink model with 3 cells and 2 edge users in each cell. By mean of orthogonal-triangular decomposes of the joint channel matrix, the equivalent system model is derived, which has the same degrees of freedom with the original model. Based on the new model, the paper proposes a linear interference alignment algorithm to align and eliminate inter-cell and intra-cell interference according to different interferences from different cells. Before the signals have not received by the users, the proposed algorithm takes advantages of the channel decomposition to eliminate half of inter-cell interference, so as to reduce the number of the receiving matrix and simplify the complexity in receivers. The simulation results show that for given the antenna configuration, system capacity and energy efficiency have significantly improved with the increase in the data stream. In the case of the same degree of freedom, the number of antennas has a greater impact on system capacity and energy efficiency.
This paper considers MIMO cellular downlink model with 3 cells and 2 edge users in each cell. By mean of orthogonal-triangular decomposes of the joint channel matrix, the equivalent system model is derived, which has the same degrees of freedom with the original model. Based on the new model, the paper proposes a linear interference alignment algorithm to align and eliminate inter-cell and intra-cell interference according to different interferences from different cells. Before the signals have not received by the users, the proposed algorithm takes advantages of the channel decomposition to eliminate half of inter-cell interference, so as to reduce the number of the receiving matrix and simplify the complexity in receivers. The simulation results show that for given the antenna configuration, system capacity and energy efficiency have significantly improved with the increase in the data stream. In the case of the same degree of freedom, the number of antennas has a greater impact on system capacity and energy efficiency.
2013, 35(5): 1037-1043.
doi: 10.3724/SP.J.1146.2012.01727
Abstract:
In order to solve over-segmentation issue and improve computing efficiency, this paper proposes a moving object segmentation model using binary level set based on shape constraint and local curve evolution. Firstly, the model introduces priori shape information in the traditional level set model to constrain segmentation, and the shape is obtained by object detection. Then, to improve efficiency the proposed model uses a binary level set function to replace the traditional level set function. Furthermore, the paper proposes a method of local curve evolution to address the lack of gradual progress in binary level set curve evolution. Finally, the experimental results show that an obvious performance improvement on segmentation could be obtained through the algorithm.
In order to solve over-segmentation issue and improve computing efficiency, this paper proposes a moving object segmentation model using binary level set based on shape constraint and local curve evolution. Firstly, the model introduces priori shape information in the traditional level set model to constrain segmentation, and the shape is obtained by object detection. Then, to improve efficiency the proposed model uses a binary level set function to replace the traditional level set function. Furthermore, the paper proposes a method of local curve evolution to address the lack of gradual progress in binary level set curve evolution. Finally, the experimental results show that an obvious performance improvement on segmentation could be obtained through the algorithm.
2013, 35(5): 1044-1048.
doi: 10.3724/SP.J.1146.2012.01257
Abstract:
Let R denote the ring R=Fq+uFq++uk1Fq , and be an invertible element of R. By means of the theory of ring homomorphism, the generators of all these (u1)-constacyclic codes of an arbitrary length N over the ring R are obtained. It is proved thatR[x]xN+1u is principal. The number of these (u1)-constacyclic codes is determined. The generator polynomials of the highest-order torsion codes of all these (u1)-constacyclic codes are given. As a result, the Hamming distances of all these (u1)-constacyclic codes are obtained.
Let R denote the ring R=Fq+uFq++uk1Fq , and be an invertible element of R. By means of the theory of ring homomorphism, the generators of all these (u1)-constacyclic codes of an arbitrary length N over the ring R are obtained. It is proved thatR[x]xN+1u is principal. The number of these (u1)-constacyclic codes is determined. The generator polynomials of the highest-order torsion codes of all these (u1)-constacyclic codes are given. As a result, the Hamming distances of all these (u1)-constacyclic codes are obtained.
2013, 35(5): 1049-1054.
doi: 10.3724/SP.J.1146.2012.01080
Abstract:
The design of shift sequences is the key issue of the Zero Correlation Zone (ZCZ) construction based on interleaving technique. The restriction condition is analyzed according to the correlation function and accordingly a method of ZCZ sequence set construction based on interleaving is derived. Sequence pairs with flexible ZCZ are brought out by using diverse base sequence and shift sequences which satisfy the restrict condition. ZCZ sequence pairs set with certain length and volume is constructed by uniting the new ZCZ sequence pair and orthogonal matrix. This method extends the utilization of interleaving technique which can generate sequence pairs set with some volume and flexible ZCZ.
The design of shift sequences is the key issue of the Zero Correlation Zone (ZCZ) construction based on interleaving technique. The restriction condition is analyzed according to the correlation function and accordingly a method of ZCZ sequence set construction based on interleaving is derived. Sequence pairs with flexible ZCZ are brought out by using diverse base sequence and shift sequences which satisfy the restrict condition. ZCZ sequence pairs set with certain length and volume is constructed by uniting the new ZCZ sequence pair and orthogonal matrix. This method extends the utilization of interleaving technique which can generate sequence pairs set with some volume and flexible ZCZ.
2013, 35(5): 1055-1062.
doi: 10.3724/SP.J.1146.2012.00415
Abstract:
According to the mapping relationship between the multiple secret images pixel combinations and the basis matrices, the redundant basis matrices are analyzed in the access-based multi-secret visual cryptography. A compression algorithm is proposed to decrease the size of basis matrices. The algorithm takes one column pixels as disposal unit, and satisfies the entire contrast of secret images. Based on the algorithm, new secret sharing and recovering procedures are designed for the access-based multi-secret visual cryptography. Compared with previous schemes, the present scheme can diminish the size of shares effectively, and the compression effects are obvious for the simple images.
According to the mapping relationship between the multiple secret images pixel combinations and the basis matrices, the redundant basis matrices are analyzed in the access-based multi-secret visual cryptography. A compression algorithm is proposed to decrease the size of basis matrices. The algorithm takes one column pixels as disposal unit, and satisfies the entire contrast of secret images. Based on the algorithm, new secret sharing and recovering procedures are designed for the access-based multi-secret visual cryptography. Compared with previous schemes, the present scheme can diminish the size of shares effectively, and the compression effects are obvious for the simple images.
2013, 35(5): 1063-1068.
doi: 10.3724/SP.J.1146.2012.01213
Abstract:
This paper firstly gives the formalization description of both players strategies and payoffs in the mimicry honeypot game, and constructs the payoff matrix of the fraudulent game using non-cooperative and incomplete dynamic game theory. Then the equilibrium strategies and the equilibrium conditions are inferred. The equilibrium conditions and relative factors are discussed in detail, and the comparison to traditional honeypot is also performed. The theoretic analysis depicts the effective condition for protective coloration and warning coloration mechanism in the fraudulent game, and demonstrates that the mimicry honeypot has better activeness, efficiency and fraudulence than the traditional scheme.
This paper firstly gives the formalization description of both players strategies and payoffs in the mimicry honeypot game, and constructs the payoff matrix of the fraudulent game using non-cooperative and incomplete dynamic game theory. Then the equilibrium strategies and the equilibrium conditions are inferred. The equilibrium conditions and relative factors are discussed in detail, and the comparison to traditional honeypot is also performed. The theoretic analysis depicts the effective condition for protective coloration and warning coloration mechanism in the fraudulent game, and demonstrates that the mimicry honeypot has better activeness, efficiency and fraudulence than the traditional scheme.
2013, 35(5): 1069-1075.
doi: 10.3724/SP.J.1146.2012.01053
Abstract:
The expired-time packet loss rate is a common parameter to measure the quality of service of transmission of deadline-constrained traffic in wireless networks. However, this parameter can not reflect the influence on the quality of service of video caused by the packets with different levels of importance. In this paper, the definition of the weighted expired-time packet loss rate is first introduced with the consideration of the importance of different kinds of packets, which describes the connection between the packet loss and quality of video more specifically. Based on the definition, an active queue management mechanism is proposed, which can be applied to deadline-constrained transmissions in wireless networks, and this mechanism, considering different levels of importance of video packets, drop packets actively to minimize the weighted expired-time packet loss rate during the service. Simulation results show that compared with the traditional queue management mechanism and real-time video filtering mechanism, the active packet discard mechanism can effectively reduce the weighted expired-time packet loss rate and improve the quality of service of video to about 0.5~1.5 dB Peak Signal to Noise Ratio (PSNR) gain.
The expired-time packet loss rate is a common parameter to measure the quality of service of transmission of deadline-constrained traffic in wireless networks. However, this parameter can not reflect the influence on the quality of service of video caused by the packets with different levels of importance. In this paper, the definition of the weighted expired-time packet loss rate is first introduced with the consideration of the importance of different kinds of packets, which describes the connection between the packet loss and quality of video more specifically. Based on the definition, an active queue management mechanism is proposed, which can be applied to deadline-constrained transmissions in wireless networks, and this mechanism, considering different levels of importance of video packets, drop packets actively to minimize the weighted expired-time packet loss rate during the service. Simulation results show that compared with the traditional queue management mechanism and real-time video filtering mechanism, the active packet discard mechanism can effectively reduce the weighted expired-time packet loss rate and improve the quality of service of video to about 0.5~1.5 dB Peak Signal to Noise Ratio (PSNR) gain.
2013, 35(5): 1076-1082.
doi: 10.3724/SP.J.1146.2012.01048
Abstract:
Considering the control channel bottleneck issue and hidden terminal issue for multi-channel MAC in Ad-hoc networks, a new multi-channel MAC with the Low Control Overhead (LCO-MAC) is proposed. Different from the multi-channel MAC mechanism based on the channel usage table, for LCO-MAC reference to the Request To Send/Clear To Send (RTS/CTS) channel assignment mechanism proposed by Meshhadany, each data channel is mapped to a time slot of the control section, but the difference is that LCO-MAC does not restrict the sending time of the RTS, and can send data once obtainning the channel. Simulation results show that LCO-MAC not needing too much control information for channel reservation provides an effective solution to the control channel bottleneck and multi-channel hidden terminal issue, the network throughput is also significantly enhanced.
Considering the control channel bottleneck issue and hidden terminal issue for multi-channel MAC in Ad-hoc networks, a new multi-channel MAC with the Low Control Overhead (LCO-MAC) is proposed. Different from the multi-channel MAC mechanism based on the channel usage table, for LCO-MAC reference to the Request To Send/Clear To Send (RTS/CTS) channel assignment mechanism proposed by Meshhadany, each data channel is mapped to a time slot of the control section, but the difference is that LCO-MAC does not restrict the sending time of the RTS, and can send data once obtainning the channel. Simulation results show that LCO-MAC not needing too much control information for channel reservation provides an effective solution to the control channel bottleneck and multi-channel hidden terminal issue, the network throughput is also significantly enhanced.
2013, 35(5): 1083-1089.
doi: 10.3724/SP.J.1146.2012.00344
Abstract:
With the increasing number of new protocols and the rapid growth of the network link rate, the packet parsing architecture has been greatly challenged on its flexibility and rate. While combining the idea of pipeline design and binary-trie, a new parsing architecture is proposed in this paper, namely Parsing Pipeline Architecture for Forwarding (PPAF). It flexibly analysis packet protocol by constructing Forwarding Protocol-trie, improved the processing rate by employing hardware pipeline look-up table, and solved the unbalance of node mapping storage resource by using the node to pipeline mapping algorithm. The simulation results through the NetFPGA platform suggest that PPAF is superior than the extant high speed parsing architecture in two ways: PPAF achieves ambidexterity in processing speed and resource consumption; and it can provide independent interface-based flexible protocol parsing capabilities.
With the increasing number of new protocols and the rapid growth of the network link rate, the packet parsing architecture has been greatly challenged on its flexibility and rate. While combining the idea of pipeline design and binary-trie, a new parsing architecture is proposed in this paper, namely Parsing Pipeline Architecture for Forwarding (PPAF). It flexibly analysis packet protocol by constructing Forwarding Protocol-trie, improved the processing rate by employing hardware pipeline look-up table, and solved the unbalance of node mapping storage resource by using the node to pipeline mapping algorithm. The simulation results through the NetFPGA platform suggest that PPAF is superior than the extant high speed parsing architecture in two ways: PPAF achieves ambidexterity in processing speed and resource consumption; and it can provide independent interface-based flexible protocol parsing capabilities.
2013, 35(5): 1090-1096.
doi: 10.3724/SP.J.1146.2012.01207
Abstract:
The ordinary SAR imaging algorithms are inapplicable to missile-borne terminal guidance SAR due to its properties of high-speed, dive and high-squint. Based on the precise SAR echo signal in dive model, this paper proposes a new imaging algorithm of pre-correcting range-walk in frequency domain to translate high-squint imaging into equivalent low-squint even zero-squint one to simplify imaging processing. The pre-correction of range-walk will cause the variation of Doppler rate with azimuth time, which can be solved effectively by the azimuth Non-Linear Chirp Scaling (NLCS) processing. Therefore, after the azimuth NLCS processing, the azimuth compression in the same range gate can be carried out with the same Doppler parameters so that the azimuth Depth of Focus (DOF) can be improved. Simulation experiments validate the effectiveness of this algorithm.
The ordinary SAR imaging algorithms are inapplicable to missile-borne terminal guidance SAR due to its properties of high-speed, dive and high-squint. Based on the precise SAR echo signal in dive model, this paper proposes a new imaging algorithm of pre-correcting range-walk in frequency domain to translate high-squint imaging into equivalent low-squint even zero-squint one to simplify imaging processing. The pre-correction of range-walk will cause the variation of Doppler rate with azimuth time, which can be solved effectively by the azimuth Non-Linear Chirp Scaling (NLCS) processing. Therefore, after the azimuth NLCS processing, the azimuth compression in the same range gate can be carried out with the same Doppler parameters so that the azimuth Depth of Focus (DOF) can be improved. Simulation experiments validate the effectiveness of this algorithm.
2013, 35(5): 1097-1102.
doi: 10.3724/SP.J.1146.2012.01136
Abstract:
The raw data of Synthetic Aperture Radar (SAR) is complex value, as well as the image data. Due to the random phase of each scattering cell, continuous scene has a wide signal band which makes single-aperture SAR difficult to realize down-sampling. In this paper, imaging model of cross-track sparse array SAR is analyzed. Signal reconstruction based imaging method under cross-track multi-aperture structure is investigated to eliminate random phases of scatters and reduce the signal bandwidth. Consequently, sparse array side-looking 3D imaging with larger interval space sampling is realized. The results on InSAR 2D real data verify that the bandwidth can be reduced after signal reconstruction. Besides, the simulation experiments on 3D imaging validate the effectiveness of the proposed method.
The raw data of Synthetic Aperture Radar (SAR) is complex value, as well as the image data. Due to the random phase of each scattering cell, continuous scene has a wide signal band which makes single-aperture SAR difficult to realize down-sampling. In this paper, imaging model of cross-track sparse array SAR is analyzed. Signal reconstruction based imaging method under cross-track multi-aperture structure is investigated to eliminate random phases of scatters and reduce the signal bandwidth. Consequently, sparse array side-looking 3D imaging with larger interval space sampling is realized. The results on InSAR 2D real data verify that the bandwidth can be reduced after signal reconstruction. Besides, the simulation experiments on 3D imaging validate the effectiveness of the proposed method.
2013, 35(5): 1103-1107.
doi: 10.3724/SP.J.1146.2012.01122
Abstract:
Shadow Inverse Synthetic Aperture Radar (SISAR) imaging algorithm is a kind of technology to extract the profile of a target from the forward scattering signal. The imaging accuracy depends on the target coordinate estimation precision. In this paper, the relationship between the imaging error and coordinate estimation errors is analytically derived with the forward scattering signal model under the assumption of small diffraction angles. The quantitative requirements for coordinate estimation precision are made. Finally, impacts of coordinate estimation errors on SISAR imaging accuracy are verified via simulation results.
Shadow Inverse Synthetic Aperture Radar (SISAR) imaging algorithm is a kind of technology to extract the profile of a target from the forward scattering signal. The imaging accuracy depends on the target coordinate estimation precision. In this paper, the relationship between the imaging error and coordinate estimation errors is analytically derived with the forward scattering signal model under the assumption of small diffraction angles. The quantitative requirements for coordinate estimation precision are made. Finally, impacts of coordinate estimation errors on SISAR imaging accuracy are verified via simulation results.
2013, 35(5): 1108-1113.
doi: 10.3724/SP.J.1146.2012.01148
Abstract:
A novel passive radar imaging time-domain algorithm is presented based on single-frequency continuous wave. The signal matched matrix is builded firstly, and the echo signal is expanded into a echo signal matrix. Finally the sum in time dimension of the Hadamard product of the signal matched matrix and the echo signal matrix can be got. Then the echo signal matched integration is implemented and the scattering points are focused. The algorithm resolution and factors of affecting the imaging quality are analyzed. The algorithm is performed in time domain and the interpolation in frequency domain is avoided. Based on the theoretical analysis, simulation experiment confirms the validity of the proposed algorithm.
A novel passive radar imaging time-domain algorithm is presented based on single-frequency continuous wave. The signal matched matrix is builded firstly, and the echo signal is expanded into a echo signal matrix. Finally the sum in time dimension of the Hadamard product of the signal matched matrix and the echo signal matrix can be got. Then the echo signal matched integration is implemented and the scattering points are focused. The algorithm resolution and factors of affecting the imaging quality are analyzed. The algorithm is performed in time domain and the interpolation in frequency domain is avoided. Based on the theoretical analysis, simulation experiment confirms the validity of the proposed algorithm.
2013, 35(5): 1114-1119.
doi: 10.3724/SP.J.1146.2012.01056
Abstract:
Through-wall-radar imaging is performed to image the architectural layout imaging and hidden target of building. Wall images formed through the architectural layout imaging are considered as the reference to define the relative position of the hidden target in building, which further contributes to implement wall phase compensation and multi-path suppression. In this paper, a multi-view imaging approach for the architectural layout is proposed. Firstly, multiple single-view layout images with the part of wall images are formed from multi-view synthetic aperture echoes through back-projection imaging algorithm. Then based on M-N-K detector, these single-view images are fused to one binary layout image with all wall images. Finally, median filter and Hough transformation are adopted to reduce the cavities and burrs in wall images to generate the near-tidy panorama image of architectural layout. XFDTD simulation and through-wall-radar experiment data validate the correctness and feasibility of the proposed approach.
Through-wall-radar imaging is performed to image the architectural layout imaging and hidden target of building. Wall images formed through the architectural layout imaging are considered as the reference to define the relative position of the hidden target in building, which further contributes to implement wall phase compensation and multi-path suppression. In this paper, a multi-view imaging approach for the architectural layout is proposed. Firstly, multiple single-view layout images with the part of wall images are formed from multi-view synthetic aperture echoes through back-projection imaging algorithm. Then based on M-N-K detector, these single-view images are fused to one binary layout image with all wall images. Finally, median filter and Hough transformation are adopted to reduce the cavities and burrs in wall images to generate the near-tidy panorama image of architectural layout. XFDTD simulation and through-wall-radar experiment data validate the correctness and feasibility of the proposed approach.
2013, 35(5): 1120-1127.
doi: 10.3724/SP.J.1146.2012.01198
Abstract:
A SAR image detection method is proposed based on local hybrid filter. This method combines the advantages of both local hybrid filter based NonSubsampled Direction Filter Bank-Dual-Tree Complex Wavelet Transform (NSDFB-DTCWT) and Dempster-Shafet (DS) evidence theory. First, local hybrid filter is applied to SAR image. Then edge strength is got by using Ratio Of Exponentially Weighted Averages (ROEWA) operator and edge direction is got by using Canny operator at every scale from NSDFB-DTCWT. Finally, DS theory is used to fuse the edge in all scales to get the whole edge of the original SAR image. The experimental results demonstrate its effectiveness and superiority in terms of edge positioning accuracy and integrity, and it also has a fewer false edge points.
A SAR image detection method is proposed based on local hybrid filter. This method combines the advantages of both local hybrid filter based NonSubsampled Direction Filter Bank-Dual-Tree Complex Wavelet Transform (NSDFB-DTCWT) and Dempster-Shafet (DS) evidence theory. First, local hybrid filter is applied to SAR image. Then edge strength is got by using Ratio Of Exponentially Weighted Averages (ROEWA) operator and edge direction is got by using Canny operator at every scale from NSDFB-DTCWT. Finally, DS theory is used to fuse the edge in all scales to get the whole edge of the original SAR image. The experimental results demonstrate its effectiveness and superiority in terms of edge positioning accuracy and integrity, and it also has a fewer false edge points.
2013, 35(5): 1128-1134.
doi: 10.3724/SP.J.1146.2012.01663
Abstract:
In the Inverse Synthetic Aperture Radar (ISAR) imaging of targets with rotating parts, the Doppler centroid of the radar returns are disturbed by the micro-Doppler of the rotating parts. As a result, the phase estimation and compensation accuracy will decline and the image of the targets main body may be unfocused. To solve this problem, an algorithm based on adaptive range bins selection in the fast minimum entropy phase compensation is presented. By adaptively selecting range bins which only contain the returns of the main body scatterers, the disturbance from the rotating parts can be avoided and the computational efficiency is improved. The result from a set of measured data is given to verify the validity of this algorithm.
In the Inverse Synthetic Aperture Radar (ISAR) imaging of targets with rotating parts, the Doppler centroid of the radar returns are disturbed by the micro-Doppler of the rotating parts. As a result, the phase estimation and compensation accuracy will decline and the image of the targets main body may be unfocused. To solve this problem, an algorithm based on adaptive range bins selection in the fast minimum entropy phase compensation is presented. By adaptively selecting range bins which only contain the returns of the main body scatterers, the disturbance from the rotating parts can be avoided and the computational efficiency is improved. The result from a set of measured data is given to verify the validity of this algorithm.
2013, 35(5): 1135-1141.
doi: 10.3724/SP.J.1146.2012.01157
Abstract:
Doppler rate estimation is an essential procedure in Synthetic Aperture Radar (SAR) signal processing. For the Doppler rate estimation in spotlight and sliding spotlight SAR focused with the full-aperture imaging algorithm, this paper proposes an improved Shift-And-Correlation (SAC) approach. The range-Keystone transform is introduced into the original SAC algorithm to eliminate the coupling effect between the cross-correlation peak position and the targets distance. Thus, the constraint of the focus depth can be avoided. Due to no zero padding in the proposed approach, aliasing will occur in the Doppler rate estimation. For this problem, an ambiguity number estimation method based on minimum entropy is presented. Finally, both simulation and results of real data are provided to demonstrate the effectiveness of the proposed approach.
Doppler rate estimation is an essential procedure in Synthetic Aperture Radar (SAR) signal processing. For the Doppler rate estimation in spotlight and sliding spotlight SAR focused with the full-aperture imaging algorithm, this paper proposes an improved Shift-And-Correlation (SAC) approach. The range-Keystone transform is introduced into the original SAC algorithm to eliminate the coupling effect between the cross-correlation peak position and the targets distance. Thus, the constraint of the focus depth can be avoided. Due to no zero padding in the proposed approach, aliasing will occur in the Doppler rate estimation. For this problem, an ambiguity number estimation method based on minimum entropy is presented. Finally, both simulation and results of real data are provided to demonstrate the effectiveness of the proposed approach.
2013, 35(5): 1142-1148.
doi: 10.3724/SP.J.1146.2012.01070
Abstract:
To solve the problem of target localization with sparse array in bistatic MIMO radar, a projection and Singular Value Decomposition (SVD) based Regularized Multi-vectors FOCal Undetermined System Solver (RMFOCUSS) algorithm is proposed. First the target angles with respect to receive array are estimated, and then the echoed signal is projected back to them. After an rearrangement of the projected signal, the target angles with respect to transmit array are estimated, so targets are located. SVD is utilized to reduce signal dimension and accumulate signal power, which makes traditional Compressive Sensing (CS) recovery algorithms perform better under low SNR, and computational complexity is reduced even more. Compared with existing sparse reconstruction approaches, the proposed method costs much less computation time in coping with large two dimensional scene and maintains a good performance whether the targets are relative or not.
To solve the problem of target localization with sparse array in bistatic MIMO radar, a projection and Singular Value Decomposition (SVD) based Regularized Multi-vectors FOCal Undetermined System Solver (RMFOCUSS) algorithm is proposed. First the target angles with respect to receive array are estimated, and then the echoed signal is projected back to them. After an rearrangement of the projected signal, the target angles with respect to transmit array are estimated, so targets are located. SVD is utilized to reduce signal dimension and accumulate signal power, which makes traditional Compressive Sensing (CS) recovery algorithms perform better under low SNR, and computational complexity is reduced even more. Compared with existing sparse reconstruction approaches, the proposed method costs much less computation time in coping with large two dimensional scene and maintains a good performance whether the targets are relative or not.
2013, 35(5): 1149-1155.
doi: 10.3724/SP.J.1146.2012.01167
Abstract:
This paper presents an angle estimation method for multitargets in bistatic MIMO radar system. It can estimate DOD and DOA with single snapshot in time domain. By exploiting transmisson diversity smoothing to the equivalent receiving array data, the rank of the auto-covariance matrix of the multitargets scatters is recovered, which make DOA estimation go well. Similarly, receiving diversity smoothing to the equivalent transmitting array data is implemented to facilitate the DOD estimation. A matching method for the two kinds of angle based on the maximum likelihood principle is employed. In this way, only one snapshot of echo data decorrelates the coherence among the multitargets scatters successfully. In addition, the method is not limited to any certain array shape. Numerical results are presented to demonstrate the effectiveness of the method.
This paper presents an angle estimation method for multitargets in bistatic MIMO radar system. It can estimate DOD and DOA with single snapshot in time domain. By exploiting transmisson diversity smoothing to the equivalent receiving array data, the rank of the auto-covariance matrix of the multitargets scatters is recovered, which make DOA estimation go well. Similarly, receiving diversity smoothing to the equivalent transmitting array data is implemented to facilitate the DOD estimation. A matching method for the two kinds of angle based on the maximum likelihood principle is employed. In this way, only one snapshot of echo data decorrelates the coherence among the multitargets scatters successfully. In addition, the method is not limited to any certain array shape. Numerical results are presented to demonstrate the effectiveness of the method.
2013, 35(5): 1156-1162.
doi: 10.3724/SP.J.1146.2012.01134
Abstract:
Considering the issue of low computation efficiency of Ground Penetrating Radar (GPR) object detection method, a fast algorithm is proposed based on the spatial distribution of reflected energy, which is modeled as a three-parameter hyperbola. A reflection hyperbola is first extracted by utilizing the correlation between adjacent reflections. Being weighted with reflection energy, the hyperbola is then fitted to estimate the two parameters of the reflected model. Finally, the object detection and localization task is completed with one dimensional Hough transform. The efficiency of the proposed algorithm is demonstrated both theoretically and experimentally. Compared with the traditional algorithm based on Hough transform, the proposed algorithm consumes only 1.5% computational time without sacrificing the detection and localization performance.
Considering the issue of low computation efficiency of Ground Penetrating Radar (GPR) object detection method, a fast algorithm is proposed based on the spatial distribution of reflected energy, which is modeled as a three-parameter hyperbola. A reflection hyperbola is first extracted by utilizing the correlation between adjacent reflections. Being weighted with reflection energy, the hyperbola is then fitted to estimate the two parameters of the reflected model. Finally, the object detection and localization task is completed with one dimensional Hough transform. The efficiency of the proposed algorithm is demonstrated both theoretically and experimentally. Compared with the traditional algorithm based on Hough transform, the proposed algorithm consumes only 1.5% computational time without sacrificing the detection and localization performance.
2013, 35(5): 1163-1169.
doi: 10.3724/SP.J.1146.2012.01249
Abstract:
The fourth generation fighters have the feature of stealth, supersonic and maneuvering, and detection performance of the fighters with traditional radar is bad. Considering the metric-band phased array radar, and using its feature of anti-stealth and agile beam forming ability, a novel long term coherent integration detection algorithm of the fourth generation fighters is proposed. First the modified Keystone transform is used to correct the range cell migration, and then each range cell data is transformed to the time-frequency plane, and Sandglass transform is used to solve the coupling effect between the delay and slow time of the bilinear transformation. Finally, the two dimensional FFT is performed to complete the long term coherent integration and target detection, and the effectiveness of the algorithm is verified by Monte Carlo simulation. At the same time, the algorithm does not need to search, and can be realized by high efficient operation.
The fourth generation fighters have the feature of stealth, supersonic and maneuvering, and detection performance of the fighters with traditional radar is bad. Considering the metric-band phased array radar, and using its feature of anti-stealth and agile beam forming ability, a novel long term coherent integration detection algorithm of the fourth generation fighters is proposed. First the modified Keystone transform is used to correct the range cell migration, and then each range cell data is transformed to the time-frequency plane, and Sandglass transform is used to solve the coupling effect between the delay and slow time of the bilinear transformation. Finally, the two dimensional FFT is performed to complete the long term coherent integration and target detection, and the effectiveness of the algorithm is verified by Monte Carlo simulation. At the same time, the algorithm does not need to search, and can be realized by high efficient operation.
2013, 35(5): 1170-1176.
doi: 10.3724/SP.J.1146.2012.01069
Abstract:
Combining with Edge Strength Map (ESM), an unsupervised estimation of the Equivalent Number of Looks (ENL) in SAR images is proposed. First, ESM is produced by ratio operation of anisotropic Gaussian kernel parallel windows. On the basis of simple image partition processing, the local ESM thresholds for image blocks are obtained by an effective unsupervised estimation method, and SAR image is divided into homogeneous regions and edge regions by shrinkage ESM. Under twice restrictions based on ESM, each local ENL of all pixels in homogeneous regions is estimated by irregular window with large scale. Then global ENL of SAR image is acquired by histogram statistical method for all local ENL values. Experimental results show that the proposed method can get over serious adverse effects to estimation value by the selection of local statistics methods in estimation methods based on spatial statics, and can avoid the need for apriority information on speckle and being an ineffective estimation for despeckling SAR images in estimation methods based on statistical model.
Combining with Edge Strength Map (ESM), an unsupervised estimation of the Equivalent Number of Looks (ENL) in SAR images is proposed. First, ESM is produced by ratio operation of anisotropic Gaussian kernel parallel windows. On the basis of simple image partition processing, the local ESM thresholds for image blocks are obtained by an effective unsupervised estimation method, and SAR image is divided into homogeneous regions and edge regions by shrinkage ESM. Under twice restrictions based on ESM, each local ENL of all pixels in homogeneous regions is estimated by irregular window with large scale. Then global ENL of SAR image is acquired by histogram statistical method for all local ENL values. Experimental results show that the proposed method can get over serious adverse effects to estimation value by the selection of local statistics methods in estimation methods based on spatial statics, and can avoid the need for apriority information on speckle and being an ineffective estimation for despeckling SAR images in estimation methods based on statistical model.
2013, 35(5): 1177-1184.
doi: 10.3724/SP.J.1146.2012.01082
Abstract:
In order to reduce the computation complexity resulted from large number of spectral information and to reduce the decline of classification performance resulted from data redundancy, a dimensionality reduction algorithm called non-negative sparse graph is proposed. At first, an over-complete block dictionary is constructed to realize the non-negative sparse representation of high-dimensional hyperspectral data. Then, according to the non-negative sparse representation, an inner non-negative sparsity graph and a penalty non-negative sparsity graph are built where the weights of edges are defined by a monotone decreasing function to embody the similarity degree of samples. At last, an optimal mapping from the high-dimensional space to a low-dimensional subspace can be obtained by simultaneously maximizing the distance between non-negative sparsity reconstruction samples of different classes and minimizing the distance between non-negative sparsity reconstruction samples of the same class, which makes the dimensionality reduction of high-dimensional hyperspectral data realized. Experimental results on AVIRIS 92AV3C hyperspectral data show that the proposed algorithm can obtain higher overall accuracy and Kappa coefficient with few training samples.
In order to reduce the computation complexity resulted from large number of spectral information and to reduce the decline of classification performance resulted from data redundancy, a dimensionality reduction algorithm called non-negative sparse graph is proposed. At first, an over-complete block dictionary is constructed to realize the non-negative sparse representation of high-dimensional hyperspectral data. Then, according to the non-negative sparse representation, an inner non-negative sparsity graph and a penalty non-negative sparsity graph are built where the weights of edges are defined by a monotone decreasing function to embody the similarity degree of samples. At last, an optimal mapping from the high-dimensional space to a low-dimensional subspace can be obtained by simultaneously maximizing the distance between non-negative sparsity reconstruction samples of different classes and minimizing the distance between non-negative sparsity reconstruction samples of the same class, which makes the dimensionality reduction of high-dimensional hyperspectral data realized. Experimental results on AVIRIS 92AV3C hyperspectral data show that the proposed algorithm can obtain higher overall accuracy and Kappa coefficient with few training samples.
2013, 35(5): 1185-1189.
doi: 10.3724/SP.J.1146.2012.01096
Abstract:
To solve the problem that the Synthetic Aperture Sonar (SAS) motion errors distinct variety in large slant range may reduce the imaging resolution, this paper proposes an improved motion error fitting and compensation method based on segment Displaced Phase Center (DPC) algorithm and motion errors fitting. Firstly the raw data are segmented in slant range and each segments motion error is estimated using the DPC algorithm. Then the sway and heave are estimated using the Least-Square (LS) estimation method. Finally, the raw data are compensated point by point in slant range with the estimated sway and heave. The proposed algorithm is applied to the simulated data and lake-trial dataset, and the result confirms that the proposed method improves the SAS imaging resolution.
To solve the problem that the Synthetic Aperture Sonar (SAS) motion errors distinct variety in large slant range may reduce the imaging resolution, this paper proposes an improved motion error fitting and compensation method based on segment Displaced Phase Center (DPC) algorithm and motion errors fitting. Firstly the raw data are segmented in slant range and each segments motion error is estimated using the DPC algorithm. Then the sway and heave are estimated using the Least-Square (LS) estimation method. Finally, the raw data are compensated point by point in slant range with the estimated sway and heave. The proposed algorithm is applied to the simulated data and lake-trial dataset, and the result confirms that the proposed method improves the SAS imaging resolution.
2013, 35(5): 1190-1195.
doi: 10.3724/SP.J.1146.2012.01459
Abstract:
This paper presents a novel multiscale variational image decomposition model. Based on the hierarchical multiscale variational model of Tadmor, a novel (BV, H-1) hierarchical multiscale image decomposition method is proposed, then the novel Integro-Differential Equation (IDE) is obtained by integrating in inverse scale space a succession of refined slices of the image and balancing a Laplacian of the curvature term at the finer scale. The IDE includes a monotone increasing scaling function which is shown to dictate the size of the residual image measured in the star-norm. Some theoretical properties of the novel IDE and its numerical implementation methods are given. Theoretical analysis and numerical experiments show the effectiveness of the IDE model.
This paper presents a novel multiscale variational image decomposition model. Based on the hierarchical multiscale variational model of Tadmor, a novel (BV, H-1) hierarchical multiscale image decomposition method is proposed, then the novel Integro-Differential Equation (IDE) is obtained by integrating in inverse scale space a succession of refined slices of the image and balancing a Laplacian of the curvature term at the finer scale. The IDE includes a monotone increasing scaling function which is shown to dictate the size of the residual image measured in the star-norm. Some theoretical properties of the novel IDE and its numerical implementation methods are given. Theoretical analysis and numerical experiments show the effectiveness of the IDE model.
2013, 35(5): 1196-1201.
doi: 10.3724/SP.J.1146.2012.01429
Abstract:
Sparsity-based Direction-Of-Arrival (DOA) estimation via l1-norm optimization requires fine tuning of the regularization parameter and large computational times. To alleviate these problems, this paper presents an efficient approach based on Sparse Bayesian Learning (SBL). The presented approach constructs and solves the jointly sparse DOA estimation model in real domain by making good use of the special geometry of the uniform linear array. Furthermore, the basis pruning mechanism of sparse Bayesian learning is modified to speed up the convergence rate. Simulation results demonstrate that the presented approach provides higher spatial resolution and accuracy with lower computational complexity in comparison with those l1-norm-based estimators.
Sparsity-based Direction-Of-Arrival (DOA) estimation via l1-norm optimization requires fine tuning of the regularization parameter and large computational times. To alleviate these problems, this paper presents an efficient approach based on Sparse Bayesian Learning (SBL). The presented approach constructs and solves the jointly sparse DOA estimation model in real domain by making good use of the special geometry of the uniform linear array. Furthermore, the basis pruning mechanism of sparse Bayesian learning is modified to speed up the convergence rate. Simulation results demonstrate that the presented approach provides higher spatial resolution and accuracy with lower computational complexity in comparison with those l1-norm-based estimators.
2013, 35(5): 1202-1207.
doi: 10.3724/SP.J.1146.2012.01138
Abstract:
Based on alternative iteration, an algorithm is proposed to estimate the DoAs of mixed signals and mutual coupling error for Uniform Linear Array (ULA). In the algorithm, utilizing the Toeplitz structure of the mutual coupling matrix of the ULA, a threshold-based method is presented and used to estimate initially the DoAs of the incoherent signals in mixed signals, and then the corresponding mutual coupling error is achieved. On the basis of this, the DoAs of the mixed signals is estimated and the mutual coupling error is updated in an alternative iteration mode. The overall algorithm can achieve the convergence with at most two times alternative iterations. The computer simulation indicates that the algorithm has good performance of DoA and mutual coupling error estimation even with less receive shotsnaps and in lower SNR region.
Based on alternative iteration, an algorithm is proposed to estimate the DoAs of mixed signals and mutual coupling error for Uniform Linear Array (ULA). In the algorithm, utilizing the Toeplitz structure of the mutual coupling matrix of the ULA, a threshold-based method is presented and used to estimate initially the DoAs of the incoherent signals in mixed signals, and then the corresponding mutual coupling error is achieved. On the basis of this, the DoAs of the mixed signals is estimated and the mutual coupling error is updated in an alternative iteration mode. The overall algorithm can achieve the convergence with at most two times alternative iterations. The computer simulation indicates that the algorithm has good performance of DoA and mutual coupling error estimation even with less receive shotsnaps and in lower SNR region.
2013, 35(5): 1208-1214.
doi: 10.3724/SP.J.1146.2012.00960
Abstract:
Current 1D signal trend extracting methods have such disadvantages as low efficiency, poor flexibility and so on. To overcome these problems, a new method of 1D signal fast trend extracting based on multi-scale extrema is proposed. By making full use of time sequence extrema information to establish a binary tree of multi-scale extrema, it avoids the time-consuming process of obtaining Intrinsic Mode Functions (IMFs) via iteratively sifting in traditional Empirical Mode Eecomposition (EMD) method. While obtaining similar results, it greatly improves the computation speed, and it could extract the trend of different scales directly. Simulated and practical signal experiments demonstrates the effectiveness of this approach. By comparing with traditional EMD method and trend filtering method, the results show that the approach could achieve 1 or 2 order of magnitude speedups.
Current 1D signal trend extracting methods have such disadvantages as low efficiency, poor flexibility and so on. To overcome these problems, a new method of 1D signal fast trend extracting based on multi-scale extrema is proposed. By making full use of time sequence extrema information to establish a binary tree of multi-scale extrema, it avoids the time-consuming process of obtaining Intrinsic Mode Functions (IMFs) via iteratively sifting in traditional Empirical Mode Eecomposition (EMD) method. While obtaining similar results, it greatly improves the computation speed, and it could extract the trend of different scales directly. Simulated and practical signal experiments demonstrates the effectiveness of this approach. By comparing with traditional EMD method and trend filtering method, the results show that the approach could achieve 1 or 2 order of magnitude speedups.
2013, 35(5): 1215-1221.
doi: 10.3724/SP.J.1146.2012.01132
Abstract:
The Scale Invariant Feature Transform (SIFT) has a fine algorithm performance and an extensive application to the matching algorithm of local features, but its descriptor is characterized by a high dimension and huge time consumption also gives rise to a low matching robustness when tackling similar areas. Therefore this paper puts forward an innovative Contourlet-SIFT feature matching algorithm. The SIFT key points are first extracted to conduct Contourlet transformation on peripheral areas in order to calculate the mean and standard deviation of the decomposition coefficient in each direction. Then the vector of overall texture description is constructed and the Euclidean distance of this low-dimensional vector provides references for prioritizing the matching pairs. The first 1% key points will be subject to the nearest ratio matching by the SIFT vector. The result proves that the new algorithm surpasses SIFT especially when addressing the images with great brightness difference and many similar areas. It can lift the matching speed while it parallels SIFT in its invariability of scale, rotation and visual angle.
The Scale Invariant Feature Transform (SIFT) has a fine algorithm performance and an extensive application to the matching algorithm of local features, but its descriptor is characterized by a high dimension and huge time consumption also gives rise to a low matching robustness when tackling similar areas. Therefore this paper puts forward an innovative Contourlet-SIFT feature matching algorithm. The SIFT key points are first extracted to conduct Contourlet transformation on peripheral areas in order to calculate the mean and standard deviation of the decomposition coefficient in each direction. Then the vector of overall texture description is constructed and the Euclidean distance of this low-dimensional vector provides references for prioritizing the matching pairs. The first 1% key points will be subject to the nearest ratio matching by the SIFT vector. The result proves that the new algorithm surpasses SIFT especially when addressing the images with great brightness difference and many similar areas. It can lift the matching speed while it parallels SIFT in its invariability of scale, rotation and visual angle.
2013, 35(5): 1222-1228.
doi: 10.3724/SP.J.1146.2012.01142
Abstract:
A novel non-rigid image registration algorithm is proposed based on an improved version of the traditional variational optical flow model and the extraction of the Scale-Invariant Feature Transform (SIFT) feature points. In this model, the issue of processing the regions of localized disease abnormalities and un-uniform brightness is tackled by using a data term combining the brightness conservation and gradient conservation assumptions. To solve the issue of severe image blurring and the loss of important details caused by the over-smoothing of the traditional optical flow model, an adaptive anisotropic regularization term is used. By extracting the SIFT feature points and using a multi-resolution layered refining, internal fixed-point iteration and coarse-to-fine warping strategy, the issue of registration of medical images with relatively larger deformation and also that of the details registration of medical images which can not be processed by the traditional optical flow method are well resolved. Extensive experimental results show the effectiveness of the model for non-rigid medical image registration.
A novel non-rigid image registration algorithm is proposed based on an improved version of the traditional variational optical flow model and the extraction of the Scale-Invariant Feature Transform (SIFT) feature points. In this model, the issue of processing the regions of localized disease abnormalities and un-uniform brightness is tackled by using a data term combining the brightness conservation and gradient conservation assumptions. To solve the issue of severe image blurring and the loss of important details caused by the over-smoothing of the traditional optical flow model, an adaptive anisotropic regularization term is used. By extracting the SIFT feature points and using a multi-resolution layered refining, internal fixed-point iteration and coarse-to-fine warping strategy, the issue of registration of medical images with relatively larger deformation and also that of the details registration of medical images which can not be processed by the traditional optical flow method are well resolved. Extensive experimental results show the effectiveness of the model for non-rigid medical image registration.
2013, 35(5): 1229-1235.
doi: 10.3724/SP.J.1146.2012.01205
Abstract:
According to H.264/AVC specific codec architecture, a novel watermarking algorithm with symbol encoding is proposed. Based on the analysis of requantization transcoding, the macroblock used for watermarking is selected with the combination of texture complexity and rate-distortion cost. The problem of watermark synchronization between the sender and receiver can be solved. The watermark information is embedded by slight modulating the coefficients with specific symbol encoding, instead of directly adding the watermark to the quantized coefficients. Experimental results reveal that the proposed scheme is robust against attacks such as requantization transcoding, additive white Gaussian noise, brightness and contrast adjustment, while preserving the perceptual quality.
According to H.264/AVC specific codec architecture, a novel watermarking algorithm with symbol encoding is proposed. Based on the analysis of requantization transcoding, the macroblock used for watermarking is selected with the combination of texture complexity and rate-distortion cost. The problem of watermark synchronization between the sender and receiver can be solved. The watermark information is embedded by slight modulating the coefficients with specific symbol encoding, instead of directly adding the watermark to the quantized coefficients. Experimental results reveal that the proposed scheme is robust against attacks such as requantization transcoding, additive white Gaussian noise, brightness and contrast adjustment, while preserving the perceptual quality.
2013, 35(5): 1236-1240.
doi: 10.3724/SP.J.1146.2012.01059
Abstract:
A method of calculating focal field is introduced, and the connection between focal field, which depends on the reflector geometry, and the aperture field of feed is analyzed. Based on focal field distribution, a design method of phased array feed, including element spacing determination, phased array scale, and excitations for expected radiation pattern is presented. Two phased array feeds are designed to realize shaped beams for an un-shaped Cassegrain antenna and shaped scanning beams for a prime-focus parabolic antenna respectively. The numerical analyses demonstrate the effectiveness of the approach.
A method of calculating focal field is introduced, and the connection between focal field, which depends on the reflector geometry, and the aperture field of feed is analyzed. Based on focal field distribution, a design method of phased array feed, including element spacing determination, phased array scale, and excitations for expected radiation pattern is presented. Two phased array feeds are designed to realize shaped beams for an un-shaped Cassegrain antenna and shaped scanning beams for a prime-focus parabolic antenna respectively. The numerical analyses demonstrate the effectiveness of the approach.
2013, 35(5): 1241-1246.
doi: 10.3724/SP.J.1146.2012.01256
Abstract:
This paper presents a new type of coaxial cavity loaded with a ridged cavity and a filter as the output circuit of S-band multiple-beam klystron. Theoretic analysis and equivalent circuit of the output circuit are given based on the theory of coupling resonance cavity. The uniformity of the gap field distribution in the output cavity and the gap impedance varying with frequency of output circuit are studied and simulated by means of CST-MWS simulation software. The results show that coaxial output cavity of the output circuit with coaxial compound cavity loaded with a filter not only reduces the request for the high current density emitted by the cathode, but also realizes 128 MHz/dB broad bandwidth and ensures the uniformity of the gap field distribution in the output cavity.
This paper presents a new type of coaxial cavity loaded with a ridged cavity and a filter as the output circuit of S-band multiple-beam klystron. Theoretic analysis and equivalent circuit of the output circuit are given based on the theory of coupling resonance cavity. The uniformity of the gap field distribution in the output cavity and the gap impedance varying with frequency of output circuit are studied and simulated by means of CST-MWS simulation software. The results show that coaxial output cavity of the output circuit with coaxial compound cavity loaded with a filter not only reduces the request for the high current density emitted by the cathode, but also realizes 128 MHz/dB broad bandwidth and ensures the uniformity of the gap field distribution in the output cavity.
2013, 35(5): 1247-1251.
doi: 10.3724/SP.J.1146.2012.01289
Abstract:
According to the characteristics of space target image, an novel method of space target image categorization based on local invariant features is proposed. The method extracts firstly local invariant features of each image and uses Gaussian Mixture Model (GMM) to establish global visual modes. Then co-occurrence matrix of the entire training set is constructed by matching local invariant features and visual models with maximum a posteriori probability and Probability Latent Semantic Analysis (PLSA) model is used to obtain latent class vector of images to achieve sencond representation. Finally, the SVM algorithm is used to implement image categorization. The experimental result demonstrates the effectiveness of the proposed method.
According to the characteristics of space target image, an novel method of space target image categorization based on local invariant features is proposed. The method extracts firstly local invariant features of each image and uses Gaussian Mixture Model (GMM) to establish global visual modes. Then co-occurrence matrix of the entire training set is constructed by matching local invariant features and visual models with maximum a posteriori probability and Probability Latent Semantic Analysis (PLSA) model is used to obtain latent class vector of images to achieve sencond representation. Finally, the SVM algorithm is used to implement image categorization. The experimental result demonstrates the effectiveness of the proposed method.
2013, 35(5): 1252-1256.
doi: 10.3724/SP.J.1146.2012.01075
Abstract:
A parameters estimation algorithm based on the power spectral FFT is proposed to estimate the parameters of Binary Phase Shift Keying (BPSK) signals under negative Signal-to-Noise Ratio (SNR) environment. The code width, carrier frequency and code length of BPSK signals are estimated according to the relationship between the parameters and the amplitude and phase spectrums, which are obtained through the FFT of power spectrum. This algorithm can further eliminate the noise based on power spectral FFT, so it is more suitable for low SNR environment. And the algorithm is computationally efficient and easy to realize. The experimental results demonstrate high accuracy and stronger capability on anti-noise, the estimation of the carrier frequency and the code width can increase by 9.9% and 190.9% with the cyclic spectral estimation algorithms.
A parameters estimation algorithm based on the power spectral FFT is proposed to estimate the parameters of Binary Phase Shift Keying (BPSK) signals under negative Signal-to-Noise Ratio (SNR) environment. The code width, carrier frequency and code length of BPSK signals are estimated according to the relationship between the parameters and the amplitude and phase spectrums, which are obtained through the FFT of power spectrum. This algorithm can further eliminate the noise based on power spectral FFT, so it is more suitable for low SNR environment. And the algorithm is computationally efficient and easy to realize. The experimental results demonstrate high accuracy and stronger capability on anti-noise, the estimation of the carrier frequency and the code width can increase by 9.9% and 190.9% with the cyclic spectral estimation algorithms.
2013, 35(5): 1257-1261.
doi: 10.3724/SP.J.1146.2012.01174
Abstract:
Resource allocation plays a significant role in current wireless communications. In order to improve the performance of the whole wireless network and social welfare, the limited network resources should be allocated efficiently and fairly. In this paper, a jointly guaranteed resource allocation scheme is proposed, which firstly introduces the required bandwidth and the desired bandwidth, and then combines the utility model with services that have different QoS (Quality of Service) requirements. An optimized model is built and the corresponding resource allocation scheme is proposed. The scheme could guarantee different QoS requirements, while the utility of the system could be improved. Furthermore, it can realize the fairness allocation between QoS and BE (Best Effort) traffics. The simulation results show that the proposed scheme can effectively improve the system utility compared with the traditional method, meanwhile, the goal of service oriented allocation is achieved, and the performance of the whole system is optimized.
Resource allocation plays a significant role in current wireless communications. In order to improve the performance of the whole wireless network and social welfare, the limited network resources should be allocated efficiently and fairly. In this paper, a jointly guaranteed resource allocation scheme is proposed, which firstly introduces the required bandwidth and the desired bandwidth, and then combines the utility model with services that have different QoS (Quality of Service) requirements. An optimized model is built and the corresponding resource allocation scheme is proposed. The scheme could guarantee different QoS requirements, while the utility of the system could be improved. Furthermore, it can realize the fairness allocation between QoS and BE (Best Effort) traffics. The simulation results show that the proposed scheme can effectively improve the system utility compared with the traditional method, meanwhile, the goal of service oriented allocation is achieved, and the performance of the whole system is optimized.
2013, 35(5): 1262-1266.
doi: 10.3724/SP.J.1146.2012.01169
Abstract:
With the development of semiconductor technologies and integration of chips, soft errors become the key factor influencing circuit reliability. In order to estimate the effects of soft errors, a reliability calculation method of gate-level circuit based on signals probability is proposed. All signals probabilities under soft errors are calculated first, and then the whole reliability is estimated using fault simulation. The proposed method is compared with the probabilistic transfer matrix approach and benchmark circuit experiments are finished, results show the method has the same accuracy as the Probabilistic Transfer Matrix (PTM) approach, but it needs shorter time and less space, especially suitable for calculation of reliability under specific vector and random vectors.
With the development of semiconductor technologies and integration of chips, soft errors become the key factor influencing circuit reliability. In order to estimate the effects of soft errors, a reliability calculation method of gate-level circuit based on signals probability is proposed. All signals probabilities under soft errors are calculated first, and then the whole reliability is estimated using fault simulation. The proposed method is compared with the probabilistic transfer matrix approach and benchmark circuit experiments are finished, results show the method has the same accuracy as the Probabilistic Transfer Matrix (PTM) approach, but it needs shorter time and less space, especially suitable for calculation of reliability under specific vector and random vectors.
2013, 35(5): 1267-1270.
doi: 10.3724/SP.J.1146.2012.01269
Abstract:
Single gap close coaxial resonator cavity with high order TM310 mode is designed. A new structure of output rectangular waveguide with the transverse inductance filter diaphragms is also designed. The average gap impedance variation with the frequency at the six drift tube center in the cavity are simulated and calculated. It is found that the curve of average gap impedance on frequency have two peaks, it indicates that the relative bandwidth of this output circuit is larger than that of empty waveguide.
Single gap close coaxial resonator cavity with high order TM310 mode is designed. A new structure of output rectangular waveguide with the transverse inductance filter diaphragms is also designed. The average gap impedance variation with the frequency at the six drift tube center in the cavity are simulated and calculated. It is found that the curve of average gap impedance on frequency have two peaks, it indicates that the relative bandwidth of this output circuit is larger than that of empty waveguide.
