Email alert
2007 Vol. 29, No. 12
Display Method:
2007, 29(12): 2795-2799.
doi: 10.3724/SP.J.1146.2006.00687
Abstract:
This paper presents a novel approach for seamless image stitching which is based on sparse feature matching and deformation propagation. First, an optimal partitioning which minimizes the structure error is found in the overlap region between the registered images, and the target region is selected from one side of partition boundary. Then, the salient structure feature is detected and matched along the partition boundary, which gets some sparse deformation vectors corresponding to the matched feature points and their associated edge points. By solving Poisson equations, these sparse deformation cues will then be propagated robustly and smoothly into the interior of the target region in which the deformation vectors of all points are derived. Finally, the gradient map of the target region is derived by interpolating the deformation vectors, from which the result is reconstructed. The implement is convenient and fast, and complex feature detection is needless. The proposed approach can handle significant structure and intensity misalignment in image stitching simultaneously, and eliminate both structure seam and intensity seam globally. Compared to several other methods, obvious improvement is achieved.
This paper presents a novel approach for seamless image stitching which is based on sparse feature matching and deformation propagation. First, an optimal partitioning which minimizes the structure error is found in the overlap region between the registered images, and the target region is selected from one side of partition boundary. Then, the salient structure feature is detected and matched along the partition boundary, which gets some sparse deformation vectors corresponding to the matched feature points and their associated edge points. By solving Poisson equations, these sparse deformation cues will then be propagated robustly and smoothly into the interior of the target region in which the deformation vectors of all points are derived. Finally, the gradient map of the target region is derived by interpolating the deformation vectors, from which the result is reconstructed. The implement is convenient and fast, and complex feature detection is needless. The proposed approach can handle significant structure and intensity misalignment in image stitching simultaneously, and eliminate both structure seam and intensity seam globally. Compared to several other methods, obvious improvement is achieved.
2007, 29(12): 2800-2804.
doi: 10.3724/SP.J.1146.2006.00667
Abstract:
A new fusion method for fusing two spatially registered multi-focus images is proposed in this paper. It is based on multi-resolution wavelet decomposition, Self-Organizing Feature Map (SOFM) neural networks and Evolutionary Strategies (ES). First, a normalized feature image, which represents the local region clarity difference of two source images, is extracted by redundant wavelet transform, then the feature image is clustered by SOFM learning algorithm and every pixel pair in source images is classified into a certain class which indicates different clarity differences. Finally, to each pixel pair in different classes, different fusion factors are used to fuse it; these fusion factors are determined by evolution strategies to achieve the best fusion performance. Experimental results show that the proposed method outperforms the Laplace transform and wavelet transform methods.
A new fusion method for fusing two spatially registered multi-focus images is proposed in this paper. It is based on multi-resolution wavelet decomposition, Self-Organizing Feature Map (SOFM) neural networks and Evolutionary Strategies (ES). First, a normalized feature image, which represents the local region clarity difference of two source images, is extracted by redundant wavelet transform, then the feature image is clustered by SOFM learning algorithm and every pixel pair in source images is classified into a certain class which indicates different clarity differences. Finally, to each pixel pair in different classes, different fusion factors are used to fuse it; these fusion factors are determined by evolution strategies to achieve the best fusion performance. Experimental results show that the proposed method outperforms the Laplace transform and wavelet transform methods.
2007, 29(12): 2805-2809.
doi: 10.3724/SP.J.1146.2006.00708
Abstract:
To the issues of threat assessment (TA), a technique for TA based on Adaptive Neuro-Intuitionistic Fuzzy Inference System (ANIFIS) is proposed with intuitionistic fuzzy set theory introduced into the area of information fusion. First, the properties and vulnerabilities of the existing TA methods are analyzed. A model for TA on ANIFIS with Takagi-Sugeno type is established. Then, the attribute functions, i.e. membership and nonmembership functions, and the inference rules of the system variables are devised with computational relations between layers of input and output and a synthesized computational expression of system outputs ascertained. Subsequently, an analysis of global approximation property of the model is performed with a learning algorithm of neural net devised. Finally, the validity of the technique is checked and rationality of constructed model is verified by providing TA instances with 20 typical targets. The simulated results show that this method can enhance creditability of TA and improve quality of assessment with precision of synthetic values in reasoning output.
To the issues of threat assessment (TA), a technique for TA based on Adaptive Neuro-Intuitionistic Fuzzy Inference System (ANIFIS) is proposed with intuitionistic fuzzy set theory introduced into the area of information fusion. First, the properties and vulnerabilities of the existing TA methods are analyzed. A model for TA on ANIFIS with Takagi-Sugeno type is established. Then, the attribute functions, i.e. membership and nonmembership functions, and the inference rules of the system variables are devised with computational relations between layers of input and output and a synthesized computational expression of system outputs ascertained. Subsequently, an analysis of global approximation property of the model is performed with a learning algorithm of neural net devised. Finally, the validity of the technique is checked and rationality of constructed model is verified by providing TA instances with 20 typical targets. The simulated results show that this method can enhance creditability of TA and improve quality of assessment with precision of synthetic values in reasoning output.
2007, 29(12): 2810-2813.
doi: 10.3724/SP.J.1146.2006.01267
Abstract:
The general Interacting Multiple Model (IMM) based on Kalman Filter (IMMKF) or Extended Kalman Filter (IMMEKF) can not deal with non-Guassian problems and also does not work so well as the IMM based on the particle filter for the nonlinear problems. The particle filter can deal with nonlinear/non-Guassian problems and it has been introduced to the algorithm of IMM for higher precision. However, the disadvantage of the particle filter is heavy computational load, because a particle filter usually has a lot of particles, which will increase the computational load greatly. Here an improved interacting multiple model particle filter, which combines Kalman filter and particle filter using multiple models, is proposed to improve the computational efficiency compared with the usual Interacting Multiple Model Particle Filter ( IMMPF). For nonlinear/non-Guassian problems, the new algorithm shows to possess a good performance as the IMMPF, while for linear problems it performs as well as the IMMEKF and works better than the IMMPF.
The general Interacting Multiple Model (IMM) based on Kalman Filter (IMMKF) or Extended Kalman Filter (IMMEKF) can not deal with non-Guassian problems and also does not work so well as the IMM based on the particle filter for the nonlinear problems. The particle filter can deal with nonlinear/non-Guassian problems and it has been introduced to the algorithm of IMM for higher precision. However, the disadvantage of the particle filter is heavy computational load, because a particle filter usually has a lot of particles, which will increase the computational load greatly. Here an improved interacting multiple model particle filter, which combines Kalman filter and particle filter using multiple models, is proposed to improve the computational efficiency compared with the usual Interacting Multiple Model Particle Filter ( IMMPF). For nonlinear/non-Guassian problems, the new algorithm shows to possess a good performance as the IMMPF, while for linear problems it performs as well as the IMMEKF and works better than the IMMPF.
2007, 29(12): 2814-2818.
doi: 10.3724/SP.J.1146.2006.00434
Abstract:
It is significant to convert planar images into stereo images. Hou et al. (2002) presented a method to realize the function and it is called Hou method. In this method, the parameters such as random variables are used to control the procedure, but the stereo effect that the parameters take on had not been discussed. Using Hou method, the planar image with psychological depth cue is converted into stereo on standard computer monitor. Quantitative criterions to evaluate the stereo effect after conversion are given. And further more, the stereo effect that the random variables take on is discussed. The results of the experiment show that, when every random variable in matrix has the same uniform distribution, different values of these random variables take little effect on stereo effect; when the distance between the monitor screen and the observer is a well chosen value (e.g. 1m), the number of the sub-block images takes little effect on stereo effect.
It is significant to convert planar images into stereo images. Hou et al. (2002) presented a method to realize the function and it is called Hou method. In this method, the parameters such as random variables are used to control the procedure, but the stereo effect that the parameters take on had not been discussed. Using Hou method, the planar image with psychological depth cue is converted into stereo on standard computer monitor. Quantitative criterions to evaluate the stereo effect after conversion are given. And further more, the stereo effect that the random variables take on is discussed. The results of the experiment show that, when every random variable in matrix has the same uniform distribution, different values of these random variables take little effect on stereo effect; when the distance between the monitor screen and the observer is a well chosen value (e.g. 1m), the number of the sub-block images takes little effect on stereo effect.
2007, 29(12): 2819-2821.
doi: 10.3724/SP.J.1146.2006.00767
Abstract:
Texture is often divided into lots of isolate areas when a texture image is segmented. This paper develops a new method for texture image segmentation. It uses Gabor wavelet to extract texture features, and then Chan and Vese model is employed to segment a texture image. Compared to other texture segmentation methods using Chan and Vese models, the proposed method has two advantages. First, by combining the gray levels of pixels and texture information of an image, this method can be used for segmentation of a texture image or a none-texture image. Second, the hierarchical method proposed in this paper can avoid the problem due to the choice of initial conditions. The segmentation tests for remote sensing and natural texture images prove the proposed segmentation method is accurate and efficient.
Texture is often divided into lots of isolate areas when a texture image is segmented. This paper develops a new method for texture image segmentation. It uses Gabor wavelet to extract texture features, and then Chan and Vese model is employed to segment a texture image. Compared to other texture segmentation methods using Chan and Vese models, the proposed method has two advantages. First, by combining the gray levels of pixels and texture information of an image, this method can be used for segmentation of a texture image or a none-texture image. Second, the hierarchical method proposed in this paper can avoid the problem due to the choice of initial conditions. The segmentation tests for remote sensing and natural texture images prove the proposed segmentation method is accurate and efficient.
2007, 29(12): 2822-2826.
doi: 10.3724/SP.J.1146.2006.00535
Abstract:
A novel robust adaptive beamforming based on diagonal loading is proposed in this paper, aiming at robustness against steering vector mismatches in spatially colored noise fields. The received data is prewhitened first according to the noise covariance, with a corresponding ellipsoidal steering vector uncertainty set defined at the same time. Then the loading level is determined based on worst-case performance optimization. Other than searching for the optimal loading by iteration, a simple closed-form solution is suggested after some approximations. Besides its low computational cost, the closed-form solution reveals how different factors affect the optimal loading. Numerical examples confirm the excellent performance of the proposed approach as well as the accurate approximation of the closed-form solution to the actual optimal loading.
A novel robust adaptive beamforming based on diagonal loading is proposed in this paper, aiming at robustness against steering vector mismatches in spatially colored noise fields. The received data is prewhitened first according to the noise covariance, with a corresponding ellipsoidal steering vector uncertainty set defined at the same time. Then the loading level is determined based on worst-case performance optimization. Other than searching for the optimal loading by iteration, a simple closed-form solution is suggested after some approximations. Besides its low computational cost, the closed-form solution reveals how different factors affect the optimal loading. Numerical examples confirm the excellent performance of the proposed approach as well as the accurate approximation of the closed-form solution to the actual optimal loading.
2007, 29(12): 2827-2831.
doi: 10.3724/SP.J.1146.2006.00734
Abstract:
Often in practice the directional knowledge of signal of interest or interferences is imprecise, the adaptive beamformer with linear constraint is hypersensitive to the array imperfection which leads to the output Signal-to-Interference-plus-Noise Ratio (SINR) decreases when the input Signal-to-Noise Ratio (SNR) exceeds a certain threshold. In order to improve the robustness of the beamformer, a new method for estimating the artificial interference power under sidelobe soft constraint is proposed. Based on the prior coarse knowledge of Direction of Arrival (DOA) of the interferences and using the subspace projection and the null area extension techniques, the artificial interference power can be obtained through the iterative procedure under the soft constraint of the maximum peak undulation in sidelobe area. The method is not sensitive to the look direction mismatch and the sensors gain error with low sidelobe level and deep null. Numerical examples demonstrate its effectiveness and robustness comparing with other methods.
Often in practice the directional knowledge of signal of interest or interferences is imprecise, the adaptive beamformer with linear constraint is hypersensitive to the array imperfection which leads to the output Signal-to-Interference-plus-Noise Ratio (SINR) decreases when the input Signal-to-Noise Ratio (SNR) exceeds a certain threshold. In order to improve the robustness of the beamformer, a new method for estimating the artificial interference power under sidelobe soft constraint is proposed. Based on the prior coarse knowledge of Direction of Arrival (DOA) of the interferences and using the subspace projection and the null area extension techniques, the artificial interference power can be obtained through the iterative procedure under the soft constraint of the maximum peak undulation in sidelobe area. The method is not sensitive to the look direction mismatch and the sensors gain error with low sidelobe level and deep null. Numerical examples demonstrate its effectiveness and robustness comparing with other methods.
2007, 29(12): 2832-2835.
doi: 10.3724/SP.J.1146.2006.00765
Abstract:
The MODE-TOEP algorithm is proposed to estimate DOA of coherent signals on a Uniform Circular Array (UCA). Firstly, the mode excitation method is used to transform the UCA in element space into a virtual ULA (VULA) in mode space. Then a Toeplitz matrix can be reconstructed to estimate DOA of coherent signals successfully. MODE-TOEP algorithm has less computational burden due to non-spatial smoothing. Simulation results show that MODE-TOEP algorithm is effective and has a better performance compared with conventional MODE-SS algorithm.
The MODE-TOEP algorithm is proposed to estimate DOA of coherent signals on a Uniform Circular Array (UCA). Firstly, the mode excitation method is used to transform the UCA in element space into a virtual ULA (VULA) in mode space. Then a Toeplitz matrix can be reconstructed to estimate DOA of coherent signals successfully. MODE-TOEP algorithm has less computational burden due to non-spatial smoothing. Simulation results show that MODE-TOEP algorithm is effective and has a better performance compared with conventional MODE-SS algorithm.
2007, 29(12): 2836-2839.
doi: 10.3724/SP.J.1146.2006.01207
Abstract:
A state space model for chaotic signal blind separation is first suggested, and then combined the Rao-Blackwellisaion (RB) strategy and incorporated roughening noise method, a particle filtering based online blind separation algorithm is proposed. To further decrease the residual noise in the online recovered signals, a novel delay estimation method which can effectively reduce the cross influence of source signals compared to the traditional delay-weight method is proposed.
A state space model for chaotic signal blind separation is first suggested, and then combined the Rao-Blackwellisaion (RB) strategy and incorporated roughening noise method, a particle filtering based online blind separation algorithm is proposed. To further decrease the residual noise in the online recovered signals, a novel delay estimation method which can effectively reduce the cross influence of source signals compared to the traditional delay-weight method is proposed.
2007, 29(12): 2840-2842.
doi: 10.3724/SP.J.1146.2006.00317
Abstract:
In this paper a new multiple targets angle tracking algorithm is proposed. First, high-resolution method can be used to derive the number of targets and the targets, initial angles. Then the angle of each target at different time could be estimated by updating the covariance matrix estimated through the sensor outputs. In this algorithm, the angles estimated at conjoint time are auto associated, so the process of data association could be omitted. Furthermore, this paper presents two modified methods: one is that the predicted angles are introduced to decrease the errors caused by two approximations, and the other is that the memory track are used when some targets are crossed. Simulation results show the algorithm has high tracing performance.
In this paper a new multiple targets angle tracking algorithm is proposed. First, high-resolution method can be used to derive the number of targets and the targets, initial angles. Then the angle of each target at different time could be estimated by updating the covariance matrix estimated through the sensor outputs. In this algorithm, the angles estimated at conjoint time are auto associated, so the process of data association could be omitted. Furthermore, this paper presents two modified methods: one is that the predicted angles are introduced to decrease the errors caused by two approximations, and the other is that the memory track are used when some targets are crossed. Simulation results show the algorithm has high tracing performance.
2007, 29(12): 2843-2847.
doi: 10.3724/SP.J.1146.2006.00710
Abstract:
For dual-aperture Synthetic Aperture Radar (SAR), Displaced Phase Center Antenna (DPCA) in image domain may dramatically suppress the static clutter background. However, even via DPCA processing, the noise and the residual clutter may be inevitably existed and seriously affect weak target detection. This paper proposed a novel method to further suppress interference by using two-dimensional multi-resolution analysis combined with DPCA, which may remarkably improve the detection performance of weak moving target. Finally, the detailed numerical experiments are also presented based on the proposed method to demonstrate the suppressing performance of the clutter and noise.
For dual-aperture Synthetic Aperture Radar (SAR), Displaced Phase Center Antenna (DPCA) in image domain may dramatically suppress the static clutter background. However, even via DPCA processing, the noise and the residual clutter may be inevitably existed and seriously affect weak target detection. This paper proposed a novel method to further suppress interference by using two-dimensional multi-resolution analysis combined with DPCA, which may remarkably improve the detection performance of weak moving target. Finally, the detailed numerical experiments are also presented based on the proposed method to demonstrate the suppressing performance of the clutter and noise.
2007, 29(12): 2848-2852.
doi: 10.3724/SP.J.1146.2006.00516
Abstract:
When the parameter estimation of Damped Exponentials (DE) based scattering center model is considered in real life situation, the noise is always non-Gaussian and has long-tailed distribution. The estimation error in such noise may be large if the estimator under gauss-noise assumption is used. In this paper, the M-estimation method is used to obtain robust parameter estimation of DE-model in long-tailed clutter. Firstly, the shortcoming of Prony-based M-estimation method is analyzed. Then, other two effective methods to realize M-estimation of DE-model are proposed. A comparative study of these two methods is carried on about their estimation performance. Lastly, Monte-Carlo simulation test is performed to validate the proposed methods. The results show that the two proposed methods both give better estimation results in long-tailed K-Distribution clutter.
When the parameter estimation of Damped Exponentials (DE) based scattering center model is considered in real life situation, the noise is always non-Gaussian and has long-tailed distribution. The estimation error in such noise may be large if the estimator under gauss-noise assumption is used. In this paper, the M-estimation method is used to obtain robust parameter estimation of DE-model in long-tailed clutter. Firstly, the shortcoming of Prony-based M-estimation method is analyzed. Then, other two effective methods to realize M-estimation of DE-model are proposed. A comparative study of these two methods is carried on about their estimation performance. Lastly, Monte-Carlo simulation test is performed to validate the proposed methods. The results show that the two proposed methods both give better estimation results in long-tailed K-Distribution clutter.
2007, 29(12): 2853-2857.
doi: 10.3724/SP.J.1146.2006.00531
Abstract:
Based on the assumption that wavelet coefficients obey Generalized Gaussian Distribution (GGD), this paper adopts Maximum Likelihood (ML) principle to estimate wavelet coefficients variance of common images in sub-bands. The proposed estimator is product of a sub-band adjustable factor and a power mean factor. Compared to the recently proposed SI-AdaptShr, LAWMAP and other wavelet-based methods, better de-noising results may be obtained for the proposed method. Furthermore, a simplified algorithm is also formed to de-speckle SAR images. It is shown that the new method may remarkably reduce the calculation amount and helpful for the post-processing of large scale SAR images.
Based on the assumption that wavelet coefficients obey Generalized Gaussian Distribution (GGD), this paper adopts Maximum Likelihood (ML) principle to estimate wavelet coefficients variance of common images in sub-bands. The proposed estimator is product of a sub-band adjustable factor and a power mean factor. Compared to the recently proposed SI-AdaptShr, LAWMAP and other wavelet-based methods, better de-noising results may be obtained for the proposed method. Furthermore, a simplified algorithm is also formed to de-speckle SAR images. It is shown that the new method may remarkably reduce the calculation amount and helpful for the post-processing of large scale SAR images.
2007, 29(12): 2858-2862.
doi: 10.3724/SP.J.1146.2006.00466
Abstract:
This paper analyses the influence of acceleration on dual frequency ranging of multiple frequency CW radar, a new real-time acceleration correction algorithm based on partial dechirping is introduced. Principles and computational costs of the algorithm are discussed in details, and it is also compared with acceleration estimation and correction algorithm based on DAF, simulation results prove its feasibility and validdity.
This paper analyses the influence of acceleration on dual frequency ranging of multiple frequency CW radar, a new real-time acceleration correction algorithm based on partial dechirping is introduced. Principles and computational costs of the algorithm are discussed in details, and it is also compared with acceleration estimation and correction algorithm based on DAF, simulation results prove its feasibility and validdity.
2007, 29(12): 2863-2866.
doi: 10.3724/SP.J.1146.2006.00706
Abstract:
Ground Moving Target Indication (GMTI) is one of hotspot of Distributed Small Satellites SAR (DSS-SAR) research, but its very difficult to detect ground moving target from background clutter because of the existence of baseline correlation which result in difficulty to separate the phase produced by terrain height and ground moving target. This paper analyzes the reason of baseline correlation and proposes a promising method to eliminate the interferometric phase induced by cross-track baseline. The computer simulations show that it is efficient to eliminate cross-track baseline interferometric phase.
Ground Moving Target Indication (GMTI) is one of hotspot of Distributed Small Satellites SAR (DSS-SAR) research, but its very difficult to detect ground moving target from background clutter because of the existence of baseline correlation which result in difficulty to separate the phase produced by terrain height and ground moving target. This paper analyzes the reason of baseline correlation and proposes a promising method to eliminate the interferometric phase induced by cross-track baseline. The computer simulations show that it is efficient to eliminate cross-track baseline interferometric phase.
2007, 29(12): 2867-2870.
doi: 10.3724/SP.J.1146.2006.00717
Abstract:
Based on the Doppler Distributed Clutter (DDC) model, this paper studies the distribution of eigenvalues of the clutter covariance matrix and proposes a robust algorithm of Doppler parameters estimation for airborne radar. Compared with existing methods, the proposed algorithm has comparable performance for Doppler centroid estimation and shows superiority for Doppler spread estimation. The proposed algorithm is competent for most practical applications of airborne radar such as moving target detection (MTD) using Pulse-Doppler (PD) radar, especially for few sampling case. Numerical simulations prove the effectiveness of the proposed algorithm.
Based on the Doppler Distributed Clutter (DDC) model, this paper studies the distribution of eigenvalues of the clutter covariance matrix and proposes a robust algorithm of Doppler parameters estimation for airborne radar. Compared with existing methods, the proposed algorithm has comparable performance for Doppler centroid estimation and shows superiority for Doppler spread estimation. The proposed algorithm is competent for most practical applications of airborne radar such as moving target detection (MTD) using Pulse-Doppler (PD) radar, especially for few sampling case. Numerical simulations prove the effectiveness of the proposed algorithm.
2007, 29(12): 2871-2875.
doi: 10.3724/SP.J.1146.2006.00735
Abstract:
This paper proposes a dimensionality reduction and compression method of hyperspectral images based on Independent Component Analysis(ICA) for hyperspectral image analysis. At first hyperspectral features are extracted using ICA and dimensionality reduction is accomplished. Then, dimensionality reduction images are compressed by the predictive code and adaptive arithmetic code. The experimental results by using 220 bands and 64 bands hyperspectral data show that the method achieved higher compression ratio, more strong analysis capability and lower peak signal-to-noise ratio than dimensionality reduction based on Principal Components Analysis(PCA).
This paper proposes a dimensionality reduction and compression method of hyperspectral images based on Independent Component Analysis(ICA) for hyperspectral image analysis. At first hyperspectral features are extracted using ICA and dimensionality reduction is accomplished. Then, dimensionality reduction images are compressed by the predictive code and adaptive arithmetic code. The experimental results by using 220 bands and 64 bands hyperspectral data show that the method achieved higher compression ratio, more strong analysis capability and lower peak signal-to-noise ratio than dimensionality reduction based on Principal Components Analysis(PCA).
2007, 29(12): 2876-2879.
doi: 10.3724/SP.J.1146.2006.00453
Abstract:
A hyperspectral image lossless compression algorithm based on multi-band prediction is proposed in this paper. Considering that as the spectral resolution of hyperspectral images increases, its spectral correlation becomes more and more significant, a group of prediction coefficients are derived, which are able to make use of multi-band to predict the current band linearly. And then a fast algorithm is given to solve the coefficients of the predictor. Experiments show that, with AVIRIS, the ratio of the algorithm is competitive and it runs fast, so it works efficiently in practice.
A hyperspectral image lossless compression algorithm based on multi-band prediction is proposed in this paper. Considering that as the spectral resolution of hyperspectral images increases, its spectral correlation becomes more and more significant, a group of prediction coefficients are derived, which are able to make use of multi-band to predict the current band linearly. And then a fast algorithm is given to solve the coefficients of the predictor. Experiments show that, with AVIRIS, the ratio of the algorithm is competitive and it runs fast, so it works efficiently in practice.
2007, 29(12): 2880-2884.
doi: 10.3724/SP.J.1146.2006.02073
Abstract:
Weight calculating and feature reduction are key preprocesses in text classification. Firstly, those useless to classify texts are filtered according the category document frequency distribution difference of each feature; and then in order to overcome the limitations of TF-IDF weighting formula a novel weighting formula called TF-CDF is presented . Calculate the weight of each feature according to TF-CDF and build the Vector Space Model (VSM) for the entire corpus. To select significant features, a feature selection approach based on Variable Precision Rough Set (VPRS) is also proposed and implement with some SQL sentences combining the definitions of VPRS with the advantages of SQL sentences. Finally, some experiments based on different weighting formulas and feature selection methods are conducted using libSVM as text classifier. The experimental results show that the novel feature filtering, weighting formula and feature selection method improve the performance of text classification.
Weight calculating and feature reduction are key preprocesses in text classification. Firstly, those useless to classify texts are filtered according the category document frequency distribution difference of each feature; and then in order to overcome the limitations of TF-IDF weighting formula a novel weighting formula called TF-CDF is presented . Calculate the weight of each feature according to TF-CDF and build the Vector Space Model (VSM) for the entire corpus. To select significant features, a feature selection approach based on Variable Precision Rough Set (VPRS) is also proposed and implement with some SQL sentences combining the definitions of VPRS with the advantages of SQL sentences. Finally, some experiments based on different weighting formulas and feature selection methods are conducted using libSVM as text classifier. The experimental results show that the novel feature filtering, weighting formula and feature selection method improve the performance of text classification.
2007, 29(12): 2885-2890.
doi: 10.3724/SP.J.1146.2006.00507
Abstract:
This paper presents a new algorithm using title category semantic recognition for text categorization. The algorithm generates feature space based on its category, picks up category semantic words of the title to produce candidate category and finally classifies it under these candidate categories. The experimental results firmly prove that the new algorithm performs better with fewer candidates and higher precision. Further research introduces category space representation efficiency to verify the validity of the new algorithm and proves that it can achieve great improvement in text representation.
This paper presents a new algorithm using title category semantic recognition for text categorization. The algorithm generates feature space based on its category, picks up category semantic words of the title to produce candidate category and finally classifies it under these candidate categories. The experimental results firmly prove that the new algorithm performs better with fewer candidates and higher precision. Further research introduces category space representation efficiency to verify the validity of the new algorithm and proves that it can achieve great improvement in text representation.
2007, 29(12): 2891-2895.
doi: 10.3724/SP.J.1146.2006.00707
Abstract:
To satisfy the requirement of high location speed and stability, the paper presents an improved algorithm for 3-D moving emitter passive location and tracking using the frequency and spatial measurements by single non-moving observer. This method firstly takes full advantage of the spatial and frequency changing information to estimate the velocity vector of the emitter via the Pseudo-Linear Kalman Filter (PLKF) algorithm, which can be used to obtain the position vector, the Kalman filter is to improve the location accuracy by taking the estimation of the position vector as measurements; the computer simulations show that the method has high location accuracy and fast convergence speed.
To satisfy the requirement of high location speed and stability, the paper presents an improved algorithm for 3-D moving emitter passive location and tracking using the frequency and spatial measurements by single non-moving observer. This method firstly takes full advantage of the spatial and frequency changing information to estimate the velocity vector of the emitter via the Pseudo-Linear Kalman Filter (PLKF) algorithm, which can be used to obtain the position vector, the Kalman filter is to improve the location accuracy by taking the estimation of the position vector as measurements; the computer simulations show that the method has high location accuracy and fast convergence speed.
2007, 29(12): 2896-2898.
doi: 10.3724/SP.J.1146.2006.01940
Abstract:
An adaptive threshold signal detection method based on the ration of signal power and correlation signal power for packet burst communication systems is studied. To avoid false alarm burrs in practical application, feeble offset is added in decision threshold. To raise the detection precision, a slipped compare window is introduced when executing decision. Theory analysis and simulation results show that the improved scheme can archive nice adaptive threshold signal detection performances.
An adaptive threshold signal detection method based on the ration of signal power and correlation signal power for packet burst communication systems is studied. To avoid false alarm burrs in practical application, feeble offset is added in decision threshold. To raise the detection precision, a slipped compare window is introduced when executing decision. Theory analysis and simulation results show that the improved scheme can archive nice adaptive threshold signal detection performances.
2007, 29(12): 2899-2901.
doi: 10.3724/SP.J.1146.2006.00637
Abstract:
F2+uF2 is a ring with four elements which shares some good properties of both Z4 and F4 Coding theory over this ring has recently received a great deal of interest among coding theorists. This paper gives the theory of Galois extensions over F2+uF2,and shows that the automorphism groups of these Galois extensions are different from the corresponding groups over Z4.Trace codes and subring subcodes over Galois extensions are defined, and it is proved that the trace codes of dual codes of linear codes are the dual codes of subring subcodes.
F2+uF2 is a ring with four elements which shares some good properties of both Z4 and F4 Coding theory over this ring has recently received a great deal of interest among coding theorists. This paper gives the theory of Galois extensions over F2+uF2,and shows that the automorphism groups of these Galois extensions are different from the corresponding groups over Z4.Trace codes and subring subcodes over Galois extensions are defined, and it is proved that the trace codes of dual codes of linear codes are the dual codes of subring subcodes.
2007, 29(12): 2902-2906.
doi: 10.3724/SP.J.1146.2006.00646
Abstract:
By analysis and the sufficient utilization of characteristics of the cyclic shift identity matrices in a cycle, the construction based on conditional-filling with shift matrices for LDPC codes is presented, the upper and lower bounds of girth of the check matrix which had been constructed by this scheme is also proposed, and a concrete conditional-filling construction is thus proposed, the simulations prove that the construction can reduce the amount of least cycle of the check matrix remarkably and improve the performance to a larger extent.
By analysis and the sufficient utilization of characteristics of the cyclic shift identity matrices in a cycle, the construction based on conditional-filling with shift matrices for LDPC codes is presented, the upper and lower bounds of girth of the check matrix which had been constructed by this scheme is also proposed, and a concrete conditional-filling construction is thus proposed, the simulations prove that the construction can reduce the amount of least cycle of the check matrix remarkably and improve the performance to a larger extent.
2007, 29(12): 2907-2911.
doi: 10.3724/SP.J.1146.2006.00824
Abstract:
Be aimed at lower complexity and RAM requirement of Low Density Parity Check (LDPC) decoder, a new class of concatenated codes called Parallel Interleaved Concatenated Gallager Code (PICGC), based on Turbo architecture and LDPC codes, is presented. In this paper, design, encoding and decoding algorithms of PICGC are studied. The RAM requirement for PICGC decoder is analyzed and compared to LDPC decoder, and an upper bound of memory-saving ratio is derived. The theoretical analysis and simulation results demonstrate that PICGC can reduce decoding complexity and RAM requirement significantly and maintain decoding delay with little sacrifice in performance in comparison to conventional LDPC codes. PICGC is an effective and feasible channel coding scheme.
Be aimed at lower complexity and RAM requirement of Low Density Parity Check (LDPC) decoder, a new class of concatenated codes called Parallel Interleaved Concatenated Gallager Code (PICGC), based on Turbo architecture and LDPC codes, is presented. In this paper, design, encoding and decoding algorithms of PICGC are studied. The RAM requirement for PICGC decoder is analyzed and compared to LDPC decoder, and an upper bound of memory-saving ratio is derived. The theoretical analysis and simulation results demonstrate that PICGC can reduce decoding complexity and RAM requirement significantly and maintain decoding delay with little sacrifice in performance in comparison to conventional LDPC codes. PICGC is an effective and feasible channel coding scheme.
2007, 29(12): 2912-2914.
doi: 10.3724/SP.J.1146.2006.00683
Abstract:
In this paper, the structure character of linear codes on ring F2+uF2 and the relation between the linear code and its residue code, torsion code, quotient code are studied. The depth distribution and depth spectrum of linear codes (especially the cyclic codes ) are presented.
In this paper, the structure character of linear codes on ring F2+uF2 and the relation between the linear code and its residue code, torsion code, quotient code are studied. The depth distribution and depth spectrum of linear codes (especially the cyclic codes ) are presented.
2007, 29(12): 2915-2918.
doi: 10.3724/SP.J.1146.2006.00732
Abstract:
In order to improve the system performance and reduce the hardware complexity further, this paper proposed a multiple-input-multiple-output (MIMO) scheme, combining Transmitting Antenna Selection and receiver Minimum-selection Maximum Ratio Combining (TAS/MMRC) scheme. A though and exact analysis for TAS/MMRC is carried out and the analytical results are verified by simulations in the end.
In order to improve the system performance and reduce the hardware complexity further, this paper proposed a multiple-input-multiple-output (MIMO) scheme, combining Transmitting Antenna Selection and receiver Minimum-selection Maximum Ratio Combining (TAS/MMRC) scheme. A though and exact analysis for TAS/MMRC is carried out and the analytical results are verified by simulations in the end.
2007, 29(12): 2919-2922.
doi: 10.3724/SP.J.1146.2006.02086
Abstract:
A scheme is put forward in this article for the joint utilization of both Modulation-Induced Cyclostationarity (MIC) and oversampling-induced cyclostationarity so that blind identification and equalization of SISO FIR channels can be performed. The advantage of the proposed method is that only one antenna is required at both the transmit and the receive ends and that heavy interferences can be rejected. The effectiveness of the algorithms are confirmed by the simulation results.
A scheme is put forward in this article for the joint utilization of both Modulation-Induced Cyclostationarity (MIC) and oversampling-induced cyclostationarity so that blind identification and equalization of SISO FIR channels can be performed. The advantage of the proposed method is that only one antenna is required at both the transmit and the receive ends and that heavy interferences can be rejected. The effectiveness of the algorithms are confirmed by the simulation results.
2007, 29(12): 2923-2925.
doi: 10.3724/SP.J.1146.2006.00678
Abstract:
An uplink power iterative algorithm is proposed for CDMA network planning. In the iteration, the mobile station is dynamically assigned to the home station which can minimize the mobile station transmitted power according to the load of the base stations. So the transmitted power of the mobile station and the received power of the base station is converged. The convergence of the algorithm is analyzed in therotical, and simulation results demonstrate the correctness of theoretical analysis. The presented algorithm can be implemented easily by computer. Thus it can be used for practical network planning.
An uplink power iterative algorithm is proposed for CDMA network planning. In the iteration, the mobile station is dynamically assigned to the home station which can minimize the mobile station transmitted power according to the load of the base stations. So the transmitted power of the mobile station and the received power of the base station is converged. The convergence of the algorithm is analyzed in therotical, and simulation results demonstrate the correctness of theoretical analysis. The presented algorithm can be implemented easily by computer. Thus it can be used for practical network planning.
2007, 29(12): 2926-2929.
doi: 10.3724/SP.J.1146.2006.00669
Abstract:
The TD-SCDMA system is code-limited. A code-reuse scheme is proposed to make the limited spreading codes support more subscribers. The smart antenna technique is used to realize the scheme. It takes advantaged of the big differences of space characteristic among subscribes to meet the demand of reuse of same spreading codes for several high data rate of users. This can improve utilization of code resources. In comparisons of system performances using several beam-forming algorithms, the results show that performance of the system using the code-reuse scheme is almost the same as that of traditional system; Moreover, an improved algorithm based on tradition null-steering method is proposed, which can improve system performance, especially for the high data rate downlink with a few subscribers.
The TD-SCDMA system is code-limited. A code-reuse scheme is proposed to make the limited spreading codes support more subscribers. The smart antenna technique is used to realize the scheme. It takes advantaged of the big differences of space characteristic among subscribes to meet the demand of reuse of same spreading codes for several high data rate of users. This can improve utilization of code resources. In comparisons of system performances using several beam-forming algorithms, the results show that performance of the system using the code-reuse scheme is almost the same as that of traditional system; Moreover, an improved algorithm based on tradition null-steering method is proposed, which can improve system performance, especially for the high data rate downlink with a few subscribers.
2007, 29(12): 2930-2933.
doi: 10.3724/SP.J.1146.2006.00709
Abstract:
The leakage of Channel Impulse Response (CIR) in the time domain channel estimation for the practical OFDM systems is analyzed. A novel channel estimation algorithm in the time domain is proposed for OFDM. The frequency domain windowing is used to reduce the leakage of CIR. The noise of CIR is direct cut according to a threshold. Comparison has been done between the proposed algorithm and the existing time domain channel estimation algorithms in the non-sample-spaced multipath fading channel. Simulation results show the decreasing of channel estimation MSE and the improvement of BER performance by the proposed algorithm.
The leakage of Channel Impulse Response (CIR) in the time domain channel estimation for the practical OFDM systems is analyzed. A novel channel estimation algorithm in the time domain is proposed for OFDM. The frequency domain windowing is used to reduce the leakage of CIR. The noise of CIR is direct cut according to a threshold. Comparison has been done between the proposed algorithm and the existing time domain channel estimation algorithms in the non-sample-spaced multipath fading channel. Simulation results show the decreasing of channel estimation MSE and the improvement of BER performance by the proposed algorithm.
2007, 29(12): 2934-2937.
doi: 10.3724/SP.J.1146.2006.00764
Abstract:
Aimed at the practical OFDM based systems with virtual subcarriers, a low complexity DFT based channel estimation method is proposed. The performance and algorithm complexity of the proposed method are compared with the classic Least Square (LS) and linear minimum mean square error algorithms. The DFT based method can well reduce the noise effect and obtain a large gain compared to LS algorithm. The proposed method has a good tradeoff between performance and complexity, which is promising for practical applications.
Aimed at the practical OFDM based systems with virtual subcarriers, a low complexity DFT based channel estimation method is proposed. The performance and algorithm complexity of the proposed method are compared with the classic Least Square (LS) and linear minimum mean square error algorithms. The DFT based method can well reduce the noise effect and obtain a large gain compared to LS algorithm. The proposed method has a good tradeoff between performance and complexity, which is promising for practical applications.
2007, 29(12): 2938-2942.
doi: 10.3724/SP.J.1146.2006.00727
Abstract:
In multiantenna systems exploiting linear zero-forcing receivers, the Signal-to-Noise Ratio (SNR) on each of the simultaneous sub-streams is not independent with each other. This makes the performance analysis of the multiuser diversity much more difficult. By means of hypothesis testing and the maximum likelihood parameter estimation, this paper uses the Gamma distribution to approximate the probability distribution of the maximum, the minimum and the mean value of the SNRs on all the streams. Based on this, the performance of multiuser diversity for multiantenna systems is analyzed. Simulation shows that the aforementioned approximation closely matches the real distribution and the analysis provides a good estimation on the multiuser diversity gain.
In multiantenna systems exploiting linear zero-forcing receivers, the Signal-to-Noise Ratio (SNR) on each of the simultaneous sub-streams is not independent with each other. This makes the performance analysis of the multiuser diversity much more difficult. By means of hypothesis testing and the maximum likelihood parameter estimation, this paper uses the Gamma distribution to approximate the probability distribution of the maximum, the minimum and the mean value of the SNRs on all the streams. Based on this, the performance of multiuser diversity for multiantenna systems is analyzed. Simulation shows that the aforementioned approximation closely matches the real distribution and the analysis provides a good estimation on the multiuser diversity gain.
2007, 29(12): 2943-2947.
doi: 10.3724/SP.J.1146.2006.02097
Abstract:
The code acquisition is an important issue of DS-UWB system design. Because the MAXimum selection/Threshold Comparison(MAX/TC) criterion for code acquisition has good performance in DS-UWB system, in this paper, a MAX/TC criterion for code acquisition algorithm using the information of the auxiliary sequence, which can be used to reduce the acquisition time of the coarse Pseudo-Noise (PN) sequences, is presented. Then the theoretical analysis is given and simulation tests of proposed acquisition algorithm are carried out in UWB indoor channel environment. The simulation results show that with the proposed algorithm, the mean acquisition time can be reduced considerably while the acquisition performance is maintained.
The code acquisition is an important issue of DS-UWB system design. Because the MAXimum selection/Threshold Comparison(MAX/TC) criterion for code acquisition has good performance in DS-UWB system, in this paper, a MAX/TC criterion for code acquisition algorithm using the information of the auxiliary sequence, which can be used to reduce the acquisition time of the coarse Pseudo-Noise (PN) sequences, is presented. Then the theoretical analysis is given and simulation tests of proposed acquisition algorithm are carried out in UWB indoor channel environment. The simulation results show that with the proposed algorithm, the mean acquisition time can be reduced considerably while the acquisition performance is maintained.
2007, 29(12): 2948-2951.
doi: 10.3724/SP.J.1146.2006.01770
Abstract:
Considering the defect existing in the previous document, the channel capacity formula of MIMO-OFDM systems is derived again, and the effects of the number of sub-carriers, delay spread, angle spread and antenna spacing on the capacity are investigated. The theory analysis and simulation results descript the characters of the channel capacity of MIMO-OFDM systems.
Considering the defect existing in the previous document, the channel capacity formula of MIMO-OFDM systems is derived again, and the effects of the number of sub-carriers, delay spread, angle spread and antenna spacing on the capacity are investigated. The theory analysis and simulation results descript the characters of the channel capacity of MIMO-OFDM systems.
2007, 29(12): 2952-2955.
doi: 10.3724/SP.J.1146.2006.00714
Abstract:
This paper focuses on the rate allocation algorithms on the reverse link (RL) of CDMA2000 system. Two conventional algorithms, the SCHedule (SCH) algorithm and the Dedicated Rate-Controlled (DRC) algorithm are compared and analyzed. Based on their pros and cons, a modified version of DRC algorithm is proposed, which is called Multi-step-Up DRC (MUDRC). While holding the advantages of DRC，the proposed algorithm can combat for the inherent slow ramp-up nature of DRC and enable a mobile to increase its rate by multiple steps when needed. Simulation results show that the proposed algorithm can efficiently utilize the radio resources and improve the users performance of delay. In addition, the performance gain is achieved by using the same amount of signaling overhead as that of DRC, which is much less than that of SCH. The proposed scheme can be generalized to other similar communication systems.
This paper focuses on the rate allocation algorithms on the reverse link (RL) of CDMA2000 system. Two conventional algorithms, the SCHedule (SCH) algorithm and the Dedicated Rate-Controlled (DRC) algorithm are compared and analyzed. Based on their pros and cons, a modified version of DRC algorithm is proposed, which is called Multi-step-Up DRC (MUDRC). While holding the advantages of DRC，the proposed algorithm can combat for the inherent slow ramp-up nature of DRC and enable a mobile to increase its rate by multiple steps when needed. Simulation results show that the proposed algorithm can efficiently utilize the radio resources and improve the users performance of delay. In addition, the performance gain is achieved by using the same amount of signaling overhead as that of DRC, which is much less than that of SCH. The proposed scheme can be generalized to other similar communication systems.
2007, 29(12): 2956-2959.
doi: 10.3724/SP.J.1146.2006.00632
Abstract:
Signaling transmission is the key technology for adaptive frequency hopping radio. In order to improve its safety and reliability, this paper provides a constrained signaling frequency hopping sequence by analyzing the performances and the limitations of the unconstrained signaling frequency-hopping sequence and the wide gap signaling frequency hopping sequence in fractional band jamming. Theory analysis and the simulation results show that the performance of the proposed sequence is better than those of the former two sequences.
Signaling transmission is the key technology for adaptive frequency hopping radio. In order to improve its safety and reliability, this paper provides a constrained signaling frequency hopping sequence by analyzing the performances and the limitations of the unconstrained signaling frequency-hopping sequence and the wide gap signaling frequency hopping sequence in fractional band jamming. Theory analysis and the simulation results show that the performance of the proposed sequence is better than those of the former two sequences.
2007, 29(12): 2960-2964.
doi: 10.3724/SP.J.1146.2006.00515
Abstract:
The relative motion direction of spot-beams decides the handover relation between cells for users in Low Earth Orbit (LEO) satellite systems, and directly affects the performance of handovers for users. User mobility models are developed for two typical cell motion patterns, mobility parameters such as cell dwell time, handover probabilities and average number of handover requests are analyzed. The impact of cell mobility on users handovers has been investigated by simulations for a channel allocation technique. It has been shown that model II has the better performance than model I, and is more applicable for LEO satellite systems with small cell overlaps, reference can be supplied for spot-beams coverage layout.
The relative motion direction of spot-beams decides the handover relation between cells for users in Low Earth Orbit (LEO) satellite systems, and directly affects the performance of handovers for users. User mobility models are developed for two typical cell motion patterns, mobility parameters such as cell dwell time, handover probabilities and average number of handover requests are analyzed. The impact of cell mobility on users handovers has been investigated by simulations for a channel allocation technique. It has been shown that model II has the better performance than model I, and is more applicable for LEO satellite systems with small cell overlaps, reference can be supplied for spot-beams coverage layout.
2007, 29(12): 2965-2969.
doi: 10.3724/SP.J.1146.2006.00644
Abstract:
A fast video object segmentation method working in MPEG compressed domain is presented in this paper. Moving object masks in P frames are extracted by exploiting features obtained by partial decoding. To increase object boundary precision, for each P frame, a 1/16 sub image is constructed using DC and three AC coefficients, and motion compensation information, then a fast mean shift clustering algorithm is used to divide the image into regions with coherence luminance and obtain high precision region boundaries. For reducing the influence of motion vector noise, a MRF-based statistical labeling method is exploited to classify regions into two classes: moving object and background. The proposed algorithm can get a boundary precision of 44 sub-block with a high processing speed. For CIF video streams, the algorithm can run at a speed of 40 frames per second in a Pentium IV 2GHz platform.
A fast video object segmentation method working in MPEG compressed domain is presented in this paper. Moving object masks in P frames are extracted by exploiting features obtained by partial decoding. To increase object boundary precision, for each P frame, a 1/16 sub image is constructed using DC and three AC coefficients, and motion compensation information, then a fast mean shift clustering algorithm is used to divide the image into regions with coherence luminance and obtain high precision region boundaries. For reducing the influence of motion vector noise, a MRF-based statistical labeling method is exploited to classify regions into two classes: moving object and background. The proposed algorithm can get a boundary precision of 44 sub-block with a high processing speed. For CIF video streams, the algorithm can run at a speed of 40 frames per second in a Pentium IV 2GHz platform.
2007, 29(12): 2970-2973.
doi: 10.3724/SP.J.1146.2006.00784
Abstract:
There is great complexity in H.264 encoder. Therefore, it is difficult to implement an encoder in DSP platform which can run in real-time environment. In this paper, the coding architecture of H.264 Baseline Profile is introduced first, and some reasons for limiting speed are presented according to the analysis of the coding architecture. Besides, several implementation techniques for optimization are presented based on the characteristics of TMS320DM642 chip. The H.264 Baseline Profile encoder is finally real-time implemented based on DSP platform. The testing results demonstrate that encoder has sped up remarkably after optimization, which can meet the real-time encoding requirements for the video sequences with CIF format.
There is great complexity in H.264 encoder. Therefore, it is difficult to implement an encoder in DSP platform which can run in real-time environment. In this paper, the coding architecture of H.264 Baseline Profile is introduced first, and some reasons for limiting speed are presented according to the analysis of the coding architecture. Besides, several implementation techniques for optimization are presented based on the characteristics of TMS320DM642 chip. The H.264 Baseline Profile encoder is finally real-time implemented based on DSP platform. The testing results demonstrate that encoder has sped up remarkably after optimization, which can meet the real-time encoding requirements for the video sequences with CIF format.
2007, 29(12): 2974-2977.
doi: 10.3724/SP.J.1146.2006.00702
Abstract:
The probabilistic models for the bilateral stop/go generator are established. The properties such as Markov property, ergodic property and stationarity of the internal state sequence are studied. It is proved that the distributions of 0 and 1 in the output sequence are imbalanced, which implies that this generator can not be used as keystream generator directly. The rate of coincidence between the output sequence and corresponding LFSR sequences, together with their first derivatives are analyzed. The limit properties of the output sequence are also considered.
The probabilistic models for the bilateral stop/go generator are established. The properties such as Markov property, ergodic property and stationarity of the internal state sequence are studied. It is proved that the distributions of 0 and 1 in the output sequence are imbalanced, which implies that this generator can not be used as keystream generator directly. The rate of coincidence between the output sequence and corresponding LFSR sequences, together with their first derivatives are analyzed. The limit properties of the output sequence are also considered.
2007, 29(12): 2978-2980.
doi: 10.3724/SP.J.1146.2006.00676
Abstract:
CICQ (Combined Input and Cross-point-Queued switch) is a fabric system with buffers in cross-points, which is a hotspot in switching area. In this paper, we analyze the current situation, and presented the LEF-LBF scheduling algorithm based on the two conceptions of congestion degree and unblocking degree to serve the urgent cells first. The simulation results show that LFF-LBF can offer very good performance on average delay and stability for uniform traffic with Bernoulli and burst arrivals.
CICQ (Combined Input and Cross-point-Queued switch) is a fabric system with buffers in cross-points, which is a hotspot in switching area. In this paper, we analyze the current situation, and presented the LEF-LBF scheduling algorithm based on the two conceptions of congestion degree and unblocking degree to serve the urgent cells first. The simulation results show that LFF-LBF can offer very good performance on average delay and stability for uniform traffic with Bernoulli and burst arrivals.
2007, 29(12): 2981-2985.
doi: 10.3724/SP.J.1146.2006.00705
Abstract:
In this paper, a novel cooperative routing algorithm is proposed based on a relay selection strategy, where the routing packets on potential relays are exploited to collect the channel state information. Then a distributed algorithm is taken to select the best relay from potential relay set with which a cooperative routing protocol is presented for Ad hoc networks. In layered wireless networks, the implementation of relay selection involves time synchronization and crosslayer optimization. The simulative results show that the proposed algorithm outgoes non-cooperative DSDV routing protocol with considerable gain of deliver ratio and latency.
In this paper, a novel cooperative routing algorithm is proposed based on a relay selection strategy, where the routing packets on potential relays are exploited to collect the channel state information. Then a distributed algorithm is taken to select the best relay from potential relay set with which a cooperative routing protocol is presented for Ad hoc networks. In layered wireless networks, the implementation of relay selection involves time synchronization and crosslayer optimization. The simulative results show that the proposed algorithm outgoes non-cooperative DSDV routing protocol with considerable gain of deliver ratio and latency.
2007, 29(12): 2986-2990.
doi: 10.3724/SP.J.1146.2006.01270
Abstract:
An adaptive strategy based stable link selection algorithm is proposed in this paper, in which the stable neighbor metric and local movement metric are defined. On the basis, by use of the weighting method and weighting coefficients of the two metrics that are acquired by the fuzzy membership function, the stability probability of each link can be computed adaptively so as to find the most stable link or route in a route discovery. The simulation results show that the algorithm can effectively adapt to the network conditions, and outperforms the longevity factor based algorithm and the residual lifetime based algorithm in selecting stable links.
An adaptive strategy based stable link selection algorithm is proposed in this paper, in which the stable neighbor metric and local movement metric are defined. On the basis, by use of the weighting method and weighting coefficients of the two metrics that are acquired by the fuzzy membership function, the stability probability of each link can be computed adaptively so as to find the most stable link or route in a route discovery. The simulation results show that the algorithm can effectively adapt to the network conditions, and outperforms the longevity factor based algorithm and the residual lifetime based algorithm in selecting stable links.
2007, 29(12): 2991-2995.
doi: 10.3724/SP.J.1146.2006.00743
Abstract:
This paper proposes a new method to provide QoS (Quality of Service) in IEEE 802.11e EDCA (Enhanced Distributed Channel Access). Several slots are assembled together to form a super slot, and every slot in the super slot is assigned for a given priority traffic to send packets. The slot assignment is done according to the traffic priority. This new method guarantees the high priority stations have better throughput, less MAC delay and less packet dropping probability than the low priority stations. Compared with the contention window differentiation proposed by IEEE 802.11e EDCA draft, this new method has many advantages such as increasing the throughput, decreasing the packet dropping probability and reducing variety of the high priority traffic throughput when the number of stations increases. This new QoS method is better than contention window differentiation and applying this method can greatly improve the performance of IEEE 802.11e EDCA.
This paper proposes a new method to provide QoS (Quality of Service) in IEEE 802.11e EDCA (Enhanced Distributed Channel Access). Several slots are assembled together to form a super slot, and every slot in the super slot is assigned for a given priority traffic to send packets. The slot assignment is done according to the traffic priority. This new method guarantees the high priority stations have better throughput, less MAC delay and less packet dropping probability than the low priority stations. Compared with the contention window differentiation proposed by IEEE 802.11e EDCA draft, this new method has many advantages such as increasing the throughput, decreasing the packet dropping probability and reducing variety of the high priority traffic throughput when the number of stations increases. This new QoS method is better than contention window differentiation and applying this method can greatly improve the performance of IEEE 802.11e EDCA.
2007, 29(12): 2996-3000.
doi: 10.3724/SP.J.1146.2006.00672
Abstract:
Greedy Forwarding (GF) is an important routing strategy among those routing protocols based on location information in Wireless Sensor Networks (WSN). But there is routing void problem while using greedy forwarding. This paper deduces the probability of a node to be a routing void node theoretically within deterministic deployment and random deployment WSNs, and get the probability of encountering the routing void node while setting up a path randomly. From the analysis the conclusion can be come that when the average number of neighbors is greater than 10, the modified GF algorithm with backoff method is good enough to meet the requirement of WSN applications.
Greedy Forwarding (GF) is an important routing strategy among those routing protocols based on location information in Wireless Sensor Networks (WSN). But there is routing void problem while using greedy forwarding. This paper deduces the probability of a node to be a routing void node theoretically within deterministic deployment and random deployment WSNs, and get the probability of encountering the routing void node while setting up a path randomly. From the analysis the conclusion can be come that when the average number of neighbors is greater than 10, the modified GF algorithm with backoff method is good enough to meet the requirement of WSN applications.
2007, 29(12): 3001-3005.
doi: 10.3724/SP.J.1146.2006.00420
Abstract:
Focusing on throughput and transmission reliability of multimedia delay-sensitive traffic, the dynamic programming and cross-layer design methods are introduced to design QoS routing algorithm in wireless mesh networks. Moreover, a novel Convex Programming Routing Metric QoS (CPRMQS) routing algorithm is proposed as well as an integrated metric (CPRM) QoS model from MAC layer information based on the assumed network model, and then the steps and example using dynamic programming method are presented to solve routing optimization problem. Finally the feasibility of the proposed algorithm is verified with protocol simulation, and the performance evaluations of proposed protocol extended based on DSR are given, including throughput and delay etc.
Focusing on throughput and transmission reliability of multimedia delay-sensitive traffic, the dynamic programming and cross-layer design methods are introduced to design QoS routing algorithm in wireless mesh networks. Moreover, a novel Convex Programming Routing Metric QoS (CPRMQS) routing algorithm is proposed as well as an integrated metric (CPRM) QoS model from MAC layer information based on the assumed network model, and then the steps and example using dynamic programming method are presented to solve routing optimization problem. Finally the feasibility of the proposed algorithm is verified with protocol simulation, and the performance evaluations of proposed protocol extended based on DSR are given, including throughput and delay etc.
United Voting Dynamic Cluster Routing Algorithm Based on Residual-Energy in Wireless Sensor Networks
2007, 29(12): 3006-3010.
doi: 10.3724/SP.J.1146.2006.00715
Abstract:
For monitoring burst events in reactive wireless sensor networks, a new United Voting Dynamic Cluster (UVDC) routing algorithm based on residual-energy is proposed. Cluster head is periodically selected according to residual-energy among the nodes located in the event area, meanwhile a structure of cluster tree rooted at the cluster head is formed within the event area to gather and aggregate data. The data transmitted in network is reduced. Simulation results show that UVDC decreases the average energy consumption, as well as balances the energy dissipation effectively, and extends the lifetime of network.
For monitoring burst events in reactive wireless sensor networks, a new United Voting Dynamic Cluster (UVDC) routing algorithm based on residual-energy is proposed. Cluster head is periodically selected according to residual-energy among the nodes located in the event area, meanwhile a structure of cluster tree rooted at the cluster head is formed within the event area to gather and aggregate data. The data transmitted in network is reduced. Simulation results show that UVDC decreases the average energy consumption, as well as balances the energy dissipation effectively, and extends the lifetime of network.
2007, 29(12): 3011-3014.
doi: 10.3724/SP.J.1146.2006.00712
Abstract:
To filter false positive alerts generated by Intrusion Detection Systems (IDS), 19 related attributes for distinguishing false positive alerts from true alerts are summarized according to the root and timeliness of intrusion alerts, and an approach to filter these false positive alerts based on RS-SVM (Rough Set and Support Vector Machine) theory is proposed. First, redundant attributes are removed and 10 attributes are obtained utilizing rough set theory in the proposed approach. Then the problem of filtering false positive alerts on the dataset with those 10 attributes is transformed to classification problem, and the classifier is constructed using support vector machine theory. The experimental data is the alert dataset raised by Snort, a network intrusion detection system, monitoring the Defense Advanced Research Projects Agency 1999 intrusion evaluation data (DARPA99). The experimental results show that the proposed approach can reduce about 98% false positive alerts at the cost of increasing about 1.6% false negative alerts. The results of this method are better than those of the other methods that adopt the same dataset and same IDS reported in the literature.
To filter false positive alerts generated by Intrusion Detection Systems (IDS), 19 related attributes for distinguishing false positive alerts from true alerts are summarized according to the root and timeliness of intrusion alerts, and an approach to filter these false positive alerts based on RS-SVM (Rough Set and Support Vector Machine) theory is proposed. First, redundant attributes are removed and 10 attributes are obtained utilizing rough set theory in the proposed approach. Then the problem of filtering false positive alerts on the dataset with those 10 attributes is transformed to classification problem, and the classifier is constructed using support vector machine theory. The experimental data is the alert dataset raised by Snort, a network intrusion detection system, monitoring the Defense Advanced Research Projects Agency 1999 intrusion evaluation data (DARPA99). The experimental results show that the proposed approach can reduce about 98% false positive alerts at the cost of increasing about 1.6% false negative alerts. The results of this method are better than those of the other methods that adopt the same dataset and same IDS reported in the literature.
2007, 29(12): 3015-3019.
doi: 10.3724/SP.J.1146.2006.00510
Abstract:
In this paper, a novel Quasi Particle Swarm Optimization(Quasi-PSO)based on Linear Decreasing Weight Particle Swarm Optimization(LDW-PSO) is presented. In order to improve the optimization speed and convergence performance, the novel algorithm is built by employing the idea of crossover of the genetic algorithm and adopting an antithetic algorithmic model. The simulation results show that it has achieved relatively high performance in the radiation pattern synthesis of antenna arrays.
In this paper, a novel Quasi Particle Swarm Optimization(Quasi-PSO)based on Linear Decreasing Weight Particle Swarm Optimization(LDW-PSO) is presented. In order to improve the optimization speed and convergence performance, the novel algorithm is built by employing the idea of crossover of the genetic algorithm and adopting an antithetic algorithmic model. The simulation results show that it has achieved relatively high performance in the radiation pattern synthesis of antenna arrays.
2007, 29(12): 3020-3023.
doi: 10.3724/SP.J.1146.2006.00756
Abstract:
A novel Split-Ring Resonator (SRR) Defected Ground Structure (DGS) is presented. Compared with the conventional dumbbell DGS, this structure has a sharp band-gap and flat low-pass characteristic. A detailed analysis of the equivalent circuit of the cell model and the relationship between physical sizes and electrical parameters is made. Then an improved SRR DGS cell model loaded with open stubs is proposed in order to improve the out-band suppression. Based on this structure, a compact S-band microstrip lowpass filter is designed and fabricated. The measured results show that, this filter has a low insertion loss of below 0.5dB and a high out-band suppression of more than 30dB.
A novel Split-Ring Resonator (SRR) Defected Ground Structure (DGS) is presented. Compared with the conventional dumbbell DGS, this structure has a sharp band-gap and flat low-pass characteristic. A detailed analysis of the equivalent circuit of the cell model and the relationship between physical sizes and electrical parameters is made. Then an improved SRR DGS cell model loaded with open stubs is proposed in order to improve the out-band suppression. Based on this structure, a compact S-band microstrip lowpass filter is designed and fabricated. The measured results show that, this filter has a low insertion loss of below 0.5dB and a high out-band suppression of more than 30dB.
2007, 29(12): 3024-3026.
doi: 10.3724/SP.J.1146.2006.00739
Abstract:
On the basis of studying chirp signal detection and parameter estimation using Radon-ambiguity transform and fractional Fourier transform, a novel iterative algorithm is put forward, which is proved much higher calculation speed through simulation.
On the basis of studying chirp signal detection and parameter estimation using Radon-ambiguity transform and fractional Fourier transform, a novel iterative algorithm is put forward, which is proved much higher calculation speed through simulation.
2007, 29(12): 3027-3030.
doi: 10.3724/SP.J.1146.2006.02094
Abstract:
Based on the basic group-based MAximum a Posterior(MAP) detector, a novel detection scheme is proposed that uses a posterior information from the detected groups instead of a priori information from the decoder. The new scheme can improve the estimation precision of the interference. Further an iterative detector that uses a posterior information from last detection is presented. The iterative detector can fully use the output information from the detector and have a performance improvement further. Simulation results show that the new detectors have a significant performance gain over basic group-based MAP algorithm in uncoded systems and have a faster convergence in coded systems.
Based on the basic group-based MAximum a Posterior(MAP) detector, a novel detection scheme is proposed that uses a posterior information from the detected groups instead of a priori information from the decoder. The new scheme can improve the estimation precision of the interference. Further an iterative detector that uses a posterior information from last detection is presented. The iterative detector can fully use the output information from the detector and have a performance improvement further. Simulation results show that the new detectors have a significant performance gain over basic group-based MAP algorithm in uncoded systems and have a faster convergence in coded systems.
2007, 29(12): 3031-3034.
doi: 10.3724/SP.J.1146.2006.00627
Abstract:
Future network is characterized by the integration of heterogeneous wireless access technologies. The IETF Mobile IP (MIP) integrates these different networks into a single IP-based platform. This paper focuses on the analysis of movement detection mechanisms for MIP: LCS, ECS. A cross-layer fast detection method in heterogeneous overlay networks is proposed. Analysis and simulation show that with the information of link layer, the proposed scheme can achieve better performance in terms of handoff latency and packet loss.
Future network is characterized by the integration of heterogeneous wireless access technologies. The IETF Mobile IP (MIP) integrates these different networks into a single IP-based platform. This paper focuses on the analysis of movement detection mechanisms for MIP: LCS, ECS. A cross-layer fast detection method in heterogeneous overlay networks is proposed. Analysis and simulation show that with the information of link layer, the proposed scheme can achieve better performance in terms of handoff latency and packet loss.
