Email alert
2017 Vol. 39, No. 9
Display Method:
2017, 39(9): 2033-2039.
doi: 10.11999/JEIT161116
Abstract:
Focusing on the complexity of network structure and the indeterminacy of community partition, this paper puts forward a novel fuzzy clustering method for uncovering community structures. In contrast to previous studies, the proposed method disposes the similarity of connecting vertices with fuzzy relation. Based on local interactive information, it considers the fuzzy relation between vertices and the transitive similarity in network topology to divide vertices into communities. In addition, multiresolution communities can be detected by adjusting fuzzy parameter. In order to avoid subjectivity in the selection of cluster number, a new modularity is introduced to evaluate the effectiveness of the clustering analysis. It is proved by experiments that the method is ef?cient and stable to detect underlying communities.
Focusing on the complexity of network structure and the indeterminacy of community partition, this paper puts forward a novel fuzzy clustering method for uncovering community structures. In contrast to previous studies, the proposed method disposes the similarity of connecting vertices with fuzzy relation. Based on local interactive information, it considers the fuzzy relation between vertices and the transitive similarity in network topology to divide vertices into communities. In addition, multiresolution communities can be detected by adjusting fuzzy parameter. In order to avoid subjectivity in the selection of cluster number, a new modularity is introduced to evaluate the effectiveness of the clustering analysis. It is proved by experiments that the method is ef?cient and stable to detect underlying communities.
2017, 39(9): 2040-2047.
doi: 10.11999/JEIT161266
Abstract:
With the expanding scale of social networks, the label classification of nodes in the network is no longer single but various, which prompts the multi-label classification in social networks to become an important research area. The previous research focuses on how to improve the precision of the predicted labels, while ignoring the system overhead caused by obtaining the node information, such as time consumption and computing memory occupancy. Now, as both expansion and complexity of the networks are increasing, the problem of previously neglected system overhead is becoming the more and the more serious. It increases not only the cost but also the difficulty of predicting labels. In this paper, an NSGA2-based multi-label seed selection algorithm in network environments (NAMESEA) is proposed to improve the accuracy of label prediction on the condition that reducing both the time consume and the memory occupancy. Compared with other multi-label prediction algorithms on multiple real datasets, NAMES
With the expanding scale of social networks, the label classification of nodes in the network is no longer single but various, which prompts the multi-label classification in social networks to become an important research area. The previous research focuses on how to improve the precision of the predicted labels, while ignoring the system overhead caused by obtaining the node information, such as time consumption and computing memory occupancy. Now, as both expansion and complexity of the networks are increasing, the problem of previously neglected system overhead is becoming the more and the more serious. It increases not only the cost but also the difficulty of predicting labels. In this paper, an NSGA2-based multi-label seed selection algorithm in network environments (NAMESEA) is proposed to improve the accuracy of label prediction on the condition that reducing both the time consume and the memory occupancy. Compared with other multi-label prediction algorithms on multiple real datasets, NAMES
2017, 39(9): 2048-2055.
doi: 10.11999/JEIT160975
Abstract:
This paper presents a Deep Belief Nets (DBN) model and a multi-modality feature extraction method to extend features, dimensionalities of short text for Chinese microblogging sentiment classification. Besides traditional features sets for document classification, comments for certain posts are also extracted as part of the microblogging features according to the relationship between commenters and posters through constructing microblogging social network as input information. Multi-modality features are combined and adopted as the input vector for DBN. A DBN model, which is stacked with several layers of Restricted Boltzmann Machine (RBM), is implemented to initialize the structure of neural network. The RBM layers can take probability distribution samples of input data to learn hidden syntactic structures for better feature representation. A Classification RBM (ClassRBM) layer, which is stacked on top of the former RBM layers, is adapted to achieve the final sentiment classification. The results demonstrate that, with proper structure and parameter, the performance of the proposed deep learning method on sentiment classification is better than the state of the art surface learning models such as SVM or NB, which proves that DBN is suitable for short-length document classification with the proposed feature dimensionality extension method.
This paper presents a Deep Belief Nets (DBN) model and a multi-modality feature extraction method to extend features, dimensionalities of short text for Chinese microblogging sentiment classification. Besides traditional features sets for document classification, comments for certain posts are also extracted as part of the microblogging features according to the relationship between commenters and posters through constructing microblogging social network as input information. Multi-modality features are combined and adopted as the input vector for DBN. A DBN model, which is stacked with several layers of Restricted Boltzmann Machine (RBM), is implemented to initialize the structure of neural network. The RBM layers can take probability distribution samples of input data to learn hidden syntactic structures for better feature representation. A Classification RBM (ClassRBM) layer, which is stacked on top of the former RBM layers, is adapted to achieve the final sentiment classification. The results demonstrate that, with proper structure and parameter, the performance of the proposed deep learning method on sentiment classification is better than the state of the art surface learning models such as SVM or NB, which proves that DBN is suitable for short-length document classification with the proposed feature dimensionality extension method.
2017, 39(9): 2056-2062.
doi: 10.11999/JEIT161348
Abstract:
It is an important measure to utilize the topic model to recognize the users focuses on social networks, such as blog, online community, and microblog. Considering the particularity of topic recognizing of short texts on the social network platform, this paper develops an AW-LDA model based on mixed-weight combined strategy according to the relevance of short texts context. This model virtually combines short texts, which are in line with contextual-related conditions, and endows different short texts with different weights according to the related extent. It proposes a new method of recognizing short texts topics. According to the experiments on data of BBS and Weibo communities, the results show that the model can effectively recognize social network users focuses on different subjects and it proposes a new idea about solving the topic recognition problem of short texts.
It is an important measure to utilize the topic model to recognize the users focuses on social networks, such as blog, online community, and microblog. Considering the particularity of topic recognizing of short texts on the social network platform, this paper develops an AW-LDA model based on mixed-weight combined strategy according to the relevance of short texts context. This model virtually combines short texts, which are in line with contextual-related conditions, and endows different short texts with different weights according to the related extent. It proposes a new method of recognizing short texts topics. According to the experiments on data of BBS and Weibo communities, the results show that the model can effectively recognize social network users focuses on different subjects and it proposes a new idea about solving the topic recognition problem of short texts.
2017, 39(9): 2063-2070.
doi: 10.11999/JEIT161360
Abstract:
Influence blocking maximization is currently a focused issue in the research area of social networks. This paper considers the issue of influence blocking maximization with uncertain negative influence sources. First, in order to increase efficiency of blocking seeds mining algorithms, the approximate estimation method of influence propagation of negative seeds under the competitive linear threshold model is discussed. Based on the estimation, a blocking seeds mining algorithm for finite uncertain negatively influence sources is proposed to maximize expected influence blocking utility. Second, for the case of huge amount of negatively influence sources with uncertainty, a blocking seeds mining algorithm based on the sampling average approximation approach is proposed to balance the tradeoffs between scalability and effectiveness of the influence blocking maximization. Finally, experiments are carried on real data sets of social networks to verify the feasibility and scalability of the proposed algorithms.
Influence blocking maximization is currently a focused issue in the research area of social networks. This paper considers the issue of influence blocking maximization with uncertain negative influence sources. First, in order to increase efficiency of blocking seeds mining algorithms, the approximate estimation method of influence propagation of negative seeds under the competitive linear threshold model is discussed. Based on the estimation, a blocking seeds mining algorithm for finite uncertain negatively influence sources is proposed to maximize expected influence blocking utility. Second, for the case of huge amount of negatively influence sources with uncertainty, a blocking seeds mining algorithm based on the sampling average approximation approach is proposed to balance the tradeoffs between scalability and effectiveness of the influence blocking maximization. Finally, experiments are carried on real data sets of social networks to verify the feasibility and scalability of the proposed algorithms.
2017, 39(9): 2071-2080.
doi: 10.11999/JEIT170102
Abstract:
Community detection method is significant to character statistics of complex network. Community detection in directed structured network is an attractive research problem while most previous approaches attempt to divide undirected networks into communities while there has appeared many large scale directed social network such as WeChat circle of friends and Sina Micro-Blog. To solve the problem that low quality of model, low efficiency of execution and high deviation of precision from the conventional community detection algorithm on large-scale social network and directed network, this paper provides an approach that starts with the triangle structure of community basis and models the local information transfer to detect community in large-scale directed social network. Basing on the directed vector theory in probability graph and the high information transfer gain of vertex in directed network, this paper constructs the Information Transfer Gain (ITG) method and the corresponding target functions for evaluating the quality of a specific partition in community detection algorithm. Then the combine of ITG with the target function to compose the new community detection algorithm for directed network. Extensive experiments in synthetic signed network and real-life large networks derived from online social media, it is proved that the proposed method is more accurate and faster than several traditional community detection methods such as FastGN, OSLOM and Infomap.
Community detection method is significant to character statistics of complex network. Community detection in directed structured network is an attractive research problem while most previous approaches attempt to divide undirected networks into communities while there has appeared many large scale directed social network such as WeChat circle of friends and Sina Micro-Blog. To solve the problem that low quality of model, low efficiency of execution and high deviation of precision from the conventional community detection algorithm on large-scale social network and directed network, this paper provides an approach that starts with the triangle structure of community basis and models the local information transfer to detect community in large-scale directed social network. Basing on the directed vector theory in probability graph and the high information transfer gain of vertex in directed network, this paper constructs the Information Transfer Gain (ITG) method and the corresponding target functions for evaluating the quality of a specific partition in community detection algorithm. Then the combine of ITG with the target function to compose the new community detection algorithm for directed network. Extensive experiments in synthetic signed network and real-life large networks derived from online social media, it is proved that the proposed method is more accurate and faster than several traditional community detection methods such as FastGN, OSLOM and Infomap.
2017, 39(9): 2081-2088.
doi: 10.11999/JEIT170030
Abstract:
To explore the distribution and correlation from massive Twitter data helps the accurate personalized recommendation. On-Line Analytical Processing (OLAP) provides an intuitive form that is suitable for people to explore the Twitter data. The key of applying OLAP to Twitter data is how to mine and build dimension hierarchy of tweeter interests. Different from the existing approaches that can extract interests of tweeters with only one level, an approach to the extraction of dimension hierarchy of interests for OLAP is proposed. Firstly, it retrieves the Twitter data through RestAPI. Afterwards, it detects the interests and sub-interests using an improved (Latent Dirichlet Allocation, LDA) model. Based on the extracted interests and sub-interests it finally constructs the dimension hierarchy of interests. The experiment verifies its effectiveness and scalability, and demonstrates it can extract dimension hierarchy of tweeters interests for OLAP more effectively than LDA and hLDA.
To explore the distribution and correlation from massive Twitter data helps the accurate personalized recommendation. On-Line Analytical Processing (OLAP) provides an intuitive form that is suitable for people to explore the Twitter data. The key of applying OLAP to Twitter data is how to mine and build dimension hierarchy of tweeter interests. Different from the existing approaches that can extract interests of tweeters with only one level, an approach to the extraction of dimension hierarchy of interests for OLAP is proposed. Firstly, it retrieves the Twitter data through RestAPI. Afterwards, it detects the interests and sub-interests using an improved (Latent Dirichlet Allocation, LDA) model. Based on the extracted interests and sub-interests it finally constructs the dimension hierarchy of interests. The experiment verifies its effectiveness and scalability, and demonstrates it can extract dimension hierarchy of tweeters interests for OLAP more effectively than LDA and hLDA.
2017, 39(9): 2089-2096.
doi: 10.11999/JEIT170246
Abstract:
Sybil attackers can improve their own influence in social networks by creating a large number of illegal illusive identities then affect the social individuals choice of relays and steal individuals privacy, which seriously threatens the interests of social individuals. Based on the analysis of the Sybils behaviors, a Sybil detection mechanism applied to social networks is proposed in this paper. The influence of nodes is calculated according to static similarity and dynamic similarity and then selecting the suspicious nodes based on the influence. Next, using the Hidden Markov Model (HMM) to infer the true identity of suspicious nodes by observing their abnormal behaviors, thus detecting the Sybil more precisely. Analysis results show that the proposed mechanism can effectively improve the recognition rate and reduce the false detection rate of the Sybil and thereby protecting the privacy and interests of social individuals better.
Sybil attackers can improve their own influence in social networks by creating a large number of illegal illusive identities then affect the social individuals choice of relays and steal individuals privacy, which seriously threatens the interests of social individuals. Based on the analysis of the Sybils behaviors, a Sybil detection mechanism applied to social networks is proposed in this paper. The influence of nodes is calculated according to static similarity and dynamic similarity and then selecting the suspicious nodes based on the influence. Next, using the Hidden Markov Model (HMM) to infer the true identity of suspicious nodes by observing their abnormal behaviors, thus detecting the Sybil more precisely. Analysis results show that the proposed mechanism can effectively improve the recognition rate and reduce the false detection rate of the Sybil and thereby protecting the privacy and interests of social individuals better.
2017, 39(9): 2097-2107.
doi: 10.11999/JEIT161192
Abstract:
Groups are important mesoscopic organizations of Online Social Networks (OSNs). Group detection not only has important theoretical significance, but also has a wide range of applications. It promotes the application and development of online social networks. In this paper, group detection technology in online social networks is studied. Based on analyzing the formation mechanism of social groups, the online social network groups is defined and the group detection problem is introduced. According to different features adopted by group detection methods, the methods based on the attribute features only and those based on combination of attribute features and structure features are analyzed, respectively. Especially, it reviews the malicious behavior group detection methods by analyzing their feature selection mechanisms and detection models in detail. Finally, further research direction of group detection in online social networks is prospected.
Groups are important mesoscopic organizations of Online Social Networks (OSNs). Group detection not only has important theoretical significance, but also has a wide range of applications. It promotes the application and development of online social networks. In this paper, group detection technology in online social networks is studied. Based on analyzing the formation mechanism of social groups, the online social network groups is defined and the group detection problem is introduced. According to different features adopted by group detection methods, the methods based on the attribute features only and those based on combination of attribute features and structure features are analyzed, respectively. Especially, it reviews the malicious behavior group detection methods by analyzing their feature selection mechanisms and detection models in detail. Finally, further research direction of group detection in online social networks is prospected.
2017, 39(9): 2108-2118.
doi: 10.11999/JEIT161273
Abstract:
Recently, social network applications develop dramatically. Social network related social behaviors are one of the most important areas, which receive broadly attentions from researchers in academics. This survey paper analyzes the social behaviors comprehensively with respect to the causes of social behaviors, the performance of social behaviors and the influence of social behaviors. More specifically, after analyzing the basic concepts of behaviors in social networks, the three most important causes of behaviors in social networks are firstly introduced, such as user adoption, user loyalty and user trust. With these causes, the performance of user behaviors in social networks can be analyzed with respect to three common behaviors, including general usage behaviors, content creation behaviors and content consumption behaviors. Finally, the research on the influence of social network behaviors is presented, which includes the most important aspects, such as influence evaluation and behavior induction. The systematical analysis of social behaviors points out the future directions of related research in next steps.
Recently, social network applications develop dramatically. Social network related social behaviors are one of the most important areas, which receive broadly attentions from researchers in academics. This survey paper analyzes the social behaviors comprehensively with respect to the causes of social behaviors, the performance of social behaviors and the influence of social behaviors. More specifically, after analyzing the basic concepts of behaviors in social networks, the three most important causes of behaviors in social networks are firstly introduced, such as user adoption, user loyalty and user trust. With these causes, the performance of user behaviors in social networks can be analyzed with respect to three common behaviors, including general usage behaviors, content creation behaviors and content consumption behaviors. Finally, the research on the influence of social network behaviors is presented, which includes the most important aspects, such as influence evaluation and behavior induction. The systematical analysis of social behaviors points out the future directions of related research in next steps.
2017, 39(9): 2119-2126.
doi: 10.11999/JEIT161285
Abstract:
Wide-width bit permutation is a very commonly used operation in symmetric cryptographic algorithms. However, current word-oriented general microprocessors are inefficient to cope with the complex bit-level permutation operations. To solve this problem, two schemes for 2N-2N and kN-kN permutations are proposed respectively, including two extended instructions BEX and BEX-ROT. Furthermore, the efficient hardware implementation of the instructions are studied, and then a unified hardware circuit named RERS (Reconfigurable Extract and Rotation Shifter) is proposed with a corresponding reconfigurable routing algorithm. The RERS can share hardware resources to achieve the purpose of reducing area. The experimental results show that the proposed schemes can truly decrease the number of instructions for accomplishing an arbitrary wide-width bit permutation (instructions reduced by 10 times), which greatly accelerate the performance of microprocessors. At the same time, the overhead of hardware resources and delay caused by the two extended instructions is very low, which will not affect the normal operating frequency of the original microprocessors.
Wide-width bit permutation is a very commonly used operation in symmetric cryptographic algorithms. However, current word-oriented general microprocessors are inefficient to cope with the complex bit-level permutation operations. To solve this problem, two schemes for 2N-2N and kN-kN permutations are proposed respectively, including two extended instructions BEX and BEX-ROT. Furthermore, the efficient hardware implementation of the instructions are studied, and then a unified hardware circuit named RERS (Reconfigurable Extract and Rotation Shifter) is proposed with a corresponding reconfigurable routing algorithm. The RERS can share hardware resources to achieve the purpose of reducing area. The experimental results show that the proposed schemes can truly decrease the number of instructions for accomplishing an arbitrary wide-width bit permutation (instructions reduced by 10 times), which greatly accelerate the performance of microprocessors. At the same time, the overhead of hardware resources and delay caused by the two extended instructions is very low, which will not affect the normal operating frequency of the original microprocessors.
2017, 39(9): 2127-2133.
doi: 10.11999/JEIT170062
Abstract:
In order to ensure the confidentiality and authentication in different network environments, the security model of IDPKC-to-CLPKC heterogeneous signcryption is defined from IDentity-based Public Key Cryptography (IDPKC) to CertificateLess Public Key Cryptography (CLPKC), and a concrete IDPKC-to-CLPKC heterogeneous signcryption scheme is presented. The system parameters in IDPKC and CLPKC are independent on each other in the scheme, which can meet the practical requirements. Based on the assumptions of Gap Bilinear Diffie-Hellman (GBDH), Computational Diffie-Hellman (CDH) and q-Strong Diffie-Hellman (q-SDH), the scheme is proved to satisfy the confidentiality and unforgeability in the random oracle model. Moreover, the scheme is also proved to satisfy the properties of ciphertext anonymity, which means the attacker can not judge the identities of the sender and the receiver. Therefore, the scheme can effectively protect the privacy of both identities.
In order to ensure the confidentiality and authentication in different network environments, the security model of IDPKC-to-CLPKC heterogeneous signcryption is defined from IDentity-based Public Key Cryptography (IDPKC) to CertificateLess Public Key Cryptography (CLPKC), and a concrete IDPKC-to-CLPKC heterogeneous signcryption scheme is presented. The system parameters in IDPKC and CLPKC are independent on each other in the scheme, which can meet the practical requirements. Based on the assumptions of Gap Bilinear Diffie-Hellman (GBDH), Computational Diffie-Hellman (CDH) and q-Strong Diffie-Hellman (q-SDH), the scheme is proved to satisfy the confidentiality and unforgeability in the random oracle model. Moreover, the scheme is also proved to satisfy the properties of ciphertext anonymity, which means the attacker can not judge the identities of the sender and the receiver. Therefore, the scheme can effectively protect the privacy of both identities.
2017, 39(9): 2134-2140.
doi: 10.11999/JEIT161347
Abstract:
In order to solve the load imbalance problem caused by uncertainty of traffic in heterogeneous dense cellular networks, this paper proposes a load balance algorithm through small cell range expansion. The proposed algorithm is based on Partially Observable Markov Decision Process (POMDP). By observing the packets of system user during the perceptual cycle, the next cycle system possible load state can be dopted. Then, the Dynamic Cell Range Expansion (DCRE) offset value is dynamically adjusted to take action in advance, reaching the purpose of optimizing the system load balance. To solve the problem efficiently, a heuristic algorithm is used to approximate and quickly get the suboptimal solution. Simulation results show that the proposed method can achieve load balance optimization in dense hetrogeneous network, and improve the system user throughput and resource utilization rate.
In order to solve the load imbalance problem caused by uncertainty of traffic in heterogeneous dense cellular networks, this paper proposes a load balance algorithm through small cell range expansion. The proposed algorithm is based on Partially Observable Markov Decision Process (POMDP). By observing the packets of system user during the perceptual cycle, the next cycle system possible load state can be dopted. Then, the Dynamic Cell Range Expansion (DCRE) offset value is dynamically adjusted to take action in advance, reaching the purpose of optimizing the system load balance. To solve the problem efficiently, a heuristic algorithm is used to approximate and quickly get the suboptimal solution. Simulation results show that the proposed method can achieve load balance optimization in dense hetrogeneous network, and improve the system user throughput and resource utilization rate.
2017, 39(9): 2141-2147.
doi: 10.11999/JEIT161312
Abstract:
In order to solve the problem of unhomogeneities of estimation error and expensive computing of existing algorithms, an iterative frequency estimation algorithm based on interpolated zoom spectrum is proposed. Firstly, fast Fourier transform algorithm is applied to get the frequency corresponding to the peak spectral amplitude of the half-length signal. The unbiased estimation of frequency of the signal is then given based on the zoom spectra, which are calculated with the half-length signal. The zoom spectra are updated with the complete signal and the frequency is estimated with the updated zoom spectra, lastly. Computing cost analysis proves the superiority of the algorithms when length of signal is long compared with the algorithms in the references. Simulation result verifies good performance of distribution of estimation error and estimation error of the proposed algorithm is closer to the Cramer-Rao lower bound at the circumstance of high signal to noise ratio.
In order to solve the problem of unhomogeneities of estimation error and expensive computing of existing algorithms, an iterative frequency estimation algorithm based on interpolated zoom spectrum is proposed. Firstly, fast Fourier transform algorithm is applied to get the frequency corresponding to the peak spectral amplitude of the half-length signal. The unbiased estimation of frequency of the signal is then given based on the zoom spectra, which are calculated with the half-length signal. The zoom spectra are updated with the complete signal and the frequency is estimated with the updated zoom spectra, lastly. Computing cost analysis proves the superiority of the algorithms when length of signal is long compared with the algorithms in the references. Simulation result verifies good performance of distribution of estimation error and estimation error of the proposed algorithm is closer to the Cramer-Rao lower bound at the circumstance of high signal to noise ratio.
2017, 39(9): 2148-2154.
doi: 10.11999/JEIT161304
Abstract:
Depth information is proved to be an important part of human vision. However, most saliency detection methods based on 2D images do not make good use of depth information, thus an effective saliency detection method for RGB-D image is presented. It extracts color feature combined with depth saliency feature and detects salient objects based on photographic composition prior and background prior. First, original depth map is preprocessed to form depth saliency feature by background vertex area, photographic composition intersections, and compactness method. Then the association matrix is constructed by the adjacency weight of comprehensive feature. Manifold ranking is running from foreground view to form foreground saliency map based on photographic composition prior and fusion of depth saliency feature and color feature. In order to correct the error caused by assumption, the boundary connectivity is used to suppress background from background view. Final saliency map builds on fusion of foreground and background saliency map. Experiments compared with 4 different methods on RGB-D1000 database show that the proposed method has better precision-recall curve and outperforms the state- of-the-art methods.
Depth information is proved to be an important part of human vision. However, most saliency detection methods based on 2D images do not make good use of depth information, thus an effective saliency detection method for RGB-D image is presented. It extracts color feature combined with depth saliency feature and detects salient objects based on photographic composition prior and background prior. First, original depth map is preprocessed to form depth saliency feature by background vertex area, photographic composition intersections, and compactness method. Then the association matrix is constructed by the adjacency weight of comprehensive feature. Manifold ranking is running from foreground view to form foreground saliency map based on photographic composition prior and fusion of depth saliency feature and color feature. In order to correct the error caused by assumption, the boundary connectivity is used to suppress background from background view. Final saliency map builds on fusion of foreground and background saliency map. Experiments compared with 4 different methods on RGB-D1000 database show that the proposed method has better precision-recall curve and outperforms the state- of-the-art methods.
2017, 39(9): 2155-2161.
doi: 10.11999/JEIT161391
Abstract:
In order to solve the defects of poor error tolerance and large calculation amount in current Turbo-code encoder recognition algorithms, a new algorithm based on conformity of parity-check equation is proposed. Firstly, according to small code constraint length, the coding polynomial database is built, then each polynomial of average conformity of parity-check equation is calculated by traversing the database, as a result, the Turbo-code encoder can be realized, because the correct polynomial can make the conformity maximum. The algorithm has small amount of calculation because of finite traversal, which is only related to amount of intercepted data, besides, this algorithm has good error tolerance by soft decisions. The simulation results show that the correct ratio of recognition can reach 90% at SNR of 0 dB by the 10 intercepted code blocks, and the length of each block is 100.
In order to solve the defects of poor error tolerance and large calculation amount in current Turbo-code encoder recognition algorithms, a new algorithm based on conformity of parity-check equation is proposed. Firstly, according to small code constraint length, the coding polynomial database is built, then each polynomial of average conformity of parity-check equation is calculated by traversing the database, as a result, the Turbo-code encoder can be realized, because the correct polynomial can make the conformity maximum. The algorithm has small amount of calculation because of finite traversal, which is only related to amount of intercepted data, besides, this algorithm has good error tolerance by soft decisions. The simulation results show that the correct ratio of recognition can reach 90% at SNR of 0 dB by the 10 intercepted code blocks, and the length of each block is 100.
2017, 39(9): 2162-2168.
doi: 10.11999/JEIT161317
Abstract:
To overcome the shortcoming of Intuitionistic Fuzzy C-Means (IFCM) that it does not take into account the spatial information, a new Kernel-based algorithm with Weighted Spatial Information (KWSI_IFCM) is proposed. Firstly, the constraint of weighted spatial neighborhood information is added. Secondly, instead of Euclidean distance, kernel-induced function is used to measure the distance between pixels and cluster centers. Thirdly, a new clustering objective function is created and then the iterative expressions of new membership and clustering centers are obtained by optimizing the new function. The quantitative analysis of image segmentation results using the new algorithm, other similar methods and a binarization method based on salient transition region shows that the new algorithm can get the F-measure value with 0.9776. The experimental results demonstrate that the proposed algorithm can obtain higher stability and segmentation accuracy than similar fuzzy C-mean algorithm.
To overcome the shortcoming of Intuitionistic Fuzzy C-Means (IFCM) that it does not take into account the spatial information, a new Kernel-based algorithm with Weighted Spatial Information (KWSI_IFCM) is proposed. Firstly, the constraint of weighted spatial neighborhood information is added. Secondly, instead of Euclidean distance, kernel-induced function is used to measure the distance between pixels and cluster centers. Thirdly, a new clustering objective function is created and then the iterative expressions of new membership and clustering centers are obtained by optimizing the new function. The quantitative analysis of image segmentation results using the new algorithm, other similar methods and a binarization method based on salient transition region shows that the new algorithm can get the F-measure value with 0.9776. The experimental results demonstrate that the proposed algorithm can obtain higher stability and segmentation accuracy than similar fuzzy C-mean algorithm.
2017, 39(9): 2169-2174.
doi: 10.11999/JEIT170040
Abstract:
Owing to its flexible frequency decomposition ability, nonuniform filter banks are widely applied to speech and image signal processing. However, the nonuniform Discrete Fourier Transform (DFT) modulated filter bank can not be constructed by directly merging certain subbands of the uniform one. In order to overcome this deficiency, a novel construction approach is proposed, in which the subband filters of the nonuniform filter bank are obtained by jointly employing the subband merging and phase modulation of the uniform one. The resultant nonuniform filter bank exhibits the very approximate overall performance as the uniform one. Moreover, the conditions are derived for the nonuniform DFT modulated filter banks to possess satisfactory frequency characteristics. Both the theoretical analysis and simulation results show the effectiveness of the proposed method.
Owing to its flexible frequency decomposition ability, nonuniform filter banks are widely applied to speech and image signal processing. However, the nonuniform Discrete Fourier Transform (DFT) modulated filter bank can not be constructed by directly merging certain subbands of the uniform one. In order to overcome this deficiency, a novel construction approach is proposed, in which the subband filters of the nonuniform filter bank are obtained by jointly employing the subband merging and phase modulation of the uniform one. The resultant nonuniform filter bank exhibits the very approximate overall performance as the uniform one. Moreover, the conditions are derived for the nonuniform DFT modulated filter banks to possess satisfactory frequency characteristics. Both the theoretical analysis and simulation results show the effectiveness of the proposed method.
2017, 39(9): 2175-2181.
doi: 10.11999/JEIT161132
Abstract:
In order to solve the problem that traditional uniform FFT beamforming algorithm reduces the resolution of 3-D sonar imaging, this paper presents a sub-region FFT beamforming algorithm. In the far field, the imaging area is partitioned into multiple regions using genetic algorithm as the optimization method. The objective of the optimization process is to minimize the number of partitions, with the constraints being the imaging resolution. In each region, a beam direction is selected to obtain the demodulated output when each receiving element receives the directional echo as the original data, and the traditional uniform FFT beamforming is performed in the region. The FFT computation process is optimized to reduce the computational complexity of the new algorithm to meet the real-time requirements of 3D imaging sonar. Simulation and experimental results show that the imaging resolution of the sub-region FFT beamforming algorithm is significantly higher than that of the traditional uniform FFT beamforming algorithm, and satisfies the real-time requirement.
In order to solve the problem that traditional uniform FFT beamforming algorithm reduces the resolution of 3-D sonar imaging, this paper presents a sub-region FFT beamforming algorithm. In the far field, the imaging area is partitioned into multiple regions using genetic algorithm as the optimization method. The objective of the optimization process is to minimize the number of partitions, with the constraints being the imaging resolution. In each region, a beam direction is selected to obtain the demodulated output when each receiving element receives the directional echo as the original data, and the traditional uniform FFT beamforming is performed in the region. The FFT computation process is optimized to reduce the computational complexity of the new algorithm to meet the real-time requirements of 3D imaging sonar. Simulation and experimental results show that the imaging resolution of the sub-region FFT beamforming algorithm is significantly higher than that of the traditional uniform FFT beamforming algorithm, and satisfies the real-time requirement.
2017, 39(9): 2182-2189.
doi: 10.11999/JEIT161300
Abstract:
As a new high resolution ground imaging technique, Circular Synthetic Aperture Radar (CSAR) is capable of long time observation and retrieving the information of the targets. By taking the long time observation advantage, a new moving target detecting method is developed based on the single channel CSAR. Moving target signal model is analyzed with equal range equal Doppler equations, then the method is described. This method regards the original SAR image as combination of two images, one is background without moving target, the other one is foreground with moving target. Because the stationary background changes slower than the moving target signal. Thus median filter can be applied to a sequence of overlapping subaperture SAR images along the time to generate the background image. Next, each original frame is used to subtract the background image. Then the moving targets can be obtained. The proposed method is demonstrated on airborne CSAR moving targets dataset.
As a new high resolution ground imaging technique, Circular Synthetic Aperture Radar (CSAR) is capable of long time observation and retrieving the information of the targets. By taking the long time observation advantage, a new moving target detecting method is developed based on the single channel CSAR. Moving target signal model is analyzed with equal range equal Doppler equations, then the method is described. This method regards the original SAR image as combination of two images, one is background without moving target, the other one is foreground with moving target. Because the stationary background changes slower than the moving target signal. Thus median filter can be applied to a sequence of overlapping subaperture SAR images along the time to generate the background image. Next, each original frame is used to subtract the background image. Then the moving targets can be obtained. The proposed method is demonstrated on airborne CSAR moving targets dataset.
2017, 39(9): 2190-2196.
doi: 10.11999/JEIT161056
Abstract:
Based on a two-dimensional phase coding, a novel range ambiguity suppression method is proposed. By transmitting two-dimensional phase coded signals and demodulating the received signals, the two-dimensional spectrum of the range ambiguity is shifted along both range and azimuth directions. Considering the range frequency and Doppler oversampling, the spectrum of the range ambiguity located outside the data region of spectral support can be filtered out, which suppresses the range ambiguity. It allows for decreasing of the Range Ambiguity to Signal Ratio (RASR) and upgrading of the SAR image quality. Simulation results for point targets and distributed targets validate the effectiveness of the proposed method.
Based on a two-dimensional phase coding, a novel range ambiguity suppression method is proposed. By transmitting two-dimensional phase coded signals and demodulating the received signals, the two-dimensional spectrum of the range ambiguity is shifted along both range and azimuth directions. Considering the range frequency and Doppler oversampling, the spectrum of the range ambiguity located outside the data region of spectral support can be filtered out, which suppresses the range ambiguity. It allows for decreasing of the Range Ambiguity to Signal Ratio (RASR) and upgrading of the SAR image quality. Simulation results for point targets and distributed targets validate the effectiveness of the proposed method.
2017, 39(9): 2197-2202.
doi: 10.11999/JEIT161394
Abstract:
In the image sequence obtained by the high frame rate Video Synthetic Aperture Radar (VideoSAR) mode, the Doppler shift results in some shadows of the moving targets in their actual position, and a strong correlation exists between adjacent frames. Based on the above rationale, this paper proposes an approach to detecting moving targets shadow in VideoSAR imagery. First, the Scale-Invariant Feature Transform (SIFT) with RANdom SAmple Consensus (RANSAC) registration algorithm is used to compensate for the change of background of each frame, and the CattePM model is employed to suppress the speckle noise effectively. Then, in order to separate the targets and the background and generate binary images automatically, a threshold segmentation algorithm, called maximizing the Tsallis entropy, is applied. Finally, shadow detection is accomplished by the background difference with three frame difference method, and the detection results are marked on the corresponding position in the original frame. Experimental results utilizing the VideoSAR imaging fragment published by Sandia National Laboratories show that multiple moving vehicles are detected effectively, hence the validity of the approach is demonstrated.
In the image sequence obtained by the high frame rate Video Synthetic Aperture Radar (VideoSAR) mode, the Doppler shift results in some shadows of the moving targets in their actual position, and a strong correlation exists between adjacent frames. Based on the above rationale, this paper proposes an approach to detecting moving targets shadow in VideoSAR imagery. First, the Scale-Invariant Feature Transform (SIFT) with RANdom SAmple Consensus (RANSAC) registration algorithm is used to compensate for the change of background of each frame, and the CattePM model is employed to suppress the speckle noise effectively. Then, in order to separate the targets and the background and generate binary images automatically, a threshold segmentation algorithm, called maximizing the Tsallis entropy, is applied. Finally, shadow detection is accomplished by the background difference with three frame difference method, and the detection results are marked on the corresponding position in the original frame. Experimental results utilizing the VideoSAR imaging fragment published by Sandia National Laboratories show that multiple moving vehicles are detected effectively, hence the validity of the approach is demonstrated.
2017, 39(9): 2203-2211.
doi: 10.11999/JEIT161209
Abstract:
The three-dimensional (3D) imaging of airborne sparse array SAR is studied. In order to avoid data redundancy, high Pulse Repetition Frequency (PRF) and motion compensation caused by aperture synthesis, the Modified Uniformly Redundant Arrays (MURA) is introduced for random modulation on echo data along the array antenna sampling plane. The MURA positive and negative coded echo signals are processed to form 3D complex images, respectively, and by interferometry of two images the 3D frequency spectrum of the signal is concentrated in low frequency band. Therefore, using low-pass filter or Compressed Sensing (CS) processing in frequency domain, 3D image with sufficient performance is obtained under the condition of sparse sampling in the frequency domain. The method has the characteristics of high utilization rate of echo data and low repetition frequency of the system. Simulation results and experimental data of the anechoic chamber verify the effectiveness of the proposed method.
The three-dimensional (3D) imaging of airborne sparse array SAR is studied. In order to avoid data redundancy, high Pulse Repetition Frequency (PRF) and motion compensation caused by aperture synthesis, the Modified Uniformly Redundant Arrays (MURA) is introduced for random modulation on echo data along the array antenna sampling plane. The MURA positive and negative coded echo signals are processed to form 3D complex images, respectively, and by interferometry of two images the 3D frequency spectrum of the signal is concentrated in low frequency band. Therefore, using low-pass filter or Compressed Sensing (CS) processing in frequency domain, 3D image with sufficient performance is obtained under the condition of sparse sampling in the frequency domain. The method has the characteristics of high utilization rate of echo data and low repetition frequency of the system. Simulation results and experimental data of the anechoic chamber verify the effectiveness of the proposed method.
2017, 39(9): 2212-2218.
doi: 10.11999/JEIT161333
Abstract:
In radar system, the target-induced interference often arises due to multipath effect or non-ideal transmit waveform of Multiple-Input Multiple-Output (MIMO) radar. For the issue of detecting a target with target-induced interference, the detectors are proposed based on the design criterion of Wald test both in the homogeneous environment and partially homogeneous environment. The proposed detectors are proved to be effective for suppressing the target-induced interference and they can ensure the desirable Constant False Alarm Rate (CFAR) property with respect to the unknown parameters of the noise. Simulation results show that the proposed detectors can suppress the interference effectively when the interference subspace is known, and can suppress the interference lying in the orthogonal complement space of the nominal signal subspace when the interference is completely unknown.
In radar system, the target-induced interference often arises due to multipath effect or non-ideal transmit waveform of Multiple-Input Multiple-Output (MIMO) radar. For the issue of detecting a target with target-induced interference, the detectors are proposed based on the design criterion of Wald test both in the homogeneous environment and partially homogeneous environment. The proposed detectors are proved to be effective for suppressing the target-induced interference and they can ensure the desirable Constant False Alarm Rate (CFAR) property with respect to the unknown parameters of the noise. Simulation results show that the proposed detectors can suppress the interference effectively when the interference subspace is known, and can suppress the interference lying in the orthogonal complement space of the nominal signal subspace when the interference is completely unknown.
2017, 39(9): 2219-2224.
doi: 10.11999/JEIT161318
Abstract:
For calibration of direction-dependent gain-phase errors, with a few precisely calibrated instrumental sensors, a method that jointly estimates the direction-dependent gain-phase errors and the target azimuth by spatial sparsity of the signal is proposed. The array manifold that perturbed by direction-dependent gain-phase errors is denoted by the multiplication form of ideally array manifold and a gain-phase errors coefficient matrix, then the received signal is represented by sparse form. The calibration for gain-phase error problem is formulated as a dual optimization problem, through alternating iterative optimization method to acquire the optimal solution of the two optimization variables, so as to realize the signal incident angle and azimuth dependent amplitude and phase errors of the optimized calculation. In this paper, the proposed algorithm has better performance than the existing algorithm, performance of the proposed algorithm is approximate to the Cramer-Rao low bound. The simulation experiments verify the effectiveness and superiority of the proposed algorithm.
For calibration of direction-dependent gain-phase errors, with a few precisely calibrated instrumental sensors, a method that jointly estimates the direction-dependent gain-phase errors and the target azimuth by spatial sparsity of the signal is proposed. The array manifold that perturbed by direction-dependent gain-phase errors is denoted by the multiplication form of ideally array manifold and a gain-phase errors coefficient matrix, then the received signal is represented by sparse form. The calibration for gain-phase error problem is formulated as a dual optimization problem, through alternating iterative optimization method to acquire the optimal solution of the two optimization variables, so as to realize the signal incident angle and azimuth dependent amplitude and phase errors of the optimized calculation. In this paper, the proposed algorithm has better performance than the existing algorithm, performance of the proposed algorithm is approximate to the Cramer-Rao low bound. The simulation experiments verify the effectiveness and superiority of the proposed algorithm.
2017, 39(9): 2225-2231.
doi: 10.11999/JEIT161204
Abstract:
A new feature extraction method based on time-frequency analysis is proposed for aircraft targets classification under low signal-to-noise ratio. This method uses the variances of time-domain modulation periods of jet aircraft, propeller aircraft and helicopter to extract the variation of entropy in the time-frequency domain and gives a way to select optimal window lengths. Experimental result based on simulated data and measured data demonstrates that the proposed method can significantly improve the classification probability of aircraft targets under low signal-to-noise ratio.
A new feature extraction method based on time-frequency analysis is proposed for aircraft targets classification under low signal-to-noise ratio. This method uses the variances of time-domain modulation periods of jet aircraft, propeller aircraft and helicopter to extract the variation of entropy in the time-frequency domain and gives a way to select optimal window lengths. Experimental result based on simulated data and measured data demonstrates that the proposed method can significantly improve the classification probability of aircraft targets under low signal-to-noise ratio.
2017, 39(9): 2232-2237.
doi: 10.11999/JEIT161363
Abstract:
The key invoation technology of Synthetic Aperture Radar ALtimeter (SARAL) are Doppler-beam sharppen and delay/doppler range compensation. The combination of these two technologies makes it has high along-track resolution and high precision in height measurement. On the basis of 2 m sea wave height and 2 km ground grid, the accuracy of sea surface height measured by SARAL will reach 2 cm. In order to verify the accuracy of SARAL, based on a thorough study of the SARAL height measurement principle and traditional radar altimeter calibration method, a calibration method for SARAL based on Global Navigation Satellite System (GNSS) buoy calibration is developed. The method uses temporal and spatial consistency matching and multi- baseline joint solution to improve the accuracy of sea surface height measurement, then it is applied to the first airborne test of SARAL and the first airborne flight experiment data processing. By analyzing the results of airborne flight experiment data, at the same time of validation SARAL measurement precision, the feasibility of the SARAL calibration method is preliminarily verified.
The key invoation technology of Synthetic Aperture Radar ALtimeter (SARAL) are Doppler-beam sharppen and delay/doppler range compensation. The combination of these two technologies makes it has high along-track resolution and high precision in height measurement. On the basis of 2 m sea wave height and 2 km ground grid, the accuracy of sea surface height measured by SARAL will reach 2 cm. In order to verify the accuracy of SARAL, based on a thorough study of the SARAL height measurement principle and traditional radar altimeter calibration method, a calibration method for SARAL based on Global Navigation Satellite System (GNSS) buoy calibration is developed. The method uses temporal and spatial consistency matching and multi- baseline joint solution to improve the accuracy of sea surface height measurement, then it is applied to the first airborne test of SARAL and the first airborne flight experiment data processing. By analyzing the results of airborne flight experiment data, at the same time of validation SARAL measurement precision, the feasibility of the SARAL calibration method is preliminarily verified.
2017, 39(9): 2238-2244.
doi: 10.11999/JEIT161223
Abstract:
Angular glints are the main error sources for radar angular measurement in the terminal guidance, which may result larger tracking error or even lead to losses of the tracking if mishandled. Although the angle glint can be suppressed to some extent through Higher Range Resolution Profile (HRRP) processing, the tracking errors due to multiple scattering centers can not be thoroughly eliminated. It is found in this paper that the tracking errors are closely related with the attributes of scattering centers. The influence on angular measurements induced by scattering centers with various attributes is investigated theoretically and numerically in this paper. For high reliability of numerical results, the scattering responses of extended targets are simulated by the scattered fields computed by the full-wave numerical method in this paper. The reached conclusion of this paper can provide a theoretical reference for the techniques in order to improve the tracing accuracy of extended targets with multiple scattering centers.
Angular glints are the main error sources for radar angular measurement in the terminal guidance, which may result larger tracking error or even lead to losses of the tracking if mishandled. Although the angle glint can be suppressed to some extent through Higher Range Resolution Profile (HRRP) processing, the tracking errors due to multiple scattering centers can not be thoroughly eliminated. It is found in this paper that the tracking errors are closely related with the attributes of scattering centers. The influence on angular measurements induced by scattering centers with various attributes is investigated theoretically and numerically in this paper. For high reliability of numerical results, the scattering responses of extended targets are simulated by the scattered fields computed by the full-wave numerical method in this paper. The reached conclusion of this paper can provide a theoretical reference for the techniques in order to improve the tracing accuracy of extended targets with multiple scattering centers.
2017, 39(9): 2245-2251.
doi: 10.11999/JEIT161308
Abstract:
The large scale parallel electromagnetic computation based on the supercomputer is of great significance for solving complicate electromagnetic problems in practical engineering. However, the probability of the process crash event caused by node failure in the supercomputer is much higher than that in the regular computer. Considering the incapable action for traditional electromagnetic computation to overcome the process crash event, an efficient fault-tolerance algorithm for large scale parallel high order Method of Moments (MoM) is proposed in this paper. According to the parallel higher order method of moments algorithm available, a scene protection algorithm and a scene recovery algorithm with high efficiency and reliability are designed, based on the disk cache and direct memory access technique. The efficiency of this algorithm lies on the feature of the fixed site protection, which makes it possible for the algorithm to work normal and ordered even encountering crash failure, while the original algorithm can only restart from the beginning. The numerical simulations demonstrate the efficiency of the fault-tolerant algorithm in dealing with the process crash, which improves greatly the reliability of the large scale parallel high order MoM.
The large scale parallel electromagnetic computation based on the supercomputer is of great significance for solving complicate electromagnetic problems in practical engineering. However, the probability of the process crash event caused by node failure in the supercomputer is much higher than that in the regular computer. Considering the incapable action for traditional electromagnetic computation to overcome the process crash event, an efficient fault-tolerance algorithm for large scale parallel high order Method of Moments (MoM) is proposed in this paper. According to the parallel higher order method of moments algorithm available, a scene protection algorithm and a scene recovery algorithm with high efficiency and reliability are designed, based on the disk cache and direct memory access technique. The efficiency of this algorithm lies on the feature of the fixed site protection, which makes it possible for the algorithm to work normal and ordered even encountering crash failure, while the original algorithm can only restart from the beginning. The numerical simulations demonstrate the efficiency of the fault-tolerant algorithm in dealing with the process crash, which improves greatly the reliability of the large scale parallel high order MoM.
2017, 39(9): 2252-2258.
doi: 10.11999/JEIT161277
Abstract:
The accuracy of three dimensional electric field measurement is greatly influenced by the coupling interference between components of Three Dimension (3D) electric field applied to the electric field sensor. In order to reduce the coupling interference and to realize accurate measurements, a decoupling calibration method based on Genetic Algorithm (GA) of 3D electric field sensor is proposed. Different from the traditional calibration method based on the matrix inversion technology, GA is utilized to obtain the optimal decoupling calibration matrix by setting the objective function and evolution operators, which avoids the calculation errors in the complex matrix inversion. A calibration apparatus, which can make an arbitrary angle between the sensor and the electric field, is designed, and a calculation model of theoretical electric field values is built. Experiments are conducted to verify the decoupling calibration methods based on the matrix inversion technology and GA. The experimental results show that the proposed method can effectively eliminate coupling interferences, improve the calibration precision and realize accurate measurements.
The accuracy of three dimensional electric field measurement is greatly influenced by the coupling interference between components of Three Dimension (3D) electric field applied to the electric field sensor. In order to reduce the coupling interference and to realize accurate measurements, a decoupling calibration method based on Genetic Algorithm (GA) of 3D electric field sensor is proposed. Different from the traditional calibration method based on the matrix inversion technology, GA is utilized to obtain the optimal decoupling calibration matrix by setting the objective function and evolution operators, which avoids the calculation errors in the complex matrix inversion. A calibration apparatus, which can make an arbitrary angle between the sensor and the electric field, is designed, and a calculation model of theoretical electric field values is built. Experiments are conducted to verify the decoupling calibration methods based on the matrix inversion technology and GA. The experimental results show that the proposed method can effectively eliminate coupling interferences, improve the calibration precision and realize accurate measurements.
2017, 39(9): 2259-2265.
doi: 10.11999/JEIT170023
Abstract:
Truth table reduction is one of the key problems in the analysis and design of digital logic circuits, FCA (Formal Concept Analysis) is a tool for data analysis and rule extraction from formal contexts. In this paper, MIMO (Multiple-Input Multiple-Output) truth table is transformed into formal decision context, thus the reduction problem of truth table is transformed into the simplest rule extraction process of formal decision context. Then, a parallel reduction algorithm for MIMO truth table based on FCA is proposed. The correctness, efficiency and rapidity of the new algorithm are illustrated by the theoretical proof, example demonstration and complexity analysis of the proposed algorithm.
Truth table reduction is one of the key problems in the analysis and design of digital logic circuits, FCA (Formal Concept Analysis) is a tool for data analysis and rule extraction from formal contexts. In this paper, MIMO (Multiple-Input Multiple-Output) truth table is transformed into formal decision context, thus the reduction problem of truth table is transformed into the simplest rule extraction process of formal decision context. Then, a parallel reduction algorithm for MIMO truth table based on FCA is proposed. The correctness, efficiency and rapidity of the new algorithm are illustrated by the theoretical proof, example demonstration and complexity analysis of the proposed algorithm.
2017, 39(9): 2266-2271.
doi: 10.11999/JEIT161202
Abstract:
A mixed-mode Test-Per-Clock Built In Self Test (BIST) scheme is proposed. The test consists of two parts: the free Linear Feedback Shift Register (LFSR) pseudo-random test mode and the deterministic test pattern based on controlled LFSR. Pseudo random test mode is used to quickly detect pseudo-random susceptible faults and reduce the deterministic data storage. Controlled LFSR test mode uses the control bits directly stored in the ROM to generate a deterministic test of the remaining faults. Based on the theoretical analysis of the proposed mixed-mode BIST test structure, a pseudo-random test sequence selection method and a deterministic test generation method based on controlled linear shifter are proposed. Simulation results on benchmark circuits show that the proposed method can obtain the complete single stuck-at fault coverage and has good stability in test generation. Compared with other methods, it has simpler Test Pattern Generator (TPG) design and lower test cost as well as shorter test application time.
A mixed-mode Test-Per-Clock Built In Self Test (BIST) scheme is proposed. The test consists of two parts: the free Linear Feedback Shift Register (LFSR) pseudo-random test mode and the deterministic test pattern based on controlled LFSR. Pseudo random test mode is used to quickly detect pseudo-random susceptible faults and reduce the deterministic data storage. Controlled LFSR test mode uses the control bits directly stored in the ROM to generate a deterministic test of the remaining faults. Based on the theoretical analysis of the proposed mixed-mode BIST test structure, a pseudo-random test sequence selection method and a deterministic test generation method based on controlled linear shifter are proposed. Simulation results on benchmark circuits show that the proposed method can obtain the complete single stuck-at fault coverage and has good stability in test generation. Compared with other methods, it has simpler Test Pattern Generator (TPG) design and lower test cost as well as shorter test application time.
2017, 39(9): 2272-2276.
doi: 10.11999/JEIT161284
Abstract:
To achieve intelligent and autonomous positioning for the vehicle, this paper presents a location estimation model based on the transformation from grid cells to place cells. Combining with the firing characteristic of the grid cells, place cells, and the information transformation between them, this location estimation model is divided into three parts, including the learning and memorizing of spatial environment, the perception of the motion state and the estimation of the spatial location. The principle and the specific steps of each part are discussed. Finally, the proposed model is applied to vehicles positioning by simulation. Simulation validates that the proposed model is feasible to achieve vehicles autonomous positioning, and the positioning performance can be adjusted by changing the parameters of grid cells and place cells included in the model.
To achieve intelligent and autonomous positioning for the vehicle, this paper presents a location estimation model based on the transformation from grid cells to place cells. Combining with the firing characteristic of the grid cells, place cells, and the information transformation between them, this location estimation model is divided into three parts, including the learning and memorizing of spatial environment, the perception of the motion state and the estimation of the spatial location. The principle and the specific steps of each part are discussed. Finally, the proposed model is applied to vehicles positioning by simulation. Simulation validates that the proposed model is feasible to achieve vehicles autonomous positioning, and the positioning performance can be adjusted by changing the parameters of grid cells and place cells included in the model.
2017, 39(9): 2277-2281.
doi: 10.11999/JEIT161177
Abstract:
A unified construction of Guassian integer sequence sets with Zero Correlation Zone (ZCZ) is presented. Based on difference sets, optimal or almost optimal ZCZ Gaussian integer sequence sets are constructed using shift sequences, whose ZCZ length and alphabets can be flexibly chosen. Since the study of difference sets has achieved abundant?accomplishment, then the presented method will produce an abundance of ZCZ Gaussian integer sequence sets for CDMA systems.
A unified construction of Guassian integer sequence sets with Zero Correlation Zone (ZCZ) is presented. Based on difference sets, optimal or almost optimal ZCZ Gaussian integer sequence sets are constructed using shift sequences, whose ZCZ length and alphabets can be flexibly chosen. Since the study of difference sets has achieved abundant?accomplishment, then the presented method will produce an abundance of ZCZ Gaussian integer sequence sets for CDMA systems.
2017, 39(9): 2282-2286.
doi: 10.11999/JEIT161381
Abstract:
The spectral subtraction speech enhancement is extensively used due to its simplicity and easy to implement. The principle of this method is to subtract the estimated magnitude of the noise from the magnitude of the noisy signal, but the phase of the noisy signal is unchanged. This conventional method produces the estimating error because it exploits the noisy phase, especially in low SNR, and it produces musical noise because of the inaccuracy of the noise estimation. This paper proposes a multi-band spectral subtraction algorithm based on maximum posteriori phase estimation. Experimental results show that the proposed method can get better performance than the conventional method especially in low SNR.
The spectral subtraction speech enhancement is extensively used due to its simplicity and easy to implement. The principle of this method is to subtract the estimated magnitude of the noise from the magnitude of the noisy signal, but the phase of the noisy signal is unchanged. This conventional method produces the estimating error because it exploits the noisy phase, especially in low SNR, and it produces musical noise because of the inaccuracy of the noise estimation. This paper proposes a multi-band spectral subtraction algorithm based on maximum posteriori phase estimation. Experimental results show that the proposed method can get better performance than the conventional method especially in low SNR.
