Email alert
2016 Vol. 38, No. 2
Display Method:
2016, 38(2): 255-261.
doi: 10.11999/JEIT150532
Abstract:
In heterogeneous cellular networks, small cells provide for the handover of users from the macro cell and organize itself during the transmission. This paper studies the downlink outage performance of heterogeneous cellular networks with an interference cancellation scheme employed at macro base station, which aims to eliminate the cross-tier interference from macro base station to small cell users. Then, the downlink performance of heterogeneous cellular networks is investigated. Expressions of the Probability Density Function (PDF) and Cumulative Distribution Function (CDF) of the received SNRs of both macro and small cell users are derived and closed-form expressions of overall outage probability of the system are provided. Both analytical results and simulations show that the overall performance of the heterogeneous cellular networks is improved with interference cancellation.
In heterogeneous cellular networks, small cells provide for the handover of users from the macro cell and organize itself during the transmission. This paper studies the downlink outage performance of heterogeneous cellular networks with an interference cancellation scheme employed at macro base station, which aims to eliminate the cross-tier interference from macro base station to small cell users. Then, the downlink performance of heterogeneous cellular networks is investigated. Expressions of the Probability Density Function (PDF) and Cumulative Distribution Function (CDF) of the received SNRs of both macro and small cell users are derived and closed-form expressions of overall outage probability of the system are provided. Both analytical results and simulations show that the overall performance of the heterogeneous cellular networks is improved with interference cancellation.
2016, 38(2): 262-268.
doi: 10.11999/JEIT150699
Abstract:
Femtocell is a promising technology to enhance indoor coverage and system capacity. However, the co-tier and cross-tier interference impair greatly the network performance for spectrum-sharing OFDMA femtocell networks. To mitigate the co-tier/cross-tier interference, a cluster-based resource allocation algorithm is proposed. The proposed algorithm consists of two parts: In the first part, an improved Hungarian algorithm is first adopted to assign sub-channels to the Macro Users Equipments (MUEs). Then the averagely assigned power is reallocated in linear water-filling fashion in order to ensure the transmission of MUEs. In the other part, Simulated Annealing algorithm is first used to cluster femtocells into disjoint groups. Then under the condition of avoiding causing interference to MUEs, femtocells perform sub-channel and power allocation according to the rate requirements of Femtocell User Equipments (FUEs) to maximize spectrum efficiency. Simulation results show that the proposed algorithm not only guarantees the data rate requirements of users, but also improves the spectrum efficiency.
Femtocell is a promising technology to enhance indoor coverage and system capacity. However, the co-tier and cross-tier interference impair greatly the network performance for spectrum-sharing OFDMA femtocell networks. To mitigate the co-tier/cross-tier interference, a cluster-based resource allocation algorithm is proposed. The proposed algorithm consists of two parts: In the first part, an improved Hungarian algorithm is first adopted to assign sub-channels to the Macro Users Equipments (MUEs). Then the averagely assigned power is reallocated in linear water-filling fashion in order to ensure the transmission of MUEs. In the other part, Simulated Annealing algorithm is first used to cluster femtocells into disjoint groups. Then under the condition of avoiding causing interference to MUEs, femtocells perform sub-channel and power allocation according to the rate requirements of Femtocell User Equipments (FUEs) to maximize spectrum efficiency. Simulation results show that the proposed algorithm not only guarantees the data rate requirements of users, but also improves the spectrum efficiency.
2016, 38(2): 269-275.
doi: 10.11999/JEIT150443
Abstract:
The base station will initiate handoff blindly and cause high failure rate of handoff if the knowledge of the terminal,s motion trend is absent. An optimized algorithm is proposed to optmize existing vertical handoff algorithm in the LTE-WiMAX heterogeneous wireless network system. The proposed algorithm uses the motion trend quantification to estimate goal cells and restrict unnecessary handoff so as to increase success rate of handoff. The computer simulation results in fading channel show that the optimized algorithm can reduce the failure rate of handoff during the handoff process and enhance the handoff performance of network.
The base station will initiate handoff blindly and cause high failure rate of handoff if the knowledge of the terminal,s motion trend is absent. An optimized algorithm is proposed to optmize existing vertical handoff algorithm in the LTE-WiMAX heterogeneous wireless network system. The proposed algorithm uses the motion trend quantification to estimate goal cells and restrict unnecessary handoff so as to increase success rate of handoff. The computer simulation results in fading channel show that the optimized algorithm can reduce the failure rate of handoff during the handoff process and enhance the handoff performance of network.
2016, 38(2): 276-281.
doi: 10.11999/JEIT150599
Abstract:
This paper proposes a new Non-Data-Aided (NDA) scheme to estimate the channel order and noise variance in the low Signal to Noise Ratio (SNR) region of Orthogonal Frequency Division Multiplexing (OFDM) signals. In this scheme, a new cost function is derived based on the joint Maximum Geometric Mean (MGM) which relies on both the Cyclic Prefix (CP) redundancy and channel memory. Compared with the schemes which only rely on the CP, more accurate estimations of channel order and noise variance can be obtained from this joint MGM cost function. Simulation results show that the proposed channel order estimator gets approximately 10 dB SNR gain in the low SNR region. Meanwhile, the proposed noise variance estimator outperforms significantly the other existing NDA algorithms, and suppresses the performance deterioration when SNR below 20 dB.
This paper proposes a new Non-Data-Aided (NDA) scheme to estimate the channel order and noise variance in the low Signal to Noise Ratio (SNR) region of Orthogonal Frequency Division Multiplexing (OFDM) signals. In this scheme, a new cost function is derived based on the joint Maximum Geometric Mean (MGM) which relies on both the Cyclic Prefix (CP) redundancy and channel memory. Compared with the schemes which only rely on the CP, more accurate estimations of channel order and noise variance can be obtained from this joint MGM cost function. Simulation results show that the proposed channel order estimator gets approximately 10 dB SNR gain in the low SNR region. Meanwhile, the proposed noise variance estimator outperforms significantly the other existing NDA algorithms, and suppresses the performance deterioration when SNR below 20 dB.
2016, 38(2): 282-287.
doi: 10.11999/JEIT150594
Abstract:
To deal with problems of large Peak-to-Average Power Ratio (PAPR) and computation complexity in existing channel estimation algorithm based on compressed sensing, Golay complementary sequence is utilized to estimate sparse channel as a deterministic training sequence. Estimation model and algorithm are provided in detail. The PAPR of this method is deduced and its performance, PAPR and computation complexity are compared with Gaussian random sequence. The simulation result indicates that Golay sequence and Gaussian random sequence can reconstruct nonzero tap coefficients. But Golay sequence outperforms Gaussian random sequence both in the Mean Square Error (MSE) and Match Rate (MR) for estimating sparse channel impulse response. And the computation and PAPR are reduced significantly in the OFDM system.
To deal with problems of large Peak-to-Average Power Ratio (PAPR) and computation complexity in existing channel estimation algorithm based on compressed sensing, Golay complementary sequence is utilized to estimate sparse channel as a deterministic training sequence. Estimation model and algorithm are provided in detail. The PAPR of this method is deduced and its performance, PAPR and computation complexity are compared with Gaussian random sequence. The simulation result indicates that Golay sequence and Gaussian random sequence can reconstruct nonzero tap coefficients. But Golay sequence outperforms Gaussian random sequence both in the Mean Square Error (MSE) and Match Rate (MR) for estimating sparse channel impulse response. And the computation and PAPR are reduced significantly in the OFDM system.
2016, 38(2): 288-293.
doi: 10.11999/JEIT150572
Abstract:
Considering the characteristics of underwater acoustic communication, random linear fountain codes with maximum likelihood decoding are studied to correct erasure errors in the short packet transmission. In existing maximum likelihood decoding methods, processing begins when all the necessary blocks are available, resulting to the unacceptable decoding delay. An increment Gaussian elimination method is proposed to decrease the decoding delay by utilizing the time-slots of every block. The computation complexity is analyzed based on the principle of the probability distribution of the summation of binary random variables. The real-time ability of the proposed method is verified on the low-cost DSP chip for the underwater acoustic modem. The method is applicable to underwater transmissions of images, and sense data.
Considering the characteristics of underwater acoustic communication, random linear fountain codes with maximum likelihood decoding are studied to correct erasure errors in the short packet transmission. In existing maximum likelihood decoding methods, processing begins when all the necessary blocks are available, resulting to the unacceptable decoding delay. An increment Gaussian elimination method is proposed to decrease the decoding delay by utilizing the time-slots of every block. The computation complexity is analyzed based on the principle of the probability distribution of the summation of binary random variables. The real-time ability of the proposed method is verified on the low-cost DSP chip for the underwater acoustic modem. The method is applicable to underwater transmissions of images, and sense data.
2016, 38(2): 294-302.
doi: 10.11999/JEIT150267
Abstract:
To predict the effects induced by stereo image content on visual health, a new objective Visual Comfort Assessment (VCA) method of stereo image is proposed based on scene modes. Natural scene is abstracted as multiple scene modes according to two position states of foreground object and background region. One is the convex-concave to screen, and the other is the whether locate on zone of comfortable viewing. In the process of mode selection, disparity map is utilized to segment scene into foreground object and background region adaptively. Then, the scenes mode can be determined by disparity angle features of both foreground object and background region. In the modeling stage, disparity angle features of foreground object and background region, width angle and sinuosity features of foreground object are utilized to build objective VCA models in various scene modes. The experimental results tested on IVY database show that high consistency exists between the proposed model and subjective perception that Pearson linear correlation coefficient is higher than 0.91, Spearman rank-order correlation coefficient is higher than 0.90, Kendall rank-order correlation coefficient is higher than 0.74, Mean Absolute Error (MAE) is lower than 0.24 and Root Mean Squared Error (RMSE) is lower than 0.32. Compared with other existing methods, the proposed model has the better assessment performance and is much closer to the subjective assessment scores.
To predict the effects induced by stereo image content on visual health, a new objective Visual Comfort Assessment (VCA) method of stereo image is proposed based on scene modes. Natural scene is abstracted as multiple scene modes according to two position states of foreground object and background region. One is the convex-concave to screen, and the other is the whether locate on zone of comfortable viewing. In the process of mode selection, disparity map is utilized to segment scene into foreground object and background region adaptively. Then, the scenes mode can be determined by disparity angle features of both foreground object and background region. In the modeling stage, disparity angle features of foreground object and background region, width angle and sinuosity features of foreground object are utilized to build objective VCA models in various scene modes. The experimental results tested on IVY database show that high consistency exists between the proposed model and subjective perception that Pearson linear correlation coefficient is higher than 0.91, Spearman rank-order correlation coefficient is higher than 0.90, Kendall rank-order correlation coefficient is higher than 0.74, Mean Absolute Error (MAE) is lower than 0.24 and Root Mean Squared Error (RMSE) is lower than 0.32. Compared with other existing methods, the proposed model has the better assessment performance and is much closer to the subjective assessment scores.
2016, 38(2): 303-309.
doi: 10.11999/JEIT150687
Abstract:
Compared with the original Legendre-Sidelnikov sequence, the generalized Legendre-Sidelnikov sequence has a better balanced property. For its autocorrelation distribution, however, only some special cases are known. In this paper, using the character sums, the autocorrelation distribution of the generalized binary Legendre-Sidelnikov sequence is determined completely. The result shows that the generalized Legendre-Sidelnikov sequence possesses a better autocorrelation distribution if p3 (mod 4) andqp .
Compared with the original Legendre-Sidelnikov sequence, the generalized Legendre-Sidelnikov sequence has a better balanced property. For its autocorrelation distribution, however, only some special cases are known. In this paper, using the character sums, the autocorrelation distribution of the generalized binary Legendre-Sidelnikov sequence is determined completely. The result shows that the generalized Legendre-Sidelnikov sequence possesses a better autocorrelation distribution if p3 (mod 4) andqp .
2016, 38(2): 310-317.
doi: 10.11999/JEIT150459
Abstract:
Intermittently connected wireless networks transmit data through opportunities caused by nodes movements. But malicious nodes in the network can attack and delete data by falsifying information about their routings, in order to impact the performance of networks. A data forwarding mechanism with blackhole attacking detection in intermittently connected wireless networks is proposed in this paper. By evaluating 4 trust properties include value of honesty, credit, indirect trust, and data forwarding rate, characteristics of attackers are summarized directly behaviors. And using the theory of rough set to decrease the rate of mistakes caused by indeterminate information, so that to determine the reliability of nodes accurately, and choose relay nodes reasonably. Results show that, the proposed mechanism can effectively find out attackers while enhancing the reliability of data transmission, and it also can defense some other non-cooperative nodes in the networks. Thus, the utilization of network resource is improved.
Intermittently connected wireless networks transmit data through opportunities caused by nodes movements. But malicious nodes in the network can attack and delete data by falsifying information about their routings, in order to impact the performance of networks. A data forwarding mechanism with blackhole attacking detection in intermittently connected wireless networks is proposed in this paper. By evaluating 4 trust properties include value of honesty, credit, indirect trust, and data forwarding rate, characteristics of attackers are summarized directly behaviors. And using the theory of rough set to decrease the rate of mistakes caused by indeterminate information, so that to determine the reliability of nodes accurately, and choose relay nodes reasonably. Results show that, the proposed mechanism can effectively find out attackers while enhancing the reliability of data transmission, and it also can defense some other non-cooperative nodes in the networks. Thus, the utilization of network resource is improved.
2016, 38(2): 318-324.
doi: 10.11999/JEIT150636
Abstract:
A design method of Zero Correlation Zone (ZCZ) punctured quaternary periodic complementary sequence pairs set is proposed. Based on perfect punctured binary sequences pairs and orthogonal matrix, ZCZ punctured sequence pairs sets can be constructed. Then, by interleaving iteration technique ZCZ punctured periodic complementary sequence pairs sets are obtained. Finally, ZCZ punctured quaternary periodic complementary sequence pairs sets are constructed using novel inverse Gray mapping. The results reach the theoretical bounds and further expand the selectable space of the spread spectrum sequences.
A design method of Zero Correlation Zone (ZCZ) punctured quaternary periodic complementary sequence pairs set is proposed. Based on perfect punctured binary sequences pairs and orthogonal matrix, ZCZ punctured sequence pairs sets can be constructed. Then, by interleaving iteration technique ZCZ punctured periodic complementary sequence pairs sets are obtained. Finally, ZCZ punctured quaternary periodic complementary sequence pairs sets are constructed using novel inverse Gray mapping. The results reach the theoretical bounds and further expand the selectable space of the spread spectrum sequences.
2016, 38(2): 325-331.
doi: 10.11999/JEIT150626
Abstract:
Information-Centric Network (ICN) architectures seek to provide the necessary foundations for a more cost-efficient content acquirement and content distribution using universal in-network caching, also universal in-network caching is a key design principle of many such architectures. Given that caching capacity of ICN is relatively small in comparison to the amount of forwarded content, a key aspect is balanced distribution of content among the available caches. The in-network caching resolution scheme is proposed in this paper, based on content popularity and nodes centrality, called PCBCS. It reduces caching redundancy and in turn, make more efficient utilization of available cache resources along a delivery path through selective caching of content passing. The proposed algorithm is compared with universal on-path caching and Leave Copy Down (LCD), also Prob (copy with probability) scheme with parameter of 0.7 and 0.3. The results show reduction of up to 30% in server hits, and up to 20% in the number of hops required to hit cached contents, but, most importantly, reduction of cache replacements up to 40% in comparison to universal caching.
Information-Centric Network (ICN) architectures seek to provide the necessary foundations for a more cost-efficient content acquirement and content distribution using universal in-network caching, also universal in-network caching is a key design principle of many such architectures. Given that caching capacity of ICN is relatively small in comparison to the amount of forwarded content, a key aspect is balanced distribution of content among the available caches. The in-network caching resolution scheme is proposed in this paper, based on content popularity and nodes centrality, called PCBCS. It reduces caching redundancy and in turn, make more efficient utilization of available cache resources along a delivery path through selective caching of content passing. The proposed algorithm is compared with universal on-path caching and Leave Copy Down (LCD), also Prob (copy with probability) scheme with parameter of 0.7 and 0.3. The results show reduction of up to 30% in server hits, and up to 20% in the number of hops required to hit cached contents, but, most importantly, reduction of cache replacements up to 40% in comparison to universal caching.
2016, 38(2): 332-339.
doi: 10.11999/JEIT150487
Abstract:
A Multi-hop Hybrid Cooperative Geographic Routing (MHCGR) algorithm with outage-probability- constrained is proposed to reduce the path length for routing in wireless sensor networks. The cooperative links using different cooperative strategies are analyzed. With theoretical analysis, the decode-amplify-and-forward hybrid cooperative strategy can further expand the transmission distance. The ideal maximum cooperative transmission distance and the location of ideal relay node are proved for per-hop cooperative transmission link. Based on the BeaconLess Geographic Routing (BLGR) algorithm, the MHCGR algorithm uses the location information of nodes to select the optimum relay node and optimum forward node for each hop. Then MHCGR algorithm forms the cooperative route from the source node to destination node by mentioned cooperative strategy. Simulation results show that, compared with the ENBGCR algorithm and the MPCR algorithm using DF strategy, the MHCGR algorithm can reduce the number of routing hop, and reduce the overall transmission power routing.
A Multi-hop Hybrid Cooperative Geographic Routing (MHCGR) algorithm with outage-probability- constrained is proposed to reduce the path length for routing in wireless sensor networks. The cooperative links using different cooperative strategies are analyzed. With theoretical analysis, the decode-amplify-and-forward hybrid cooperative strategy can further expand the transmission distance. The ideal maximum cooperative transmission distance and the location of ideal relay node are proved for per-hop cooperative transmission link. Based on the BeaconLess Geographic Routing (BLGR) algorithm, the MHCGR algorithm uses the location information of nodes to select the optimum relay node and optimum forward node for each hop. Then MHCGR algorithm forms the cooperative route from the source node to destination node by mentioned cooperative strategy. Simulation results show that, compared with the ENBGCR algorithm and the MPCR algorithm using DF strategy, the MHCGR algorithm can reduce the number of routing hop, and reduce the overall transmission power routing.
2016, 38(2): 340-346.
doi: 10.11999/JEIT150721
Abstract:
Crowdsourcing traffic monitoring is a promising application, which exploits ubiquitous mobile devices to upload GPS samples to obtain live road traffic. However, uploading the sensitive location information raises significant privacy issues. By analyzing the upload behavior of mobile users, this paper designs a privacy preserving traffic data collection mechanism. Using the relationships among the traffic service quality, privacy loss and the upload behavior, an incomplete information game is built to analyze the upload behavior of users. Based on the existence and uniqueness of Nash equilibrium in this game, a user-centric privacy preserving traffic data collection mechanism is proposed, which can maximize the utilities of users, and this mechanism has a feature of incentive compatible. Finally, the experimental results on real world traffic data confirm the effectiveness of privacy protecting and the feature of incentive compatible.
Crowdsourcing traffic monitoring is a promising application, which exploits ubiquitous mobile devices to upload GPS samples to obtain live road traffic. However, uploading the sensitive location information raises significant privacy issues. By analyzing the upload behavior of mobile users, this paper designs a privacy preserving traffic data collection mechanism. Using the relationships among the traffic service quality, privacy loss and the upload behavior, an incomplete information game is built to analyze the upload behavior of users. Based on the existence and uniqueness of Nash equilibrium in this game, a user-centric privacy preserving traffic data collection mechanism is proposed, which can maximize the utilities of users, and this mechanism has a feature of incentive compatible. Finally, the experimental results on real world traffic data confirm the effectiveness of privacy protecting and the feature of incentive compatible.
2016, 38(2): 347-353.
doi: 10.11999/JEIT150380
Abstract:
To lower the sizes of keys, a certificateless encryption scheme is put forward by using a trapdoor sampling algorithm over a selected NTRU lattice to extract partial private keys and using Ring Learning With Errors (RLWE) problem to generate public keys. Its security is based on both assumptions of the decisional ring learning with errors problem and the decisional Small Polynomial Ratio (SPR) problem. To further improve efficiency, a certificateless parallel encryption scheme with more efficient algorithms only using arithmetic in integers is also given by respectively using the Chinese Remainder Theorem (CRT) to decompose the enlarged plaintext space into the product of distinct prime ideals and to break down the ring, over which encryption operations work, for obtaining the Chinese Remainder basis. The given results show that the proposed schemes are characterized by low computation complexity and small communication complexity.
To lower the sizes of keys, a certificateless encryption scheme is put forward by using a trapdoor sampling algorithm over a selected NTRU lattice to extract partial private keys and using Ring Learning With Errors (RLWE) problem to generate public keys. Its security is based on both assumptions of the decisional ring learning with errors problem and the decisional Small Polynomial Ratio (SPR) problem. To further improve efficiency, a certificateless parallel encryption scheme with more efficient algorithms only using arithmetic in integers is also given by respectively using the Chinese Remainder Theorem (CRT) to decompose the enlarged plaintext space into the product of distinct prime ideals and to break down the ring, over which encryption operations work, for obtaining the Chinese Remainder basis. The given results show that the proposed schemes are characterized by low computation complexity and small communication complexity.
2016, 38(2): 354-360.
doi: 10.11999/JEIT150702
Abstract:
This paper proposes a novel scheme of reversible steganography in encrypted domain based on Learning With Errors (LWE). The original data is encrypted by the cryptographic algorithms with LWE. Then additional data could be embedded into the cipher text. With embedded cipher text, the additional data can be extracted by using data-hiding key, and the original data can be recovered by using encryption key, and the processes of extraction and decryption are separable. By deducing the error probability of the scheme, the standard deviation of noise sequence which directly related to the schemes correctness is mainly discussed, and reasonable range of the standard deviation is obtained by experiments. The probability distribution function of the embedded cipher text is deduced, that proves the embedded cipher text is not detective. The proposed scheme based on encrypted domain can apply to different kinds of media vehicle such as text, image or audio. Experimental results demonstrate that the proposed scheme can not only achieve statistical security without degrading the quality of encryption or data embedding, but realize that 1 bit original data can maximally load 1 bit additional data in encrypted domain.
This paper proposes a novel scheme of reversible steganography in encrypted domain based on Learning With Errors (LWE). The original data is encrypted by the cryptographic algorithms with LWE. Then additional data could be embedded into the cipher text. With embedded cipher text, the additional data can be extracted by using data-hiding key, and the original data can be recovered by using encryption key, and the processes of extraction and decryption are separable. By deducing the error probability of the scheme, the standard deviation of noise sequence which directly related to the schemes correctness is mainly discussed, and reasonable range of the standard deviation is obtained by experiments. The probability distribution function of the embedded cipher text is deduced, that proves the embedded cipher text is not detective. The proposed scheme based on encrypted domain can apply to different kinds of media vehicle such as text, image or audio. Experimental results demonstrate that the proposed scheme can not only achieve statistical security without degrading the quality of encryption or data embedding, but realize that 1 bit original data can maximally load 1 bit additional data in encrypted domain.
2016, 38(2): 361-366.
doi: 10.11999/JEIT150653
Abstract:
Concerning the resource-limited RFID tags, this paper presents a lightweight mutual authentication scheme based on Hash function, combining with the pseudo-random number and shared secret mechanisms, and implements the mutual authentication among the end database, reader and the tags. The anti-attack performance and the overhead of the scheme are analyzed in detail. Afterwards, the protocol security model is formalized using BAN logical analysis method. Theoretical analysis shows that the proposed authentication scheme could achieve the desired security goals, has good anti-attack performance and high efficiency. It can be applied to big population RFID since its low overhead for RFID tags.
Concerning the resource-limited RFID tags, this paper presents a lightweight mutual authentication scheme based on Hash function, combining with the pseudo-random number and shared secret mechanisms, and implements the mutual authentication among the end database, reader and the tags. The anti-attack performance and the overhead of the scheme are analyzed in detail. Afterwards, the protocol security model is formalized using BAN logical analysis method. Theoretical analysis shows that the proposed authentication scheme could achieve the desired security goals, has good anti-attack performance and high efficiency. It can be applied to big population RFID since its low overhead for RFID tags.
2016, 38(2): 367-372.
doi: 10.11999/JEIT 150566
Abstract:
Motivated by the success of sparse representation in speaker recognition,?a good?dictionary?plays an important role in?sparse representation. In this paper, the structured dictionary learning is introduced to speaker recognition based on the Fisher criterion. In the process of learning the discrimination dictionary, each sub-dictionary of the learned dictionary corresponds to a class label, so the reconstruction error of the same training samples is small. Meanwhile, the sparse coding coefficients have small with-class scatter and big between-class scatter. On the NIST SRE 2003 database, the experimental results indicate that the proposed method achieves an Equal Error Rate (EER) of 7.62%, and the i-vector system based on cosine distance scoring gives an EER of 6.7%. Moreover, an EER of 5.07% is obtained by combining two systems.
Motivated by the success of sparse representation in speaker recognition,?a good?dictionary?plays an important role in?sparse representation. In this paper, the structured dictionary learning is introduced to speaker recognition based on the Fisher criterion. In the process of learning the discrimination dictionary, each sub-dictionary of the learned dictionary corresponds to a class label, so the reconstruction error of the same training samples is small. Meanwhile, the sparse coding coefficients have small with-class scatter and big between-class scatter. On the NIST SRE 2003 database, the experimental results indicate that the proposed method achieves an Equal Error Rate (EER) of 7.62%, and the i-vector system based on cosine distance scoring gives an EER of 6.7%. Moreover, an EER of 5.07% is obtained by combining two systems.
2016, 38(2): 373-380.
doi: 10.11999/JEIT150544
Abstract:
Adaboost.M1 requires each weak classifier,s accuracy rate more than 1/2. But it is difficult to find a weak classifier which accuracy rate more than 1/2 in a multiple classification issues. Some scholars put forward the Stagewise Additive Modeling using a Multi-class Exponential loss function (SAMME) algorithm, it reduces the weak classifier accuracy requirements, from more than 1/2 to more than 1/k (k is the category number). SAMME algorithm reduces the difficulty to find weak classifier. But, due to the SAMME algorithm is no guarantee that the effectiveness of the weak classifier, which does not ensure that the final classifier improves classification accuracy. This paper analyzes the multi-class Adaboost algorithm by graphic method and math method, and then a new kind of multi-class classification method is proposed which not only reduces the weak classifier accuracy requirements, but also ensures the effectiveness of the weak classifier. In the benchmark experiments on UCI data sets show that the proposed algorithm are better than the SAMME, and achieves the effect of Adaboost.M1.
Adaboost.M1 requires each weak classifier,s accuracy rate more than 1/2. But it is difficult to find a weak classifier which accuracy rate more than 1/2 in a multiple classification issues. Some scholars put forward the Stagewise Additive Modeling using a Multi-class Exponential loss function (SAMME) algorithm, it reduces the weak classifier accuracy requirements, from more than 1/2 to more than 1/k (k is the category number). SAMME algorithm reduces the difficulty to find weak classifier. But, due to the SAMME algorithm is no guarantee that the effectiveness of the weak classifier, which does not ensure that the final classifier improves classification accuracy. This paper analyzes the multi-class Adaboost algorithm by graphic method and math method, and then a new kind of multi-class classification method is proposed which not only reduces the weak classifier accuracy requirements, but also ensures the effectiveness of the weak classifier. In the benchmark experiments on UCI data sets show that the proposed algorithm are better than the SAMME, and achieves the effect of Adaboost.M1.
2016, 38(2): 381-388.
doi: 10.11999/JEIT150403
Abstract:
In order to solve the problem of degraded images captured in hazy weather, a single image dehazing method based on interval estimation is proposed. From the atmospheric scattering model, the minimal filtering and gray-scale opening operation are used to estimate the value of atmospheric light based on dark channel prior theory. At the same time, the initial estimated value of medium transmission is defined. Then, the white balance is performed and the atmospheric scattering model is simplified. Secondly, the simplified atmospheric scattering model and initial estimated value of medium transmission are used to estimate the dark channel value of scene albedo, which is adopted to obtain the coarse estimated value of medium transmission. The final estimated value of medium transmission is obtained by getting through image fusion, joint bilateral filtering and range adjustment. Finally, the simplified atmospheric scattering model and tone mapping are adopted to get the restored image. Experimental results show that the proposed algorithm has a high computation speed, effectively improves the clarity and contrast of restored image, and obtains good color fidelity.
In order to solve the problem of degraded images captured in hazy weather, a single image dehazing method based on interval estimation is proposed. From the atmospheric scattering model, the minimal filtering and gray-scale opening operation are used to estimate the value of atmospheric light based on dark channel prior theory. At the same time, the initial estimated value of medium transmission is defined. Then, the white balance is performed and the atmospheric scattering model is simplified. Secondly, the simplified atmospheric scattering model and initial estimated value of medium transmission are used to estimate the dark channel value of scene albedo, which is adopted to obtain the coarse estimated value of medium transmission. The final estimated value of medium transmission is obtained by getting through image fusion, joint bilateral filtering and range adjustment. Finally, the simplified atmospheric scattering model and tone mapping are adopted to get the restored image. Experimental results show that the proposed algorithm has a high computation speed, effectively improves the clarity and contrast of restored image, and obtains good color fidelity.
2016, 38(2): 389-393.
doi: 10.11999/JEIT150543
Abstract:
Interacting algorithm is widely used in multi-model target tracking, but it is rarely used in multi-system target tracking. In this paper, the interacting idea is used as a reference, and an interacting multi-system tracking algorithm is proposed. The direct interaction between systems is finished based on their former state estimation. Then system probabilities are updated using innovation and its covariance from the parallel filters. Finally, weighted fusing results are achieved on the updated probabilities. The simulation result of tracking a maneuvering target shows that system probability can be adjusted based on its performance immediately, and the tracking performance can be improved effectively.
Interacting algorithm is widely used in multi-model target tracking, but it is rarely used in multi-system target tracking. In this paper, the interacting idea is used as a reference, and an interacting multi-system tracking algorithm is proposed. The direct interaction between systems is finished based on their former state estimation. Then system probabilities are updated using innovation and its covariance from the parallel filters. Finally, weighted fusing results are achieved on the updated probabilities. The simulation result of tracking a maneuvering target shows that system probability can be adjusted based on its performance immediately, and the tracking performance can be improved effectively.
2016, 38(2): 394-399.
doi: 10.11999/JEIT150589
Abstract:
The data maybe miss due to problems in the acquisition, compression or transmission process of multi- channel audio signal. In order to take audiences real auditory sense, an approach of signal recovery based on low rank tensor completion is proposed. First, multi-channel audio signal is represented as a signal tensor. Second, tensor completion is formulated as a convex optimization problem. A closed form for augmented Lagrangian function is obtained via relaxation technique and separation of variables technique. At last, the audio tensor is recovered by alternating iteration. In experiments of varying number of missing entries, the comparisons show that the proposed method is more accurate than linear prediction and CANDECOMP/PARAFAC weighted optimization. The results of multiple stimuli with hidden reference and anchor indicate that low rank tensor completion method is validated for multi-channel audio signal recovery. The better auditory effects are obtained by recovered audio.
The data maybe miss due to problems in the acquisition, compression or transmission process of multi- channel audio signal. In order to take audiences real auditory sense, an approach of signal recovery based on low rank tensor completion is proposed. First, multi-channel audio signal is represented as a signal tensor. Second, tensor completion is formulated as a convex optimization problem. A closed form for augmented Lagrangian function is obtained via relaxation technique and separation of variables technique. At last, the audio tensor is recovered by alternating iteration. In experiments of varying number of missing entries, the comparisons show that the proposed method is more accurate than linear prediction and CANDECOMP/PARAFAC weighted optimization. The results of multiple stimuli with hidden reference and anchor indicate that low rank tensor completion method is validated for multi-channel audio signal recovery. The better auditory effects are obtained by recovered audio.
2016, 38(2): 400-407.
doi: 10.11999/JEIT150494
Abstract:
To solve the problem of low adaptability in existing dehazing algorithms caused by the randomness and complexity of atmospheric environment, an adaptive closed-loop dehazing algorithm based on the feedback mechanism is proposed. Firstly, parameters in the proposed algorithm are initialized according to human visual system based characteristic cognitive assessment. Secondly, the estimation of dehazing strength is given as the feedback to correct parameters of local contrast adjustment method, and then adaptively improve the local contrast of image after removing additive light. Finally, the terminating condition is set according to the naturalness of image after dehazing to determine whether to output the result. Experimental results show that the proposed algorithm can adaptively improve the contrast of hazy images with a variety of degradation types and degrees, and the evaluation of information entropy and definition of dehazing results is better than those of other existing algorithms.
To solve the problem of low adaptability in existing dehazing algorithms caused by the randomness and complexity of atmospheric environment, an adaptive closed-loop dehazing algorithm based on the feedback mechanism is proposed. Firstly, parameters in the proposed algorithm are initialized according to human visual system based characteristic cognitive assessment. Secondly, the estimation of dehazing strength is given as the feedback to correct parameters of local contrast adjustment method, and then adaptively improve the local contrast of image after removing additive light. Finally, the terminating condition is set according to the naturalness of image after dehazing to determine whether to output the result. Experimental results show that the proposed algorithm can adaptively improve the contrast of hazy images with a variety of degradation types and degrees, and the evaluation of information entropy and definition of dehazing results is better than those of other existing algorithms.
2016, 38(2): 408-412.
doi: 10.11999/JEIT150628
Abstract:
Troposphere slant delay is the main error source in two way troposphere time transfer. But there is not an accurate model to estimate the slant delay caused by tropospheric in this system. In order to estimate accurately slant delay, the method of ray tracing is presented. Computing mode of refractive index in Hopfield model is introduced to overcome the methods dependence on radiosonde data. Meteorologic data of three observation stations of 35 to 37N in 2010~2012 are selected to improve the applicability of the Hopfield model, the results suggest that precision is less than 35 mm. Then, in order to calculate the tropospheric delay under different angle of incidence (0~ 5) through modified model, three parts observation stations are distinguished by different length, modified model is used to estimate to slant delay of those parts. In the process, meteorologic data of those stations in 2012 is selected. The results suggest that max delay is 24.94~45.37 m in a single way. In two way time transfer, when the delay can counteract 90% or 95%, time delay is 3.1~5.7 ns or 1.5~2.9 ns.
Troposphere slant delay is the main error source in two way troposphere time transfer. But there is not an accurate model to estimate the slant delay caused by tropospheric in this system. In order to estimate accurately slant delay, the method of ray tracing is presented. Computing mode of refractive index in Hopfield model is introduced to overcome the methods dependence on radiosonde data. Meteorologic data of three observation stations of 35 to 37N in 2010~2012 are selected to improve the applicability of the Hopfield model, the results suggest that precision is less than 35 mm. Then, in order to calculate the tropospheric delay under different angle of incidence (0~ 5) through modified model, three parts observation stations are distinguished by different length, modified model is used to estimate to slant delay of those parts. In the process, meteorologic data of those stations in 2012 is selected. The results suggest that max delay is 24.94~45.37 m in a single way. In two way time transfer, when the delay can counteract 90% or 95%, time delay is 3.1~5.7 ns or 1.5~2.9 ns.
2016, 38(2): 413-417.
doi: 10.11999/JEIT150563
Abstract:
To recognize automatically road network from remote sensing image, an improved fuzzy connectedness method is proposed by combining traditional fuzzy connectedness theory with wavelet modulus maximum algorithm. The wavelet modulus maximum image edge detection algorithm is used to solve the problem of selecting seed points automatically in traditional fuzzy connectedness theory. On this basis, traditional fuzzy similarity computational formula is simplified. This can reduce the cost of calculation greatly without reducing the recognition accuracy. Three high-resolution remote sensing images from the satellite Quickbird are processed in the experiments to prove the effectiveness of the proposed method. The results show that the proposed road network recognition method has high accuracy and rapid computation speed.
To recognize automatically road network from remote sensing image, an improved fuzzy connectedness method is proposed by combining traditional fuzzy connectedness theory with wavelet modulus maximum algorithm. The wavelet modulus maximum image edge detection algorithm is used to solve the problem of selecting seed points automatically in traditional fuzzy connectedness theory. On this basis, traditional fuzzy similarity computational formula is simplified. This can reduce the cost of calculation greatly without reducing the recognition accuracy. Three high-resolution remote sensing images from the satellite Quickbird are processed in the experiments to prove the effectiveness of the proposed method. The results show that the proposed road network recognition method has high accuracy and rapid computation speed.
2016, 38(2): 418-424.
doi: 10.11999/JEIT150570
Abstract:
The calibration reference point targets Radar Cross Section (RCS) is a inherent property depending on frequency and incidence angle, for traditional SAR systems, which can be approximately regarded as a constant under the condition of narrowband and narrow beam. However, for high-resolution SAR systems, replacing the RCS in the case of wideband and wide beam by the RCS of central frequency and azimuth aspect, will result in an inaccurate radiometric calibration output. In this paper, correction methods of reference point targets in the echo domain or in the complex image domain are presented. From the experimental result of simulation and ground-based SAR, it can be seen that the absolute calibration factor varies over 1.2 dB before and after the reference point target correction. The results of real data show that point targets in the SAR image are more symmetric after correction of reference targets radiometric characteristic, and the main lobe in azimuth becomes narrower, which is more close to ideal point target impulse response in time domain, thus validating the effectiveness of the correction algorithm.
The calibration reference point targets Radar Cross Section (RCS) is a inherent property depending on frequency and incidence angle, for traditional SAR systems, which can be approximately regarded as a constant under the condition of narrowband and narrow beam. However, for high-resolution SAR systems, replacing the RCS in the case of wideband and wide beam by the RCS of central frequency and azimuth aspect, will result in an inaccurate radiometric calibration output. In this paper, correction methods of reference point targets in the echo domain or in the complex image domain are presented. From the experimental result of simulation and ground-based SAR, it can be seen that the absolute calibration factor varies over 1.2 dB before and after the reference point target correction. The results of real data show that point targets in the SAR image are more symmetric after correction of reference targets radiometric characteristic, and the main lobe in azimuth becomes narrower, which is more close to ideal point target impulse response in time domain, thus validating the effectiveness of the correction algorithm.
2016, 38(2): 425-433.
doi: 10.11999/JEIT150320
Abstract:
The traditional OFDM radar is usually without regard to transmit communication information. A new radar transmitting pattern based on OFDM is designed to realize the integration of radar and communication. And a new method based on compensated communication information is proposed to achieve joint high-resolution estimation of targets ranges and velocities. In the designed radar transmitting pattern, the radar transmits pulse consisting of multi-OFDM symbols and the communication function is realized within the pulse. During coherent processing interval, the subspace projection method is used to obtain the joint super-resolution estimation of ranges and velocities of targets after the echo data is compensated using communication information and induced non-coherent. Theoretical analysis and simulation results show that the proposed method can obtain the joint super-resolution estimation of targets distances and velocities under the condition of guaranteeing the communication function.
The traditional OFDM radar is usually without regard to transmit communication information. A new radar transmitting pattern based on OFDM is designed to realize the integration of radar and communication. And a new method based on compensated communication information is proposed to achieve joint high-resolution estimation of targets ranges and velocities. In the designed radar transmitting pattern, the radar transmits pulse consisting of multi-OFDM symbols and the communication function is realized within the pulse. During coherent processing interval, the subspace projection method is used to obtain the joint super-resolution estimation of ranges and velocities of targets after the echo data is compensated using communication information and induced non-coherent. Theoretical analysis and simulation results show that the proposed method can obtain the joint super-resolution estimation of targets distances and velocities under the condition of guaranteeing the communication function.
2016, 38(2): 434-441.
doi: 10.11999/JEIT150316
Abstract:
Range alignment is an important step in the preprocessing of wideband radar echoes. However for the echoes of space coning target with micro-motions that only has several scatter centers, the range alignment methods available do not work well. Range alignment is the key step in the parameters estimation algorithms for wideband radar echoes of targets with micro-motions. This paper proposes a range alignment method based on the polarization scattering characteristics. The proposed method first estimates the number, locations and polarization scattering matrices of the scatter centers in the wideband radar echoes using the combing Pole Capon and Amplitude and Phase EStimation (P-CAPES) super-resolution method. Then, polarization similarities of the scatter centers in the adjacent echoes are calculated with the polarization scattering matrices and are used to range alignment. Experiments verify the effectiveness of the proposed method by using electromagnetic data.
Range alignment is an important step in the preprocessing of wideband radar echoes. However for the echoes of space coning target with micro-motions that only has several scatter centers, the range alignment methods available do not work well. Range alignment is the key step in the parameters estimation algorithms for wideband radar echoes of targets with micro-motions. This paper proposes a range alignment method based on the polarization scattering characteristics. The proposed method first estimates the number, locations and polarization scattering matrices of the scatter centers in the wideband radar echoes using the combing Pole Capon and Amplitude and Phase EStimation (P-CAPES) super-resolution method. Then, polarization similarities of the scatter centers in the adjacent echoes are calculated with the polarization scattering matrices and are used to range alignment. Experiments verify the effectiveness of the proposed method by using electromagnetic data.
2016, 38(2): 442-449.
doi: 10.11999/JEIT150637
Abstract:
Transmitted waveform can be designed to improve the SINR performance of colocated MIMO in the preference of interference. However, the optimized waveforms generally have high auto-correlation sidelobes which worsen the detection performance of weak targets at the receiver. To solve this problem, a method of partially correlation signal design for MIMO radar in the presence of interference is proposed in this paper. A set of orthogonal waveforms with constant modulus is weighted at the transmit antenna with the constraint of Peak-to-Average Power Ratios (PAPR), and the objective function is constructed by maximizing the Signal to Interference plus Noise Ratio (SINR) of the receive system. The Sequential Optimization Algorithm based on the Quasi-Newton Method (SOA-QNM) is proposed to find the optimal weights to improve the SINR of the system. Simulation results show that the proposed method can suppress the interference effectively and the emitted power of the transmitted signal can be adaptively concentrated on the direction of the target to improve the SINR performance.
Transmitted waveform can be designed to improve the SINR performance of colocated MIMO in the preference of interference. However, the optimized waveforms generally have high auto-correlation sidelobes which worsen the detection performance of weak targets at the receiver. To solve this problem, a method of partially correlation signal design for MIMO radar in the presence of interference is proposed in this paper. A set of orthogonal waveforms with constant modulus is weighted at the transmit antenna with the constraint of Peak-to-Average Power Ratios (PAPR), and the objective function is constructed by maximizing the Signal to Interference plus Noise Ratio (SINR) of the receive system. The Sequential Optimization Algorithm based on the Quasi-Newton Method (SOA-QNM) is proposed to find the optimal weights to improve the SINR of the system. Simulation results show that the proposed method can suppress the interference effectively and the emitted power of the transmitted signal can be adaptively concentrated on the direction of the target to improve the SINR performance.
2016, 38(2): 450-454.
doi: 10.11999/JEIT150464
Abstract:
In order to improve the penetration ability of the cone target, [d6]a method is proposed[d7] to generate the array ISAR distributed along the range and azimuth direction by the intermittent sampling repeater jamming in fast and slow time. The ISAR imaging method of the precession cone target under binary radar is[d8] analyzed; aimed at the linear frequency modulation continuous wave, the basic theory of the array ISAR generated by intermittent sampling repeater jamming in fast and slow time is analyzed[d9], and the distribution character of the array ISAR false[d10] target is also[d11] analyzed. Finally, with the electromagnetic simulation data, the array ISAR,s distribution character influenced by the parameters of intermittent sampling repeater jamming in fast and slow time is[d12] discussed, which can provide reference for deception jamming for ISAR.
In order to improve the penetration ability of the cone target, [d6]a method is proposed[d7] to generate the array ISAR distributed along the range and azimuth direction by the intermittent sampling repeater jamming in fast and slow time. The ISAR imaging method of the precession cone target under binary radar is[d8] analyzed; aimed at the linear frequency modulation continuous wave, the basic theory of the array ISAR generated by intermittent sampling repeater jamming in fast and slow time is analyzed[d9], and the distribution character of the array ISAR false[d10] target is also[d11] analyzed. Finally, with the electromagnetic simulation data, the array ISAR,s distribution character influenced by the parameters of intermittent sampling repeater jamming in fast and slow time is[d12] discussed, which can provide reference for deception jamming for ISAR.
2016, 38(2): 455-463.
doi: 10.11999/JEIT150581
Abstract:
This paper focuses on the multifractal properties of sea clutter in power spectrum domain. To overcome the deficiencies of Fourier transform analysis, the power spectrum of the sea clutter is obtained by AutoRegressive (AR) spectrum estimation. The AR model is a linear predictive model, which estimates the power spectrum of sea clutter from its autocorrelation matrix and has a higher frequency resolution than Fourier analysis. This paper concentrates on analyzing the multifractal property of the power spectrum based on AR spectral estimation and its application to weak target detection. Firstly, Fractional Brownian Motion (FBM) is taken as an example to prove the multifractal property of the power spectrum. Then, real measured X-band data is used to verify the multifractal property of the AR spectrum of sea clutter by MultiFractal Detrended Fluctuation Analysis (MF-DFA) method. Finally, the generalized Hurst exponent of AR spectrum and its influence factors are analyzed, and a novel detection method based on local AR generalized Hurst exponent is proposed. The results show that the proposed method is effective for weak target detection in sea clutter background. Compared to the existing fractal method and the traditional CFAR method, the proposed method has a better detection performance in low SCR condition.
This paper focuses on the multifractal properties of sea clutter in power spectrum domain. To overcome the deficiencies of Fourier transform analysis, the power spectrum of the sea clutter is obtained by AutoRegressive (AR) spectrum estimation. The AR model is a linear predictive model, which estimates the power spectrum of sea clutter from its autocorrelation matrix and has a higher frequency resolution than Fourier analysis. This paper concentrates on analyzing the multifractal property of the power spectrum based on AR spectral estimation and its application to weak target detection. Firstly, Fractional Brownian Motion (FBM) is taken as an example to prove the multifractal property of the power spectrum. Then, real measured X-band data is used to verify the multifractal property of the AR spectrum of sea clutter by MultiFractal Detrended Fluctuation Analysis (MF-DFA) method. Finally, the generalized Hurst exponent of AR spectrum and its influence factors are analyzed, and a novel detection method based on local AR generalized Hurst exponent is proposed. The results show that the proposed method is effective for weak target detection in sea clutter background. Compared to the existing fractal method and the traditional CFAR method, the proposed method has a better detection performance in low SCR condition.
2016, 38(2): 464-471.
doi: 10.11999/JEIT150683
Abstract:
In many multi-target tracking scenarios, the amplitude of target returns are stronger than those coming from false alarms. This amplitude information can be used to improve the multi-target state estimation by obtaining more accurate target and false-alarm likelihoods. In this paper, a novel multi-Bernoulli filtering algorithm is proposed, which is based on the random finite set and incorporate the amplitude information. The amplitude likelihood functions are derived to incorporate the amplitude information into the multi-Bernoulli filter in the update step. In addition, a Gaussian Mixture (GM) implementation for the linear model and a Sequential Monte Carlo (SMC) implementation for the non-linear model are proposed. Simulation results for Gaussian Mixture and Sequential Monte Carlo implementations show that the proposed filter demonstrates a significant improvement than conventional multi-Bernoulli filter in the estimation accuracy of both the number of targets and their states.
In many multi-target tracking scenarios, the amplitude of target returns are stronger than those coming from false alarms. This amplitude information can be used to improve the multi-target state estimation by obtaining more accurate target and false-alarm likelihoods. In this paper, a novel multi-Bernoulli filtering algorithm is proposed, which is based on the random finite set and incorporate the amplitude information. The amplitude likelihood functions are derived to incorporate the amplitude information into the multi-Bernoulli filter in the update step. In addition, a Gaussian Mixture (GM) implementation for the linear model and a Sequential Monte Carlo (SMC) implementation for the non-linear model are proposed. Simulation results for Gaussian Mixture and Sequential Monte Carlo implementations show that the proposed filter demonstrates a significant improvement than conventional multi-Bernoulli filter in the estimation accuracy of both the number of targets and their states.
2016, 38(2): 472-477.
doi: 10.11999/JEIT150523
Abstract:
According to the theoretical performance for parameter estimators of a signal combined Pseudo-Random Binary-phase Code (PRBC) and Linear Frequency Modulation (LFM), this paper analyzes the impact of step-by- step method on the signal estimation problem and gives the analytical expressions of Modified Cramer-Rao Lower Bound (MCRLB) for parameter estimation of PRBC-LFM signal in white Gaussian noise. Using the statistical characteristics of PRBCs, the Fishers information matrix is derived for parameters including initial frequency, chirp rate, initial phase and code width. The MCRLBs are therefore calculated. The MCRLBs are analyzed by comparison and the validity is demonstrated by numerical simulation experiments.
According to the theoretical performance for parameter estimators of a signal combined Pseudo-Random Binary-phase Code (PRBC) and Linear Frequency Modulation (LFM), this paper analyzes the impact of step-by- step method on the signal estimation problem and gives the analytical expressions of Modified Cramer-Rao Lower Bound (MCRLB) for parameter estimation of PRBC-LFM signal in white Gaussian noise. Using the statistical characteristics of PRBCs, the Fishers information matrix is derived for parameters including initial frequency, chirp rate, initial phase and code width. The MCRLBs are therefore calculated. The MCRLBs are analyzed by comparison and the validity is demonstrated by numerical simulation experiments.
2016, 38(2): 478-485.
doi: 10.11999/JEIT150598
Abstract:
Scattering centers are important features of electromagnetic scattering at high frequencies. The aspect dependency of the amplitude and location of a scattering center has a significant influence on radar imaging and target recognition. Compared with other radar images, the Time-Frequency Representation (TFR) of radar returns more clearly presents features of scattering centers. In this paper, the signatures of TFR of different types of scattering centers are investigated theoretically and numerically. In numerical experiments, the scattering responses of several typical targets are computed by the full-wave numerical method. The conclusion of this paper on signatures of TFR of scattering centers can provide a theoretical reference for the feature extraction and target recognition from TFRs.
Scattering centers are important features of electromagnetic scattering at high frequencies. The aspect dependency of the amplitude and location of a scattering center has a significant influence on radar imaging and target recognition. Compared with other radar images, the Time-Frequency Representation (TFR) of radar returns more clearly presents features of scattering centers. In this paper, the signatures of TFR of different types of scattering centers are investigated theoretically and numerically. In numerical experiments, the scattering responses of several typical targets are computed by the full-wave numerical method. The conclusion of this paper on signatures of TFR of scattering centers can provide a theoretical reference for the feature extraction and target recognition from TFRs.
2016, 38(2): 486-494.
doi: 10.11999/JEIT150401
Abstract:
The traditional numerical method of calculating electromagnetic scattering from the dielectric sea surface requires large amounts of memory and computation time as irradiated area increasing rapidly at low grazing angles. The method of Sparse Matrix Canonical Grid (SMCG) computes the product of the Taylor expanded flat surface matrix and the surface current column vector in far field by the Fast Fourier Transform (FFT), which decreases the computation complexity efficiently. According to the properties of the Greens functions of lossy dielectric and free space, the Physics-Based Two-Grid (PBTG) calculates surface field solutions on the both of dense and coarse grids, which reduces the amounts of memory required. Predictions of an exact numerical model using SMCG-PBTG based on Monte Carlo simulation are compared with experimental data. Experimental data is obtained from wave tank experiments in which the backscattering patterns of 1D sea surfaces with PM spectrum at S- and Ku-band are measured. The sea surfaces corresponding to low and moderate windspeed can be directly simulated in wave tank, and the scale model provides an alternative approach for measuring scattering from sea surfaces corresponding to high windspeed. A comparison of the absolute value of the backscattering coefficient shows the theory and experiment to be in good agreement. Results show that the correlation lengths and scattering behaviors are significantly different under the different windspeed.
The traditional numerical method of calculating electromagnetic scattering from the dielectric sea surface requires large amounts of memory and computation time as irradiated area increasing rapidly at low grazing angles. The method of Sparse Matrix Canonical Grid (SMCG) computes the product of the Taylor expanded flat surface matrix and the surface current column vector in far field by the Fast Fourier Transform (FFT), which decreases the computation complexity efficiently. According to the properties of the Greens functions of lossy dielectric and free space, the Physics-Based Two-Grid (PBTG) calculates surface field solutions on the both of dense and coarse grids, which reduces the amounts of memory required. Predictions of an exact numerical model using SMCG-PBTG based on Monte Carlo simulation are compared with experimental data. Experimental data is obtained from wave tank experiments in which the backscattering patterns of 1D sea surfaces with PM spectrum at S- and Ku-band are measured. The sea surfaces corresponding to low and moderate windspeed can be directly simulated in wave tank, and the scale model provides an alternative approach for measuring scattering from sea surfaces corresponding to high windspeed. A comparison of the absolute value of the backscattering coefficient shows the theory and experiment to be in good agreement. Results show that the correlation lengths and scattering behaviors are significantly different under the different windspeed.
2016, 38(2): 495-508.
doi: 10.11999/JEIT150874
Abstract:
Compressive Sensing (CS) theory, based on the sparsity of interested signal, samples degree-of-freedom of signal. CS is expected to improve the performance of imaging radar in the following aspects: improving the quality of imaging, simplifying the designing of radar hardware, shortening the imaging time and compressing data. This paper first combines the analysis of radar imaging with the three aspects of CS, namely the sparsity of interested signal, the compressive sampling and optimization method. Thereafter a particular and comprehensive review of CS theory in imaging radar is summarized, mainly including the relationship between sparsity of the scene and imaging, compressive sampling methods, fast and accurate reconstruction of the scene and the applications to different imaging radar systems. Finally, the unresolved problems in current research and further study directions are pointed out.
Compressive Sensing (CS) theory, based on the sparsity of interested signal, samples degree-of-freedom of signal. CS is expected to improve the performance of imaging radar in the following aspects: improving the quality of imaging, simplifying the designing of radar hardware, shortening the imaging time and compressing data. This paper first combines the analysis of radar imaging with the three aspects of CS, namely the sparsity of interested signal, the compressive sampling and optimization method. Thereafter a particular and comprehensive review of CS theory in imaging radar is summarized, mainly including the relationship between sparsity of the scene and imaging, compressive sampling methods, fast and accurate reconstruction of the scene and the applications to different imaging radar systems. Finally, the unresolved problems in current research and further study directions are pointed out.
