Email alert
2007 Vol. 29, No. 2
Display Method:
2007, 29(2): 255-259.
doi: 10.3724/SP.J.1146.2005.00960
Abstract:
SNR maximum likelihood estimator based on data directed decision was evaluated first, and was concluded as a biased estimator. An unbiased estimator can be acquired through high order moments of received data. Two SNR estimators based on iteration method were developed in this paper. One is solving the QPSK NDA maximum likelihood equation by gradient approximation, which has better convergence performance than other methods. The other is based on the statistical parameters of received data, which not only makes received data buffering unnecessary, but also reduces iteration calculation greatly compared with other iteration methods, and more important, not at the cost of estimation performance degradation. Finally, the result of several SNR estimators in performance and calculation were compared and analyzed.
SNR maximum likelihood estimator based on data directed decision was evaluated first, and was concluded as a biased estimator. An unbiased estimator can be acquired through high order moments of received data. Two SNR estimators based on iteration method were developed in this paper. One is solving the QPSK NDA maximum likelihood equation by gradient approximation, which has better convergence performance than other methods. The other is based on the statistical parameters of received data, which not only makes received data buffering unnecessary, but also reduces iteration calculation greatly compared with other iteration methods, and more important, not at the cost of estimation performance degradation. Finally, the result of several SNR estimators in performance and calculation were compared and analyzed.
2007, 29(2): 260-264.
doi: 10.3724/SP.J.1146.2005.00952
Abstract:
An improved iterative detection algorithm is proposed in this paper, which introduces a new method of computing soft information based on the idea of semidefinite relaxation rather than on a list. Also a head-first list search algorithm is put forward to improve the performance of list sequential detector. Simulations show that the head-first list sequential detector outperforms the basic list sequential detector at the cost of little increasing complexity and the proposed iterative detector achieves the same performance of head-first list sequential detector but decreases the complexity.
An improved iterative detection algorithm is proposed in this paper, which introduces a new method of computing soft information based on the idea of semidefinite relaxation rather than on a list. Also a head-first list search algorithm is put forward to improve the performance of list sequential detector. Simulations show that the head-first list sequential detector outperforms the basic list sequential detector at the cost of little increasing complexity and the proposed iterative detector achieves the same performance of head-first list sequential detector but decreases the complexity.
2007, 29(2): 265-268.
doi: 10.3724/SP.J.1146.2005.00813
Abstract:
Relayed transmission is an efficient method to improve the power efficiency and system capacity. In this paper, subcarrier pair selection and power allocation in broadband OFDM relaying system is addressed. Firstly, an equivalent channel gain model for subcarrier pairs is developed for amplify-and-forward and decode-and-forward relaying modes, respectively. Based on this, a subcarrier pair selection method is given by applying the Hungarian algorithm. Finally, the overall transmit power is allocated to the selected subcarrier pairs using the water-filling approach. Simulation results show that the proposed algorithm achieves a higher system capacity than both the random subcarrier pair selection and uniform power allocation scheme.
Relayed transmission is an efficient method to improve the power efficiency and system capacity. In this paper, subcarrier pair selection and power allocation in broadband OFDM relaying system is addressed. Firstly, an equivalent channel gain model for subcarrier pairs is developed for amplify-and-forward and decode-and-forward relaying modes, respectively. Based on this, a subcarrier pair selection method is given by applying the Hungarian algorithm. Finally, the overall transmit power is allocated to the selected subcarrier pairs using the water-filling approach. Simulation results show that the proposed algorithm achieves a higher system capacity than both the random subcarrier pair selection and uniform power allocation scheme.
2007, 29(2): 269-273.
doi: 10.3724/SP.J.1146.2005.00509
Abstract:
A new algorithm for searching the optimal puncture matrices of the rate compatible punctured Turbo(RCPT) codes is proposed. The algorithm is based on density evolution and its main goal is to analyze the thresholds for different puncture matrices. The thresholds are calculated using Gaussian approximation and simulation. Compared to a conventional distance spectrum approach, this new approach has two advantages. The first advantage is that the new algorithm is independent of the interleaver. The second is that when the Turbo code is very long, the complexity of the calculation will not increase too much for the proposed approach. In AWGN channels, when the code is long, the examples shows that the optimal puncture matrices are the same for both approaches while the proposed approach has reduced computational complexity. This paper also analyzes the puncture matrices of the code used in 3GPP and provides the optimal matrices and thresholds. The result shows that the rate matching algorithm given in 3GPP corresponds to some RCPT matrixes when using puncturing and these matrixes are the same as those optimal matrices given in this paper for certain rates.
A new algorithm for searching the optimal puncture matrices of the rate compatible punctured Turbo(RCPT) codes is proposed. The algorithm is based on density evolution and its main goal is to analyze the thresholds for different puncture matrices. The thresholds are calculated using Gaussian approximation and simulation. Compared to a conventional distance spectrum approach, this new approach has two advantages. The first advantage is that the new algorithm is independent of the interleaver. The second is that when the Turbo code is very long, the complexity of the calculation will not increase too much for the proposed approach. In AWGN channels, when the code is long, the examples shows that the optimal puncture matrices are the same for both approaches while the proposed approach has reduced computational complexity. This paper also analyzes the puncture matrices of the code used in 3GPP and provides the optimal matrices and thresholds. The result shows that the rate matching algorithm given in 3GPP corresponds to some RCPT matrixes when using puncturing and these matrixes are the same as those optimal matrices given in this paper for certain rates.
2007, 29(2): 274-277.
doi: 10.3724/SP.J.1146.2005.01579
Abstract:
Convergence of Serially Concatenated Continuous Phase Modulation (SCCPM) is studied. Based on the theory of dynamic system, phenomena of positive feedback during iterative detection are analyzed, and improved methods are proposed from the views of decrease and detection of positive feedback. An iterative method based on weighed extrinsic information exchange is proposed, which can improve the convergence of SCCPM effectively. A dynamic iterative method based on the number of change of hard decision is designed; which can detect the positive feedback as well as reduce the average number of iteration. Simulation results show the proposed methods can decrease positive feedback and improve the performance of SCCPM effectively.
Convergence of Serially Concatenated Continuous Phase Modulation (SCCPM) is studied. Based on the theory of dynamic system, phenomena of positive feedback during iterative detection are analyzed, and improved methods are proposed from the views of decrease and detection of positive feedback. An iterative method based on weighed extrinsic information exchange is proposed, which can improve the convergence of SCCPM effectively. A dynamic iterative method based on the number of change of hard decision is designed; which can detect the positive feedback as well as reduce the average number of iteration. Simulation results show the proposed methods can decrease positive feedback and improve the performance of SCCPM effectively.
2007, 29(2): 278-282.
doi: 10.3724/SP.J.1146.2005.00614
Abstract:
This paper discusses two traceback algorithms for Viterbi decoder. The realization methods for the traceback algorithms with FPGA are given through optimization for the hardware architecture. The comparison between the two realization methods is given. Finally, the two realization methods are applied to Viterbi decoder, and both simulation and hardware test show that the presented implementation methods are correct.
This paper discusses two traceback algorithms for Viterbi decoder. The realization methods for the traceback algorithms with FPGA are given through optimization for the hardware architecture. The comparison between the two realization methods is given. Finally, the two realization methods are applied to Viterbi decoder, and both simulation and hardware test show that the presented implementation methods are correct.
2007, 29(2): 283-286.
doi: 10.3724/SP.J.1146.2005.00743
Abstract:
Constant modulus algorithm is a most popular algorithm and is widely used for blind equalization of non-constant envelope signals and constant envelope signals. But it exhibits very slow convergence speed and large residual error. A modified variable step size algorithm is proposed based on constant modulus algorithm. Its theoretical analysis is made and it is simulated with binary sequence and quadrature amplitude modulation signals. Theoretical analysis and simulation results show that it has the characteristics of fast convergence speed and small residual error.
Constant modulus algorithm is a most popular algorithm and is widely used for blind equalization of non-constant envelope signals and constant envelope signals. But it exhibits very slow convergence speed and large residual error. A modified variable step size algorithm is proposed based on constant modulus algorithm. Its theoretical analysis is made and it is simulated with binary sequence and quadrature amplitude modulation signals. Theoretical analysis and simulation results show that it has the characteristics of fast convergence speed and small residual error.
2007, 29(2): 287-290.
doi: 10.3724/SP.J.1146.2005.01499
Abstract:
A based-MSE semi-blind multiuser detection algorithm of asynchronous fast-frequency hopping multiply access communication system is proposed, which exploit the difference of arrive time of asynchronous user and coordinate of signal energy of desired user. The detector implement decision through distinguish most energy row and the minimum Mean Square Error of decode T-F matrix. The algorithm can blind detect without other interference user address, or semi-blind detect utilize some known user address and improve performance of system. Simulation results show that proposed detector has better performance and lower complication, and strong capability of resistant false alarm and deletion.
A based-MSE semi-blind multiuser detection algorithm of asynchronous fast-frequency hopping multiply access communication system is proposed, which exploit the difference of arrive time of asynchronous user and coordinate of signal energy of desired user. The detector implement decision through distinguish most energy row and the minimum Mean Square Error of decode T-F matrix. The algorithm can blind detect without other interference user address, or semi-blind detect utilize some known user address and improve performance of system. Simulation results show that proposed detector has better performance and lower complication, and strong capability of resistant false alarm and deletion.
2007, 29(2): 291-295.
doi: 10.3724/SP.J.1146.2005.00711
Abstract:
In the spread spectrum communication, the characteristics of spreading sequences are important to the anti-jamming,anti multi-path fading,the low probability of intercept and the acquisition of the system. The characteristics of the spreading codes have tremendous impact on the performance of the spread spectrum system. Specifically, the characteristics of the spreading codes also play an important part in the Self-Encoded Spread Spectrum (SESS) communication system. In this paper, the characteristics of the spreading codes generated by AR filter in SESS are analyzed in detail. The results of the computer simulation show that the characteristics of self-encoded spread spectrum sequences obtained with AR filter meet the requirements of SESS system.
In the spread spectrum communication, the characteristics of spreading sequences are important to the anti-jamming,anti multi-path fading,the low probability of intercept and the acquisition of the system. The characteristics of the spreading codes have tremendous impact on the performance of the spread spectrum system. Specifically, the characteristics of the spreading codes also play an important part in the Self-Encoded Spread Spectrum (SESS) communication system. In this paper, the characteristics of the spreading codes generated by AR filter in SESS are analyzed in detail. The results of the computer simulation show that the characteristics of self-encoded spread spectrum sequences obtained with AR filter meet the requirements of SESS system.
2007, 29(2): 296-300.
doi: 10.3724/SP.J.1146.2005.00902
Abstract:
The performance of Transmit Antenna Selection (TAS) with imperfect feedback is analysed in this paper. An asymptotic method is presented to approach the Symbol Error Probability(SEP) of TAS with imperfect feedback. This method can be applied to arbitrary Orthogonal Space Time Block Code (OSTBC) scheme with TAS, and the generalized formula of SEP is also presented for Alamouti TAS scheme with imperfect feedback. It is concluded that the transmit diversity order is equal to the largest norm-amplitude ordinal number of the antenna within the selected antenna subset. It is validated by Monte-Carlo simulation that the presented asymptotic SEP match well with the theoretical SEP at high SNR.
The performance of Transmit Antenna Selection (TAS) with imperfect feedback is analysed in this paper. An asymptotic method is presented to approach the Symbol Error Probability(SEP) of TAS with imperfect feedback. This method can be applied to arbitrary Orthogonal Space Time Block Code (OSTBC) scheme with TAS, and the generalized formula of SEP is also presented for Alamouti TAS scheme with imperfect feedback. It is concluded that the transmit diversity order is equal to the largest norm-amplitude ordinal number of the antenna within the selected antenna subset. It is validated by Monte-Carlo simulation that the presented asymptotic SEP match well with the theoretical SEP at high SNR.
2007, 29(2): 301-304.
doi: 10.3724/SP.J.1146.2005.00886
Abstract:
This paper investigates the impact of feedback period and delay on the performance of Multiple-Input Multiple-Output (MIMO) systems with adaptive modulation, and analyzes the characteristics of the equivalent parallel fading subchannels obtained by Singular Value Decomposition (SVD). Its receiver detection method and a simplified detection order are given as well. The results show that the correlation of the subchannels fading remains large within a certain period of time. System's performance can be improved significantly when the Zero Forcing (ZF) and Interference Cancellation (IC) techniques are adopted at the receiver.
This paper investigates the impact of feedback period and delay on the performance of Multiple-Input Multiple-Output (MIMO) systems with adaptive modulation, and analyzes the characteristics of the equivalent parallel fading subchannels obtained by Singular Value Decomposition (SVD). Its receiver detection method and a simplified detection order are given as well. The results show that the correlation of the subchannels fading remains large within a certain period of time. System's performance can be improved significantly when the Zero Forcing (ZF) and Interference Cancellation (IC) techniques are adopted at the receiver.
2007, 29(2): 305-309.
doi: 10.3724/SP.J.1146.2006.00599
Abstract:
In distributed radio mobile communications system, Remote Antenna Unit (RAU) selection scheme will influence the system performance. In this paper, a selection model is presented for RAU selection in distributed radio system based on soft handoff theories. The selection strategy is based on received signal strength (RSS) measurements and described as a multiple state Markov chain. The RSS threshold is determined by velocity of MS. Performance analysis results show that numeric data achieved from the model are in good agreement with simulation results, which verifies the validity of the model.
In distributed radio mobile communications system, Remote Antenna Unit (RAU) selection scheme will influence the system performance. In this paper, a selection model is presented for RAU selection in distributed radio system based on soft handoff theories. The selection strategy is based on received signal strength (RSS) measurements and described as a multiple state Markov chain. The RSS threshold is determined by velocity of MS. Performance analysis results show that numeric data achieved from the model are in good agreement with simulation results, which verifies the validity of the model.
2007, 29(2): 310-313.
doi: 10.3724/SP.J.1146.2005.00740
Abstract:
This paper proposes a method of channel estimation based on adaptive filters in wireless MIMO-OFDM systems. Though this method does not require any prior knowledge of channel statistics, it can track and estimate time-varying channel instantaneously with the help of adaptive filters. Simulation results have proved this method based on adaptive filters has better MSE and BER performances than another method based on the LS (Least Square) algorithm. The method based on LMS (Least Mean Square) filter has low computational complexity. Moreover, the method based on RLS (Recursive Least Square) filter has better performance and faster convergence than the one based on LMS filter.
This paper proposes a method of channel estimation based on adaptive filters in wireless MIMO-OFDM systems. Though this method does not require any prior knowledge of channel statistics, it can track and estimate time-varying channel instantaneously with the help of adaptive filters. Simulation results have proved this method based on adaptive filters has better MSE and BER performances than another method based on the LS (Least Square) algorithm. The method based on LMS (Least Mean Square) filter has low computational complexity. Moreover, the method based on RLS (Recursive Least Square) filter has better performance and faster convergence than the one based on LMS filter.
2007, 29(2): 314-318.
doi: 10.3724/SP.J.1146.2005.00907
Abstract:
This paper analyzes the parallel estimation method of the fast fading channel in the present of impulse noise. In wireless system, the fast fading channel is typically modeled as the AR (Auto-Regressive) process. LS (Least Square) algorithm and adaptive Kalman filter are used to estimate the AR parameters and the channel impulse response respectively. The performance of these algorithms, however, is very sensitive to the impulse noise. In this paper, the enhanced RLM algorithm and adaptive Kalman filter are proposed and employed to jointly estimate the AR parameters and the channel impulse response under the impulse noise. Simulation results show that the coupled estimator using the enhanced RLM algorithm and Kalman filter has better convergence ability than conventional algorithms.
This paper analyzes the parallel estimation method of the fast fading channel in the present of impulse noise. In wireless system, the fast fading channel is typically modeled as the AR (Auto-Regressive) process. LS (Least Square) algorithm and adaptive Kalman filter are used to estimate the AR parameters and the channel impulse response respectively. The performance of these algorithms, however, is very sensitive to the impulse noise. In this paper, the enhanced RLM algorithm and adaptive Kalman filter are proposed and employed to jointly estimate the AR parameters and the channel impulse response under the impulse noise. Simulation results show that the coupled estimator using the enhanced RLM algorithm and Kalman filter has better convergence ability than conventional algorithms.
2007, 29(2): 319-322.
doi: 10.3724/SP.J.1146.2006.01024
Abstract:
This paper proposes a generalized selection cooperative diversity and its improved scheme with threshold selection for multi-hop wireless networks. The regenerative capacity performances for the proposed schemes are analyzed and derived over Rayleigh fading channels. The numerical and simulation results show that generalized selection cooperative diversity can offer the full-route diversity, however, its improved scheme not only provides an efficient selection scheme for cooperation nodes but achieves a good tradeoff between the system performance and the processing cost of the destination, and it is very helpful to reduce the power consumption and the hardware complexity of the destination.
This paper proposes a generalized selection cooperative diversity and its improved scheme with threshold selection for multi-hop wireless networks. The regenerative capacity performances for the proposed schemes are analyzed and derived over Rayleigh fading channels. The numerical and simulation results show that generalized selection cooperative diversity can offer the full-route diversity, however, its improved scheme not only provides an efficient selection scheme for cooperation nodes but achieves a good tradeoff between the system performance and the processing cost of the destination, and it is very helpful to reduce the power consumption and the hardware complexity of the destination.
2007, 29(2): 323-327.
doi: 10.3724/SP.J.1146.2005.00600
Abstract:
This paper focuses on how to maintain the wireless network connectivity and performance while reducing the maintenance overhead as some nodes fail. A topology maintenance algorithm based on the shortest path tree is proposed. In this algorithm, the neighbors of a faulty node are first triggered to respond (i.e., rerun the topology control algorithm). Without extra communication overhead, each responding node determines the network connectivity by the contents of Hello messages sent by its neighbors. If the responding nodes could not ensure that the network is connected, the other reachable nodes from the fault node are further triggered. Simulation results show that the proposed scheme can efficiently maintain the network connectivity with low overhead and achieve acceptable performance in terms of both power efficiency and power stretch factor.
This paper focuses on how to maintain the wireless network connectivity and performance while reducing the maintenance overhead as some nodes fail. A topology maintenance algorithm based on the shortest path tree is proposed. In this algorithm, the neighbors of a faulty node are first triggered to respond (i.e., rerun the topology control algorithm). Without extra communication overhead, each responding node determines the network connectivity by the contents of Hello messages sent by its neighbors. If the responding nodes could not ensure that the network is connected, the other reachable nodes from the fault node are further triggered. Simulation results show that the proposed scheme can efficiently maintain the network connectivity with low overhead and achieve acceptable performance in terms of both power efficiency and power stretch factor.
2007, 29(2): 328-332.
doi: 10.3724/SP.J.1146.2005.01617
Abstract:
The paper proposes a new MAC protocol to improve the capacity of 802.11b. IEEE 802.11b provides very high raw bandwidth. However, the overheads introduced by the physical and MAC layers are also increasingly substantial. Moreover, there are two kinds of PLCP (physical layer convergence protocol) PPDU (PLCP protocol data unit), but many researchers did not take full advantage of short PLCP PPDU to improve the capacity of 802.11b. A new physical-adaptation algorithm is proposed, which is able to adaptively choose long or short PLCP PPDU and compute network allocation vector (NAV). The proposed MAC protocol and 802.11b were evaluated by NS. Performance results reveal that the proposed MAC protocol performs better than the 802.11b in term of average end-to-end delay, goodput and average successful delivered rate, specially for high load and high raw bandwidth.
The paper proposes a new MAC protocol to improve the capacity of 802.11b. IEEE 802.11b provides very high raw bandwidth. However, the overheads introduced by the physical and MAC layers are also increasingly substantial. Moreover, there are two kinds of PLCP (physical layer convergence protocol) PPDU (PLCP protocol data unit), but many researchers did not take full advantage of short PLCP PPDU to improve the capacity of 802.11b. A new physical-adaptation algorithm is proposed, which is able to adaptively choose long or short PLCP PPDU and compute network allocation vector (NAV). The proposed MAC protocol and 802.11b were evaluated by NS. Performance results reveal that the proposed MAC protocol performs better than the 802.11b in term of average end-to-end delay, goodput and average successful delivered rate, specially for high load and high raw bandwidth.
2007, 29(2): 333-339.
doi: 10.3724/SP.J.1146.2005.00792
Abstract:
It is a key issue in wireless multi-hop Ad hoc networks to develop a Media Access Control (MAC) protocol by which the contending nodes can share the wireless channel fairly and can fully utilize the network capacity. Considering the characteristics of wireless multi-hop Ad hoc networks, such as multi-hop, networks topology changing frequently, and some nodes contending the wireless channel, this paper builds the model of wireless Ad hoc networks and defines two MAC performance parameters: fairness and network capacity utilization ratio. Moreover, this paper proposes a Fair MAC (FMAC) protocol by modifying the IEEE 802.11 DCF. The simulation results prove that FMAC can make the contending nodes share the wireless channel fairly and fully utilize the network capacity.
It is a key issue in wireless multi-hop Ad hoc networks to develop a Media Access Control (MAC) protocol by which the contending nodes can share the wireless channel fairly and can fully utilize the network capacity. Considering the characteristics of wireless multi-hop Ad hoc networks, such as multi-hop, networks topology changing frequently, and some nodes contending the wireless channel, this paper builds the model of wireless Ad hoc networks and defines two MAC performance parameters: fairness and network capacity utilization ratio. Moreover, this paper proposes a Fair MAC (FMAC) protocol by modifying the IEEE 802.11 DCF. The simulation results prove that FMAC can make the contending nodes share the wireless channel fairly and fully utilize the network capacity.
2007, 29(2): 340-344.
doi: 10.3724/SP.J.1146.2005.00727
Abstract:
In order to implement multi-destination and multi-path anycast routing on the heavy load network, an anycast routing based on improved ant colony algorithm is proposed in this paper according to the similar between ant searching path and anycast routing. When the pheromone trails is simulated by the history information of the forwarded packet, it is added that the food distributing smells is simulated by the destination address flooding load information, so every node can receive the latest information of links and servers. Tables of probabilities are created by nodes based on pheromone, smell and visibility. Tables act as the basis that the succeeding ants choose the routes. This routing algorithm is simulated by NS-2. The test result indicates that it can reduce the delay of transmission, decrease the fluctuating range of severs load, implement the load-balancing of links, increase the capability of networks transmission, and improve the scalability of anycast service.
In order to implement multi-destination and multi-path anycast routing on the heavy load network, an anycast routing based on improved ant colony algorithm is proposed in this paper according to the similar between ant searching path and anycast routing. When the pheromone trails is simulated by the history information of the forwarded packet, it is added that the food distributing smells is simulated by the destination address flooding load information, so every node can receive the latest information of links and servers. Tables of probabilities are created by nodes based on pheromone, smell and visibility. Tables act as the basis that the succeeding ants choose the routes. This routing algorithm is simulated by NS-2. The test result indicates that it can reduce the delay of transmission, decrease the fluctuating range of severs load, implement the load-balancing of links, increase the capability of networks transmission, and improve the scalability of anycast service.
2007, 29(2): 345-349.
doi: 10.3724/SP.J.1146.2005.00607
Abstract:
Peer to Peer (P2P) computing technology provides the basic operation facility for collaborations among nodes of open and dynamic network (esp. Internet) and complicated distributed business processing. However, it is difficult to accomplish such works successfully because of some inherent facts of P2P environment, such as peers could randomly join in or quit the P2P network, which make it hard to choose appropriate task undertakers before the project is constructed. In this paper, a flexible and dynamic collaboration alliance mechanism of P2P based on mobile multi-agent technology is proposed. The collaboration alliance mechanism can be deployed in complex and changeful environments to overcome the difficult problem of peers collaboration, which is working as the following ways: driven by one demand, the leader peer constructs the project; after the work undertakers are all selected, the temporary and optimized collaboration alliance is formed; according to the workflow, the leader peer and other work peers collaborate to accomplish the tasks of the project based on their commitments. The mechanism is also utilized to build the agile Supply Chain Management (SCM) system to verify its practical value.
Peer to Peer (P2P) computing technology provides the basic operation facility for collaborations among nodes of open and dynamic network (esp. Internet) and complicated distributed business processing. However, it is difficult to accomplish such works successfully because of some inherent facts of P2P environment, such as peers could randomly join in or quit the P2P network, which make it hard to choose appropriate task undertakers before the project is constructed. In this paper, a flexible and dynamic collaboration alliance mechanism of P2P based on mobile multi-agent technology is proposed. The collaboration alliance mechanism can be deployed in complex and changeful environments to overcome the difficult problem of peers collaboration, which is working as the following ways: driven by one demand, the leader peer constructs the project; after the work undertakers are all selected, the temporary and optimized collaboration alliance is formed; according to the workflow, the leader peer and other work peers collaborate to accomplish the tasks of the project based on their commitments. The mechanism is also utilized to build the agile Supply Chain Management (SCM) system to verify its practical value.
2007, 29(2): 350-354.
doi: 10.3724/SP.J.1146.2005.00733
Abstract:
In the dense complex signal radar environment, the estimation of the parameters Pulse Repetition Interval (PRI) and deinterleaving is a tedious problem. This paper presents a new method, named Square Sine Wave Interpolating Algorithm (SSWIA ) for deinterleaving based on a Time-Of-Arrival (TOA) analysis, which works well in the dense complex signal radar environment. The core of this method is that it turns the unequal interval TOA sequence into a continuous signal. Then the periodic TOA sequence in the mixed TOA sequence will be extract by using FFT, filtering and zero-cross detection technology. This algorithm is very successful in high-pulse-density radar environments and for complex signal types. The simulation results show that it has better precision of the estimation of parameter PRI, more quick speed of deinterleaving the mixed TOA sequence, less sensitive to interfering pulse, more robust to missed pulses than the conventional published techniques.
In the dense complex signal radar environment, the estimation of the parameters Pulse Repetition Interval (PRI) and deinterleaving is a tedious problem. This paper presents a new method, named Square Sine Wave Interpolating Algorithm (SSWIA ) for deinterleaving based on a Time-Of-Arrival (TOA) analysis, which works well in the dense complex signal radar environment. The core of this method is that it turns the unequal interval TOA sequence into a continuous signal. Then the periodic TOA sequence in the mixed TOA sequence will be extract by using FFT, filtering and zero-cross detection technology. This algorithm is very successful in high-pulse-density radar environments and for complex signal types. The simulation results show that it has better precision of the estimation of parameter PRI, more quick speed of deinterleaving the mixed TOA sequence, less sensitive to interfering pulse, more robust to missed pulses than the conventional published techniques.
2007, 29(2): 355-359.
doi: 10.3724/SP.J.1146.2006.00679
Abstract:
This paper proposes a fused algorithm based on fuzzy integral. This method firstly obtains belief function using the wavelet coefficient of multispectral images and panchromatic images ,then constructs fuzzy density adaptively using the deviation of local window of wavelet coefficient, moreover obtains fused wavelet coefficient by fuzzy integral, finally obtains fused image by inverse wavelet transform. The experiment shows that while the developed method can keep spectral quality of fused image, it can update the spatial detail quality of fused image when appropriate fuzzy density is chosen.
This paper proposes a fused algorithm based on fuzzy integral. This method firstly obtains belief function using the wavelet coefficient of multispectral images and panchromatic images ,then constructs fuzzy density adaptively using the deviation of local window of wavelet coefficient, moreover obtains fused wavelet coefficient by fuzzy integral, finally obtains fused image by inverse wavelet transform. The experiment shows that while the developed method can keep spectral quality of fused image, it can update the spatial detail quality of fused image when appropriate fuzzy density is chosen.
2007, 29(2): 360-364.
doi: 10.3724/SP.J.1146.2005.00654
Abstract:
This paper presents a signal based Doppler rate estimation and compensation method to form a well focused Ultra-Wide Band Synthetic Aperture Radar (UWB-SAR) image. A refined MapDriftt(MD) autofocus technique is promoted to improve its original performance in UWB-SAR. In the refined MD, some implementation skills are adopted to accommodate the broadside sparking and the long aperture. Further more, the estimated Doppler rate error is interpolated and compensated in an iteration manner. As a result, higher order phase errors are compensated and this enable the refined MD technique to be capable in applications with large motion error, for example in UAV SAR. Finally the refined MD technique is verified by real UWB-SAR data imaging.
This paper presents a signal based Doppler rate estimation and compensation method to form a well focused Ultra-Wide Band Synthetic Aperture Radar (UWB-SAR) image. A refined MapDriftt(MD) autofocus technique is promoted to improve its original performance in UWB-SAR. In the refined MD, some implementation skills are adopted to accommodate the broadside sparking and the long aperture. Further more, the estimated Doppler rate error is interpolated and compensated in an iteration manner. As a result, higher order phase errors are compensated and this enable the refined MD technique to be capable in applications with large motion error, for example in UAV SAR. Finally the refined MD technique is verified by real UWB-SAR data imaging.
2007, 29(2): 365-369.
doi: 10.3724/SP.J.1146.2005.00540
Abstract:
Analog Time Delay Units (TDUs) are usually used for wideband phased array beamforming, but quatilization error and high cost restrict its applications. This paper presents a new wideband digital beamforming method using subarrays and Linear Frequency Modulation (LFM) signal. The algorithm implements stretch processing and narrowband filtering, and achieves digital beamforming by weighting inter-subarrays. Lowpass filtering limits the bandwidth of signal prior to digitization and signal processing, reducing the required speed of subsequent digitization and the computing burden. Furthermore, the presence of wideband interference will degrade performance of one dimension profile. In order to suppress the wideband interference, a new wideband jammer nulling algorithm is proposed based on digital beamforming above. The methods can work well in the case of strong interferences. Finally, influences of distance error and grating-lobe are analyzed and computer simulations demonstrate the effectiveness of the proposed method.
Analog Time Delay Units (TDUs) are usually used for wideband phased array beamforming, but quatilization error and high cost restrict its applications. This paper presents a new wideband digital beamforming method using subarrays and Linear Frequency Modulation (LFM) signal. The algorithm implements stretch processing and narrowband filtering, and achieves digital beamforming by weighting inter-subarrays. Lowpass filtering limits the bandwidth of signal prior to digitization and signal processing, reducing the required speed of subsequent digitization and the computing burden. Furthermore, the presence of wideband interference will degrade performance of one dimension profile. In order to suppress the wideband interference, a new wideband jammer nulling algorithm is proposed based on digital beamforming above. The methods can work well in the case of strong interferences. Finally, influences of distance error and grating-lobe are analyzed and computer simulations demonstrate the effectiveness of the proposed method.
2007, 29(2): 370-374.
doi: 10.3724/SP.J.1146.2005.00562
Abstract:
A novel method of target aspect estimation in SAR imagery based on Hough transform and major axis extraction is proposed here, which fully exploits the information contained in the major axis of the target. Firstly, the accurate contour of the target in SAR imagery is extracted after the image segmentation based on Markov Random Field(MRF)model. Secondly, the straight lines on the contour are found by Hough transform and the major axis is extracted by the detection of feature points on the contour. Finally, a fine estimation can be acquired by the combination of estimation results from both Hough transform and major axis extraction. The validity of the method is indicated by experimental results with real SAR data sets.
A novel method of target aspect estimation in SAR imagery based on Hough transform and major axis extraction is proposed here, which fully exploits the information contained in the major axis of the target. Firstly, the accurate contour of the target in SAR imagery is extracted after the image segmentation based on Markov Random Field(MRF)model. Secondly, the straight lines on the contour are found by Hough transform and the major axis is extracted by the detection of feature points on the contour. Finally, a fine estimation can be acquired by the combination of estimation results from both Hough transform and major axis extraction. The validity of the method is indicated by experimental results with real SAR data sets.
2007, 29(2): 375-378.
doi: 10.3724/SP.J.1146.2005.00523
Abstract:
The Immune Algorithm (IA) is proposed with analogies to the concept and the theory of immunity in life science. It inherits the advantages of Genetic Algorithm (GA) and avoids the deterioration-phenomenon. Spatial matrix describes the probabilities that one area is the neighbor of other areas. Because different areas in SAR images have different textural features and have big contrast, the spatial matrix of SAR image has the character that the probability of the areas of same kind is bigger and the probability of the areas of different kinds is smaller. By using this feature of the spatial matrix of SAR image as the vaccine, this paper employs IA to search the best segmentation-threshold. Simulation results show that this method is effective for SAR image segmentation and the performance of the method is better than the present algorithm.
The Immune Algorithm (IA) is proposed with analogies to the concept and the theory of immunity in life science. It inherits the advantages of Genetic Algorithm (GA) and avoids the deterioration-phenomenon. Spatial matrix describes the probabilities that one area is the neighbor of other areas. Because different areas in SAR images have different textural features and have big contrast, the spatial matrix of SAR image has the character that the probability of the areas of same kind is bigger and the probability of the areas of different kinds is smaller. By using this feature of the spatial matrix of SAR image as the vaccine, this paper employs IA to search the best segmentation-threshold. Simulation results show that this method is effective for SAR image segmentation and the performance of the method is better than the present algorithm.
2007, 29(2): 379-382.
doi: 10.3724/SP.J.1146.2005.00824
Abstract:
Ratio Of Exponentially Weighted Averages (ROEWA )operator is an edge detector suitable for SAR images. However, the detector can not compute the direction of edges accurately. Due to the excellent directional selectivity of Gabor filters, this paper puts forward a method of utilizing Gabor filter to computer edge direction. And by this way, the edge detector is improved. Experimental results prove that the performance of the improved method is satisfactory.
Ratio Of Exponentially Weighted Averages (ROEWA )operator is an edge detector suitable for SAR images. However, the detector can not compute the direction of edges accurately. Due to the excellent directional selectivity of Gabor filters, this paper puts forward a method of utilizing Gabor filter to computer edge direction. And by this way, the edge detector is improved. Experimental results prove that the performance of the improved method is satisfactory.
2007, 29(2): 383-386.
doi: 10.3724/SP.J.1146.2005.00932
Abstract:
The paper concerns the error analysis of image geometric distortion correction for missile-borne side-looking SAR. Because of the decrease of altitude during SAR operation, the SAR image obtained by common used imaging algorithm is seriously distorted. In the paper, the geometric distortion correction method of the SAR image related to subaperture RD algorithm is presented. Especially, the errors of geometric distortion correction are analyzed. The effectiveness of the method and error analysis of the geometric distortion correction are demonstrated by simulation.
The paper concerns the error analysis of image geometric distortion correction for missile-borne side-looking SAR. Because of the decrease of altitude during SAR operation, the SAR image obtained by common used imaging algorithm is seriously distorted. In the paper, the geometric distortion correction method of the SAR image related to subaperture RD algorithm is presented. Especially, the errors of geometric distortion correction are analyzed. The effectiveness of the method and error analysis of the geometric distortion correction are demonstrated by simulation.
2007, 29(2): 387-390.
doi: 10.3724/SP.J.1146.2005.00859
Abstract:
The beam of meter-wave radar is rather board, the exist of multi-path signals coherent with the direct arrival signal will cause badly affect to the altitude measurement, so it is more difficult for meter-wave radar to measure the altitude of the target. The general MUSIC algorithm is a super-resolution algorithm that can handle coherent signals, so it can be used in the height-finding of meter-wave radar. Based on the general MUSIC algorithm, a new algorithm for altitude measurement in the height-finding environment is presented. The new algorithm not only has a low computation complexity, but also has a good performance. Computer simulation results demonstrate the merits of the new algorithm.
The beam of meter-wave radar is rather board, the exist of multi-path signals coherent with the direct arrival signal will cause badly affect to the altitude measurement, so it is more difficult for meter-wave radar to measure the altitude of the target. The general MUSIC algorithm is a super-resolution algorithm that can handle coherent signals, so it can be used in the height-finding of meter-wave radar. Based on the general MUSIC algorithm, a new algorithm for altitude measurement in the height-finding environment is presented. The new algorithm not only has a low computation complexity, but also has a good performance. Computer simulation results demonstrate the merits of the new algorithm.
2007, 29(2): 391-396.
doi: 10.3724/SP.J.1146.2005.00814
Abstract:
This paper investigates the parallel processing of spaceborne ScanSAR in detail. Firstly, it analyzes the character of ScanSAR imaging and gives the parallel processing flow based on the traditional Host-Slave (HS) mode. Then, a new parallel technology based on Multiple Node-Multiple Progress (MNMP) mode is proposed, it can improve the deficiency of HS mode. The simulation result indicates the parallel processing based on MNMP mode improves the imaging time. So the way is feasible and superiority.
This paper investigates the parallel processing of spaceborne ScanSAR in detail. Firstly, it analyzes the character of ScanSAR imaging and gives the parallel processing flow based on the traditional Host-Slave (HS) mode. Then, a new parallel technology based on Multiple Node-Multiple Progress (MNMP) mode is proposed, it can improve the deficiency of HS mode. The simulation result indicates the parallel processing based on MNMP mode improves the imaging time. So the way is feasible and superiority.
2007, 29(2): 397-400.
doi: 10.3724/SP.J.1146.2005.00766
Abstract:
A novel approach of Ground Moving Target Indication (GMTI) based on subaperture is proposed in this paper. In this approach, the aperture of an antenna is divided into two subapertures. When the radar is in operation, the two subapertures pass the same ground patch in tandem with an interval of half of synthetic aperture time. During this interval, the position of a stationary target keeps unchanged, but that of a moving one is altered. Then, the moving target is detected by comparing images derived from the two subapertures, and its motion parameters can be estimated with its trajectory. In order to avoid the moving target being smeared and preserve its trajectory, azimuth focusing process is omitted. Compared with present GMTI approaches, the new one has many advantages, only single channel antenna, low cost, less computation load, compatible with SAR mode, etc.
A novel approach of Ground Moving Target Indication (GMTI) based on subaperture is proposed in this paper. In this approach, the aperture of an antenna is divided into two subapertures. When the radar is in operation, the two subapertures pass the same ground patch in tandem with an interval of half of synthetic aperture time. During this interval, the position of a stationary target keeps unchanged, but that of a moving one is altered. Then, the moving target is detected by comparing images derived from the two subapertures, and its motion parameters can be estimated with its trajectory. In order to avoid the moving target being smeared and preserve its trajectory, azimuth focusing process is omitted. Compared with present GMTI approaches, the new one has many advantages, only single channel antenna, low cost, less computation load, compatible with SAR mode, etc.
2007, 29(2): 401-404.
doi: 10.3724/SP.J.1146.2005.00738
Abstract:
Ship ISAR imaging is studied in the paper based on the airborne SAR real data. SAR imaging deals with static targets. A moving ship appears blurred in the SAR image. For the purpose of getting high resolution ship images, ISAR imagery may be introduced into the imaging due to the non-cooperative character of the ship target. In the paper, the extraction of ship data from SAR real data is described firstly. Then, effective ISAR motion compensation methods appropriate to ship imaging are discussed. Finally, clear images of several typical ships are presented. This work provides effective processing methods for ship imaging when the airborne radar works in SAR mode.
Ship ISAR imaging is studied in the paper based on the airborne SAR real data. SAR imaging deals with static targets. A moving ship appears blurred in the SAR image. For the purpose of getting high resolution ship images, ISAR imagery may be introduced into the imaging due to the non-cooperative character of the ship target. In the paper, the extraction of ship data from SAR real data is described firstly. Then, effective ISAR motion compensation methods appropriate to ship imaging are discussed. Finally, clear images of several typical ships are presented. This work provides effective processing methods for ship imaging when the airborne radar works in SAR mode.
2007, 29(2): 405-409.
doi: 10.3724/SP.J.1146.2005.00827
Abstract:
Firstly, a Bistatic Synthetic Aperture Sonar (Bi-SAS) configuration with fixed transmitter is given in this paper, and then the azimuth and range resolution, the Pulse Repetition Frequency (PRF) and the mapping rate of the Bistatic-SAS are studied in detail. The analytical solutions are obtained, consequently, the computer simulation are presented in the end. The simulation results validate the correctness of the theory and show the validity of the method in the Bistatic-SAS image.
Firstly, a Bistatic Synthetic Aperture Sonar (Bi-SAS) configuration with fixed transmitter is given in this paper, and then the azimuth and range resolution, the Pulse Repetition Frequency (PRF) and the mapping rate of the Bistatic-SAS are studied in detail. The analytical solutions are obtained, consequently, the computer simulation are presented in the end. The simulation results validate the correctness of the theory and show the validity of the method in the Bistatic-SAS image.
2007, 29(2): 410-413.
doi: 10.3724/SP.J.1146.2005.00818
Abstract:
This paper introduces a new adaptive filtering algorithm under the -stable noise conditions, which is obtained by minimizing the weighted average p-norm of error function. Two new versions of momentum LMP algorithm can be derived from the proposed one. Computer simulations on the channel equalization under the -stable noise conditions show that the proposed algorithms provide better estimation performance than those of the existing LMP algorithm.
This paper introduces a new adaptive filtering algorithm under the -stable noise conditions, which is obtained by minimizing the weighted average p-norm of error function. Two new versions of momentum LMP algorithm can be derived from the proposed one. Computer simulations on the channel equalization under the -stable noise conditions show that the proposed algorithms provide better estimation performance than those of the existing LMP algorithm.
2007, 29(2): 414-417.
doi: 10.3724/SP.J.1146.2005.00873
Abstract:
A method for joint estimating frequency, elevation, bearing and range of near-field narrowband sources is presented. The method uses eigenvalues together with the corresponding eigenvectors and four-order cumulant to estimate 4-D parameters of sources, which does not need spectral peak searching, and it can be applied to additive Guassian noise environments. The effectiveness is verified by the computer simulations.
A method for joint estimating frequency, elevation, bearing and range of near-field narrowband sources is presented. The method uses eigenvalues together with the corresponding eigenvectors and four-order cumulant to estimate 4-D parameters of sources, which does not need spectral peak searching, and it can be applied to additive Guassian noise environments. The effectiveness is verified by the computer simulations.
2007, 29(2): 418-420.
doi: 10.3724/SP.J.1146.2005.00834
Abstract:
A fast recursive algorithm is proposed in this paper for the realization of Discrete Cosine Transforms (DCT) with arbitrary length jointly using Clenshaw recurrence formula and the symmetry of DCT. Compared with other exiting recursive algorithms and the method of arithmetic Fourier transform for computing DCT, the proposed algorithm holds a lower computation complexity. With regular digital filters structures, the algorithm is also effective for VLSI implementation.
A fast recursive algorithm is proposed in this paper for the realization of Discrete Cosine Transforms (DCT) with arbitrary length jointly using Clenshaw recurrence formula and the symmetry of DCT. Compared with other exiting recursive algorithms and the method of arithmetic Fourier transform for computing DCT, the proposed algorithm holds a lower computation complexity. With regular digital filters structures, the algorithm is also effective for VLSI implementation.
2007, 29(2): 421-424.
doi: 10.3724/SP.J.1146.2005.00826
Abstract:
The ridgelet transform is a new directional multi-resolution transform，which is more suitable for describing signals with high dimensional singularities. Finite ridgelet transform is a discrete version of ridgelet transform, which is as numerically precise as the continuous ridgelet transform and has low computational complexity. However, finite ridgelet transform is only suitable for images of prime-pixels length, which is a limitation of its application in image processing. In this paper, an improved finite ridgelet algorithm which is suitable for images of dyadic length is proposed. This method not only expands the application of finite ridgelet, but also simplifies the algorithm.
The ridgelet transform is a new directional multi-resolution transform，which is more suitable for describing signals with high dimensional singularities. Finite ridgelet transform is a discrete version of ridgelet transform, which is as numerically precise as the continuous ridgelet transform and has low computational complexity. However, finite ridgelet transform is only suitable for images of prime-pixels length, which is a limitation of its application in image processing. In this paper, an improved finite ridgelet algorithm which is suitable for images of dyadic length is proposed. This method not only expands the application of finite ridgelet, but also simplifies the algorithm.
2007, 29(2): 425-428.
doi: 10.3724/SP.J.1146.2005.00901
Abstract:
A new matching metricalignment metric of feature points is defined, and a new approach to image registration based on alignment metric of feature points is proposed. Feature points and their angles are detected by using wavelet multi-scale product, then the alignment metric of all feature points are calculated, so matching points can be obtained. The experimental results illustrate the algorithm has a good performance of accuracy and efficiency.
A new matching metricalignment metric of feature points is defined, and a new approach to image registration based on alignment metric of feature points is proposed. Feature points and their angles are detected by using wavelet multi-scale product, then the alignment metric of all feature points are calculated, so matching points can be obtained. The experimental results illustrate the algorithm has a good performance of accuracy and efficiency.
2007, 29(2): 429-433.
doi: 10.3724/SP.J.1146.2005.00910
Abstract:
A very low bit rate speech encoding algorithm in 600bps is proposed in this paper for application in anti-jamming communication. Super-frame structure with 2 base frames and 4 interpolate frames is used, the LPC coefficients of interpolate frames is quantized with 4 stages residual matrix quantization based on optimal 1-order linear prediction. In the decoder, a closed loop estimation of the impulse magnitude is proposed to improve the naturalness of speech and the articulation of nasals. This speech coder achieves good quality of speech and the DRT score is 88.55。
A very low bit rate speech encoding algorithm in 600bps is proposed in this paper for application in anti-jamming communication. Super-frame structure with 2 base frames and 4 interpolate frames is used, the LPC coefficients of interpolate frames is quantized with 4 stages residual matrix quantization based on optimal 1-order linear prediction. In the decoder, a closed loop estimation of the impulse magnitude is proposed to improve the naturalness of speech and the articulation of nasals. This speech coder achieves good quality of speech and the DRT score is 88.55。
2007, 29(2): 434-438.
doi: 10.3724/SP.J.1146.2005.00899
Abstract:
To protect the privacy during the real-time speech communication, in this paper, a novel fast modification method for real-time speech is presented, in which the transformation method of Pseudo Log Area Ratio (PLAR) curve and Pitch-Synchronous-OverLap-and-Add (LP-PSOLA) algorithm based on Linear Prediction are respectively adopted to modify the spectrum parameters and the prosodic parameters of the speech, then the personal characteristics of the speech are modified. In addition, for Synchronous-OverLap-and-Add(SOLA) method popularly used for Time-Scale Modification (TSM) has a lot of computation load, it can not be used for processing the real-time speech. Therefore, in this paper, a novel TSM algorithmAdaptive Synchronous OverLap and Add(ASOLA) presented by authors workgroup is adopted for TSM of the speech, the personal characteristics of which have been modified, to assure the real-time property of the speech. Finally, the fast modification method for real-time speech is used to protect the privacy during the real-time speech communication, and the test shows that, the speech synthesized by this method has high quality, and this approach can assure the real-time speech communication.
To protect the privacy during the real-time speech communication, in this paper, a novel fast modification method for real-time speech is presented, in which the transformation method of Pseudo Log Area Ratio (PLAR) curve and Pitch-Synchronous-OverLap-and-Add (LP-PSOLA) algorithm based on Linear Prediction are respectively adopted to modify the spectrum parameters and the prosodic parameters of the speech, then the personal characteristics of the speech are modified. In addition, for Synchronous-OverLap-and-Add(SOLA) method popularly used for Time-Scale Modification (TSM) has a lot of computation load, it can not be used for processing the real-time speech. Therefore, in this paper, a novel TSM algorithmAdaptive Synchronous OverLap and Add(ASOLA) presented by authors workgroup is adopted for TSM of the speech, the personal characteristics of which have been modified, to assure the real-time property of the speech. Finally, the fast modification method for real-time speech is used to protect the privacy during the real-time speech communication, and the test shows that, the speech synthesized by this method has high quality, and this approach can assure the real-time speech communication.
2007, 29(2): 439-442.
doi: 10.3724/SP.J.1146.2005.00865
Abstract:
This paper addresses the problem of single-channel speech enhancement using spectral subtraction method. The proposed approach is directed toward combining modified a priori SNR based spectral subtraction algorithm with iterative spectral subtraction algorithm. Experimental results show that the method is quite effective in speech enhancement as it can cancel a large mount of background noise with little distortion to speech signals，and that the method greatly lowered music noise.
This paper addresses the problem of single-channel speech enhancement using spectral subtraction method. The proposed approach is directed toward combining modified a priori SNR based spectral subtraction algorithm with iterative spectral subtraction algorithm. Experimental results show that the method is quite effective in speech enhancement as it can cancel a large mount of background noise with little distortion to speech signals，and that the method greatly lowered music noise.
2007, 29(2): 443-446.
doi: 10.3724/SP.J.1146.2005.00566
Abstract:
Based on mulitsensor single model dynamic systems, a state fusion estimation algorithm is presented. Multisensors observe the same target, where different sensors may have different sampling rates and the ratio between them may be positive rational numbers. The algorithm is in real-time, and the optimal in the sense of linear minimum covariance. It is proved that the fused estimate is more accurate than the Kalman filtering result based on single sensors. The fused estimation error covariance will increase if any of the sensors information is neglected. The feasibility and effectiveness of the algorithm are shown through simulation results.
Based on mulitsensor single model dynamic systems, a state fusion estimation algorithm is presented. Multisensors observe the same target, where different sensors may have different sampling rates and the ratio between them may be positive rational numbers. The algorithm is in real-time, and the optimal in the sense of linear minimum covariance. It is proved that the fused estimate is more accurate than the Kalman filtering result based on single sensors. The fused estimation error covariance will increase if any of the sensors information is neglected. The feasibility and effectiveness of the algorithm are shown through simulation results.
2007, 29(2): 447-450.
doi: 10.3724/SP.J.1146.2005.00757
Abstract:
To develop an effective digital media protection method, an algorithm for digital video watermarking is proposed based on binary operation in 3D wavelet transform domain. For realizing the extraction of watermark image, a blind detection and restoration algorithm is presented, which does not require the original video. The proposed watermarking embedding and extracting algorithm uses only four image frames for performing wavelet decomposition per operation with lower memory demands and short delay, which suits for watermark online embedding and extracting and real-time implementation with hardware. The experimental results illustrate that the proposed watermarking algorithm is robust to video degradation and distortion attacks, such as Gaussian noise, MPEG coding, frame averaging, frame dropping and frame swapping.
To develop an effective digital media protection method, an algorithm for digital video watermarking is proposed based on binary operation in 3D wavelet transform domain. For realizing the extraction of watermark image, a blind detection and restoration algorithm is presented, which does not require the original video. The proposed watermarking embedding and extracting algorithm uses only four image frames for performing wavelet decomposition per operation with lower memory demands and short delay, which suits for watermark online embedding and extracting and real-time implementation with hardware. The experimental results illustrate that the proposed watermarking algorithm is robust to video degradation and distortion attacks, such as Gaussian noise, MPEG coding, frame averaging, frame dropping and frame swapping.
2007, 29(2): 451-454.
doi: 10.3724/SP.J.1146.2005.00746
Abstract:
An iterative factorization method based on linear subspace for projective reconstruction is presented in the paper. It relies on the facts that the rows in the matrix including all the image points span the same linear subspace as the rows in the matrix including space points and the fact that any basis of the subspace can be regarded as projective reconstruction. The projective reconstruction and the depth factors are obtained based on linear iteration. The experiments with both simulate and real data show that the method is efficient, robust and has good property of convergence and small reprojection errors.
An iterative factorization method based on linear subspace for projective reconstruction is presented in the paper. It relies on the facts that the rows in the matrix including all the image points span the same linear subspace as the rows in the matrix including space points and the fact that any basis of the subspace can be regarded as projective reconstruction. The projective reconstruction and the depth factors are obtained based on linear iteration. The experiments with both simulate and real data show that the method is efficient, robust and has good property of convergence and small reprojection errors.
2007, 29(2): 455-459.
doi: 10.3724/SP.J.1146.2005.00790
Abstract:
In this paper, a new distance called person-similarity weighted distance is proposed for person-independent facial expression recognition. In the new method, expression features associated with all persons in training set are extracted by Higher-Order Singular Value Decomposition (HOSVD) firstly. Then, based on the assumption similar persons have similar facial expression, the person-similarity weighted distance is calculated to measure the similarity between the test expression and standard expressions. By the weighting process, the distance can remove the differences caused by individual and becomes less person-dependent. Experimental results show the superiority of proposed method over the existing methods.
In this paper, a new distance called person-similarity weighted distance is proposed for person-independent facial expression recognition. In the new method, expression features associated with all persons in training set are extracted by Higher-Order Singular Value Decomposition (HOSVD) firstly. Then, based on the assumption similar persons have similar facial expression, the person-similarity weighted distance is calculated to measure the similarity between the test expression and standard expressions. By the weighting process, the distance can remove the differences caused by individual and becomes less person-dependent. Experimental results show the superiority of proposed method over the existing methods.
2007, 29(2): 460-464.
doi: 10.3724/SP.J.1146.2006.00196
Abstract:
In order to segment the moving objects accurately, an edge based moving objects segmentation algorithm is proposed in this paper. First of all, the initial edge model of the moving object is obtained on the basis of the adaptive change detection method and Canny edge detection. Secondly, the binary edge model is tracked and updated to accommodate rotation and changes in shape of the tracked object. Finally, the outer contour of the moving object is determined based on the mathematical morphological operation and an improved active contour which uses the gradient vector as the external force guides the initial outer contour moving to the actual contour. The proposed algorithm is robust to the entire motion and local deformation of object, adaptable to the complex background and can obtain the accurate moving object contour.
In order to segment the moving objects accurately, an edge based moving objects segmentation algorithm is proposed in this paper. First of all, the initial edge model of the moving object is obtained on the basis of the adaptive change detection method and Canny edge detection. Secondly, the binary edge model is tracked and updated to accommodate rotation and changes in shape of the tracked object. Finally, the outer contour of the moving object is determined based on the mathematical morphological operation and an improved active contour which uses the gradient vector as the external force guides the initial outer contour moving to the actual contour. The proposed algorithm is robust to the entire motion and local deformation of object, adaptable to the complex background and can obtain the accurate moving object contour.
2007, 29(2): 465-468.
doi: 10.3724/SP.J.1146.2005.00963
Abstract:
Based on the feature distance of generalized dimension, a speech endpoint detection method is proposed in order to detect the noisy-corrupted speech efficiently. Through calculating the generalized dimension by covering the signal with n-dimension boxes, three dimension feature vectors including the box dimension, the information dimension and the correlation dimension are got. Then dimension feature distance could be calculated and used to make a classification for the speech signal. Experimental results show that compared with the detection using one dimension feature only, the proposed method is more robust to the endpoint detection of speech signal containing different noise and SNR, especially for the lower SNR signal.
Based on the feature distance of generalized dimension, a speech endpoint detection method is proposed in order to detect the noisy-corrupted speech efficiently. Through calculating the generalized dimension by covering the signal with n-dimension boxes, three dimension feature vectors including the box dimension, the information dimension and the correlation dimension are got. Then dimension feature distance could be calculated and used to make a classification for the speech signal. Experimental results show that compared with the detection using one dimension feature only, the proposed method is more robust to the endpoint detection of speech signal containing different noise and SNR, especially for the lower SNR signal.
2007, 29(2): 469-472.
doi: 10.3724/SP.J.1146.2005.00749
Abstract:
There is a basic choice in the form of covariance matrix to be used with Gaussian mixture model in text-independent speaker identification. In general, diagonal covariance matrix is chose, which implies strong assumption that elements of the feature vector are independent, because full covariance matrix suffers from too many parameters and large computational requirement. Unfortunately, in most application the assumption is not reasonable. In order to make feature vectors more suit to be modeled with diagonal covariance, features are usually de-correlated in feature space or model space. In this paper, an improved model-based PCA transformation algorithm is presented to de-correlate the elements of feature vectors. In this algorithm, principal component analysis is directly made for covariance of Gaussians. Also, the number of parameter is deduced through tying the PCA transformation between Gaussians. Experiments on the MSRA mandarin task show that the algorithm could achieve above 35% identification error reduction over the best diagonal covariance models.
There is a basic choice in the form of covariance matrix to be used with Gaussian mixture model in text-independent speaker identification. In general, diagonal covariance matrix is chose, which implies strong assumption that elements of the feature vector are independent, because full covariance matrix suffers from too many parameters and large computational requirement. Unfortunately, in most application the assumption is not reasonable. In order to make feature vectors more suit to be modeled with diagonal covariance, features are usually de-correlated in feature space or model space. In this paper, an improved model-based PCA transformation algorithm is presented to de-correlate the elements of feature vectors. In this algorithm, principal component analysis is directly made for covariance of Gaussians. Also, the number of parameter is deduced through tying the PCA transformation between Gaussians. Experiments on the MSRA mandarin task show that the algorithm could achieve above 35% identification error reduction over the best diagonal covariance models.
2007, 29(2): 473-477.
doi: 10.3724/SP.J.1146.2005.00915
Abstract:
In the research of spatial hearing and implementation of virtual auditory space, it is important to accurately model the latent acoustical clues in HRTFs(Head-Related Transfer Functions) or HRIRs(Head-Related Impulse Responses) related to certain position of sound source. As an essential preprocessing step, this work introduced a new smoothing means based on trous algorithm with translation-invariant and reconstruction from modulus maxima, and managed to carry through adaptive non-linear approximation in the field of wavelet transformation. The simulation results show that, under the uniform threshold, the performance of the new way is averagely 8.3dB better than that of traditional PCA(Principal Component Analysis) method, and 2.4dB than that of wavelet method using Mallat algorithm.
In the research of spatial hearing and implementation of virtual auditory space, it is important to accurately model the latent acoustical clues in HRTFs(Head-Related Transfer Functions) or HRIRs(Head-Related Impulse Responses) related to certain position of sound source. As an essential preprocessing step, this work introduced a new smoothing means based on trous algorithm with translation-invariant and reconstruction from modulus maxima, and managed to carry through adaptive non-linear approximation in the field of wavelet transformation. The simulation results show that, under the uniform threshold, the performance of the new way is averagely 8.3dB better than that of traditional PCA(Principal Component Analysis) method, and 2.4dB than that of wavelet method using Mallat algorithm.
2007, 29(2): 478-481.
doi: 10.3724/SP.J.1146.2005.00913
Abstract:
This paper presents the all-phase filter design methods based on two kinds of symmetric frequency sampling. Under the condition of no window and single window, it is proved that the all-phase transfer curves designed by conventional symmetry frequency vector and even symmetry frequency vector can pass through the frequency sampling points of odd multiples and even multiples of /N, respectively. In this way all-phase filter can control critical frequency to a relatively precise extent. Simulation experiment proves that the performances of these two kinds of all-phase filter are superior to that of filter designed by neural network methods.
This paper presents the all-phase filter design methods based on two kinds of symmetric frequency sampling. Under the condition of no window and single window, it is proved that the all-phase transfer curves designed by conventional symmetry frequency vector and even symmetry frequency vector can pass through the frequency sampling points of odd multiples and even multiples of /N, respectively. In this way all-phase filter can control critical frequency to a relatively precise extent. Simulation experiment proves that the performances of these two kinds of all-phase filter are superior to that of filter designed by neural network methods.
2007, 29(2): 482-485.
doi: 10.3724/SP.J.1146.2005.00821
Abstract:
To investigate the problem of classical analytical method, which can not be used at Low-Grazing Angle (LGA) for rough surface scattering, a new high-order analytical method based on the high-order series expansion of the two-dimensional electric field and the profile function of rough surface is adopted. By this arithmetic, the co-polarized backscattering coefficient at LGA is compared with high-order Small Perturbation Method (SPM) and classical SPM, the grazing effect at different roughness is discussed, the modified effect for Gaussian rough surface at LGA is analyzed by high-order SPM and high-order Kirchhoff Approximation (KA). The result shows that the high-order analytical method is valid for LGA problem. At last，the vertical-to-horizontal Polarization Ratio (PR) with different power of secant function of grazing incident angle and scattering angle is considered particularly, and the classical expression of PR is rectified.
To investigate the problem of classical analytical method, which can not be used at Low-Grazing Angle (LGA) for rough surface scattering, a new high-order analytical method based on the high-order series expansion of the two-dimensional electric field and the profile function of rough surface is adopted. By this arithmetic, the co-polarized backscattering coefficient at LGA is compared with high-order Small Perturbation Method (SPM) and classical SPM, the grazing effect at different roughness is discussed, the modified effect for Gaussian rough surface at LGA is analyzed by high-order SPM and high-order Kirchhoff Approximation (KA). The result shows that the high-order analytical method is valid for LGA problem. At last，the vertical-to-horizontal Polarization Ratio (PR) with different power of secant function of grazing incident angle and scattering angle is considered particularly, and the classical expression of PR is rectified.
2007, 29(2): 486-488.
doi: 10.3724/SP.J.1146.2006.00134
Abstract:
The dihedral corner reflector with wing and fuselage joint of aircraft is a very strong electromagnetic source in a wide angular range. But traditional RCS computation method is only suit for the ideal dihedral corner reflector. This paper adopts Geometrical Optics (GO) and Physical Optics (PO) methods to compute double and thrice bounce contributions of a rectangle dihedral corner reflector. And on the base of illumination width, a concept of Equivalent Illumination Area (EIA) is presented. Through the analysis of EIA, a modified computational method is given, which can analyze a dihedral corner reflector with arbitrary shape of planes and with shadow objects. To valid this method, RCS of wing and fuselage joints of a certain hypersonic vehicle is computed, and the result validates its correctness.
The dihedral corner reflector with wing and fuselage joint of aircraft is a very strong electromagnetic source in a wide angular range. But traditional RCS computation method is only suit for the ideal dihedral corner reflector. This paper adopts Geometrical Optics (GO) and Physical Optics (PO) methods to compute double and thrice bounce contributions of a rectangle dihedral corner reflector. And on the base of illumination width, a concept of Equivalent Illumination Area (EIA) is presented. Through the analysis of EIA, a modified computational method is given, which can analyze a dihedral corner reflector with arbitrary shape of planes and with shadow objects. To valid this method, RCS of wing and fuselage joints of a certain hypersonic vehicle is computed, and the result validates its correctness.
2007, 29(2): 489-491.
doi: 10.3724/SP.J.1146.2005.00629
Abstract:
The principal objective of the present work is to investigate the statistic properties of SAR return signals. The Synthetic Aperture Radar (SAR) return signals from 3-dimensional Gauss random rough conducting surfaces is efficiently and accurately computed using Multi-Level Fast Multipole Algorithm (MLFMA) combined with traditional Monte Carlo method. According to the real situation of SAR, incidents elevation angles are fixed at 45, and relevant azimuth angles vary from 88 to 92. Return signals are divided into several groups in terms of geometry coefficients. Then the statistic properties of these obtained signals are investigated by computing the means of return signals amplitudes and phases. Several useful conclusions are acquired.
The principal objective of the present work is to investigate the statistic properties of SAR return signals. The Synthetic Aperture Radar (SAR) return signals from 3-dimensional Gauss random rough conducting surfaces is efficiently and accurately computed using Multi-Level Fast Multipole Algorithm (MLFMA) combined with traditional Monte Carlo method. According to the real situation of SAR, incidents elevation angles are fixed at 45, and relevant azimuth angles vary from 88 to 92. Return signals are divided into several groups in terms of geometry coefficients. Then the statistic properties of these obtained signals are investigated by computing the means of return signals amplitudes and phases. Several useful conclusions are acquired.
2007, 29(2): 492-495.
doi: 10.3724/SP.J.1146.2006.00028
Abstract:
In this paper, the finite radial thickness tape helix is replaced by the same thickness vacuum layer laid in an infinite thin tape helix. The effect of position of the infinite thin helix on the dispersion and coupling impedance is studied by field analysis and computer simulation. The computed and simulated results are compared to those measured. As a result, if the infinite thin helix lay in the middle of the vacuum, the three results of dispersion and coupling impedance are good agreements when laying an infinite thin tape helix in the middle of the same thickness vacuum layer.
In this paper, the finite radial thickness tape helix is replaced by the same thickness vacuum layer laid in an infinite thin tape helix. The effect of position of the infinite thin helix on the dispersion and coupling impedance is studied by field analysis and computer simulation. The computed and simulated results are compared to those measured. As a result, if the infinite thin helix lay in the middle of the vacuum, the three results of dispersion and coupling impedance are good agreements when laying an infinite thin tape helix in the middle of the same thickness vacuum layer.
2007, 29(2): 496-498.
doi: 10.3724/SP.J.1146.2005.00755
Abstract:
The shapes of real defect outline usually play an extremely important role in the yield prediction and inductive fault analysis of Integrated Circuits (IC). In this paper, the multifractal characterizations of real defect outlines are discussed, and the multifractal spectrum of one typical real defect outline is estimated by the method of Wavelet Transform Modulus Maximum(WTMM), the results obtained in this paper will be useful for a fine characterization and computer simulation of the defects on wafer.
The shapes of real defect outline usually play an extremely important role in the yield prediction and inductive fault analysis of Integrated Circuits (IC). In this paper, the multifractal characterizations of real defect outlines are discussed, and the multifractal spectrum of one typical real defect outline is estimated by the method of Wavelet Transform Modulus Maximum(WTMM), the results obtained in this paper will be useful for a fine characterization and computer simulation of the defects on wafer.
2007, 29(2): 506-508.
doi: 10.3724/SP.J.1146.2005.00627
Abstract:
In this paper, the resonance of fractal Frequency Selective Surfaces(FSS) is studied in particular. Because fractals contain several scales of the starting geometry and each of them resonant at a scaled frequency of the original one, the dualband and dual-polarized FSS can be realized plusing the symmetry of the geometry. Several FSSs have been designed by the periodic Method Of Moments(MOM), and simulation and measurement are basically in agreement.
In this paper, the resonance of fractal Frequency Selective Surfaces(FSS) is studied in particular. Because fractals contain several scales of the starting geometry and each of them resonant at a scaled frequency of the original one, the dualband and dual-polarized FSS can be realized plusing the symmetry of the geometry. Several FSSs have been designed by the periodic Method Of Moments(MOM), and simulation and measurement are basically in agreement.
2007, 29(2): 499-505.
doi: 10.3724/SP.J.1146.2005.01199
Abstract:
Super-Resolution (SR) reconstruction is to estimate High-Resolution (HR) images from Low-Resolution (LR) image sequence, which has been a great focus for compressed video. This paper firstly presents the base of SR reconstruction for compressed video by building the relationships between the HR and LR images and surveying the models of quantization noise and motion vector. Then, the typical algorithms such as MAP, POCS and IBP are described in detail with experimental results. The computation complexity and real-time approaches are also investigated. Finally, aiming at the drawbacks, the research rang of prospects are demonstrated through pointing out the focus on degradation mode, motion estimation, reconstruction algorithm and real-time implementations.
Super-Resolution (SR) reconstruction is to estimate High-Resolution (HR) images from Low-Resolution (LR) image sequence, which has been a great focus for compressed video. This paper firstly presents the base of SR reconstruction for compressed video by building the relationships between the HR and LR images and surveying the models of quantization noise and motion vector. Then, the typical algorithms such as MAP, POCS and IBP are described in detail with experimental results. The computation complexity and real-time approaches are also investigated. Finally, aiming at the drawbacks, the research rang of prospects are demonstrated through pointing out the focus on degradation mode, motion estimation, reconstruction algorithm and real-time implementations.
