Spectrum prediction is considered as a key technology to assist spectrum decision.Despite the great efforts that have been put on the construction of spectrum prediction,achieving accurate spectrum prediction emphasiz...Spectrum prediction is considered as a key technology to assist spectrum decision.Despite the great efforts that have been put on the construction of spectrum prediction,achieving accurate spectrum prediction emphasizes the need for more advanced solutions.In this paper,we propose a new multichannel multi-step spectrum prediction method using Transformer and stacked bidirectional LSTM(Bi-LSTM),named TSB.Specifically,we use multi-head attention and stacked Bi-LSTM to build a new Transformer based on encoder-decoder architecture.The self-attention mechanism composed of multiple layers of multi-head attention can continuously attend to all positions of the multichannel spectrum sequences.The stacked Bi-LSTM can learn these focused coding features by multi-head attention layer by layer.The advantage of this fusion mode is that it can deeply capture the long-term dependence of multichannel spectrum data.We have conducted extensive experiments on a dataset generated by a real simulation platform.The results show that the proposed algorithm performs better than the baselines.展开更多
The seismic records of target response spectrum used in the time-history analysis should be allowed to meet the norms. However, the previous fitting methods of target spectrum are mostly for the situations that the ta...The seismic records of target response spectrum used in the time-history analysis should be allowed to meet the norms. However, the previous fitting methods of target spectrum are mostly for the situations that the target spectrum is a smooth curve. In many cases, it needs to match unsmooth target spectrum for single determined response spectrum. An adjustment of time history via wavelet packet transform was presented, which is able to fit unsmooth target spectrum. It was found that there is a certain bias between the band center frequency of the component of seismic record after wavelet packet decomposition and the peak frequency of response spectra of wavelet packet components. For this reason, five strategies were presented to select iteration points, and the effects of the five strategies were compared with two calculation examples. It was turned out that the peak frequency of the response spectrum of wavelet packet component can lead to good fitting effect when it is selected as the iteration point. In the iteration process, it shows great promise in fitting non-smooth target spectrum and has a trend of converge.展开更多
A high performance fast-Fourier-transform (FFT) spectrum analyzer, which is developed for measure spin noise spectrums, is presented in this paper. The analyzer is implemented with a field-programmable-gate-arrays (FP...A high performance fast-Fourier-transform (FFT) spectrum analyzer, which is developed for measure spin noise spectrums, is presented in this paper. The analyzer is implemented with a field-programmable-gate-arrays (FPGA) chip for data and command management. An analog-to-digital-convertor chip is integrated for analog signal acquisition. In order to meet the various requirements of measuring different types of spin noise spectrums, multiple operating modes are designed and realized using the reprogrammable FPGA logic resources. The FFT function is fully managed by the programmable resource inside the FPGA chip. A 1 GSa/s sampling rate and a 100 percent data coverage ratio with non-dead-time are obtained. 30534 FFT spectrums can be acquired per second, and the spectrums can be on-board accumulated and averaged. Digital filters, multi-stage reconfigurable data reconstruction modules, and frequency down conversion modules are also implemented in the FPGA to provide flexible real-time data processing capacity, thus the noise floor and signals aliasing can be suppressed effectively. An efficiency comparison between the FPGA-based FFT spectrum analyzer and the software-based FFT is demonstrated, and the high performance FFT spectrum analyzer has a significant advantage in obtaining high resolution spin noise spectrums with enhanced efficiency.展开更多
Electromagnetic Spectrum(EMS)recognition is vital in spectrum control,interference location,electronic countermeasures,etc.However,samples of high-value targets are incredibly scarce,even single,and are easily overwhe...Electromagnetic Spectrum(EMS)recognition is vital in spectrum control,interference location,electronic countermeasures,etc.However,samples of high-value targets are incredibly scarce,even single,and are easily overwhelmed by noise and numerous low-value targets,resulting in poor recognition accuracy using traditional methods.Furthermore,the great similarity between samples from the same manufacturer,model,and batch,makes Specific Emitter Identification(SEI)with the EMS especially challenging.Based on the powerful extension and extraction ability of the Fractional Fourier Transform(FrFT)for detailed features,this paper proposes a novel algorithm for the EMS recognition under a single-sample condition.The proposed method constructs a feature matrix FrFT-M from the results of the FrFT under specific orders for each sample.Then,the most relevant item,obtained by analyzing the correlations among FrFT-Ms between the unidentified sample and known samples,determines the optimal recognition.Three simple tests are conducted,including two simulations considering fifteen basic waveforms and six typical radar signals,and one experiment using STM32 microcontroller boards.The detection results of simulated and experimental data show that the accuracies of all three cases are higher than 86%,even for samples of the same model.Our method is promising and may have significant value in other fields.展开更多
A method allowing a desirable matrix spectrum to be constructed as an alternative to the method using matrix transformation to the Frobenius form is stated. It can be applied to implement control algorithms for techni...A method allowing a desirable matrix spectrum to be constructed as an alternative to the method using matrix transformation to the Frobenius form is stated. It can be applied to implement control algorithms for technical systems without executing the variables transformation procedures that are needed for deriving a Frobenius matrix. The method can be used for simulation of systems with different spectrums for choosing an alternative that satisfies to the distinct demands.展开更多
Traditional lapped transform domain excision techniques obtain good performance at the expense of increased processing delay. Extension of transform domain filtering techniques to the lapped biorthogonal transform dom...Traditional lapped transform domain excision techniques obtain good performance at the expense of increased processing delay. Extension of transform domain filtering techniques to the lapped biorthogonal transform domain can help solve the problem. By incorporating biorthogonality into the lapped transforms, more flexibility is obtained in the design of windows. Thus transform bases with better stopband attenuation can be generated by designing windows, but not by increasing the overlapping factor. In this paper, a new modulated lapped biorthogonal transform (MLBT) with optimized windows is introduced for efficient compression of multi-tone interfering signal energy. The bit error rate (BER) performance of the receiver employing the proposed MLBT excision technique is analyzed and compared with that of the lapped transform domain excision-based receivers. Simulation results demonstrate the improved performance and increased robustness of the proposed technique.展开更多
A novel time-frequency domain interference excision technique is proposed. The technique is based on adaptive biorthogonal local discrete cosine trans form (BLDCT). It uses a redundant library of biorthogonal local d...A novel time-frequency domain interference excision technique is proposed. The technique is based on adaptive biorthogonal local discrete cosine trans form (BLDCT). It uses a redundant library of biorthogonal local discrete cosine bases and an efficient concave cost function to match the transform basis to the interfering signal. The main advantage of the algorithm over conventional trans form domain excision algorithms is that the basis functions are not fixed but ca n be adapted to the time-frequency structure of the interfering signal. It is w e ll suited to transform domain compression and suppression of various types of in terference. Compared to the discrete wavelet transform (DWT) that provides logar ithmic division of the frequency bands, the adaptive BLDCT can provide more flex ible frequency resolution. Thus it is more insensitive to variations of jamming frequency. Simulation results demonstrate the improved bit error rate (BER) perf ormance and the increased robustness of the receiver.展开更多
The quaternion Fourier transform plays a vital role in the representation of two-dimensional signals. This paper characterizes spectrum of quaternion-valued signals on the quaternion Fourier transform domain by the pa...The quaternion Fourier transform plays a vital role in the representation of two-dimensional signals. This paper characterizes spectrum of quaternion-valued signals on the quaternion Fourier transform domain by the partial derivative.展开更多
The proposed method inserts a watermark into the spectral components of the data using techniques analogous to spread spectrum communications,hiding a narrow band signal in a wideband channel.The watermark is difficul...The proposed method inserts a watermark into the spectral components of the data using techniques analogous to spread spectrum communications,hiding a narrow band signal in a wideband channel.The watermark is difficult for an attacker to remove.It is also robust to common signal and geometric distortion such as JPEG compression,cropping and scaling.In addition,the watermark can be extracted accurately without source host image at the receiver.展开更多
Field spectrum pretreatment experiments were carried out, and denoising numerical experiment via lifting wavelet transform （LWT） was designed, and several famous test signals including blocks, bumps, heavy sine and ...Field spectrum pretreatment experiments were carried out, and denoising numerical experiment via lifting wavelet transform （LWT） was designed, and several famous test signals including blocks, bumps, heavy sine and doppler were processed via Lw＇r in these experiment. And the field spectrum was processed via Lw＇r. Experiments proved that SNRG-tO-SNRN curves have similar feature and they all have a peak. And SNRG of almost all employed wavelets have higher value with SNRN between 0 and 20 dB. When signal is at high SNR, the SNRG is very little, and the MSED of denoised signal became little by little. LWT is more suite to denoise the low SNR or heavy noise contaminated signals. Bior4.4 have wider SNRN interval for denoising comparing with other five wavelets, includ- ing haar, db6, sym6, bior2.2 and bior3.3. Original field spectrum is processed by 3 stage liftings based on bior4.4 to denoise the trivial noise-contaminated regions. On processing the water band signal, logarithm transform is firstly taken. And then the spectrum is denoised via LWT based on bior4.4. The results show that an excellent denoised spectrum can be get, especially between 350 nm and 1 800 nm, and between 1 960 nm to 2 500 nm. While there is still a bump around 1 900 nm, this maybe due to the spectrum machine＇s limited precision.展开更多
A method is presented for calculating a matrix spectrum with a given set of eigenvalues. It can be used to build systems with different spectrums with the aim of choosing desired alternative. It enables a practical im...A method is presented for calculating a matrix spectrum with a given set of eigenvalues. It can be used to build systems with different spectrums with the aim of choosing desired alternative. It enables a practical implementation of control algorithms without resorting to transformation of variables.展开更多
基金supported in part by the National Natural Science Foundation of China under Grants 62231015,62427801in part by Jiangsu Province Frontier Leading Technology Basic Research Project BK20232030.
文摘Spectrum prediction is considered as a key technology to assist spectrum decision.Despite the great efforts that have been put on the construction of spectrum prediction,achieving accurate spectrum prediction emphasizes the need for more advanced solutions.In this paper,we propose a new multichannel multi-step spectrum prediction method using Transformer and stacked bidirectional LSTM(Bi-LSTM),named TSB.Specifically,we use multi-head attention and stacked Bi-LSTM to build a new Transformer based on encoder-decoder architecture.The self-attention mechanism composed of multiple layers of multi-head attention can continuously attend to all positions of the multichannel spectrum sequences.The stacked Bi-LSTM can learn these focused coding features by multi-head attention layer by layer.The advantage of this fusion mode is that it can deeply capture the long-term dependence of multichannel spectrum data.We have conducted extensive experiments on a dataset generated by a real simulation platform.The results show that the proposed algorithm performs better than the baselines.
基金Projects(41272304,51374244,41372278,51304241)supported by the National Natural Science Foundation of ChinaProject(2010CB732004)supported by the National Basic Research Program of China
文摘The seismic records of target response spectrum used in the time-history analysis should be allowed to meet the norms. However, the previous fitting methods of target spectrum are mostly for the situations that the target spectrum is a smooth curve. In many cases, it needs to match unsmooth target spectrum for single determined response spectrum. An adjustment of time history via wavelet packet transform was presented, which is able to fit unsmooth target spectrum. It was found that there is a certain bias between the band center frequency of the component of seismic record after wavelet packet decomposition and the peak frequency of response spectra of wavelet packet components. For this reason, five strategies were presented to select iteration points, and the effects of the five strategies were compared with two calculation examples. It was turned out that the peak frequency of the response spectrum of wavelet packet component can lead to good fitting effect when it is selected as the iteration point. In the iteration process, it shows great promise in fitting non-smooth target spectrum and has a trend of converge.
基金Project supported by the Strategic Priority Research Program of the Chinese Academy of Sciences(Grant No.XDC07020200)the National Key R&D Program of China(Grant Nos.2018YFA0306600 and 2016YFB0501603)+3 种基金the National Natural Science Foundation of China(Grant No.11927811)the Chinese Academy of Sciences(Grants Nos.GJJSTD20170001 and QYZDY-SSW-SLH004)Anhui Initiative in Quantum Information Technologies,China(Grant No.AHY050000)the Fundamental Research Funds for the Central Universities,China.
文摘A high performance fast-Fourier-transform (FFT) spectrum analyzer, which is developed for measure spin noise spectrums, is presented in this paper. The analyzer is implemented with a field-programmable-gate-arrays (FPGA) chip for data and command management. An analog-to-digital-convertor chip is integrated for analog signal acquisition. In order to meet the various requirements of measuring different types of spin noise spectrums, multiple operating modes are designed and realized using the reprogrammable FPGA logic resources. The FFT function is fully managed by the programmable resource inside the FPGA chip. A 1 GSa/s sampling rate and a 100 percent data coverage ratio with non-dead-time are obtained. 30534 FFT spectrums can be acquired per second, and the spectrums can be on-board accumulated and averaged. Digital filters, multi-stage reconfigurable data reconstruction modules, and frequency down conversion modules are also implemented in the FPGA to provide flexible real-time data processing capacity, thus the noise floor and signals aliasing can be suppressed effectively. An efficiency comparison between the FPGA-based FFT spectrum analyzer and the software-based FFT is demonstrated, and the high performance FFT spectrum analyzer has a significant advantage in obtaining high resolution spin noise spectrums with enhanced efficiency.
基金supported in part by the National Natural Science Foundation of China(No.62293495)。
文摘Electromagnetic Spectrum(EMS)recognition is vital in spectrum control,interference location,electronic countermeasures,etc.However,samples of high-value targets are incredibly scarce,even single,and are easily overwhelmed by noise and numerous low-value targets,resulting in poor recognition accuracy using traditional methods.Furthermore,the great similarity between samples from the same manufacturer,model,and batch,makes Specific Emitter Identification(SEI)with the EMS especially challenging.Based on the powerful extension and extraction ability of the Fractional Fourier Transform(FrFT)for detailed features,this paper proposes a novel algorithm for the EMS recognition under a single-sample condition.The proposed method constructs a feature matrix FrFT-M from the results of the FrFT under specific orders for each sample.Then,the most relevant item,obtained by analyzing the correlations among FrFT-Ms between the unidentified sample and known samples,determines the optimal recognition.Three simple tests are conducted,including two simulations considering fifteen basic waveforms and six typical radar signals,and one experiment using STM32 microcontroller boards.The detection results of simulated and experimental data show that the accuracies of all three cases are higher than 86%,even for samples of the same model.Our method is promising and may have significant value in other fields.
文摘A method allowing a desirable matrix spectrum to be constructed as an alternative to the method using matrix transformation to the Frobenius form is stated. It can be applied to implement control algorithms for technical systems without executing the variables transformation procedures that are needed for deriving a Frobenius matrix. The method can be used for simulation of systems with different spectrums for choosing an alternative that satisfies to the distinct demands.
文摘Traditional lapped transform domain excision techniques obtain good performance at the expense of increased processing delay. Extension of transform domain filtering techniques to the lapped biorthogonal transform domain can help solve the problem. By incorporating biorthogonality into the lapped transforms, more flexibility is obtained in the design of windows. Thus transform bases with better stopband attenuation can be generated by designing windows, but not by increasing the overlapping factor. In this paper, a new modulated lapped biorthogonal transform (MLBT) with optimized windows is introduced for efficient compression of multi-tone interfering signal energy. The bit error rate (BER) performance of the receiver employing the proposed MLBT excision technique is analyzed and compared with that of the lapped transform domain excision-based receivers. Simulation results demonstrate the improved performance and increased robustness of the proposed technique.
基金Project supported by the National Natural Science Foundation of China(Grant No.6017201860372007)
文摘A novel time-frequency domain interference excision technique is proposed. The technique is based on adaptive biorthogonal local discrete cosine trans form (BLDCT). It uses a redundant library of biorthogonal local discrete cosine bases and an efficient concave cost function to match the transform basis to the interfering signal. The main advantage of the algorithm over conventional trans form domain excision algorithms is that the basis functions are not fixed but ca n be adapted to the time-frequency structure of the interfering signal. It is w e ll suited to transform domain compression and suppression of various types of in terference. Compared to the discrete wavelet transform (DWT) that provides logar ithmic division of the frequency bands, the adaptive BLDCT can provide more flex ible frequency resolution. Thus it is more insensitive to variations of jamming frequency. Simulation results demonstrate the improved bit error rate (BER) perf ormance and the increased robustness of the receiver.
文摘The quaternion Fourier transform plays a vital role in the representation of two-dimensional signals. This paper characterizes spectrum of quaternion-valued signals on the quaternion Fourier transform domain by the partial derivative.
基金Foundation item:Supported by Wuhan University Innovation Founda-tion(904270072)
文摘The proposed method inserts a watermark into the spectral components of the data using techniques analogous to spread spectrum communications,hiding a narrow band signal in a wideband channel.The watermark is difficult for an attacker to remove.It is also robust to common signal and geometric distortion such as JPEG compression,cropping and scaling.In addition,the watermark can be extracted accurately without source host image at the receiver.
文摘Field spectrum pretreatment experiments were carried out, and denoising numerical experiment via lifting wavelet transform （LWT） was designed, and several famous test signals including blocks, bumps, heavy sine and doppler were processed via Lw＇r in these experiment. And the field spectrum was processed via Lw＇r. Experiments proved that SNRG-tO-SNRN curves have similar feature and they all have a peak. And SNRG of almost all employed wavelets have higher value with SNRN between 0 and 20 dB. When signal is at high SNR, the SNRG is very little, and the MSED of denoised signal became little by little. LWT is more suite to denoise the low SNR or heavy noise contaminated signals. Bior4.4 have wider SNRN interval for denoising comparing with other five wavelets, includ- ing haar, db6, sym6, bior2.2 and bior3.3. Original field spectrum is processed by 3 stage liftings based on bior4.4 to denoise the trivial noise-contaminated regions. On processing the water band signal, logarithm transform is firstly taken. And then the spectrum is denoised via LWT based on bior4.4. The results show that an excellent denoised spectrum can be get, especially between 350 nm and 1 800 nm, and between 1 960 nm to 2 500 nm. While there is still a bump around 1 900 nm, this maybe due to the spectrum machine＇s limited precision.
文摘A method is presented for calculating a matrix spectrum with a given set of eigenvalues. It can be used to build systems with different spectrums with the aim of choosing desired alternative. It enables a practical implementation of control algorithms without resorting to transformation of variables.
