With the electromagnetic environment becoming more and more complex and the analysis demand of the radar emitter intropulse signal presenting more and more urgent, a modified method of the radar emitter intrapulse signal blind sorting under wavelet denoising is proposed. This study aims to improve the weak adaptability to the noise of the fast independent component analysis (FastICA) algorithm and its blind source separating performance. In this method, a pre-processing of noise based on the modified wavelet denoising is added. Then the FastICA algorithm is used to sort the unknown radar emitter intrapulse signal for the next intrapulse signal analysis. Simulations and analysis indicate that the modified method improves the signal to noise ratio of the received intermediate signals and the blind sorting performance.

Conventional radar emitter identification methods are faced with intensive noise and complex electromagnetic environment, which degrades the performance. Aiming at the feature extraction and identification at low signal-to-noise ratio (SNR), an improved algorithm is developed for the analysis and identification. First, the approach forms the time-frequency distribution and preprocesses the 2-dimensional image with whitening and normalization operations. Then, convolutional neural network (CNN) is introduced to generate high-level and abstract representations. Next, random vector functional link (RVFL) is trained to promote the fast feature learning. Finally, identification is implemented by picking out the maximum of RVFL output. 6 types of simulated emitters are used to validate the proposed algorithm. Experimental results show that the selected types of radar waveforms with additive Gaussian white noise can reach 85

针对存在脉冲无意调制特征时的特定辐射源识别问题,在时域对个体辐射源建模,并分析了无意特征在特征空间形成的分量对个体识别的作用。首先在时频域提取了奇异值作为特征描述,接着建立了联合协作表示模型并引入Tikhonov矩阵约束增加类间区分度,提取了特有分量表示系数,最后依据最小分类残差实现有监督的特定辨识。仿真中利用具有脉内无意调制的3种辐射源进行分析,实验结果表明提出的方法能有效识别无意幅度和相位调制特征,且在信噪比为10dB时能达到80%以上的正确识别率,具有泛化建模意义。

Aimed at the deficiency of conventional parameter-level methods in radar specific emitter identification (SEI), which heavily rely on empirical experience and cannot adapt to the waveform change, a novel algorithm is proposed to extract specific features and identify in Hilbert-Huang transform domain. Firstly, 2-dimensional physical representation of emitters is formed with Hilbert-Huang transform (HHT). Based on this, 4 types of multi-view features are constructed, and the feature space is spanned by elaborating the extraction. Principal components, between-class similarity, spectrum entropy, and deep architecture are used to describe the subtle features. Finally, support vector machine (SVM) is selected as the classifier to realize identification to alleviate the small sample problem. Experimental results show that the proposed algorithm realizes specific identification using 4 intentional modulations of simulated data. The selected 4 types of unintentional representations are feasible to discriminate identical emitters. Additionally, the proposed algorithm obtains higher accuracy than typical signal-level methods in the signal-to-noise ratio (SNR) range [0, 20] dB.

The convolution code with large constraint length plays an irreplaceable role in deep space communication and ultra-low frequency communication. Therefore, it is very important to find and test convolution code with large constraint length. Convolutional codes are widely used in deep space communication systems because of their high coding gain and simple and reliable encoders. The performance and implementation difficulty of convolutional codes mainly depend on the constraint length of the decimal codes and the coding efficiency. Enlarging and improving the coding gain of convolutional codes will greatly increase the complexity of decoders. The method of mathematical derivation and verification by mathematical tools is not suitable for convolutional codes with large large constraint length.. In order to improve the efficiency of inspection, the method of parallel multi-core computing needs to be introduced into the evil code test. Tests show that the FPGA-based parallel inspection method can improve the test efficiency by geometric multiples.

Presently, the extraction of hand-crafted features is still the dominant method in radar emitter recognition. To solve the complicated problems of selection and updation of empirical features, we present a novel automatic feature extraction structure based on deep learning. In particular, a convolutional neural network (CNN) is adopted to extract high-level abstract representations from the time-frequency images of emitter signals. Thus, the redundant process of designing discriminative features can be avoided. Furthermore, to address the performance degradation of a single platform, we propose the construction of an ensemble learning-based architecture for multi-platform fusion recognition. Experimental results indicate that the proposed algorithms are feasible and effective, and they outperform other typical feature extraction and fusion recognition methods in terms of accuracy. Moreover, the proposed structure could be extended to other prevalent ensemble learning alternatives.

For the separated MIMO radar system,a constrained weighted least squares algorithm is proposed by using time delay measurements.By introducing the variable which is the distance between the target and the reference station,the target location equation is pseudolinearized and the cost function is derived.And then,by further study,the relationship between the variable and the target position is explored and used as the constraint condition.Finally,the problem of locating the target is transformed from the nonlinear equation into the quadratic program problems with quadratic constraints,and by use of the Lagrange multiplier method,the target position is solved in a closed-form.Simulation results verify that the proposed algorithm can attain the Cramer-Rao Lower Bound in a relatively high level of noise and that it is robust.

Aimed at the deficiency of traditional feature extraction techniques in radar emitter recognition, a novel deep feature extraction and recognition architecture is proposed. To fit into the model, the time-domain emitters are transformed into unique time-frequency images correspondingly. Since auto-encoders restrict the input data to be vector-form and convolutional model is hard to optimize, denoising auto-coders are stacked in a convolutional manner in the pre-training stage and the proposed framework is trained by greedy layer-wise algorithm. The optimized network parameters are employed to initialize convolutional neural networks. By layers of mapping and pooling, deep time-frequency features are extracted, which are fed into the collaborative representation-based classifier to implement classification task. Experimental results on simulated data validate the feasibility of the proposed architecture. Furthermore, compared with conventional shallow algorithms, the proposed one can obtain higher recognition accuracy and more robust performance. Taking advantage of collaborative representation, the proposed algorithm is more applicable to the small-sample-size case.

To cope with the complex electromagnetic environment and varied signal styles, a novel method based on the energy cumulant of short time Fourier transform and reinforced deep belief network is proposed to gain a higher correct recognition rate for radar emitter intra-pulse signals at a low signal-to-noise ratio. The energy cumulant of short time Fourier transform is attained by calculating the accumulations of each frequency sample value with the different time samples. Before this procedure, the time frequency distribution via short time Fourier transform is processed by base noise reduction. The reinforced deep belief network is proposed to employ the input feature vectors for training to achieve the radar emitter recognition and classification. Simulation results manifest that the proposed method is feasible and robust in radar emitter recognition even at a low SNR.

Aimed at the impact of noise in receiving channels and signal distortion caused by sensors,simply improving the recognition performance of a single sensor no long meeting the demands,a collaborative represen-tation based radar emitter fusion recognition of multi-sensor method is proposed.Firstly,a completed dictionary is constructed with off-line sample signals in the training phase,on which collaborative coefficients of multiple receiving signals and classification residuals are obtained.Then,multi-sensor classification residuals and the D-S theory are combined by designing the basic probability assignment function reasonably,and consequently the fu-sion recognition result is acquired according to the maximum belief rule.Simulation experiments are performed by adopting 6 types of conventional radar emitters,the results validate the effectiveness of the proposed method and show that the method not only improves the performance in comparison with the single sensor,but is robust to noise and applicable to small-sample-size recognition.

According to the military demands and latest research progress in radar emitter recognition,three core aspects of emitter recognition are taken as the research object and survey is therefore developed on the status and development of relevant algorithms. For the low signal-to-noise ratio( SNR) ,deficiency of clas-sifier and limitations of single sensor,detailed analysis is presented on emitter feature extraction,classifica-tion and multi-source fusion recognition in terms of ideas and performance. In addition,current hot pattern recognition algorithms are introduced and references are analyzed. Finally,other remained problems in this field and the prospect of future research direction are demonstrated.

For the restrictions of emitter identification based on single sensor, simply improving the performance of single surveillance platform no longer meets the practical demands in low signal-to-noise ratio(SNR). Thus an algorithm of emitter identification of multisensory fusion based on collaborative representation and Boosting is proposed. By virtue of redundancy and complementarity of multisensory data information, feature extraction is implemented with time-frequency analysis, and multi-branch residuals are obtained through collaborative representation-based classifiers. The minimum classification residuals are acquired according to the weights in the Boosting training phase, in consequence the decision-level fusion identification is implemented. Simulation results show the effectiveness of the proposed algorithm, and show more robustness to noise when the signal-to-noise ratio is relatively low. Meanwhile, the proposed framework is easy to be conducted.

To improve the recognition rate of radar emitters with complex signal system in an awful electromagnetic environment, a new recognition method based on short time Fourier transform (STFT) and convolutional neural networks (CNN) was proposed. In this method, STFT obtains the time-frequency distribution of radar emitter in-pulse modulated signals and CNN extracted the features of different radar signals with the processed data. Before the time-frequency distribution (TFD) arrays were input to the CNN, a base noise reduction was conducted after six-time zero-means scaling. In the end of the classification, an additional operation was made to distinguish NS and BPSK. Simulations were implemented to present the high recognition performance of the method.

Aimed at the deficiency of traditional techniques of radar emitter feature extraction which rely heavily on artificial experience,a novel emitter identification algorithm based on joint deep time-frequency features is proposed.Time-domain signals are transformed into the 2-D time-frequency domain,and dimensionality reduction is implemented with random projection and principal component analysis with respect to sustaining subspace and energy.In the phase of pre-training,the deep model is layer-wise trained with unlabelled samples and network parameters are fine-tuned with label information.Finally the identification task is achieved with a logistic regression classifier.6 types of emitter signals are adopted in simulation experiments to validate the effectiveness of the proposed algorithm,the experimental results indicating that the joint deep features help to obtain higher identification accuracy and that the algorithm is more efficient.

针对低信噪比(SNR)条件下传统辐射源识别算法性能下降的问题,提出了基于压缩协作表示的识别算法,分别从特征提取和分类器设计两方面进行描述.首先将时域辐射源信号变换到二维时频域,通过图像处理方法提取高维特征列向量.经随机矩阵压缩到一定维度后,输入到提出的压缩协作表示分类器中得到识别结果.进而,对协作表示系数进行非负约束,提出了更符合实际应用场景的算法.仿真结果验证了所提算法的可行性与有效性,且在低信噪比条件下稳健性强、抗噪声干扰性能好、计算量较小、易于工程实现.

Aimed at the problem of underdetermined blind identification , an algorithm based on generalized generating function decomposition is proposed , which no longer imposes sparsity restrictions on source signals . First , the second derivative matrices of the generalized generating function are stacked to the third‐order tensor form , from which the number of source signals can be blindly estimated . Then the tensor is decomposed with singular value decomposition , and the mixture matrix is estimated by the joint diagonalization method . Simulation results validate the effectiveness of the proposed algorithm , and show that the proposed algorithm can acquire a better estimation precision than other classical algorithms with the same SNRs in the conditions of well‐posed and underdetermined mixtures , meanwhile it extends the field of blind source separation application via the generalized generating function restricted only to the well‐posed case .

When signal samples are severely contaminated by interference noise, good emitter recognition can't be achieved in most cases simply by extracting distinctive features and improving the performance of a single classifier. Firstly, vectorized time-frequency features are extracted, and then representation coefficients are obtained in the frame of collaborative representation. Then, a decision-level fusion of multiple sensors is implemented under the maximum activity rule and recognition results are acquired by selecting the minimum residual. The simulation experiments validate the feasibility of the proposed algorithm and show that the recognition rate of fusion is higher than a single classifier, which indicates the good recognition performance.

Aimed at the performance decline of conventional radar emitter signals recognition techniques with the presence of noise and impulsive interference, a robust recognition algorithm based on Cauchy random projection is proposed. By virtue of sparse characteristics of impulse system radar and impulsive interference, the more robust ℓ1 norm random projection is combined with the state-of-art sparse representation recognition. As a consequence, the recognition problem is converted into solving ℓ1 norm minimization. The SaS interference model is built, and the recognition performance between ℓ2 and ℓ1 random mapping. Simulation results show the recognition validity of the proposed method in the range of impulse radar signals, and that it's more robust than sparse recognition method based on ℓ2 norm random dimensionality-reduction, which is more applicable for the practical applications.

在短波数字化接收机中,通过控制射频前端可变的衰减器或放大器,可以实现模拟增益控制.但无论是压控还是数控衰减器和放大器,器件的非线性都会引起模拟增益分配的非线性,从而影响解调的质量.本文针对这种非线性设计了一种反馈型的自动增益校正方法,并在DSP和FPGA的硬件平台上进行效果测试,结果表明校正后增益误差有了3.2dB的改善.

在短波数字化接收机中,频谱扫描是对天线端的射频信号进行频谱分析.由于扫频速度难以满足工程需要,本文运用有限状态机的思想,在基于DSP和FPGA的硬件平台上,设计实现频谱扫描功能,并通过合理调度DSP的线程压缩扫频时间,满足实际应用需求.