Considering the problem of synchronous Frequency Hopping(FH) network station sorting, an Underdetermined Blind Source Separation(UBSS) algorithm based on time-frequency domain single source point detection is proposed. Firstly, the algorithm performs time-frequency transform on the observed signal, and uses adaptive threshold denoising algorithm to eliminate the background noise of the time-frequency matrix. It can increase the algorithm anti-noise performance. Then, single source point detection is performed according to the absolute azimuth difference of the signal. It can effectively ensure the sufficient sparsity of a single source point. The hybrid matrix estimation is completed by the improved fuzzy C value clustering algorithm. It can reduce the influence of noise and sample set distribution differences and improve the estimation accuracy. Finally, the source signal is reconstructed and restored by a variable step size Sparsity Adaptive Subspace Pursuit(SASP) algorithm. The simulation experiments show that the proposed algorithm has higher recovery accuracy of the frequency hopping signal under the condition of low Signal to Noise Ratio (SNR), and can effectively complete the blind separation of the synchronous frequency hopping signal.

Automatic Dependent Surveillance-Broadcast (ADS-B) technology is faces serious security risk of spoofing due to the characteristics of broadcasting the clear text. In view of the detection of delay-forwarding messages and the positioning of corresponding jammers, according to the distribution characteristics of the interval between radiation positions of messages received by ADS-B ground station, a method is proposed for detecting delay-forwarding messages by using the difference between the instantaneous velocity of sequential messages and the average velocity of interest of period. According to the spatial relationship between the positions of the original messages of those delay-forwarding messages and the jammer, the positions of the delay-forwarding messages received by ADS-B ground station and their received time stamps are used to locate the position of jammer with the multilateration technique. The simulation results show that the difference between the instantaneous velocity and the average velocity can effectively detect the presence or absence of delay-forwarding messages, the positioning accuracy of jammer is related to the timing accuracy of ADS-B ground station, the number of messages used for positioning, the distance between jammer and track and the azimuth of jammer. The better positioning performance of jammer can be obtained by using the message positions in side-looking area. The proposed method can help the ADS-B ground station independently to detect the delay forwarding interference and to locate the corresponding jammer.

On the basis of the comprehensive consideration in the design of the indexs of the adaptive filter algorithm convergence speed, steady-state error, computational complexity and tracking performance, a kind of Versoria function normalized adaptive filtering algorithm is proposed in this paper. The class Versoria function is used instead of Sigmoid function as step iterative formula, introducing variable step size based on the relevant error adjustment principle, the stability of the algorithm is enhanced greatly. At the same time, the convergence speed and tracking performance of the algorithm is promoted and the computational complexity of the algorithm is reduced. The influence of the parameter

,

and

different value of the step function of algorithm is analyzed on Matlab platform. Compared with the Sigmoid function variable step size LMS algorithm and variable step size LMS algorithm based on Versoria function, and the simulation results show that this algorithm has faster convergence speed, better tracking ability, smaller steady-state error and strong robustness.

To overcome the flaw that the sensor selection methods based on either of Bayesian Fisher information matrix or mutual information could not provide coincident results, the multiple objective optimal technology is developed for sensor selection by minimizing the number of sensors, maximizing corresponding Bayesian Fisher information matrix and mutual information of the selected sensors. Then, the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) approach is proposed to find the candidate that can better trade off the cost and two performance metrics. Comparison results demonstrate that the proposed method can find a better sensor group, and ultimately, its overall localization performance is more stable and accurate.

Impulsive noise causes nonnegative algorithms to yield excessive error during iterations, which will damage the stability of the algorithm and causes performance degradation. In the paper, a NonNegative Least Mean Square algorithm based on the Sigmoid framework (SNNLMS) is proposed. The algorithm embeds the conventional nonnegative cost function into the Sigmoid framework to receive a new cost function. The new cost function has the characteristics of suppressing the impact of impulse noise. In addition, in order to enhance the robustness of the SNNLMS algorithm under sparse system identification, the Inversely-Proportional Sigmoid NonNegative Least Mean Square (IP-SNNLMS) is proposed based on the inversely-proportional function. Simulation results demonstrate that the SNNLMS algorithm effectively solves the problem of misadjustment caused by impulsive noise. IP-SNNLMS enhances the robustness of the algorithm and improves the defect of the convergence rate of the SNNLMS algorithm under the sparse system identification.

Airborne Distributed Coherent Aperture Radar (DCAR) has the advantages of wide observation range, high maneuverability and flexible deployment. However, airborne DCAR is confronted with more stringent time, space and phase synchronization requirements. Therefore, an airborne DCAR signal model and its matrix representation based on Slow-Time Code Division Multiple Access (ST-CDMA) waveform are established successively. Moreover, the influence on target coherence synthesis resulted from time, space and phase synchronization errors is analyzed in detail, and a novel airborne DCAR synchronization error calibration method based on prominent points is proposed. This method utilizes the target parameter search strategy to eliminate the grid mismatch filtering firstly. Then, with the utilization of estimation approaches based on target model or repeater station model, the unit platform position error is calibrated. Finally, equivalent amplitude and phase errors are calibrated by Eigen structure methods. The validity of the proposed method to calibrate the airborne DCAR synchronization error is demonstrated by simulation experiments.

A maneuvering SAR imaging algorithm is proposed based on the separation of azimuthal motion information. After range compression, the influence of non-azimuthal motion on range walk and spatiality is removed by separating azimuthal motion information. Then, the azimuth velocity equivalent transformation is used to realize range bending correction. Finally, the azimuth compression is realized by non-uniform Fourier transform. Using this principle, the parameters of three directions in the oblique distance equation have the same properties in the three-dimensional coordinate system. The algorithm transforms the three-dimensional spatial motion into the azimuthal non-uniform motion model. Through simulation, the applicability of the algorithm to different models and other motion states is verified. The algorithm is simple, stable and applicable.

Minimum Mean Square Error (MMSE) algorithm can achieve near-optimal detection performance for multi-user massive Multiple Input Multiple Output (MIMO) systems. However, the calculation of the high-dimensional matrix inversion required in MMSE algorithm causes excessively high computational complexity, which makes it difficult to implement quickly and effectively in practical applications. At the same time, for the Higher Quadrature Amplitude Modulation (HQAM), the direct use of hard decision to realize the symbol-to-bit demapper will result in an obvious performance loss. Therefore, a low-complexity soft output signal detection algorithm based on Chebyshev-Trace Iteration (CTI) is proposed for Gray-coded HQAM in this paper. The algorithm not only effectively avoids the calculation of the high dimensional matrix inversion, but also gives a simplified calculation method with the trident list searching to compute the bit Log-Likelihood Ratio (LLR) by using the bit flip property of Gray-coded modulation and binary tree architecture. The simulation results show that the proposed soft output detection algorithm needs at most 3 iterations to converge and achieve detection performance close to MMSE, which achieves a good tradeoff between the complexity and the detection performance.

Ad hoc network is a kind of self-organized network without center. The reliability of virtual user identification and channel security are reduced when SNR is low due to user energy limitation. In order to solve this problem, a joint virtual user identification and channel security en/decoding method is proposed in this paper. Transmitter-receiver-based virtual user identification code is generated by xoring the orthogonal address code of transmitter with the pseudo random address code of receiver and encrypted with channel security code as key to acquire orthogonal random secure sequence so as to improve channel security. In order to realize spread spectrum as well as improve transmission efficiency, transmitted data is divided into 6-bit symbols, each symbol is mapped with an orthogonal random secure sequence. Receivers adopt subspace-based method to process received data and establish a judgment model to identify virtual users. Simulation results indicate that the proposed method obtains 1.6 dB

gains compared with existing methods when miss alarm rate of virtual user identification is 10^–3.

Transmission delay and packet loss rate are critical issues in reliable transmission of power communication services. A minimum path selection routing control strategy for software-defined power communication networks is proposed. Combining the characteristics of the centralized control structure of the software-defined power communication network, a Link Bandwidth Occupancy Predictive model based on Graph Convolutional Network (LBOP-GCN) is built to analyze the route paths bandwidth occupancy in the next period. The selectivity (Q) of different transmission paths from the source node is calculated to the destination node is calculated by using Triangle Modular Operator (TMO) to fuse the transmission delay of the path, the path bandwidth occupancy at the current moment and the path bandwidth occupancy at the next moment. Then the path with the lowest Q value is used as the flow table of the OpenFlow switch delivered by the Software Defined Network (SDN) controller. Experiments show that the proposed routing control strategy can effectively reduce service transmission delay and packet loss rate.

Wireless Sensor Network (WSN) has the characteristics of scale-free network, usually works in an unattended open environment, and is vulnerable to a variety of deliberate attacks. The attack causes the network to break down, and even causes the whole network to be paralyzed. In this paper, the scale-free network in complex network is taken as the research object, and a scale-free wireless sensor network model is constructed. Using the advantages of Fireworks algorithm and Particle Swarm Optimization (PSO) algorithm, such as search ability and population diversity, the FW-PSO (FireWorks and Particle Swarm Optimization) algorithm is proposed, which has good performance in global search ability and convergence speed. For the scale-free network model, FW-PSO algorithm is used to optimize the network topology. Under different attack strategies, the performance of the network before and after the optimization is analyzed from dynamic and static invulnerability respectively. Simulation results show that, compared with other similar algorithms, the dynamic and static invulnerability of wireless sensor network optimized by the proposed algorithm has obvious advantages.

To solve the problem of Virtual Network Function (VNF) migration optimization, which is caused by the dynamic change of resource requirements of Service Function Chain (SFC) under Network Function Virtualization/ Software Defined Network (NFV/SDN) architecture, a VNF migration optimization algorithm is proposed based on deep reinforcement learning. Firstly, based on the underlying CPU, bandwidth resources and SFC end-to-end delay constraints, a Markov Decision Process (MDP) based stochastic optimization model is established. This model is used to optimize jointly network energy consumption and SFC end-to-end delay by migrating VNF. Secondly, since the state space and action space of this paper are continuous value sets, a VNF intelligent migration algorithm based on Deep Deterministic Policy Gradient (DDPG) is proposed to obtain an approximate optimal VNF migration strategy. The simulation results show that the algorithm can achieve the compromise between network energy consumption and SFC end-to-end delay, and improve the resource utilization of the physical network.

For the low recognition rate of motor imagery ElectroEncephaloGram (EEG) signals using single feature in Brain-Computer Interface (BCI) research, a feature extraction method combining brain function network and sample entropy is proposed. According to the neural mechanism appearing in Event Related Synchronization/Event Related Desynchronization (ERS/ERD) phenomenon and the contralateral mapping mechanism between cortex and limb motor imagery, the μ rhythm is denoised by wavelet packet transform. The brain function network is constructed for left hemispherical brain region and right hemispherical brain region by μ rhythm of 27 left channels and 27 right channels respectively. The mean node degree and the mean clustering coefficient are calculated as the brain function network characteristics, and the feature vectors combining the distribution and directivity are constructed by the sample entropy of C3 and C4 channels with the μ rhythm. The Support Vector Machine (SVM) is used to classify the left hand and right hand motor imagery EEG signals. The results show that the feature extraction method based on brain function network and sample entropy achieves better classification result, and the highest classification rate reached 90.27%.

Convolutional compressed sensing emerging in recent years is a new type of compressed sensing technology. By using cyclic matrix as measurement matrices, the sampling in convolutional compressed sensing can be simplified into convolution process, thus the complexity of the algorithm is greatly reduced. In this paper, a construction of measurement matrices for convolutional compressed sensing based on cyclotomic classes is proposed. The measurements are obtained by using the circulate convolution signal of the deterministic sequence and then by random subsampling. The correlation of the measurement matrix constructed in this paper is smaller than that of the existing constructions in the literature. The simulation results show that the measurement matrix constructed in this paper can recover the sparse signal better than the random Gaussian matrix under the same conditions. The proposed matrix can also be applied to channel estimation and reconstruction of two-dimensional images.

Underwater optical image processing is an important basis for underwater equipment to complete deep-sea exploration and operation tasks. Based on a brief description of the research background, significance and hotspots of underwater optical image processing, this paper gives a detailed overview of underwater imaging technology and clearness of underwater images from the aspects of improving the lighting factors and color correction of underwater images. The research progress focuses on the research status of the two most active research directions of image restoration methods and image enhancement methods based on imaging models. According to the research hotspots of underwater optical image processing, the research of underwater optical image processing is prospected from the perspectives of considering the forward refraction of light, combining underwater imaging models and image enhancement algorithms, introducing new algorithms in related fields, and improving the real-time performance of image processing.

A chaotic system with three scrolls is proposed. The dynamics characteristics of the system, such as Lyapunov index, bifurcation diagram, Poincare cross section diagram, power spectrum and equilibrium stability, are studied by numerical simulation. The analysis results show that the system has good dynamics characteristics and rich topological attractors. In addition, an experimental simulation circuit is developed based on the circuit simulation software Multisim, which has a simple structure and is easy to realize in practice. Moreover, the simulation experiment is very consistent with the theoretical analysis conclusion, confirming that the proposed chaotic system circuit can be realized physically, thus verifying the chaos generation capacity of the chaotic system. Finally, a color image encryption algorithm is designed based on DNA algorithm, and the results show that the system has high security performance.Using the new system chaotic sequence to encrypt image, the encryption histogram and correlation between adjacent pixels are analyzed, results show that the new systems are very sensitive to image key and plaintext, key space is large, chaotic system applied to image encryption has higher safety performance.

To overcome the shortcomings of low transmission rate of Noise Reduction Differential Chaos Shift Keying (NR-DCSK), a novel Noise Reduction Differential Chaos Shift Keying system based on Quadrature Modulation (QM-NRDCSK) is proposed. The generator generates two chaotic sequences, the reference signal of each channel is P-time repetition of the information-bearing signal. The information of different users is distinguished by different time slots, and the two signals are transmitted on the same frequency band by using quadrature modulation. The reference signal of each channel is averaged P times by the moving average filter at the receiving end, and then non-coherently demodulated with the information signal. The correctness of the theoretical derivation is verified by simulations in AWGN and multi-path Rayleigh fading channels, and it shows that the system can effectively improve the transmission rate and has better bit error performance while having high spectrum utilization.

This paper provides the sufficient conditions for topological conjugation between the general cubic polynomial maps and a piecewise linear chaotic map, then provides indirectly the sufficient conditions that make the cubic polynomial maps be chaotic. This paper analyzes further the uniformity, structural complexity and randomness of the piecewise linear map and cubic polynomial maps of topological conjugation. The results show that the uniformity of the piecewise linear map is better than the polynomial maps while the randomness of the polynomial maps is superior to the piecewise linear map. As for the structural complexity, there is no significant difference between the two kinds of systems, but it should be noted that the quantitative method makes a significant impact on the structure complexity of the systems.

The construction of ZCZ Aperiodic Complementary Sequence (ZACS) sets are researched based on orthogonal matrices. The proposed approach can provide optimal ZACS sets and the length of ZCZ can be chosen flexibly under the condition of Z|N. The resultant sequence sets have ideal autocorrelation properties and intra-group complementary properties. By adjusting the parameter q, different ZACS sets can be obtained. Moreover, based on the multilevel perfect sequence over integer, Gaussian integer orthogonal matrix is constructed which can be used as the initial sequence in the construction of ZACS. The sequence sets can be applied to Multi-Carrier Code Division Multiple Access (MC-CDMA) system to remove multipath interference and multiple access interference. Furthermore, it can be used as training sequence in Multiple Input Multiple Output (MIMO) channel estimation.

With the development of cloud computing, the security and search performance of ciphertext retrieval has become the focus of research. In the traditional encryption schemes, most of them only solve the problem of defending against external keyword guessing attacks but ignore the honest and curious cloud server. In order to improve the security of ciphertext, an inside keyword attack scheme based on inverted index is proposed. Firstly, the private key of the data owner is added to resist the keyword attack of the malicious cloud server when the ciphertext inversion index is built. Secondly, an efficient public key ciphertext search scheme of parallel encryption index structure is introduced to realize the parallel search task of keywords. Compared with the traditional public key searchable encryption, the proposed scheme enhances greatly the security and search efficiency of the search system.

With the popularization of various cloud services such as cloud computing and cloud storage, privacy preservation issues in the cloud environment have gradually become the focus of industrial applications. Homomorphic encryption has become an important method to solve this issue. Among them, how to construct an efficient fully homomorphic encryption scheme is one of the hotspots at present. Firstly, the development of homomorphic encryption is introduced. The homomorphic encryption schemes are analyzed and classified from different perspectives. The research progress of verifiable fully homomorphic encryption schemes is discussed in detail. By analyzing the property rights literature on homomorphic encryption that has been published in recent years, the progress in the theoretical research and application about homomorphic encryption are summarized. Secondly, the working performances of three typical homomorphic encryption libraries, Helib, SEAL and TFHE, are compared and analyzed. Finally, various application scenarios of homomorphic encryption technology are sorted out, and possible research and development directions in the future are proposed.

With the development of computer technology and parallel solving technology, domain decomposition method has been increasingly applied to various fields of computational electromagnetics. For the simulation of microwave tube permanent magnet focusing system, this paper proposes a finite element-based non-overlapping domain decomposition method, and introduces a novel transmission condition. Then the interior penalty formulation is used to derive the finite element weak form. The biggest advantage of the proposed domain decomposition method is that no extra unknowns are introduced, and the final finite element matrix is symmetric and positive definite, which makes the matrix equation suitable be solved by the preconditioned conjugate gradient method. In this paper, several microwave tube permanent magnet focusing systems are simulated and compared with the commercial software Maxwell in detail. The results show that the proposed domain decomposition method has the same accuracy as Maxwell, but has a more superior computational performance.

Image reconstruction for Mirrored Aperture Synthesis(MAS) radiometer is an image inversion process from cosine visibility function to brightness temperature, and the cosine visibility function is solved by the transformation equation. However, the transformation equation is ill-conditioned equation, and a small error in the correlation output causes a big deviation in the cosine visibility function. Therefore, the solution of the ill-conditioned equations is the key to the success of the brightness temperature reconstruction algorithm. Based on the basic principle of MAS, the ill-conditioned transformation matrix is analyzed, and the truncated singular value decomposition is applied to the solution of the transformation equation for MAS. Simulation and experiment show that this method can effectively reduce noise and improve image quality.

Considering the problem that it is difficult to obtain the initial position accurately in the low-precision Inertial Measurement Unit/Global Navigation Satellite System (IMU/GNSS) loose integrated navigation system and the course divergence is easy to travel. A two-segment continuous alignment method assisted by dual antennas is designed. Firstly, the influence of initial bearing error on heading accuracy is analyzed. Secondly, due to the characteristics of high accuracy and poor dynamic response of the GNSS direction-finding system, a least-squares attitude estimation model is derived based on the dual-antenna baseline vector for initial alignment. Finally, for the alignment between travels, the research extendes the one-dimensional measurement based on the heading difference to suppress the heading error. The design experiment explores the influence of the dual antenna baseline vector on the initial alignment and the heading accuracy between travels. The improved method can make the initial azimuth error better than 0.7°. At the same time, the heading angle between travels can be tracked more accurately. For the initial alignment of the target and the alignment between the traveling, the dual antenna can provide auxiliary information, its effect is better than the single antenna IMU/GNSS combination, and the method calculation is moderate.