17 results about "Gaussian approximation" patented technology

The invention discloses a covert communication system based on a generative diffusion model, and belongs to the technical field of covert communication. In order to solve the problems of insufficient concealment and poor anti-noise performance in the prior art, hidden transmission and reliable recovery of information are realized by constructing a diffusion generation and classification discrimination module. The sending end carries out noise addition and shaping on the original bit waveform by using a forward diffusion process to form a noise-like hidden signal; and a receiving end recovers a signal structure through a reverse denoising process, and accurately judges bit information through a classifier. According to the method, a Gaussian approximation hidden waveform library is constructed by adopting a signal pair training model which is consistent in statistics but distinguishable in structure, and random sampling mapping and boundary smooth splicing are combined, so that transmission signals are difficult to distinguish from noise in a frequency domain and a feature space; therefore, the avoidance capability, the anti-interference robustness and the information recovery accuracy under the low signal-to-noise ratio of the detection means are remarkably improved.

The invention discloses a GNN signal detection method based on an attention mechanism in a cellular-removed large-scale MIMO system, and belongs to the technical field of wireless communication. Aiming at the problem that the traditional EP algorithm adopts independent Gaussian approximation for multi-user posterior distribution and cannot effectively capture interference association among users, the method is combined with a graph neural network and a 2D MHA module, the 2D MHA module captures the interference relation between users through the time attention sub-block, the frequency attention sub-block captures the correlation between different features, the two outputs are spliced and subjected to linear projection to form enhanced feature representation, and meanwhile the fusion capacity of prior and likelihood information is enhanced through the EP module. According to the method, the low calculation complexity is kept while the detection precision is improved, and the adaptability and robustness of the algorithm in different channel and network environments are remarkably enhanced.

A method of processing wireless communication signals for joint communication and environmental awareness in a region of interest (ROI) is presented. The method alternately uses a first iterative process and a second iterative process for determining an initial representation of a voxelized environment of the ROI and for determining the transmitted signal. The first iterative process and the second iterative process may implement a Gaussian approximation belief propagation process.

The application discloses a kind of multi-energy coupling and conversion comprehensive energy system multi-state reliability evaluation method.Establish the multi-state reliability model of multi-energy conversion element considering multi-energy conversion, using clustering algorithm reliability aggregation, and mapping formation aggregated multi-state reliability model;Multi-state reliability model of multi-energy coupling element considering multi-energy coupling is established, and then using Gaussian approximation and sampling approximation reliability approximation processing in sequence obtains multi-state approximate reliability model;The aggregated multi-state reliability model and multi-state approximate reliability model are processed according to the fast reliability evaluation method to obtain the reliability of the comprehensive energy system, and the reliability evaluation is realized.The application improves the traditional reliability evaluation method from time and accuracy, reduces the calculation time, and quickly and effectively realizes the estimation of system reliability.

The application provides a CRC concatenated Polar code encoding method for key set protection, which comprises the following steps: according to a Polar code Gaussian approximation (GA) construction algorithm, selecting reliable bit channels as information bits and unreliable bit channels as frozen bits; obtaining the information bit most likely to have the first error, i.e. the key set, through a code tree structure or Monte Carlo simulation; selecting a CRC polynomial to pre-encode the key set, and then appending the obtained CRC check vector to the end of the information bit vector; and finally encoding the pre-encoded vector by using a Polar code encoding method. The method has the following advantages: compared with a traditional CRC concatenated Polar code encoding method, the method can use a smaller CRC code length by protecting the key set, and can reduce the block error rate when applied to SC-Flip decoding.

The variational parameter determination unit 81 determines a variational parameter that specifies a position where a likelihood function and a lower bound of the likelihood function to be approximated by Gaussian are in contact. The gradient direction lower bound calculation unit 82 generates a likelihood function made one-dimensional in a gradient direction at the center of a prior distribution and calculates the lower bound of the generated likelihood function. The full dimensional lower bound calculation unit 83 sets covariances in directions other than the gradient direction to an arbitrary covariance and calculates the lower bounds of the set covariances.

The invention discloses a signal detection method and device based on Gaussian approximation interference and parallel architecture, relates to the field of digital signal processing, and solves the problem of complex calculation of a traditional E-MPA algorithm. The technical scheme is characterized in that a received signal vector is modeled to obtain a signal model; modeling interference terms in the received signal model as a Gaussian random variable distribution model, and determining to obtain Gaussian information; updating the function node to obtain a logarithm domain function node message; executing a cyclic shift gating operation according to the control signal; receiving a logarithm domain function node message, and carrying out variable node updating to obtain a normalized probability domain variable node message; the probability domain variable node message is returned for iterative updating until a preset number of times is reached; after the last iteration is completed, calculation is carried out, and the log-likelihood ratio of the bit stream is output; and the effects of reducing the calculation complexity, optimizing the in-chip storage access scheduling and greatly reducing the resource overhead are achieved.

The invention discloses an anti-attack windowing Gaussian approximate fusion filtering method under unknown noise statistics. The method comprises the following steps: 1, establishing a multi-sensor nonlinear system dynamic model; 2, establishing a measurement model actually received by a local estimator; 3, designing an anti-attack windowing Gaussian approximation local filter unified framework; 4, calculating, and summing; 5, obtaining an estimator of unknown noise statistics based on a mobile windowing and random weighting estimation method; 6, determining a random weighting factor; seventhly, Gaussian weighted integral involved in the approximate local filtering unified framework is obtained, and local filtering is obtained; and 8, providing an anti-attack windowing Gaussian approximate sequential weighted covariance cross fusion filter in combination with a sequential thought and a weighted covariance cross fusion criterion. According to the method, the accuracy of the filtering performance of the fusion filtering problem of the multi-sensor nonlinear system with attack and unknown noise statistics under the Gaussian approximation filtering framework is effectively improved.

The invention discloses a polarization code decoding optimization method and a communication system, and the method comprises the following steps: A1, optimizing a Gaussian approximation method, and calculating a reliability value of a polarization channel based on the optimized Gaussian approximation method; a2, when the first continuous elimination decoding fails, determining a first bit index set based on a reliability value, a log-likelihood ratio and a channel index value, carrying out single bit flipping decoding in sequence based on the sequence of the first bit index set, and if flipping cyclic redundancy check decoding is successful in a preset flipping decoding time threshold, ending decoding; otherwise, executing the step A3; a3, combining the polarization channels in pairs, determining a second bit index set, performing two bit flipping decoding in sequence based on the sorting of the second bit index set, and if flipping cyclic redundancy check decoding is successful in a preset flipping decoding time threshold, ending decoding; and otherwise, the decoding process is ended, so that the bit flipping decoding performance is effectively improved.

A residual life prediction method based on nonlinear wear and random jump comprises the following steps: firstly, constructing a health index of multi-source information fusion, and establishing a composite degradation model integrating a power law time scale function and a non-homogeneous Poisson process so as to simultaneously represent progressive accelerated wear and time-varying burst impact characteristics of equipment; then, designing an improved two-stage parameter estimation strategy, utilizing an expected condition maximization algorithm to process impact frequency hidden variables, identifying time-varying intensity parameters in combination with maximum likelihood estimation, and realizing robust identification of complex model parameters; deriving an approximate probability density function of the residual life based on a first passing time frame and a Gaussian approximation theory, and realizing probabilistic prediction through numerical integration; finally, a sliding window updating mechanism is established, and online self-adaptive adjustment and dynamic prediction of model parameters are achieved. The method is high in nonlinear fitting capability, sensitive in impact capture, high in prediction precision and accurate in uncertainty quantification under complex working conditions and limited observation data conditions.

The variational parameter determination unit 81 determines a variational parameter that specifies a position where a likelihood function and a lower bound of the likelihood function to be approximated by Gaussian are in contact. The gradient direction lower bound calculation unit 82 generates a likelihood function made one-dimensional in a gradient direction at the center of a prior distribution and calculates the lower bound of the generated likelihood function. The full dimensional lower bound calculation unit 83 sets covariances in directions other than the gradient direction to an arbitrary covariance and calculates the lower bounds of the set covariances.

The application discloses a kind of multi-energy coupling's comprehensive energy system fast multi-state reliability improvement method.For the comprehensive energy system, a multi-state reliability model considering multi-energy coupling is established;Gaussian approximation is used to obtain a multi-state Gaussian approximation reliability model by approximating the reliability of the multi-state reliability model; a multi-state sampling approximation reliability model is obtained by approximating the reliability of the multi-state Gaussian approximation reliability model using sampling approximation; the approximate reliability of the comprehensive energy system under different multi-energy load requirements is calculated using a fast reliability evaluation algorithm for the multi-state sampling approximation reliability model; the approximate reliability is used to establish a reliability constraint condition for comprehensive energy system structure planning and standby optimization scheduling, thereby improving the reliability of the comprehensive energy system.The application improves the traditional reliability evaluation method in terms of time and accuracy, reduces the calculation time, and effectively improves the system reliability.

The invention relates to the technical field of unmanned aerial vehicle communication, particularly discloses a polarization code construction method suitable for a Rician fading channel, and provides two low-complexity polarization code construction methods suitable for a fast Rician fading channel, the first method is a simplified discrete density evolution (DDE) method, and the second method is a low-complexity polarization code construction method suitable for a fast Rician fading channel. According to the method, a probability density function PDF of a received symbol log-likelihood ratio LLR is discretized into a probability quality function PMF, optimization is achieved through PMF evolution based on minimum sum decoding, a numerical result shows that the proposed DDE method can effectively evaluate the polarization code performance and can also be used as a practical construction tool, and the method is suitable for popularization and application. According to the second method, a Rice channel is approximated to be a binary input additive Gaussian white noise B-AWGN channel with equivalent capacity, a Gaussian approximation density evolution (DEGA) method is adopted for optimization, the simulation result shows that the performance of the polarization code constructed by the Gaussian equivalent GE method is equivalent to that of the polarization code constructed by the DDE method, and meanwhile, the construction complexity is reduced.

The invention belongs to the technical field of communication channel coding, and particularly relates to a space coupling polarization code construction and decoding method under an AWGN channel, which comprises the following sequentially executed steps: S1, acquiring a reliability distribution sequence adaptive to the AWGN channel by adopting a Gaussian approximation algorithm; s2, calculating performance improvement scores under different coupling ratios by taking the bit error rate and the frame error rate of the first iteration as base lines, and selecting the ratio with the highest performance improvement score as the optimal coupling ratio; s3, generating a space coupling structure based on the optimal coupling ratio and the reliability distribution sequence; s4, optimizing the time aggregation weight based on the normalized kurtosis score, and obtaining system operation parameters; s5, performing enhanced iterative decoding, and outputting a decoding result; a quality index optimization weight coefficient is introduced, and time domain correlation among different iteration stages is fully mined.

A method for evaluating classification models that are trained by using incomplete data sets for which data is known to be missing and the missing data is known to be non-random is provided. The method includes: receiving first information that relates to data to be used for training and evaluating a performance of a classification model that is designed to make a determination with respect to a particular query; analyzing the first information to determine second information that relates to a known portion of the first information and third information that relates to missing data; estimating, based on the third information, an uncertainty that corresponds to the missing data; and calculating, based on the second information and the estimated uncertainty, a first Gaussian approximation to a performance metric that relates to the first classification model.

The invention provides a wind power plant power adjustment method and device, equipment and a storage medium. The wind power plant power adjustment method comprises the steps of generating wind speed time sequence data of a target working condition by using a Kaimal model based on an average wind speed, turbulence intensity and scale parameters of the target working condition; obtaining a conditional probability density function satisfied by the wind speed time sequence data through Gaussian approximation; performing convolution correction on a known calibration power curve based on a conditional probability density function satisfied by the wind speed time sequence data to obtain a corrected power curve under the turbulence intensity; and adjusting the output power of the wind turbine generator under the turbulence intensity based on the modified power curve and the real-time wind speed. The embodiment of the invention can effectively improve the adjustment accuracy of the output power of the wind turbine generator under the turbulence condition.

The invention provides an incremental pulsar candidate body identification method and system based on a low-rank Bayesian neural network. The method comprises the following steps: firstly, preprocessing observation data of a radio telescope to generate a multi-modal diagnostic graph; then constructing a Bayesian neural network containing one-dimensional and two-dimensional convolution branches, introducing a low-rank and diagonal Gaussian approximation technology in a variational inference process, decomposing a weight matrix into the sum of a low-rank matrix and a diagonal matrix, and greatly reducing the parameter space complexity; and meanwhile, a mixed loss function is constructed in combination with Focal Loss and KL divergence. According to the system, model parameters can be updated online along with inflow of new observation data without forgetting old knowledge, the calculation complexity is reduced from O (d2) to O (dr), memory occupation and training time are remarkably reduced, and the system is suitable for resource-limited real-time astronomical data processing terminals and large-scale patrol projects.