73 results about "Random coding" patented technology

Example embodiments of the present disclosure provide for an example method for collaborative randomized encoding for training machine-learned models with vertically split training data. The method can include maintaining a machine-learned model. The method can include generating a first encoded dataset by sealing, using a shared key, a first private dataset into a fixed joint computation. The method can include storing, by the first encoded dataset. The method can include deleting the shared key. The method can include obtaining a second encoded dataset from a second computing system. The second encoded dataset was generated by the second computing system using the shared key to seal a second private dataset into the fixed joint computation. The method can include generating an output by evaluating the fixed joint computation using the first encoded dataset and the second encoded dataset as input and updating the machine-learned model based on the output.

The invention belongs to the technical field of communication, and particularly relates to an ultra-wideband ground penetrating radar transmitting link design method based on pseudo-random coding. The pseudo-random code generation module is connected with the serial-to-parallel conversion module, the serial-to-parallel conversion module is connected with the double data transmission rate module, and the double data transmission rate module is connected with the signal emission module; the pseudo-random code generation module is used for generating high-speed pseudo-random codes; the serial-to-parallel conversion module is used for integrating four paths of parallel data into one path of serial data and outputting the serial data; the double data transmission rate module is used for converting a signal transmitted only at a rising edge into a form that the rising edge and a falling edge are sampled and transmitted; and the signal transmitting module is used for transmitting the high-speed pseudo-random codes sampled at the rising edge and the falling edge. The detection resolution of the ground penetrating radar and the signal-to-noise ratio of a transmitted signal are improved.

The invention provides a single-cell multi-omics data clustering method based on a diffusion auto-encoder, and belongs to the technical field of biological single-cell data classification. Comprising the following steps: carrying out dimension reduction processing on paired single-cell omics A and omics B data to obtain two types of omics data with consistent dimensions; integrating the features of the two types of omics data by using a feature fusion module based on a bidirectional cross attention mechanism to obtain common semantic embedding z of the cells; generating random encoder features based on the two types of omics data, and reconstructing original data based on z and the random encoder features by using two conditional diffusion models; updating parameters of the conditional diffusion model by using a loss function to obtain a common semantic embedding z'which minimizes the loss function; and dividing the cells into different cell subgroups by using z'by using a clustering algorithm. According to the method, the semantic features and random noise of the cells are separated by using decoupling learning, and the cells are clustered by using cross-omics invariant features, so that different subtypes can be accurately identified.

The invention discloses a voltage measurement method based on pulse coding and related equipment, relates to the technical field of voltage measurement, and solves the problem of low voltage measurement precision. The method comprises the following steps: generating a pseudo-random coding sequence with sparse time domain, and modulating a pilot frequency reference signal based on the pseudo-random coding sequence to obtain a low-frequency pulse modulation wave; performing pulse power amplification on the low-frequency pulse modulation wave, and migrating a signal frequency band to a high-frequency range to obtain a high-frequency pulse train; and after the high-frequency pulse train is injected into the measured line, acquiring a coupling signal, and extracting a characteristic component corresponding to the pseudo-random coding sequence from the coupling signal so as to invert the voltage value of the measured line. According to the invention, the pilot frequency reference signal based on pulse code modulation is adopted to replace a continuous sine wave, so that the power consumption of signal injection can be reduced; and the corresponding characteristic component in the coupling signal is identified through the pseudo-random coding sequence, so that interference of a strong electromagnetic signal can be avoided, and the voltage measurement precision is improved.

The present disclosure provides an image data augmentation method, apparatus, electronic device, and storage medium, which relate to the technical fields of artificial intelligence such as data mining, data synthesis, and data generation, and at least solve the technical problems of low data augmentation efficiency and poor model enhancement effect in the prior art for the method of augmenting edge data based on data enhancement or feature mining. The specific implementation solution is as follows: Obtain an original image data set, where the original image data set includes edge image data and non-edge image data, and the number of edge image data is less than the number of non-edge image data; Use a generative adversarial network to perform feature encoding on the edge image data to obtain a first encoding result, where the generative adversarial network is trained using the original image data set; Based on the first encoding result, perform an offset process on a random encoding to obtain a second encoding result; Use the generative adversarial network to decode the second encoding result to obtain augmented image data corresponding to the edge image data.

This invention discloses an end-to-end spectral reconstruction method, comprising the following steps: S1, reconstructing a hyperspectral image; S2, using the obtained hyperspectral image as input to an optimized neural network, training the neural network parameters according to the end-to-end spectral reconstruction model, and calculating the loss function; S3, backpropagating errors to the coding template to be designed, changing the design parameters of the coding template, and continuing to train the model until a high-quality hyperspectral image is reconstructed. By employing the above technical solution, the design of the optical system is combined with the reconstruction method. Based on the coded aperture snapshot system, an adaptive coding template is added according to the end-to-end spectral reconstruction method, replacing the traditional random coding template, thereby improving the quality of spectral reconstruction.

The invention provides an underwater acoustic network online fountain code method, a sender device and a receiver device, and the method adopts a dynamic rule coding strategy of rule coding and random coding to generate a rule degree 2 coding packet, a random degree 2 coding packet and a random degree 1 coding packet, and random coding dynamically adjusts coding redundancy based on channel conditions, and two coding modes are dynamically fused, so that the decoding overhead can be effectively reduced, the feedback frequency is reduced, and the overall time delay is reduced. In the decoding stage, the feedback information in the feedback packet is fully utilized to remarkably improve the decoding efficiency, the optimal coding set information contained in the feedback packet can reflect the decoding state of a receiver more efficiently, the useful probability of the coding packet in the decoding process is improved, and the overall decoding overhead is reduced. According to the invention, the sender can estimate the channel state in real time and adaptively determine the redundant data sending amount, thereby realizing effective control of the decoding overhead, improving the decoding efficiency and reducing the number of feedback times.

This application discloses a random network coding method, apparatus, storage medium, and computer device based on coding dimension interval constraints. The method includes: dividing the data to be transmitted into multiple data batches, each data batch including multiple raw data packets; for each data batch, dynamically determining the current actual coding dimension of the data batch within a preset coding dimension interval based on the current network state information; repeatedly selecting raw data packets from the data batch with a number equal to the current actual coding dimension, and linearly combining the multiple selected raw data packets using random coding coefficients to generate coded data packets; sending each generated coded data packet to a receiving end, so that when the receiving end determines that the number of received linearly independent packets has reached the current actual coding dimension, it constructs a decoding matrix based on the random coding coefficients carried in the linearly independent packets, and recovers all the raw data packets in the data batch through the decoding matrix.

The present invention provides a high-speed multi-level quantum phase drive circuit based on radio frequency transistors, comprising a random signal generator, a plurality of high-speed level converters respectively connected to the output ends of the random signal generator, and a plurality of radio frequency transistors and their resistor network units respectively connected to the output ends of the high-speed level converters; the random signal generator generates a random coding signal according to random coding logic, the random coding signal is transmitted to the high-speed level converter, and the high-speed level converter converts the random coding signal into a low-voltage positive emitter coupled logic level. The low-voltage positive emitter coupled logic level drives the corresponding radio frequency transistor to conduct, and the radio frequency transistor outputs a level of corresponding amplitude. At any time, only one radio frequency transistor is conducted. The advantages of the present invention are: the power amplification of the level signal is achieved by using the parallel connection of radio frequency transistors instead of a broadband power amplifier, which simplifies the design difficulty of the quantum phase drive circuit and greatly reduces the cost.

The present invention discloses an artificial intelligence-based electromagnetic performance design method and system for laminated structural materials, which relates to the field of artificial intelligence technology. The method includes obtaining basic electromagnetic parameters of ply units through experiments, encoding ply units with different electromagnetic absorption properties, and generating random codes for ply configurations based on time dependence. Using electromagnetic simulation software, a simulation database of random codes for ply configurations and corresponding electromagnetic absorption properties is established. The simulation database is used to train a proxy model for predicting electromagnetic absorption performance under different ply configurations to obtain a prediction model. The prediction model is used to predict the electromagnetic absorption performance of candidate ply configurations, determine whether the electromagnetic absorption performance of the candidate ply configuration is optimal, and if so, determine that the candidate ply configuration is the optimal ply configuration. This method solves the problems faced by electromagnetic absorption materials, such as the difficulty and high cost of searching a large parameter space composed of ply units and material types.

The application provides a design method and device of a wave-absorbing composite material based on a genetic algorithm, and the method comprises the following steps: based on a plurality of different types of wave-absorbing materials, random coding is performed to obtain a plurality of genetic individuals, and an initial population is formed; the reflectivity of each genetic individual in the population is calculated; the fitness of each genetic individual is evaluated; according to the number of bits corresponding to the data variable in the coding and the current population evolution generation number, the mutation rate of each genetic individual is calculated; the genetic individual with the minimum fitness is retained, and the other genetic individuals are crossed and mutated by using the mutation rate to obtain a new generation population, and the calculation of the reflectivity of each genetic individual in the population is performed on the new generation population, until the termination condition of the genetic algorithm is met, so that the genetic individuals in the new generation population obtained form the wave-absorbing composite material. According to the scheme, the mutation rate can be gradually increased to accelerate the optimization speed under the premise of ensuring the early rapid convergence, and the rapid convergence and rapid optimization of the genetic algorithm are realized.

A method includes obtaining a training text sample, the training text sample not paired with corresponding audio data, and generating a sequence of pseudo-random encoder variables. The method also includes processing, using a decoder of a sequence transduction model, the sequence of pseudo-random encoder variables to predict a probability distribution over possible output labels. The method further includes determining a loss based metric based on the training text sample and the predicted probability distribution over possible output labels, and training the decoder based on the loss metric.

Embodiments of the present application provide a kind of invalid instruction generation method, device, electronic equipment and storage medium, involve computer technical field, in the method, first instruction coding format is selected from the instruction coding format set corresponding to the preset instruction set architecture;First random code with the fixed bit section total length of first instruction coding format is equal in length;In the case where the first random code is inconsistent with each preset fixed code corresponding to the first instruction coding format, the invalid instruction corresponding to the preset instruction set architecture is generated based on the first random code.The invalid instruction corresponding to the preset instruction architecture can be directly generated in a random manner, which does not comply with the instruction coding format, which helps to improve the efficiency of generating the invalid instruction of the preset instruction set architecture.

Communication systems and methods are disclosed that utilize Probability Amplitude Shaping (PAS) and Trellis Coded Modulation (TCM) to transmit short block-length messages. While capacity-approaching codes, such as Low Density Parity Check (LDPC) codes, Turbo codes, and Polar codes, can achieve data rates approaching the Shannon limit at large block-lengths, the performance of these codes can deteriorate dramatically at short block-lengths. Communication systems configured in accordance with various embodiments of the invention can utilize classical codes to encode short block-length messages to achieve communication rates exceeding Polyanskiy's Random Coding Union bound. In several embodiments, message bits are transmitted using a TCM system in which message bits that have been previously encoded by an error detection code are encoded using a convolutional code (CC). In a number of embodiments, the error detection code and the CC are obtained via a joint optimization with respect to a Frame Error Rate bound.

The invention relates to a wavefront measurement method based on a modulation type Shack-Hartmann wavefront sensor, the modulation type Shack-Hartmann wavefront sensor comprises a composite laminated modulation device and a detector, and the composite laminated modulation device is composed of a microlens array and a binary random coding mask. The method comprises the following steps: acquiring a composite diffraction pattern which is collected by a detector and is coherently modulated by a composite laminated modulator; wherein the composite diffraction pattern has diffraction texture features, and the diffraction texture features are obtained by coding after coherent modulation is carried out on high-frequency phase information through a binary random coding mask; reconstructing a high-frequency complex wavefront based on the diffraction texture features by adopting a coherent modulation phase recovery algorithm; and outputting the high-frequency complex wavefront as a wavefront measurement result. According to the scheme provided by the invention, unified measurement of high spatial resolution and wide dynamic range can be realized under the condition of single exposure, so that the inherent limitation of the SHWFS between the spatial resolution and the dynamic range in the related technology is broken through.

The application discloses a kind of mobile phone token management method based on iOS platform, belong to information security authentication technical field, its method specifically includes: mobile phone end reads the key after official website encryption, and save to keychain, key decryption obtains a random code associated with user, according to current time and step length, dynamic code is obtained using SHA1 algorithm, dynamic code 30 seconds update once, dynamic code comparison is consistent, and it is bound successfully;When user logs in again, in addition to inputting correct account password, dynamic code also needs to be verified to enter into system, mobile phone end reads key again, compare the key previously saved in keychain, and reset to keychain according to the difference between the two, the key is saved in keychain in the application, and data will not be lost, will not be affected by network, APP exits restart or uninstallation and reinstallation, the data in keychain will not be lost, and can be used permanently, can well solve the problem that existing mobile phone SMS verification code exists attack, blast.

The invention relates to the technical field of electric power safety, and particularly discloses a cable alignment device and an alignment method thereof.The cable alignment device comprises a host unit, a far-end marker and a signal processing module, the host unit is used for generating and sending a dynamic coding signal and receiving a feedback signal of the far-end marker, and the signal processing module is used for processing the dynamic coding signal; the far-end identifier is used for receiving a host signal and returning a check code, the far-end identifier comprises a low-power-consumption wireless wake-up module, and the signal processing module is used for executing frequency hopping communication; by setting frequency hopping and pseudo-random coding modes, the signal can dynamically avoid a power frequency interference frequency band, so that the anti-interference capability in a strong electromagnetic environment is effectively improved, on-demand activation and automatic dormancy of the far-end marker are realized by setting the 125kHz low-frequency wake-up signal, the overall power consumption of the system is remarkably reduced, and the power consumption of the system is reduced. Through the arrangement of the frequency division multiplexing or time division multiplexing technology, parallel transmission and decoding of multiple paths of signals are supported, so that the detection efficiency of the multi-core cable is greatly improved.

The invention discloses a conformal coding metasurface for RCS reduction, and the metasurface comprises a bottom layer substrate, and an electromagnetic regulation and control working region of the bottom layer substrate comprises a plurality of resonance units which are arranged at one side of the bottom layer substrate in an array manner; the resonance unit comprises a composite dielectric layer and a metal pattern layer which are sequentially stacked on the bottom substrate; the metal pattern layer comprises a first metal layer, and a first active control element and a second active control element which are arranged in the middle area of the first metal layer; the electromagnetic wave regulation and control mode of the coding metasurface is space-time random coding. Through the design of the resonance unit structure, the superficial areas of the metal pattern layer and the composite dielectric layer are reduced, and the RCS reduction and conformal capability of the coding metasurface is improved.

The invention is suitable for the technical field of radars, and provides a radar control method and device, terminal equipment and a storage medium, and the method comprises the steps: controlling a radar to transmit a detection signal based on a preset transmission coding mode; the preset emission coding mode is to control the radar to emit a detection signal according to the emission code corresponding to the emission interval in the current emission period; wherein the transmitting period comprises a random coding area and an analysis area; the random coding area is used for controlling the overall offset of the analysis area in the current measurement period, the correlation interference can be effectively resisted by introducing the random coding area, and the probability of crosstalk is reduced as much as possible.

The application provides a time coding compressed sensing imaging method and device based on FPGA and electronic equipment, the method is applied to the technical field of photoelectric detection and computational imaging, and the method comprises the steps of: calibrating an FPGA delay unit based on a reference clock, constructing a delay lookup table and writing a time coding sequence of random coding and a push-scan mode. In response to a synchronization signal, coarse and fine delay tap values are analyzed, a reference pulse is generated and is synchronized to the reference clock, a delay pulse is generated through fine delay, three alignment timing control signals are generated, and the excitation gate and exposure signals are cooperatively controlled. Discrete integral observation values are collected in an exposure window, a continuous observation operator is constructed in combination with the delay lookup table, a target function is constructed based on physical priori, and the target function is iteratively solved to obtain a high signal-to-noise ratio, a transient response curve and physical parameters. The method can solve the problems of slow sampling speed and lack of software flexibility due to dependence on modulation devices in traditional hyperspectral imaging technology.

The invention discloses an unmanned aerial vehicle intelligent fault detection method based on a generative adversarial network. The method comprises the following steps: establishing a monitoring data set; constructing and training a generative adversarial network; setting a gradient information constraint mechanism which introduces an encoder into the discriminator, so that the input of the discriminator is changed into a random code obtained by mapping an input sample by the encoder, and applying upper bound constraint to mutual information between the input sample and the random code; introducing a cyclic difference balance module which introduces the gradient information constraint mechanism as a constraint in a target function in a discriminator training process, and training a discriminator to distinguish monitoring data from distribution of random codes; performing data enhancement on the monitoring data set by using the trained generative adversarial network, and constructing a classifier model; and training a classifier model by using the monitoring data set after data enhancement, so as to identify new monitoring data acquired in real time in the flight process of the unmanned aerial vehicle and output a fault type.

An optical detector for detecting a coolant leak within an information handling system includes a light source, light sensor, and controller. The light source is configured to illuminate objects within the information handling system with a pattern of light. The pattern is determined according to a randomized encoding. The light sensor is configured to generate a signal in response to absorbing light emitted from an object illuminated by the light source. The signal is encoded according to the randomized encoding. A controller operatively coupled with the light sensor is configured to generate a decoded signal by reversing the randomized encoding of the signal. The controller is configured to identify a potential coolant leak in response to the decoded signal matching a predetermined pattern that indicates a coolant leak.

The invention discloses a spatial light modulation method and device for quantum random number coding, equipment and a medium, relates to the technical field of spatial light modulation, and realizes seamless adaptation of quantum random coding and spatial light modulator hardware by converting a quantum random coding matrix corresponding to a target coding base into a driving signal adaptive to a spatial light modulator. It is ensured that the driving signal can accurately control the spatial light modulator to carry out random code modulation on the incident spatial light. The method comprises the steps that a quantum random number generator is adopted as an entropy source, verification screening operation is executed on an original quantum random number sequence output by the quantum random number generator, and an effective quantum random number sequence is obtained; entropy extraction operation is executed on the effective quantum random number sequence according to the set sampling number, and a target coding base is constructed; and converting the quantum random coding matrix corresponding to the target coding base into a driving signal adaptive to a spatial light modulator, so as to control the spatial light modulator to carry out random coding modulation on the incident spatial light through the driving signal to obtain modulated light.

The invention discloses an enterprise policy dynamic matching method and system based on big data, and the method comprises the following steps: S1, obtaining enterprise data and policy data, relates to the technical field of big data analysis, and carries out the preprocessing, feature extraction, feature fusion and nonlinear transformation of the enterprise data and the policy data, thereby obtaining an enterprise policy dynamic matching result; generating an enterprise abstract feature vector and a policy abstract feature vector, using the cosine similarity of the two as an initial matching score, and combining a historical matching success rate, an industry matching preference coefficient and a time coefficient to generate a score adjustment factor; meanwhile, a historical matching success rate, an industry matching preference coefficient, a time coefficient weight and a product of an initial matching score and a score adjustment factor are generated through combination of an entropy method and random coding to serve as a dynamic matching score, and a dynamic scoring method considering timeliness, domain correlation and historical matching experience is constructed; and the dynamic matching accuracy of the enterprise and the policy is effectively improved.

The invention relates to a selection and sorting decision-making method considering the sowing wall capacity, a service platform and a storage medium, and the method comprises the steps: carrying out the random coding and population initialization of SKUs associated with a plurality of obtained to-be-selected selection orders according to a preset coding mode, and generating a plurality of first sorting coding bodies; using a heap optimization simulated annealing algorithm to perform sorting correction preprocessing on the sub-codes corresponding to the plurality of first sorting coding bodies to generate a plurality of second sorting coding bodies, and based on the plurality of second sorting coding bodies, repeatedly performing non-dominated solution sorting and genetic evolution operation iteration by using an NSGA-II algorithm and the fitness of the corresponding sorting coding bodies, so as to obtain a sorting result; until a plurality of alternative sorting coding bodies are generated; and selecting a target sorting coding body from the plurality of alternative sorting coding bodies according to the fitness of the alternative sorting coding bodies, and taking a plurality of target sorting batches obtained by processing the target sorting coding body according to a sorting decision rule and sorting paths mapped by the target sorting batches as decision results.

The invention relates to the technical field of radar detection, in particular to a glacier detection system and method. The system comprises a signal generation module used for alternately repeating complementary code elements with a preset time width to generate a pseudo-random coded signal with a preset code length, and performing square wave modulation on the pseudo-random coded signal according to a carrier square wave period to obtain a radio frequency signal, the preset time width being an integer multiple of the carrier square wave period, and the preset time width being an integer multiple of the carrier square wave period; the preset code length is integer power of two; the signal transmitting module is used for processing the radio frequency signal to obtain a transmitting signal and transmitting the transmitting signal to the glacier; the signal receiving module is used for receiving an echo signal of the glacier to the transmitted signal and carrying out digital signal processing on the echo signal to obtain a digital echo signal; the signal analysis module is used for analyzing the digital echo signal to obtain a glacier structure, and the glacier structure comprises at least one of the glacier thickness, the glacier terrain and the ice layer structure.

The application relates to a kitchen garbage transfer station optimization site selection method and a service platform considering double targets, the method comprising the following steps: randomly encoding candidate transfer station candidate points to generate a first site selection configuration coding array, the first site selection configuration coding array comprising a plurality of first codes; performing non-dominated solution sorting and genetic evolution operation on the first codes based on a preset improved NSGA-II algorithm to generate a hybrid site selection configuration coding array; performing new population iteration on the hybrid site selection configuration coding array by using an elite reservation strategy corresponding to the improved NSGA-II algorithm to generate a second site selection configuration coding array with a plurality of second codes; determining the fitness of the second codes, performing non-dominated solution sorting, genetic evolution operation and new population iteration processing based on the elite reservation strategy on the second codes in the second site selection configuration coding array according to the fitness to generate a target site selection configuration coding array, and obtaining target codes from the target site selection configuration coding array to obtain a site selection result.

The present invention relates to a method for authenticating a Client in a Client-Server architecture. The method encodes the Client's input passcode using a Regulated Activation Network computational model to create a Hierarchical model. Depending upon a Policy shared between the Client and the Server, different combinations of encoded passcodes generated from the Hierarchical model, are selected. After every successful authentication operation, a new policy is generated. Because of these new policies, for every authentication attempt, the Client inputs the same passcode but the passcode communicated over the network is always different. The advantages of this method include random encoded passcode generation for the Client's identity verification. This randomness ensures that the generated encoded passcodes are never repeated in two successive authentication attempts. The use of Models, Policies, and the Client's passcode together makes the authentication method highly robust.

The present disclosure provides a method for determining a driving path, which can be applied to the field of computer technology. The method for determining a driving path includes: randomly encoding the user information to form an initial population set, the initial population set including multiple individual information subsets, each individual information subset including multiple user current information and corresponding path information; for each individual information subset, calculating the fitness of the individual information subset based on the user current information and the corresponding path information to obtain multiple fitnesses; stratifying the multiple individual information subsets according to the optimization parameters and fitnesses; optimizing the corresponding path information in the individual information subsets according to the layers of the multiple individual information subsets to obtain optimized corresponding path information; and determining the driving path based on the optimized corresponding path information. The present disclosure also provides an apparatus, device, storage medium, and program product for determining a driving path.

The invention discloses a power system false data injection attack detection method and device based on periodic random diagonal measurement coding, and belongs to the field of power system information security. In order to solve the problem that a traditional residual detection method cannot identify hidden false data injection attacks, the invention provides a coding-decoding + state estimation collaborative mechanism. The method comprises the steps that reversible random coding is applied to necessary measurement at a source end, so that latent attack vectors are explicit in a decoding domain; a pseudorandom sequence generator is adopted to synchronously generate a reversible matrix for measurement coding, decoding and state estimation are carried out in a control center, and attack existence is judged through residual statistics.