41 results about "Channel decoder" patented technology

The present invention relates to the technical field of cross-modal image signal reconstruction. Disclosed is a cross-modal joint source-channel coding and decoding method adaptable to changeable scenarios. The method comprises: first designing a Transformer encoder-based cross-modal channel coding and decoding optimization solution, so as to achieve the performance improvement and robustness of a channel encoder and a channel decoder; then designing a cross-modal source coding and decoding optimization solution for haptic-to-image generation based on a latent diffusion model, so that under an image signal loss scenario, haptic information is used to guide image generation; and finally, incorporating transfer learning technology, so as to reduce additional training costs caused by a system facing changeable cross-modal communication scenarios such as a changeable channel signal-to-noise ratio and different transmission tasks. Under cross-modal changeable communication scenarios, the joint source-channel coding and decoding method provided in the present invention can solve the problems of the inability of a receiving end to well complete image reconstruction, and additional model training costs caused by changeable channel environments and scenarios.

The invention provides a diffusion-driven channel adaptive point cloud semantic communication method, which belongs to the technical field of wireless communication, and comprises the following steps: obtaining a training point cloud data set, constructing a point cloud feature extraction network based on a hypergraph convolutional neural network as a semantic encoder, and constructing a channel adaptive enhancement module as a channel decoder, a channel adaptive recovery module is constructed at a receiving end as a channel decoder, a diffusion reconstruction network is constructed as a semantic decoder, end-to-end joint training is performed on the point cloud feature extraction network, the channel adaptive enhancement module, the channel adaptive recovery module and the diffusion reconstruction network, the diffusion reconstruction network predicts original point cloud distribution, and the point cloud feature extraction network and the channel adaptive enhancement module are subjected to end-to-end joint training. And a loss function is constructed based on the chamfering distance between the predicted point cloud distribution and the original point cloud. In the communication process, the reconstructed semantic features are input into the diffusion reconstruction network to reconstruct an original point cloud structure. The high-order semantic features of the point cloud can be effectively extracted so as to improve the adaptive capacity of the method to the incomplete point cloud.

The application provides a task-oriented millimeter wave link adaptive multi-modal semantic communication method and device in an automatic driving scene, electronic equipment and a storage medium, and belongs to the technical field of automatic driving. The method comprises the following steps: inputting a target image into a semantic encoder to obtain a first feature map, and then inputting the first feature map into a joint source channel encoder to obtain a second feature map; performing feature screening on the second feature map to obtain a third feature map, and then performing digital precoding and analog precoding on the third feature map to obtain a corresponding sending signal matrix; and then sending the sending signal matrix to a receiving end; performing analog combining and digital combining on the receiving signal matrix to obtain a receiving feature map, and then inputting the receiving feature map into a joint source channel decoder to obtain decoded features; performing fusion processing on the decoded features to obtain fused features, and then inputting the fused features into a semantic decoder to obtain a semantic segmentation result output by the semantic decoder. The application improves the robustness and accuracy of semantic communication.

The present disclosure provides a semantic communication method and device, electronic equipment and storage medium, relates to the technical field of communication, in particular to the technical field of semantic communication. The scheme is: obtaining high-speed mobile user data and low-speed mobile user data; using a first deep joint source channel encoder to encode the high-speed mobile user data to obtain first semantic symbols, and using a second deep joint source channel encoder to encode the low-speed mobile user data to obtain second semantic symbols; performing delay Doppler domain modulation on the first semantic symbols to generate a first signal, and performing time-frequency domain modulation on the second semantic symbols to generate a second signal; superimposing the first signal and the second signal, and sending the superimposed signal to a receiving end; and performing parallel decoding on the receiving end through a first deep joint source channel decoder and a second deep joint source channel decoder to recover the data. The present disclosure suppresses the cliff effect and improves transmission reliability.

A semantic communication system includes a base station including a semantic encoder configured to convert data to be transmitted into semantic information and a channel encoder configured to convert the converted semantic information into a channel signal and a plurality of user equipment devices (UEs) including a channel decoder configured to receive the channel signal and decode the channel signal into the semantic information and a semantic decoder configured to decode data included in the decoded semantic information based on the decoded semantic information.

This invention discloses a source-channel joint encoding and decoding method based on a cross-modal communication system, relating to the field of cross-modal image signal reconstruction technology. The method includes: First, designing a cross-modal source encoder, inputting tactile signals and image signals with category labels into the encoder to mine deep-level fusion feature information between modalities. Second, designing a channel encoder and channel decoder, directly mapping the extracted fusion features into a bitstream for channel transmission, improving the system's adaptability to channel changes. Finally, designing a corresponding source decoder at the receiving end to complete high-quality image signal reconstruction from deep-level feature information. The source-channel joint encoding and decoding method provided by this invention reduces power consumption in the network, uses a fitness function to measure the load balance between the cloud network and the server (host), transforming the load balancing problem into an optimization problem, thus making it suitable for resource optimization. Using the particle swarm optimization results as the initial population for the genetic algorithm, this invention achieves better results in terms of execution cost, load balancing, and completion time.

The invention provides a denoising method and related device for a transformer substation hardware monitoring signal, and the method comprises the steps: obtaining the transformer substation hardware monitoring signal, carrying out the signal preprocessing, and obtaining a preprocessed monitoring signal; inputting the monitoring signal into a pre-trained denoising model for denoising processing so as to obtain a denoised transformer substation hardware monitoring signal; wherein the denoising model extracts local features and global features of a signal from a monitoring signal through a dual-channel encoder; then, the local features and the global features are subjected to denoising processing and then spliced, and fusion features are obtained; performing feature reconstruction processing on the fusion features through a dual-channel decoder to obtain local reconstruction features and global reconstruction features, and splicing the local reconstruction features and the global reconstruction features to obtain final optimization features; and reconstructing the monitoring signal so as to realize de-noising processing. According to the method, the local and global features are extracted respectively, and the noise in the monitoring signal of the transformer substation hardware fitting in the complex noise environment can be removed more thoroughly through denoising, feature reconstruction and other processing.

The application discloses a kind of semantic communication methods, devices and systems for image classification tasks, the method includes transmitter obtains the semantic feature of given image and is encoded into transmitted image semantic vector based on auto-encoder;Relay node extracts the text feature of given image, and it is multimodal feature fusion with the image feature recovered by image channel decoder, then it is obtained multimodal semantic vector by multimodal channel encoder and is forwarded to receiver;Receiver receives multimodal semantic vector and executes image classification task.The semantic communication method provided by the application introduces the structure of auto-encoder into semantic communication on the one hand, greatly reduces the information redundancy, reduces the occupied communication resources, on the other hand, introduces multimodal semantic relay module, it can supplement complementary text semantic features in relay node, to help receiver better complete image classification task.

This invention provides a resource allocation method for an image semantic communication system that balances computation and transmission. The image semantic communication system includes a channel encoder, a channel decoder, a semantic decoder, and a semantic encoder with selectable slices. Multiple slices in the encoder have different network depths and widths. The method includes: calculating the computational and transmission loads required by each terminal when selecting different slices; determining the corresponding objectives and constraints based on the communication scenario; modeling an optimization model based on the objectives and constraints; solving the optimization model based on the computational and transmission loads required by each terminal when selecting different slices to obtain a target slice combination and a target resource allocation scheme. The slice combination is the set of slices selected by all terminals, and the resource allocation scheme is the computational and transmission resources allocated to each terminal. This invention, by selecting different slices for encoding and transmission, can flexibly adapt to different communication scenarios and requirements.

A multi-channel decoder circuit associated with a multi-channel decoder system is disclosed. The multi-channel decoder circuit comprises a distributed decoder circuit comprising a set of unit decoder circuits, each unit decoder circuit configured to receive one or more codewords of a plurality of codewords associated with a plurality of input channels, and decode the one or more codewords. The multi-channel decoder circuit further comprises a distribution controller circuit configured to distribute each incoming codeword of the one or more codewords to the respective unit decoder circuit of the set of unit decoder circuits within the distributed decoder circuit, based on determining a currently available unit decoder circuit within the set of unit decoder circuits.

The present invention relates to a multi-user scheduling method, device and storage medium based on a restricted channel decoder, which is applied to the field of 5G NR communication technology. The method comprises: by judging the number of user data packets, if the number of user data packets exceeds the limited number of users x processed per time slot of the FECA decoder, then caching x user data packets in a first buffer area, caching the remaining user data packets in a second buffer area, first performing CRC processing on the user data packets in the first buffer area in the current air interface time slot, and performing CRC processing on the user data packets in the second buffer area in the time slot next to the current air interface time slot; through the above scheme, the present application can process multiple uplink user data packets at the same time without generating uplink processing conflicts, that is, uplink and downlink processing can be performed in parallel in the current air interface time slot and the next time slot without interfering with each other.

The invention discloses a finite field multiple access system codebook construction method based on EP resource allocation, a minimum branch distance decoding method and a multiple access system, and belongs to the technical field of 6G communication. The problem of low reliability under a block fading channel is solved, and the problem of lack of a decoding method for superposed user information under a finite field in an uplink at present is solved. According to the invention, the codebook is constructed based on UR-EPA, the resources of the whole channel are utilized by allocating a plurality of EPs to a single bit, and the degree of freedom of the whole channel is fully utilized. The invention also designs a decoding method for superposed user information in a finite field, firstly, Turbo iteration is carried out between a basic signal estimator and a channel decoder to realize detection of information bits, and then detection is carried out by utilizing a parity check part based on complex field superposition and patterns, so that joint decoding of check bits is realized.

The invention discloses a semantic communication system suitable for an OSI network, and relates to the technical field of digital information transmission, comprising: a semantic coding module of a sending end application layer performs semantic extraction on original data based on a semantic extractor to generate a standardized semantic packet containing header information, an AI channel encoder is dynamically called for encoding through a joint encoding module of a physical layer of the sending end according to the head information and the channel state of the standardized semantic packet, and a traditional encoder is called for the self-defined semantic packet to generate a transmission signal; and a receiving end physical layer joint decoding module analyzes the head information of the transmission signal, calls an AI channel decoder to restore the standardized semantic packet after identifying the identifier of the semantic extractor, restores the self-defined semantic packet through a traditional decoder, fuses the standardized semantic packet and the self-defined semantic packet through an application layer semantic decoding module, and reconstructs original data. The method is compatible with a traditional codec, and the cross-device interoperability is ensured by using the unified semantic knowledge base.

The application provides a frequency domain differential modulation method, system and storage medium for high-reliability and low-latency mobile communication, the frequency domain differential modulation method comprising: step 1: converting B-bit information bits into J-bit coded bits; step 2: mapping the J-bit coded bits into an M-order phase shift keying symbol block, and using a frequency domain differential modulation technology to express information as a phase difference between two adjacent subcarrier signals in the same OFDM symbol; step 3: performing K-point inverse discrete Fourier transform to convert a frequency domain transmission sequence into a time domain transmission sequence, and adding a cyclic prefix in the time domain transmission sequence; step 4: a receiver receives a serial sequence, performs K-point discrete Fourier transform, detects data by comparing subcarrier signals on the same OFDM symbol, and converts the data into B-bit information bits by using a phase shift keying soft demodulator and a channel decoder. The application has the beneficial effect that high-reliability and low-latency mobile communication can be realized at extremely low signaling cost and wireless resources.

The invention provides an OTFS system signal detection method based on a Turbo structure, and the method comprises the steps: 1, carrying out the message passing algorithm detection of a channel output signal demodulated by OTFS, and enabling the output mean value and variance to serve as the initial values of MMSE detection in the first iteration; and 2, judging whether the current number of iterations reaches a preset maximum number of times, if so, exiting iteration, and outputting a final detection result. Otherwise, entering the step 3; and step 3, MMSE detection is carried out on the demodulated OTFS signal, a mean value and a variance of the first MMSE detection are provided by a message passing detection algorithm, and a subsequent mean value and a subsequent variance are obtained by operation of extrinsic information output by the LDPC. The MMSE detector calculates extrinsic information of each symbol and sends the extrinsic information to the LDPC channel decoder; 4, the LDPC decoder carries out decoding operation on the information transmitted by the MMSE detector to obtain new extrinsic information; 5, calculating a mean value and a variance of the new extrinsic information, and taking the mean value and the variance as input prior information of the MMSE detector during the next round of iteration; and 6, returning to the step 2, entering a next round of iteration, and adding 1 to the number of iterations.

The invention discloses a text semantic communication system optimization method oriented to multiple tasks. According to the method, firstly, a semantic encoder, a channel encoder, a channel decoder and a task decoder are constructed, the task decoder comprises a plurality of semantic decoders, and each semantic decoder corresponds to a different text communication task. Firstly, a semantic encoder extracts multi-task universal semantic features, after channel encoding, transmission and channel decoding, a denoising module is used for eliminating noise influence, then an MMoE model is introduced into a task decoder, specific task features are extracted through an expert network in the MMoE model, and the task features are extracted through a task decoder. And dynamically calculating weights using the gated network of the model to generate task-related feature representations. And after feature fusion, obtaining a task-specific semantic vector, and finally decoding and outputting the task-specific semantic vector. According to the method, parallel processing of multiple tasks can be realized, and the execution efficiency and accuracy of the multiple tasks are effectively improved.

This invention discloses a speech enhancement method, an electronic device, and a storage medium. The speech enhancement method is used in a speech enhancement system, which includes an encoder module, a dual-channel module, a decoder module, and a merging module. The decoder module includes an excitation decoder, a channel decoder, and a complex spectrum decoder. The method includes: extracting preset features from the original noisy speech signal using the encoder module; modeling the preset features on the frequency axis using the dual-channel module to obtain frequency axis features; modeling the preset features on the time axis using the dual-channel module to obtain time axis features; processing the time axis features using the excitation decoder and channel decoder to obtain a first denoised speech signal; processing the frequency axis features using the complex spectrum decoder to obtain a second denoised speech signal; and merging the first denoised speech signal, the second denoised speech signal, and the original noisy speech signal using the merging module to obtain a final estimated speech signal.

The application discloses a kind of based on source channel joint coding haptic signal reconstruction method, it is related to haptic signal generation technical field, including: the haptic and image signal for training is passed through source encoder, based on classification task preliminary carries out the modal intra-feature extraction and inter-modal feature fusion of haptic and image.Based on the channel encoder of neural network modeling that fusion feature is input, it is used to produce the code stream suitable for channel transmission, after channel transmission, input channel decoder, through the joint training of channel encoder and decoder, the recovered fusion feature is input into haptic decoder, and haptic signal is reconstructed.The haptic signal reconstruction method based on source channel joint coding provided in the application realizes the semantic information fusion of different modal, greatly reduces data volume, improves the robustness of signal to channel change by separately designed channel encoder and channel decoder, and reduces the delay of signal through channel, so that the reconstruction quality of haptic signal is significantly improved.

The invention discloses a cross-modal pedestrian re-identification method based on intermediate modal and multi-attention fusion, and the method comprises the steps: firstly carrying out the modal preprocessing of a visible light image and an infrared image, and inputting the images into a nonlinear modal converter; the converter comprises a bimodal information encoder and a unified channel decoder, and is used for generating an intermediate modal image with modal alignment; then, jointly inputting the generated intermediate modal image and the original visible light and infrared image into a backbone network, and extracting part aggregation features through an intra-modal multi-attention fine-grained feature aggregation module; furthermore, the distribution consistency of the middle modal image is restrained through a loss function of unified modal alignment, the inter-modal alignment effect is optimized, and the cross-modal recognition performance is improved. According to the method, image-level nonlinear mapping and a multi-attention mechanism are utilized, the problems of modal difference and intra-modal heterogeneity between visible light and infrared images are effectively relieved, and the discrimination and robustness of pedestrian identity features are enhanced.

The invention discloses an image semantic communication method and device based on semantic entropy and CSI non-intrusive fusion. The method comprises the following steps: determining estimated CSI based on an original pilot signal and a received pilot signal; when the semantic information entropy of the image semantic information output by the channel decoder is greater than or equal to the dynamic entropy threshold of the channel, performing feature extraction on the estimated CSI to obtain the feature information of the estimated CSI; respectively calculating a value vector, a key vector and a query vector of the semantic information and the feature information of the image, and carrying out vector fusion to obtain enhanced feature representation; performing semantic decoding on the enhanced feature representation to obtain a reconstructed image; according to the method, for a 6G MIMO fading channel, image semantic information and estimated CSI are fused, a more robust semantic feature can be generated, and the feature intrusion problem in a traditional method is avoided.

The application relates to a Gold sequence spread spectrum underwater acoustic modulation and demodulation method against strong multipath interference, relating to underwater acoustic communication. The method comprises the design of a transmitting end modulation and a receiving end demodulation; on the basis of a traditional Gold sequence spread spectrum modulation and demodulation technology, a molecule set method is adopted to divide the traditional GSSS code set into multiple subsets, and different subsets are adopted in the modulation process of each symbol, so that weak correlation is ensured between each symbol. The demodulation process is based on the molecule set modulation method of the modulation process, energy detection is carried out according to the corresponding local GSSS code subset symbol by symbol, the number of bits carried by each symbol is calculated, each bit soft information is calculated, the soft information is input into an interleaver and a channel decoder, and the original information is correctly demodulated. The method overcomes the problem that the traditional GSSS modulation and demodulation technology cannot be correctly demodulated in a strong multipath environment, improves the robustness of the GSSS modulation and demodulation technology to the strong multipath underwater acoustic channel, and realizes stable and robust communication in the strong multipath environment.

The invention relates to a digital semantic communication method and system for voice reconstruction and classification, and the method comprises the steps: carrying out the preprocessing of voice data, and inputting the preprocessed data into a semantic encoder formed based on a double-attention residual mechanism network; after semantic features output by the semantic encoder are processed by the channel encoder, the semantic features are mapped into digital constellation symbols by a modulator adopting a soft quantization method; after transmission through the physical channel, the modulation symbols recover features by the soft decision demodulator, and signal distortion caused by the physical channel is eliminated through the channel decoder; and the demodulated semantic features are respectively input into a semantic decoder with a symmetric architecture to realize voice reconstruction, and input into a voice classifier consisting of a lightweight feature compression network and a classification head to realize voice classification. According to the method, the feature extraction integrity is improved through a double-attention architecture, the gradient fracture problem is solved through soft quantization and soft decision, and the channel robustness is enhanced in combination with progressive signal-to-noise ratio training.

The invention discloses a channel decoding method and device and a channel decoder. The method comprises the following steps: acquiring a received signal, an actual channel noise variance and a plurality of preset channel noise variances corresponding to a plurality of preset time steps; performing matching processing based on the actual channel noise variance and a plurality of preset channel noise variances corresponding to the plurality of preset time steps, and determining an initial time step; determining a de-noising time step sequence based on the initial time step and a preset time step; based on the received signal, the de-noising time step sequence and a preset time embedding converter model, performing de-noising processing to obtain a target original signal; and performing hard decision processing based on a preset threshold and the target original signal to obtain a target decoding value corresponding to the received signal. According to the method, the received signal can be initialized to a proper state of the diffusion time axis, and the received signal is gradually denoised based on the time step sequence by using the time embedding converter model, so that the physical consistency with a forward process is realized, and the decoding accuracy is improved.

A multi-channel decoder circuit associated with a multi-channel decoder system is disclosed. The multi-channel decoder circuit comprises a distributed decoder circuit comprising a set of unit decoder circuits, each unit decoder circuit configured to receive one or more codewords of a plurality of codewords associated with a plurality of input channels, and decode the one or more codewords. The multi-channel decoder circuit further comprises a distribution controller circuit configured to distribute each incoming codeword of the one or more codewords to the respective unit decoder circuit of the set of unit decoder circuits within the distributed decoder circuit, based on determining a currently available unit decoder circuit within the set of unit decoder circuits.

The application discloses a cross-modal source channel joint coding method for coping with variable scenes, relates to the technical field of cross-modal image signal reconstruction, and comprises the following steps: firstly, a cross-modal channel coding optimization scheme of a Transformer encoder is designed to improve the performance and robustness of a channel encoder and a channel decoder; secondly, a cross-modal source coding optimization scheme of haptic-image generation based on a hidden diffusion model is designed to guide image generation by using haptic information in the case of image signal loss; finally, a transfer learning technology is introduced to reduce the additional training cost of the system in the face of variable cross-modal communication scenes, such as variable channel signal-to-noise ratios and different transmission tasks. The source channel joint coding method provided by the application can solve the problems that the receiver cannot well complete image reconstruction and additional model training costs caused by variable channel environments and scenes in the variable scene of cross-modal communication.

The invention belongs to the technical field of SoC design in an integrated circuit. The Crossbar interconnection architecture based on the AMBA AXI protocol comprises a top-layer interface, an arbitration system and a data distribution system, the top-layer interface comprises a slave interface and a master interface, the arbitration system comprises a decoding module and an arbiter, the decoding module comprises a write transaction channel decoder, a read address channel decoder and a read data channel decoder, and the data distribution system comprises a read transaction channel decoder, a read address channel decoder and a read data channel decoder. The data distribution system comprises a write transaction channel occupancy table, a read transaction channel occupancy table and a multi-path data selection module, wherein the read transaction channel occupancy table comprises a read address channel occupancy table and a read data channel occupancy table. Allowing the plurality of master devices to access the slave devices in parallel; as different master devices send communication requests of the same access mode to the same slave device, communication arbitration can be completed in one period; communication mechanisms such as Outstanding, out-of-order, Interleaf and the like in the AXI protocol can be supported and realized.

The application discloses a simple receiving multiple access method based on symbol and bit stream joint distribution, first, a transmitter distributes symbol streams and bit streams of each user based on a hierarchical quadrature amplitude modulation constellation diagram; different interference management schemes are adopted for inter-user interference in different beams and inter-user interference in the same beam to inhibit the influence of inter-user interference on system performance; channel equalization processing is performed on signals received by each user receiver, each user calculates log likelihood ratios of each bit stream allocated to each user according to a bit stream distribution result, and the calculation result is sent into a channel decoder to obtain original bit data. The above method jointly optimizes symbol and bit stream distribution schemes of each user and constellation diagram design, and can realize efficient user access in a large-scale access scene while satisfying a constraint condition of signal processing complexity of the receiver.

The invention relates to a redundant remainder system (RRNS) channel encoder, an RRNS channel decoder, an RRNS channel encoding method, an RRNS channel decoding method, electronic equipment and a storage medium. A mode set corresponding to the RRNS channel encoder comprises k information modes and n-k redundant modes; the RRNS channel encoder comprises: an input module receiving a first k-element remainder vector comprising k information remainder components; the modular operation modules from the first level to the (k-1) th level determine k weights based on the k information remainder components and the k information moduli; the kth-level modular operation module determines n-k check remainder components based on the k information moduli, the k weights and the n-k redundancy moduli; the output module determines and sends a first n-element remainder vector to the target channel according to the k information remainder components and the n-k check remainder components. According to the embodiment of the invention, the implementation complexity of the digital circuit of the RRNS channel encoder can be reduced.

The present invention discloses a multi-user source channel adaptive semantic coding method based on deep learning, comprising the following steps: S1. Constructing a multi-user semantic communication system based on a deep neural network: at the transmitting end, the signal to be transmitted is processed by a deep source encoder and a digital channel encoder, and then superimposed, numbered, and sent; at the receiving end, the received signal is processed by a digital channel decoder, and then processed by a data user decoder and a semantic user decoder to obtain source data and semantic information; S2. End-to-end distortion modeling of data users and semantic users is performed based on data regression; S3. Joint optimization of source channel coding rate, power allocation, and beamforming is performed. The present invention takes into account the multi-user broadcast channel and designs a joint optimization algorithm for source channel rate, power allocation, and beamforming to minimize the weighting and end-to-end distortion of the multi-user system.

The disclosure relates to a 5G communication system or a 6G communication system for supporting higher data rates beyond a 4G communication system such as long term evolution (LTE). Disclosed herein is a method for decoding received signal that includes obtaining an input parameter from the received signal, obtaining a tree structure based on matrix factorization of the channel matrix, pre-processing of the received signal, obtaining an initial linear solution, determining a neighbour of the linear solution based on the modulation order for a layer and a pre-defined number of the neighbour in the layer, creating a search space by pruning the tree structure, applying a decoding technique on the search space, computing a decision metric, determining a likelihood-based metric based on the decision metric, and decoding the received signal by feeding the likelihood-based metric for the at least one bit to a channel decoder.