30results about How to "Reduce communication overhead" patented technology

The invention discloses a federal recommendation privacy protection method based on ring signature and selective aggregation, which realizes anonymization privacy protection and selective aggregation in a federal learning process through an innovative double-server architecture and a hierarchical security mechanism. The method comprises the following steps: 1) deploying an aggregation server and a security server in a system initialization stage, and completing parameter initialization and key distribution; 2) the client executes local model training, and anonymization privacy protection uploading is realized by adopting a ring signature technology; 3) the security server performs selective aggregation of credibility verification, and completes grouping and parameter fusion based on user similarity; and 4) ensuring the completeness and source credibility of model updating through a hierarchical security parameter distribution mechanism. While the recommendation precision is ensured, the user identity leakage and privacy tracking are effectively prevented, the comprehensive security guarantee is provided for the federal recommendation system, and the method has the advantages of low calculation overhead and high communication efficiency.

This invention discloses a method and system for protecting privacy in wearable devices for human activity recognition, aiming to solve the problems of existing methods struggling to balance privacy and utility, and high resource consumption on the edge. The method first acquires multi-channel time-series signals from the wearable device and constructs fixed-length feature vectors. Based on a sensitive attribute inference model and the number of high-risk features, it divides the device into high-risk and low-risk feature sets. Only the high-risk feature sub-vectors are subjected to perturbation noise to generate perturbation latent representations, which are then uploaded to an edge server along with the low-risk feature set. The edge server generates the final feature vectors and assesses the risk of privacy leakage. An offline evaluation is used to construct a privacy-utility trade-off, dynamically adjusting the perturbation dimension and noise amplitude. This invention achieves targeted protection of high-risk features, maximizing the preservation of key recognition information, reducing edge computing power and communication overhead, adapting to the differentiated needs of different scenarios, and balancing privacy security and activity recognition performance.

This invention relates to the fields of federated learning and cryptography, specifically to a secure three-party aggregation method for federated learning in the Industrial Internet of Things (IIoT). The method includes an initialization phase, a training phase, an online phase, and a verification phase. By setting up an architecture with one honest server and two aggregation servers, each server receives a portion of the secret share of local model gradient parameters sent by various IoT devices, thus protecting the privacy of local data. The honest server assists the two aggregation servers in interactively executing a three-party weight calculation protocol to calculate the secret share of the aggregation weights. They also interactively execute a three-party multi-weight aggregation protocol to calculate the secret share of the gradient parameters of the current aggregation model. Furthermore, a linear homomorphic hashing method is used to verify the correctness of the gradient parameters of the current aggregation model. This method solves the problems of low efficiency, weak security, and low robustness in federated learning during model training and aggregation, and can promote data flow and utilization in IoT scenarios.

The application discloses a kind of unmanned vehicle logistics directional relay network system and its coordination method, belong to intelligent logistics and unmanned vehicle application technical field, the system includes: multiple cellular airspace units, cellular airspace unit is constituted continuous grid in service area;Deployed in the center of each cellular airspace unit one micro hub, as the ground logistics node of corresponding unit;Wherein, each micro hub includes one terminal nest and at least two relay nests;Terminal nest is used to provide the storage and withdrawal interface of user goods;Each relay nest is arranged around terminal nest, and each relay nest is pre-allocated and fixedly responsible for the bidirectional goods transfer task between the unit where the micro hub is located and a particular adjacent unit;Terminal nest and each relay nest are connected by internal logistics transmission device 103.The application fundamentally simplifies scheduling and management complexity by deep coordination of cellular airspace grid, heterogeneous node division of labor, directional routing rules and high automation.

The invention belongs to the technical field of wireless digital broadcast communication security, and relates to an Auracast broadcast privacy protection system supporting key distribution and access control, which comprises a channel establishment and authentication module for establishing an out-of-band security control channel; the precursor key issuing module is used for issuing a static session precursor key to the target receiving equipment and the broadcast source equipment through channel encryption; the session key generation module is used for deriving a slice session key in combination with the public synchronization factor and the precursor key; the data packaging and sending module is used for encrypting the audio by using the slice session key, assembling a data frame and sending the data frame; the local key synthesis module is used for extracting a synchronization factor and locally synthesizing a slice session key; the audio stream analysis and output module is used for decrypting the data and managing a key life cycle; and the secret key life cycle management module is used for monitoring the state through a channel and executing erasing. According to the invention, the problem of realizing low-delay dynamic encryption and instant permission revocation without relying on periodic key redistribution in wireless broadcasting is solved.

According to the chip verification system, the chip verification method and the related equipment, in the chip verification system, the verification environments of a plurality of first tested designs are integrated into the first verification environment, and compared with the mode that the corresponding verification environments are set for the plurality of first tested designs, the number of the verification environments can be reduced, and the verification efficiency is improved. Therefore, the integration difficulty between verification environments can be reduced, and the management and debugging difficulty can be reduced. Moreover, a first interconnection assembly located in the first verification environment or the second verification environment realizes the connection between the first tested design and the second test through the first application program interface, and a second interconnection assembly located in the first verification environment or the second verification environment realizes the connection between the first tested design and the second test through the second application program interface. The connection between the first reference model and the second reference model is realized through the second application program interface, so that the communication overhead between the first verification environment and the second verification environment is reduced, and the simulation verification efficiency is improved.

The invention relates to a video memory virtual access method and device of a GPU model and computer equipment. The method comprises the following steps: receiving an access request initiated by a CPU (Central Processing Unit) in the virtual machine, wherein the access request comprises indication information of a target base address register; based on the indication information and a preset association relationship, routing the access request to a target address range in a simulated physical address space; wherein the preset association relationship comprises a corresponding relationship between an address range in a simulated physical address space and a base address register of the GPU model, and a mapping relationship between the address range in the simulated physical address space and a host memory allocated for a video memory area of the GPU model; the GPU model is a software module for simulating a physical GPU in a virtual environment. By adopting the method, the access speed can be improved.

The present disclosure provides a new energy station power prediction method and system based on longitudinal federal decision tree, relates to the field of federated learning and information security technology, and comprises the following steps: acquiring real-time environmental characteristic data; inputting the environmental characteristic data into a power load prediction model to output a predicted power value; wherein the power load prediction model is a longitudinal federal decision tree model, and the training process of the longitudinal federal decision tree model comprises the following steps: each participant generates a key and a hyperparameter for each column of characteristics according to a security requirement, generates a random topological mapping for each column of characteristics, constructs a nonlinear conversion function based on each mapping relationship, encrypts the data in a cascade manner using the generated nonlinear conversion function and a lightweight order-preserving encryption, concentrates the encrypted data in a coordination party for ciphertext state model training, and obtains the longitudinal federal decision tree model after the training is completed. The present disclosure can improve the accuracy of joint power prediction modeling and realize more accurate scheduling optimization.

The PUF-based lightweight smart grid authentication method provided in the application comprises: generating first anonymous identity information of a smart meter, generating second anonymous identity information of a gateway, and sending the first anonymous identity information to the smart meter and the second anonymous identity information to the gateway; generating a first challenge-response pair based on a registration request of the smart meter by using a PUF function; the first challenge-response pair comprises first challenge information and first response information; generating a second challenge-response pair based on a registration request of the gateway by using the PUF function; the second challenge-response pair comprises second challenge information and second response information; performing first authentication on the smart meter and the gateway based on the challenge-response pair generated by combining the PUF function with the first anonymous identity information and the second anonymous identity information; and performing second authentication on the smart meter and the gateway based on the challenge-response pair and a random number generated by each of the two parties. The method reduces the calculation cost and the communication cost, effectively resists desynchronization attacks, and realizes forward and backward security.

The application discloses a novel flexible manufacturing system dynamic scheduling method based on multi-agent reinforcement learning, relates to the technical field of intelligent manufacturing, and comprises the following steps: constructing a cross-domain knowledge graph driven state space, fusing process knowledge and equipment capacity to perform semantic embedding; establishing an adaptive evolution intelligent agent collaborative network of a space-time heterogeneous graph, and dynamically adjusting a topological structure according to task dependence and equipment coupling; adopting a progressive collaborative value decomposition network to optimize a strategy and accurately realize credit distribution; and constructing a multi-scale closed-loop compensation decision mechanism to form deep coupling and closed-loop feedback at the equipment layer, the production line layer and the system layer, so that the scheduling efficiency and robustness of the flexible manufacturing system are significantly improved.

This application provides a map construction method, apparatus, device, and computer storage medium. The method involves encapsulating collected environmental state information into a first link state data packet according to a preset format, and broadcasting this first link state data packet to a robot network via a flooding mechanism. The robot network includes multiple intelligent robots, among which the first intelligent robot is included. The preset format includes at least a local map state field. The method also receives and parses second link state data packets uploaded by other intelligent robots in the robot network (excluding the first intelligent robot), where the second link state data packets are encapsulated using the preset format. A local global environmental state database is updated based on the first and second link state data packets, and a global map is constructed based on this database. The method provided by this application ensures consistency in global environmental perception and avoids map conflicts.

The invention discloses an operation quality evaluation method, device and equipment based on Beidou positioning, and a storage medium, and relates to the technical field of Beidou navigation and Internet of Things data fusion, and the method comprises the steps: carrying out the time sequence alignment processing of Beidou positioning terminal track data, the facility state data of Internet of Things sensor nodes, and mobile terminal image data, the aligned multi-modal data is obtained; extracting a space-time distance value between each track point in the track point position sequence data of the positioning terminal and each facility coordinate in the geographical coordinate set data of the facility from the aligned multi-modal data; determining associated facility data corresponding to each trajectory point according to the space-time distance value, and constructing a space-time constraint model of the trajectory and the facility point location according to the associated facility data; quantitative evaluation is carried out on the quality of the mobile supervision object based on the space-time constraint model, a space-time correlation evaluation result is obtained, and the space-time matching precision and correlation analysis efficiency of the Beidou positioning terminal and the Internet of Things facility in a complex urban environment are improved.

The application relates to the technical field of semiconductors, in particular to a semiconductor device performance parameter automatic detection method, which comprises the following steps: obtaining an initial detection working interval of a semiconductor device; performing initial random sampling in the initial detection working interval to obtain an initial sampling point set; performing fitting processing on the initial sampling point set by using a Gaussian process regression algorithm to construct a proxy model; and performing analysis processing on the proxy model by using a batch Bayesian optimization algorithm. The asynchronous hardware pipeline capability of a source measurement unit is utilized to perform high-speed parallel sampling, the communication overhead between an upper computer and an instrument is reduced, parallel processing of calculation and measurement tasks is realized, and physical test time consumption is compressed. In this way, the accurate performance parameters can be quickly and safely converged by using far fewer measurement points than in a traditional scanning mode, so that the test efficiency and safety are improved under the premise of ensuring the accuracy.

The invention relates to the technical field of data security, in particular to an energy power data privacy calculation protection aggregation method based on cloud calculation, which comprises the following steps of: acquiring a power consumption time sequence of each user in each jurisdiction, and calculating the power consumption time sequence of each user in each jurisdiction according to the difference between the power consumption time sequences of different users in each jurisdiction; the power consumption complexity of users in each jurisdiction and the power consumption data sensitivity of each user in each jurisdiction are obtained, so that the power consumption data encryption level coefficient of each user in each jurisdiction is obtained, the power consumption time sequence of all users in all jurisdiction is subjected to hierarchical encryption, and energy power encryption data is obtained. And carrying out energy power encryption data aggregation based on cloud calculation. According to the energy power data privacy calculation protection aggregation method based on cloud calculation, the aggregation value of the data can be reserved to the maximum extent on the premise of effectively protecting the privacy of the user.

The technical scheme of the invention discloses a system for constructing a high-precision user behavior portrait through an improved federal learning algorithm, and the system is characterized in that the system comprises a multi-source privacy protection data obtaining module; a federated learning modeling module based on dynamic weighting and contribution proving; a multi-dimensional user behavior portrait module; a personalized vehicle network interaction strategy generation module; a dynamic execution and fair excitation module; and a feedback and model self-optimization module. According to the method, deep analysis and portrait construction can be carried out on historical data of user charging and discharging behavior habits, a high-precision and high-coverage user behavior portrait model is generated through an improved federated machine learning algorithm, the real-time income of users is improved, the user viscosity is enhanced, and the power grid stability and the renewable energy consumption rate are further improved.

This invention belongs to the field of communication technology and discloses a federated learning method based on partial client participation and power constraints. The method includes: employing biased client selection, setting client addressability and continuously reporting to a central server, allowing the central server to understand the state of each client during iteration; and allocating power across communication time slots by adding a total power constraint, enabling devices to transmit their local updates to the server. This invention addresses the problems of existing technologies by reducing communication costs while maintaining model convergence performance. The proposed design can mitigate the negative impacts caused by imperfect channel conditions. Convergence of the learning process is slowed by power control, while biased client selection accelerates convergence, balancing model aggregation. This invention can guarantee fast convergence under harsh wireless conditions with low signal-to-noise ratios, and is cost-effective in terms of time and energy savings in wireless FL systems.

The embodiment of the present specification provides a multiplication execution method, a data preparation method and device based on SMPC. The multiplication execution method comprises: a first party obtains a first random multiplier and a first addition fragment according to an execution round i, the first random multiplier is generated by a first random number generator based on the execution round i, the first addition fragment is calculated by a third party based on the first random multiplier, a second random multiplier and a second addition fragment, the second random multiplier and the second addition fragment are generated by a second random number generator deployed in the third party based on the execution round i; the first party sends the execution round i to a second party; the second party obtains the second random multiplier and the second addition fragment generated by the second random number generator deployed by the second party based on the execution round i according to the execution round i; the first party uses the first random multiplier and the first addition fragment, and the second party uses the second random multiplier and the second addition fragment to jointly execute the ith multiplication based on SMPC.

This invention discloses a distributed inference method for large language models based on intra-layer parallelism and communication quantization. The method relates to the field of distributed inference technology and includes the following steps: computational resource efficiency assessment and control; network environment adaptation assessment; and inference result generation and output. This invention, in distributed inference of large language models, first assesses the efficiency of computational resource utilization; if the efficiency is insufficient, it optimizes synchronous communication and load balancing; then it assesses the network environment adaptability; if the adaptability is insufficient, it optimizes the network topology and transmission protocol; finally, it outputs the inference result. This achieves efficient resource utilization and adaptive control in heterogeneous network environments, improving the practicality of distributed inference for large language models and solving the problem in existing technologies where the existing parallel inference schemes for large language models cannot achieve a balance between high computational efficiency and low network dependence, resulting in low practicality of distributed inference for large language models.

The application discloses a kind of based on in-network aggregation's distributed model training task unloading method and device. Among them, the method is executed by the controller of data center network, the data center network includes several servers and programmable switch, including: according to the real-time network topology information in data center network, network resource information is constructed network physical topology model for each incoming distributed model training task;On the basis of the physical topology model, logical aggregation model is constructed by omitting the forwarding information of gradient data;According to the logical aggregation model and the available aggregation resources in the data center network, the task unloading strategy of distributed model training task is determined.The application is jointly optimized to worker, PS's assignment and in-network aggregation routing, to further accelerate model training, and reduce the communication overhead generated by training.

The present application relates to the field of industrial internet, and provides a federated learning method and system for training of an industrial internet edge sensor device. The method comprises: distributing global model parameters and aggregated momentum to a selected subset of edge sensor devices, and initializing to obtain local model parameters and local momentum; dividing the local model parameters into multiple parameter blocks, performing momentum updating on the current batch gradient through a momentum accumulation algorithm to obtain a momentum matrix; performing orthogonalization processing on the momentum matrix through a Newton-Schulz iteration algorithm to obtain an orthogonalization matrix; performing momentum gradient alignment scoring on the multiple parameter blocks to obtain a smooth alignment score; performing weighted updating on the orthogonalization matrix to obtain updated parameter blocks; and aggregating the parameter block increments and the updated momentum by the server to obtain updated global model parameters and updated global aggregated momentum. The present application reduces the communication frequency and improves the training efficiency of the edge sensor device.

The application discloses a blockchain sharding method combining a node contribution mechanism and a two-round consensus, and comprises the following steps: completing blockchain node initialization and random sharding by using a verifiable random function based on an elliptic curve algorithm; constructing a multi-dimensional dynamic contribution value model based on discretization calculation, comprehensively considering the voting consistency of nodes in the current round, the historical contribution average value and the identity additional score, so as to select a leader node; executing a two-round consensus protocol; and triggering a dynamic replacement mechanism for a malicious leader node or a malicious shard by comparing the consistency of a pre-commitment and the number of valid signatures of a block header.

The application discloses a method and device for encrypted communication and key error correction, which comprises the following steps: a sender and a receiver each collect and store a same-length key, respectively denoted as k A and k B ; then, the sender performs first channel coding on a sending information M A to obtain a bit sequence C A of the same length as the key k A ; then, the sender performs XOR operation on C A and k A to obtain D A , and then performs second channel coding on D A and sends the second channel coding result to the receiver; the receiver performs first decoding error correction on a received signal to obtain D B , then performs XOR operation on the checked D B and the key k B of the receiver to obtain C B ; then, the receiver performs second decoding error correction on C B to obtain M B , and re-codes the checked receiving information M B to obtain C A ; finally, the receiver performs XOR operation on C A and D B to obtain k A . The application can realize encrypted communication and key error correction by only one communication in the case that the keys of the sender and the receiver are inconsistent.

According to the method for carrying out number multiplication operation under the secret state condition and the computing device, a first participant divides N to-be-computed scalar quantities into m groups, arrangement is carried out on each group of to-be-computed scalar quantities to obtain a corresponding to-be-coded vector, each to-be-coded vector is obtained by splicing m identical first vectors and second vectors, and m is an integer greater than or equal to 1; the first vector and the second vector are each composed of a to-be-calculated scalar, the m to-be-coded vectors are coded to obtain m plaintext polynomials, non-zero coefficient items of the m plaintext polynomials are extracted and spliced to obtain a fusion polynomial, the fusion polynomial is encrypted by using a homomorphic encryption algorithm to obtain a fusion ciphertext, and the fusion ciphertext is used for decoding the fusion ciphertext. Transmitting the fused ciphertext to the second participant; and the second participant determines a cryptographic operation result of the number multiplication operation according to the fused ciphertext and m combined polynomials by using a homomorphic encryption algorithm, the combined polynomials are obtained by encoding combined vectors, and the combined vectors are obtained by arranging mutually independent to-be-calculated vectors according to the to-be-encoded vectors, so that the overhead of the number multiplication operation can be reduced.

This invention provides a parallel split learning method, apparatus, and system based on adaptive gradient correction, relating to the field of distributed machine learning computing technology. The invention receives intermediate layer gradients uploaded by each client, divides them into deep and shallow gradient components, calculates the consistency and comprehensive ranking score of each client's gradient, and generates an ordered gradient queue. Then, gradients are introduced one by one for weighted aggregation, directional stability judgment, and noise truncation to update the aggregation center. The comprehensive ranking score of high-confidence gradients is selected from the ordered gradient queue, and the global confidence score is calculated to dynamically adjust the global model parameters. A global reference direction is output and sent to each client, enabling each client to correct its gradients according to the global reference direction. The residual after gradient correction is used as a regularization constraint for the next round of local training optimization. This invention effectively suppresses gradient conflicts, improves model convergence speed and robustness, and reduces server computational overhead.

The application discloses a medical trajectory prediction system and method based on federated learning and differential privacy, aiming at the problems of difficult balance between privacy protection and performance, and large communication overhead in cross-institutional data collaborative training. The application embeds privacy constraints, performance requirements, data and calculation budgets into neural network design, deduces the optimal matching relationship between model hyperparameters and differential privacy parameters based on the differential privacy large model scaling law, adopts a double-branch decoupling model, configures differentiated privacy budgets and noise intensities for disease prediction and trajectory prediction tasks, and introduces differentiated coefficients and comorbidity protection coefficients for medical scene optimization. The communication overhead is reduced, and data heterogeneity is coped with through parameter increment uploading and three-dimensional weighted aggregation mechanism, and a closed-loop verification is formed in combination with a dynamic verification module. The application realizes high-precision, strong-privacy and low-overhead medical trajectory prediction, and is suitable for clinical auxiliary diagnosis and disease development trend prediction in a multi-institutional collaborative environment.

The invention belongs to the technical field of block chains, and particularly discloses an asynchronous Byzantine consensus fault tolerance method based on consistent broadcast and a fast path. According to the application, the transaction is packaged, and a delivery block is determined in a mode of broadcasting the packaged block to other consensus nodes by adopting a linear consistency broadcast protocol; the local timeout device is used for starting other consensus nodes; if the fast message from the target consensus node is received before timeout, verifying the correctness of the fast message; and when the verification is passed, marking the delivery block according to the submittable label, and submitting the marked delivery block through the fast path. Through the above mode, the linear consistency broadcast protocol is adopted to replace a conventional reliable broadcast protocol, the communication complexity of block broadcast is reduced, the fast path is set, the fast path is preferentially performed under the condition that the network is good or no Byzantine behavior exists, and the number of communication rounds is reduced, so that the communication overhead and delay of consensus can be effectively reduced.

The application discloses an edge data integrity verification method and system for electronic health records, relates to the fields of electronic health records (EHR), edge computing and data integrity, and comprises the following steps: before generating a request, an AppVend generates relevant keys, initializes parameters, and discloses system public key information; the request is completed global data verification and independent data verification; after receiving the request, an edge server detects whether EID and DID in the request are correct; the edge server constructs a local minimum multi-branch tree, after forming a signature, sends a response as a proof to the AppVend for re-aggregation and verification; the AppVend receives the request and checks whether EID and DID contained in the response are correct, and if not, returns a re-computed proof; the AppVend verifies an integrity verification response message returned by the edge server to the AppVend, and locates a damaged edge server and a data copy.

The application discloses an elastic federated learning method and system based on elastic weight consolidation and dynamic aggregation, which comprises the following steps: initializing model parameters, setting the maximum communication round number and the preset local training step number of the client in each communication; the server sends the current global model parameters to the selected client, and the client receives the global model parameters as the local model and updates the local model after receiving the global model parameters; the client constructs an optimization objective function and performs iterative training; the client calculates the update amount for the local model which does not complete the local training step number; the client performs sparse processing and ternary quantization processing on the calculated update amount and sends the update amount to the server; the server adjusts the aggregation weight, updates the global model, and judges whether the maximum communication round number is reached; if the maximum communication round number is not reached, the updated global model parameters are sent to the selected client again, and the local model is continuously updated; if the maximum communication round number is reached, the final global model is output.

The invention discloses a security key distribution system, method and chip based on a physical random entropy source, and the system takes a high-credibility physical random source as a root of trust, and safely pre-distributes an entropy material to a terminal dedicated chip through a credible charging station network to form a hardware-bound local entropy pool; during communication, the two parties can synchronously extract the same entropy segment from respective entropy pools without interaction based on a one-time seed and a preset synchronization algorithm, and then a session key is generated; the system supports online updating of a local entropy pool through quantum key distribution; the system constructs a complete trust transfer chain from a physical world to a digital world, solves the fundamental problem of key initialization and distribution of a large-scale cryptosystem, and has high security and high deployment efficiency.