9results about How to "Reduce communication cost" patented technology

The invention belongs to the technical field of artificial intelligence information security, particularly discloses a neural network privacy reasoning method and device based on winograd sparse convolution and differentiable activation function approximation, and aims at solving the problems that existing neural network privacy reasoning is high in calculation time delay and large in communication cost. And in combination with a multi-granularity weight pruning algorithm, the sparsity of a linear convolutional layer is maximized, and the communication cost required by the convolutional layer with dense multiplication is reduced. Meanwhile, collaborative optimization is carried out on a nonlinear activation function, a new activation function sensitivity and importance quantitative evaluation algorithm is designed, and an adaptive learnable polynomial function is introduced, so that the fitting precision of the activation function is improved, and the polynomial replacement loss is reduced; and the powerful prediction performance of the neural network model is reserved by using a multi-stage knowledge distillation method, and the calculation and communication efficiency of privacy protection neural network reasoning is effectively improved on the premise of not influencing the prediction precision.

The invention discloses an efficient secret state joint neural network inference method. The method comprises the following steps: S100, a data owner C and a model owner # imgabs0 # respectively send privacy data and a privacy model to computational nodes # imgabs1 # and # imgabs2 # in a secret sharing manner; s200, computing nodes # imgabs3 # and # imgabs4 # to run a secret state joint neural network reasoning algorithm, and respectively calculating secret shares Z0 and Z1 of reasoning results by sequentially executing secret state computing layer protocols, the secret state computing layer protocols comprising: S210: a secret state convolution layer algorithm # imgabs5 #; s220, a function layer algorithm # imgabs6 # is activated in a dense state; the invention further discloses an efficient dense state joint neural network inference system. According to the technical scheme, the ReLU algorithm with few communication rounds is designed based on the function secret sharing technology, so that the communication cost required for calculating the activation function is reduced, the convolutional neural network inference sub-protocol is designed by utilizing bilateral comparison and correct symbol discrimination, the inference accuracy is improved, and the inference result is consistent with a plaintext inference result.

The invention discloses a real-time auditing method for privacy protection of data cross-border flow. The method comprises the following steps: initializing an auditing mechanism, and generating public and private key pairs of domestic enterprises and overseas enterprises; calculating an exit data encryption key and an entry data encryption key; the domestic enterprise encrypts the outbound data through the outbound data encryption key to generate an outbound data ciphertext, and the review institution performs secret state review on the outbound data ciphertext; the overseas enterprise cooperates with the review institution to decrypt the review result, and obtains a partial decryption result of the review institution on the outbound data ciphertext when the review result is passed, and the partial decryption result and partial decryption of the overseas enterprise on the outbound data ciphertext are aggregated to obtain outbound data; the overseas enterprise encrypts the entry data through the entry data encryption key to generate an entry data ciphertext, and the review institution performs secret state review on the entry ciphertext; and the domestic enterprise cooperates with the review institution to decrypt the review result, and obtains a partial decryption result of the review institution on the entry data ciphertext when the review result is passed, and the partial decryption result and partial decryption of the domestic enterprise on the entry data ciphertext are aggregated to obtain entry data.

The embodiment of the present specification provides a two-party secure computing method and system. In the two-party secure computing method, a first party sends a computing request to a second party. The second party generates a Boolean circuit and a corresponding garbled table for computing two data, the Boolean circuit including L first input lines and L second input lines, wherein the L first input lines correspond to first data. The second party obtains n second data corresponding to n identifiers, and determines n string groups corresponding to the n second data according to a label string of the L second input lines. The second party encrypts the n string groups by using a target OT protocol of n-of-1, and sends the n encrypted results and the garbled table to the first party. The first party decrypts the n encrypted results by using the target OT protocol, and obtains a target string group corresponding to second data corresponding to a target identifier. The first party decrypts the garbled table by using the target string group, and obtains a computing result of the second data corresponding to the target identifier and the first data.

The application realizes a satellite cloud-oriented distributed file system through a method in the field of distributed systems.The system is composed of five components: a metadata management module, a StorageEngine, a Policy, a client and a satellite topology service module; the metadata management module stores metadata, the Storage Engine is a data storage module, the Policy is a strategy module, the user uses the client to interact with the satellite cloud file system, the satellite topology service module is specially designed for the satellite cloud scene to provide the satellite cloud network topology information for the satellite cloud file system, and the modules communicate with each other through a special file system interface using RPC.The method provided by the application guarantees efficient reading and reliability of metadata for the satellite communication scene, maximally reduces the communication cost and energy loss of replica placement by using the network topology characteristics of the satellite cloud, and maximally reduces the access delay.

This invention discloses an adaptive distributed support vector regression method to combat Byzantine faults. First, the objective function of the SVR model is smoothed to construct an empirical risk minimization objective function. Then, with t=1, the master machine obtains initial parameter estimates on its local data. In the t-th iteration, a MOM-based aggregation method is used to obtain the MOM estimate, which is then used as the global gradient estimate. Subsequently, a robust global gradient is obtained based on the global gradient estimate. After obtaining the robust global gradient, the master machine updates the global parameters using gradient descent. Finally, it checks whether the convergence condition is met. If the convergence condition is not met, the above steps are repeated; otherwise, the globally optimal parameter estimate is obtained. By introducing a robust aggregation mechanism based on the MOM method, it can effectively suppress the influence of abnormal machines, significantly improving the stability and robustness of the system while ensuring prediction accuracy. It can be widely applied to various practical distributed systems.

The invention discloses a password leakage detection method for protecting an account from large-scale voucher filling attack, which comprises the following steps: a client selects a prefix of a hash value of a to-be-detected voucher and a blind value corresponding to the prefix to form a request report, and sends the request report to a service provider; after receiving the request report, the service provider performs modular exponential operation on the blind value in the request report; searching a corresponding block from a database by using prefixes of all hash values in the request report, forming a reply report by a modular exponent operation result and the block, and feeding back the reply report to the client; and after the client receives the reply report, performing modular index operation on a modular index operation result in the reply report, if the modular index operation result is equal to the blind value in the database establishment stage, outputting that the voucher of the to-be-detected voucher is leaked, and otherwise, outputting that the voucher is not leaked. The password security of the account is effectively protected.

The invention discloses an adaptive differential privacy protection and communication compression joint optimization method and system for industrial federated learning, and belongs to the technical field of communication compression. Collecting equipment data to construct a local data set; performing local training and gradient calculation by using the local data set to obtain a local gradient; differential privacy protection is carried out to obtain a privacy protection gradient after noise addition; performing compression processing on the privacy protection gradient to obtain compressed data; the client uploads the compressed data to a server; the server performs inverse quantization reconstruction on the gradient of each client according to the received masks and quantization parameters to obtain a reconstruction ladder; and carrying out gradient aggregation, privacy budget allocation and accounting on the reconstruction ladder. According to the method, higher model precision and communication efficiency can be realized, and the signal-to-noise ratio is further optimized.

The application discloses a visibility determination method and device, equipment and a storage medium, and relates to the technical field of databases. The method comprises the following steps: in response to a first visibility judgment event of a target transaction being triggered, obtaining a current cache value from transaction public cache data of a node, and obtaining a first cache commit sequence number snapshot, wherein the transaction public cache data comprises a global commit sequence number in a cached metadata node; determining a first target commit sequence number snapshot of the target transaction according to the first cache commit sequence number snapshot; performing initial judgment on the visibility according to the first target commit sequence number snapshot; and in the case that the initial judgment result is visible, determining that the visibility judgment result of the first visibility judgment event is visible. Through the above technical solution, the communication cost with the metadata node can be reduced, and the efficiency of the visibility determination and the transaction execution efficiency can be improved.