Model issuing method and system

Abstract

The invention discloses a model issuing method and system, and belongs to the technical field of cloud edge cooperative computing. The method comprises the steps of obtaining a model mirror image of a to-be-issued target model, wherein the target model is a trained model; compressing the model mirror image to obtain a mirror image compression package, and setting and storing a decompression password of the mirror image compression package in the compression process of the model mirror image; the mirror image compressed package is sent to a node management platform, the node management platform sends the mirror image compressed package to a target edge node, and the target edge node is used for decompressing the mirror image compressed package through a decompression password to obtain a target model. According to the method, the model mirror image is compressed, so that the data storage space is saved, the data transmission quantity and broadband occupation are reduced, the efficiency of the model issuing process is improved, the model issuing time is shortened, meanwhile, the mirror image compression package and the decompression password are issued differently, and the model issuing safety is enhanced.

Description

Technical Field

[0001] The present application relates to the field of cloud-edge collaborative computing technology, and in particular to a model distribution method and system. Background Art

[0002] Cloud-edge collaborative computing is a computing method that closely combines cloud computing with edge computing. It realizes the sinking of cloud computing by reasonably allocating computing tasks between cloud platforms and edge nodes, and extends cloud computing and cloud analysis to edge nodes. Among them, the cloud platform is mainly responsible for optimizing the model and sending the model to the edge nodes. However, due to the scattered distribution and large number of edge nodes, the cloud platform has a large demand for model sending. At the same time, as the size of the sent model increases, the amount of data transmission in the network increases, making the model sending process slow. Therefore, a model sending method is urgently needed. Summary of the invention

[0003] The present application provides a model distribution method and system, which can improve the efficiency of the model distribution process and enhance the security of the model distribution. The technical solution is as follows:

[0004] On the one hand, a model distribution method is provided, which is applied to a cloud platform, and the method includes:

[0005] Obtain a model image of a target model to be issued, where the target model is a trained model;

[0006] Compressing the model image to obtain an image compression package, and setting and storing a decompression password for the image compression package during the compression process of the model image;

[0007] The image compression package is sent to a node management platform, and the node management platform sends the image compression package to a target edge node. The target edge node is used to decompress the image compression package using the decompression password to obtain the target model.

[0008] Optionally, sending the image compression package to a node management platform includes:

[0009] Encrypting the image compression package;

[0010] The encrypted image compression package is sent to the node management platform.

[0011] Optionally, the method further comprises:

[0012] Receiving a configuration information reading instruction sent by the node management platform through a configuration information reading interface, wherein the configuration information reading instruction carries an identifier of the node management platform;

[0013] Verifying the identity of the node management platform based on the identifier of the node management platform;

[0014] When the identity authentication of the node management platform is passed, the configuration information of the target model is sent to the node management platform.

[0015] Optionally, the method further comprises:

[0016] Receiving a password acquisition instruction sent by the target edge node through a password reading interface, wherein the password acquisition instruction carries an identifier of the target edge node;

[0017] Verifying the identity of the target edge node based on the identifier of the target edge node;

[0018] When the identity authentication of the target edge node is passed, the decompression password is sent to the target edge node.

[0019] On the other hand, a model delivery method is provided, which is applied to a node management platform, and the method includes:

[0020] Receiving a compressed image package sent by a cloud platform, wherein the compressed image package is obtained by compressing the model image of the target model to be sent;

[0021] Obtaining a configuration file of the target model, wherein the configuration file indicates that the target model needs to be sent to a target edge node;

[0022] According to the configuration file, the image compression package is sent to the target edge node, and the target edge node decompresses the image compression package using a decompression password to obtain the target model, where the decompression password is set during the compression process of the model image.

[0023] Optionally, the receiving of the image compression package sent by the cloud platform includes:

[0024] Receiving the encrypted image compression package sent by the cloud platform;

[0025] Decrypting the encrypted image compression package to obtain the decrypted image compression package;

[0026] The step of sending the image compression package to the target edge node according to the configuration file includes:

[0027] According to the configuration file, the decrypted image compression package is sent to the target edge node.

[0028] Optionally, the obtaining a configuration file of the target model includes:

[0029] Call the configuration information reading interface;

[0030] Sending a configuration information reading instruction to the cloud platform through the configuration information reading interface, wherein the configuration information reading instruction carries the identifier of the node management platform;

[0031] Receive the configuration information of the target model sent by the cloud platform, and generate a configuration file of the target model based on the configuration information of the target model.

[0032] On the other hand, a model delivery method is provided, which is applied to a target edge node, and the method includes:

[0033] Receive an image compression package sent by a node management platform, where the image compression package is obtained by compressing the model image of the target model to be sent;

[0034] Reading a decompression password, where the decompression password is set during the compression process of the model image;

[0035] The image compression package is decompressed according to the decompression password to obtain the target model.

[0036] Optionally, the reading the decompression password includes:

[0037] Call the password reading interface;

[0038] Sending a password acquisition instruction to the cloud platform through the password reading interface, wherein the password acquisition instruction carries the identifier of the target edge node;

[0039] Receive the decompression password sent by the cloud platform.

[0040] On the other hand, a model distribution system is provided, which includes a cloud platform, a node management platform and a target edge node to implement the steps of the model distribution method described above.

[0041] The technical solution provided by this application can at least bring the following beneficial effects:

[0042] The cloud platform compresses the model image of the model to be issued to obtain an image compression package, sets a compression password during the compression of the model image, and stores the encrypted decompression password. After that, the cloud platform sends the image compression package to the node management platform, which then sends it to the corresponding edge node. The edge node then obtains the decompression password from the cloud platform, and then decompresses the image compression package according to the decompression password. This method saves data storage space, reduces data transmission and bandwidth occupancy, improves the efficiency of the model issuance process, and reduces the time of model issuance by compressing the model image. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 It is a schematic diagram of an implementation environment provided by an embodiment of the present application;

[0045] Figure 2 This is a flow chart of a model issuing method provided in an embodiment of the present application;

[0046] Figure 3 It is a structural schematic diagram of a model sending device provided in an embodiment of the present application;

[0047] Figure 4 It is a structural schematic diagram of another model sending device provided in an embodiment of the present application;

[0048] Figure 5 It is a structural schematic diagram of another model sending device provided in an embodiment of the present application;

[0049] Figure 6 It is a structural diagram of a terminal device provided in an embodiment of the present application;

[0050] Figure 7 It is a structural diagram of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0052] Before explaining in detail the model distribution method provided in the embodiment of the present application, the application scenarios and implementation environment involved in the embodiment of the present application are first introduced.

[0053] Cloud-edge collaborative computing is a computing method that closely combines cloud computing with edge computing. It realizes the sinking of cloud computing by reasonably allocating computing tasks between cloud platforms and edge nodes, and extends cloud computing and cloud analysis to edge nodes. Among them, the cloud platform is mainly responsible for optimizing the model and sending the model to the edge nodes. However, due to the scattered distribution and large number of edge nodes, the cloud platform has a large demand for model sending. At the same time, due to the increasing size of the sent models, the amount of data transmission increases, making the model sending process slow. For example, in the field of oil extraction, oil production sites are usually scattered and large in number. When some parameter models for debugging equipment need to be sent to these oil production sites, the model sending process of the cloud platform is slow, and the user experience is not good.

[0054] Based on this, the embodiment of the present application provides a model distribution method, in which the cloud platform compresses the model image of the model to be distributed to obtain a compressed image package, sets a compression password during the compression of the model image, and stores the encrypted decompression password. After that, the cloud platform sends the compressed image package to the node management platform, which is then sent to the corresponding edge node by the node management platform, and the edge node obtains the decompression password from the cloud platform, thereby performing operations such as decompression of the compressed image package according to the decompression password. This method saves data storage space, reduces data transmission volume and bandwidth occupancy, improves the efficiency of the model distribution process, and reduces the time of model distribution by compressing the model image. At the same time, the compressed image package and the decompression password are not distributed at the same time, which enhances the security of model distribution.

[0055] The model delivery method provided in the embodiment of the present application can be applied to a model delivery system, which includes a cloud platform, a node management platform and an edge node. Figure 1 A schematic diagram of the structure of an implementation environment provided for an embodiment of the present application, the implementation environment includes a cloud platform 101, multiple node management platforms 102 (one node management platform is schematically illustrated in the figure), and multiple edge nodes 103 (one edge node is schematically illustrated in the figure). The cloud platform 101 can be communicated with the node management platform 102 and the edge node 103 respectively, and the node management platform 102 can be communicated with the edge node 103. The communication connection can be a wired connection or a wireless connection, which is not limited in the embodiment of the present application.

[0056] Among them, the cloud platform and node management platform can be an independent server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms, or a cloud computing service center.

[0057] The edge node can be a terminal device, which can be any electronic product that can interact with the user through one or more methods such as keyboard, touchpad, touch screen, remote control, voice interaction or handwriting device, such as PC (Personal Computer), mobile phone, smart phone, PDA (Personal Digital Assistant), wearable device, PPC (Pocket PC), tablet computer, smart car machine, smart TV, smart speaker, etc.

[0058] Those skilled in the art should understand that the above-mentioned servers and terminal devices are only examples, and other existing or future servers or terminals that may appear, if applicable to the embodiments of the present application, should also be included in the protection scope of the embodiments of the present application and are included here by reference.

[0059] It should be noted that the application scenarios and implementation environments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Ordinary technicians in this field can know that with the emergence of new application scenarios and the evolution of the implementation environment, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0060] Next, the model distribution method provided in the embodiment of the present application is explained in detail.

[0061] Figure 2 This is a flow chart of a model distribution method provided in an embodiment of the present application. Please refer to Figure 2 , the method comprises the following steps.

[0062] Step 201: The cloud platform obtains a model image of a target model to be issued, where the target model is a trained model.

[0063] The cloud platform uses the trained model as the target model, and then obtains the model image of the target model, which includes description information of the target model and / or the application environment of the target model, wherein the application environment is the software environment for the target model application. The model image can be a basic model image or an application model image, or a combination of a basic model image and an application model image, which is not limited in the embodiments of the present application.

[0064] When the model image is a base model image, the model image mainly includes description information of the application environment of the target model; when the model image is an application model image, the model image mainly includes the calculation program of the target model, that is, the target model; when the model image is a combination of the base model image and the application model image, the model image includes both the calculation program of the target model and the description information of the application environment of the target model, that is, the description information of the target model and the application environment of the target model.

[0065] As an example, assuming that the model distribution system is applied at an oil production site, the target model may be a parameter model A of the trained oil production equipment, which may be used for parameter input of the oil production equipment and for adjusting the input parameters. The cloud platform may obtain a model image A of the parameter model A, and the model image A is a combination of a basic model image and an application model image, that is, the model image A includes description information of the parameter model A and the application environment of the parameter model A.

[0066] Step 202: The cloud platform compresses the model image to obtain an image compression package, and sets and stores a decompression password for the image compression package during the compression process of the model image.

[0067] After obtaining the model image, the cloud platform can use the compression command to compress the model image to obtain an image compression package in the corresponding format. During the model image compression process, the cloud platform will provide a decompression password setting interface, through which the administrator can set the decompression password of the image compression package.

[0068] Continuing with the above example, after the cloud platform obtains the model image A of the parameter model A, it uses the "tar-zcvf" compression command to compress the model image A into a tar.gz format image compression package A. In addition, during the compression process, the administrator sets the decompression password of the image compression package A to "123456" through the decompression password setting interface provided by the cloud platform.

[0069] Optionally, after the administrator sets the decompression password of the image compression package, the cloud platform can encrypt the decompression password through an encryption algorithm, and store the encrypted decompression password of the image compression package in its own database. Of course, the cloud platform may not encrypt the decompression password and directly store it in its own database. This embodiment of the application does not limit this.

[0070] Continuing with the above example, assuming that the decompression password of the image compression package A is "123456", the cloud platform uses the AES (Advanced Encryption Standard) encryption algorithm to encrypt the decompression password, that is, uses the AES encryption function E to encrypt the decompression password, including: taking the plaintext P, i.e., the decompression password "123456", and the key K as the parameter input of the AES encryption function E, and the AES encryption function will output the ciphertext C, i.e., the encrypted decompression password "******", that is, C = E(K, P), so as to encrypt the decompression password from the plaintext "123456" to the ciphertext "******", and the cloud platform stores the encrypted decompression password "******" in its own database.

[0071] The AES encryption algorithm is a symmetric encryption algorithm. Both the encryption function and the decryption function use the same key. That is, if the cloud platform uses the AES encryption algorithm to encrypt the decompression password, the cloud platform and the target edge node agree to use the same key. The cloud platform converts the plaintext of the decompression password into ciphertext and stores it in the database. After the target edge node reads the ciphertext of the decompression password, it uses the same key and the corresponding decryption function to decrypt the ciphertext of the decompression password into plaintext.

[0072] It should be noted that the cloud platform can compress the model image into a tar.gz image compression package by using the compression command "tar-zcvf", and can also use other compression commands to compress the model image into a corresponding format of image compression package, and can encrypt the decompression password by using the AES encryption algorithm, and can also encrypt the decompression password by using other encryption algorithms, which is not limited to the embodiments of the present application.

[0073] Step 203: The cloud platform sends the image compression package to the node management platform.

[0074] In some embodiments, the cloud platform can encrypt the image compression package and send the encrypted image compression package to the node management platform, that is, the cloud platform encrypts the image compression package to be sent by using an encryption algorithm and then sends it to the node management platform. Of course, the cloud platform can also send the image compression package directly to the node management platform without encrypting it, and this embodiment of the application does not limit this.

[0075] Continuing with the above example, after the cloud platform obtains the image compression package A, it uses the SSL (Secure Sockets Layer) protocol to encrypt the image compression package A, that is, the cloud platform and the node management platform communicate through the SSL protocol to establish a secure connection. The cloud platform uses the symmetric key agreed upon with the node management platform to encrypt the image compression package A and sends the encrypted image compression package A to the node management platform.

[0076] The SSL protocol is a secure socket layer protocol that can provide data encryption, authentication and integrity protection based on the TCP (Transmission Control Protocol) protocol. The SSL protocol mainly consists of two parts: the handshake protocol and the record protocol. The handshake protocol is responsible for establishing a secure connection, exchanging public keys and certificates, and agreeing on symmetric keys; the record protocol is responsible for using symmetric keys to encrypt and decrypt data, as well as detect the integrity of data. After being encrypted by the SSL protocol, the image compression package is sent down, which increases the security of the model sending process and can prevent unauthorized hosts from impersonating DNS (Domain Name Server) or the IP (Internet Protocol) address of a trusted host to steal the transmitted data.

[0077] It should be noted that the cloud platform can use the SSL protocol to encrypt the image compression package, and can also use other encryption methods to encrypt the image compression package, which is not limited in the embodiments of the present application.

[0078] Step 204: The node management platform receives the image compression package sent by the cloud platform, and obtains the configuration file of the target model, where the configuration file indicates that the target model needs to be sent to the target edge node.

[0079] After receiving the compressed image package sent by the cloud platform, the node management platform generates a corresponding configuration file based on the configuration information of the target model, and the configuration file includes the configuration rules of the target model. For example, the configuration file includes the target edge node identifier, the start command and access address of the target model. The target edge node identifier is used to indicate the target edge node to which the target model needs to be sent, the start command is used to start the target model, and the access address is used to indicate the browser address or port path of the target model.

[0080] Continuing with the above example, after the node management platform receives the image compression package A sent by the cloud platform, it generates a corresponding configuration file based on the configuration information of the parameter model A. The configuration file contains the configuration rules of the parameter model A, that is, the IP address X of the edge node A to which the parameter model A needs to be sent, the start command of the parameter model A, and the access address.

[0081] Based on the above description, the cloud platform may encrypt the image compression package. In this case, the node management platform receives the encrypted image compression package sent by the cloud platform, decrypts the encrypted image compression package, and obtains the decrypted image compression package.

[0082] Since the cloud platform encrypts the image compression package to be sent by using an encryption algorithm, after the node management platform receives the image compression package sent by the cloud platform, it is necessary to decrypt the image compression package to obtain the decrypted image compression package.

[0083] Continuing with the above example, the cloud platform uses the SSL protocol to encrypt the image compression package A. After the node management platform receives the encrypted image compression package A sent by the cloud platform, it uses the symmetric key agreed upon with the cloud platform to decrypt the image compression package A to obtain the decrypted image compression package A.

[0084] The configuration information of the target model can be obtained by the node management platform from the cloud platform. That is, in some embodiments, the node management platform calls the configuration information reading interface, and sends a configuration information reading instruction to the cloud platform through the configuration information reading interface, and the configuration information reading instruction carries the identifier of the node management platform. The cloud platform receives the configuration information reading instruction sent by the node management platform through the configuration information reading interface, verifies the identity of the node management platform based on the identifier of the node management platform, and sends the configuration information of the target model to the node management platform if the identity authentication of the node management platform is passed. The node management platform receives the configuration information of the target model sent by the cloud platform, and generates a configuration file of the target model based on the configuration information of the target model.

[0085] The node management platform obtains the configuration information corresponding to the target model by calling the configuration information reading interface and sending a configuration information reading instruction to the cloud platform. The configuration information reading instruction carries the identifier of the node management platform. In this way, the cloud platform searches whether the identifier of the node management platform is stored in the database. The database is used to store the identifiers of platforms with the authority to obtain configuration information. If the identifier of the node management platform is stored in the database, it is determined that the identity authentication of the node management platform has passed. Otherwise, it is determined that the identity authentication of the node management platform has failed. In the case that the identity authentication of the node management platform has passed, the cloud platform sends the configuration information of the target model to the node management platform. In the case that the identity authentication of the node management platform has failed, the cloud platform does not send the configuration information of the target model to the node management platform.

[0086] The identifier of the node management platform is used to uniquely identify the node management platform. The identifier of the node management platform may be a name, a MAC (Media Access Control) address, an IP address, etc. of the node management platform.

[0087] Step 205: The node management platform sends the image compression package to the target edge node according to the configuration file.

[0088] After the node management platform generates the configuration file, it sends the image compression package to the target edge node according to the target edge node identifier contained in the configuration rule of the target model in the configuration file.

[0089] Continuing with the above example, after the node management platform generates configuration file A corresponding to parameter model A, it sends the image compression package A corresponding to parameter model A to the edge node A according to the IP address X of the edge node A to which the parameter model A needs to be sent as indicated in the configuration file.

[0090] Based on the above description, the cloud platform may encrypt the image compression package. Therefore, after the node management platform decrypts the encrypted image compression package, it sends the decrypted image compression package to the target edge node according to the configuration file.

[0091] Step 206: The target edge node receives the image compression package sent by the node management platform, reads the decompression password, and decompresses the image compression package according to the decompression password to obtain the target model.

[0092] After the node management platform sends the image compression package to the target edge node according to the configuration file, the target edge node will receive the image compression package, then read the image compression package decompression password, and use the decompression command to decompress the image compression package based on the decompression password to obtain the target model.

[0093] Continuing with the above example, edge node A receives the image compression package A sent by the node management platform, and the image compression package A is compressed into an image compression package in tar.gz format by using the compression command "tar-zcvf", reads the decompression password A corresponding to the image compression package A, and based on the decompression password A, uses the decompression command "tar-xvf" to decompress the image compression package A in tar.gz format to obtain the parameter model A, i.e., the parameter model of the oil production equipment.

[0094] The decompression password can be obtained by the target edge node from the cloud platform. That is, in some embodiments, the target edge node calls the password reading interface, and sends a password acquisition instruction to the cloud platform through the password reading interface, and the password acquisition instruction carries the identifier of the target edge node. The cloud platform receives the password acquisition instruction sent by the target edge node through the password reading interface, verifies the identity of the target edge node based on the identifier of the target edge node, and sends the decompression password to the target edge node if the identity authentication of the target edge node is passed. The target edge node receives the decompression password sent by the cloud platform.

[0095] After the target edge node receives the image compression package corresponding to the target model, it calls the password reading interface and sends a password acquisition instruction to the cloud platform through the password reading interface to obtain the decompression password of the image compression package corresponding to the target model, and the password acquisition instruction carries the identifier of the target edge node. The cloud platform receives the password acquisition instruction sent by the target edge node through the password reading interface, searches for the identifier of the target edge node in the database, and the database is used to store node identifiers with the authority to obtain the decompression password. If the cloud platform finds the identifier of the target edge node in the database, it determines that the identity authentication of the target edge node is passed, and the cloud platform sends the decompression password corresponding to the image compression package to the target edge node; if the cloud platform does not find the identifier of the target edge node in the database, it determines that the identity authentication of the target edge node fails, and the cloud platform does not send the decompression password corresponding to any image compression package to the target edge node.

[0096] The identifier of the target edge node is used to uniquely identify the target edge node, and the identifier of the target edge node may be the name, MAC address, IP address, etc. of the target edge node.

[0097] Based on the above description, the cloud platform may encrypt the decompression password through an encryption algorithm and store the encrypted decompression password in its own database. Therefore, the decompression password read by the target edge node is the encrypted decompression password. At this time, it is necessary to decrypt the encrypted decompression password and decompress the image compression package according to the decrypted decompression password, so that the target edge node can obtain the target model.

[0098] Continuing with the above example, the cloud platform uses the AES encryption algorithm to encrypt the decompression password "123456" of the image compression package A and stores it in its own database. After the edge node A reads the encrypted decompression password "******" corresponding to the image compression package A, it uses the AES decompression function D to decrypt the encrypted decompression password A, that is, the ciphertext C, that is, the encrypted decompression password "******" and the key K are used as the parameter input of the decompression function D, then the decompression function D will output the plaintext P, that is, the decompression password "123456", that is, P = D(K, C), so as to decrypt the decompression password to obtain the decompression password "123456" of the image compression package A, and then, the edge node A decompresses the image compression package A according to the decompression password "123456" to obtain the parameter model A, that is, the parameter model of the oil production equipment.

[0099] In an embodiment of the present application, the cloud platform compresses the model image of the target model to obtain an image compression package, sets a compression password during the compression of the model image, and stores the encrypted decompression password. After that, the cloud platform sends the image compression package to the node management platform, which is then sent to the target edge node by the node management platform. The target edge node then obtains the decompression password from the cloud platform, and then performs operations such as decompression of the image compression package according to the decompression password. The method provided in the embodiment of the present application saves data storage space, reduces the amount of data transmission and bandwidth occupancy, improves the efficiency of the model issuance process, and reduces the time of model issuance by compressing the model image. At the same time, the image compression package and the decompression password are not issued at the same time, which enhances the security of model issuance.

[0100] Figure 3 This is a schematic diagram of the structure of a model distribution device provided in an embodiment of the present application. The model distribution device can be implemented by software, hardware or a combination of both to become part or all of the cloud platform. Please refer to Figure 3 The device includes: an acquisition module 301, a compression module 302 and a sending module 303.

[0101] An acquisition module 301 is used to acquire a model image of a target model to be issued, where the target model is a trained model;

[0102] The compression module 302 is used to compress the model image to obtain an image compression package, and to set and store a decompression password for the image compression package during the compression process of the model image;

[0103] The first sending module 303 is used to send the image compression package to the node management platform, and the node management platform sends the image compression package to the target edge node. The target edge node is used to decompress the image compression package using a decompression password to obtain a target model.

[0104] Optionally, the first sending module 303 includes:

[0105] The encryption submodule is used to encrypt the image compression package;

[0106] The sending submodule is used to send the encrypted image compression package to the node management platform.

[0107] Optionally, the device further comprises:

[0108] A first receiving module, configured to receive a configuration information reading instruction sent by a node management platform through a configuration information reading interface, wherein the configuration information reading instruction carries an identifier of the node management platform;

[0109] A first identity authentication module, used to verify the identity of the node management platform based on the identifier of the node management platform;

[0110] The second sending module is used to send the configuration information of the target model to the node management platform when the identity authentication of the node management platform is passed.

[0111] Optionally, the device further comprises:

[0112] A second receiving module, configured to receive a password acquisition instruction sent by a target edge node through a password reading interface, wherein the password acquisition instruction carries an identifier of the target edge node;

[0113] A second identity authentication module, used to verify the identity of the target edge node based on the identifier of the target edge node;

[0114] The third sending module is used to send the decompression password to the target edge node when the identity authentication of the target edge node is passed.

[0115] In an embodiment of the present application, the cloud platform compresses the model image of the target model to obtain an image compression package, sets a compression password during the compression of the model image, and stores the encrypted decompression password. After that, the cloud platform sends the image compression package to the node management platform, which is then sent to the target edge node by the node management platform. The target edge node then obtains the decompression password from the cloud platform, and then performs operations such as decompression of the image compression package according to the decompression password. The method provided in the embodiment of the present application saves data storage space, reduces the amount of data transmission and bandwidth occupancy, improves the efficiency of the model issuance process, and reduces the time of model issuance by compressing the model image. At the same time, the image compression package and the decompression password are not issued at the same time, which enhances the security of model issuance.

[0116] Figure 4 This is a schematic diagram of the structure of another model distribution device provided in an embodiment of the present application. The model distribution device can be implemented by software, hardware or a combination of both to become part or all of the node management platform. Please refer to Figure 4 The device includes: a receiving module 401, an acquiring module 402 and a sending module 403.

[0117] The receiving module 401 is used to receive the image compression package sent by the cloud platform, where the image compression package is obtained by compressing the model image of the target model to be sent;

[0118] An acquisition module 402 is used to acquire a configuration file of a target model, where the configuration file indicates that the target model needs to be sent to a target edge node;

[0119] The sending module 403 is used to send the image compression package to the target edge node according to the configuration file, and the target edge node decompresses the image compression package using a decompression password to obtain a target model. The decompression password is set during the compression process of the model image.

[0120] Optionally, the receiving module 401 includes:

[0121] The first receiving submodule is used to receive the encrypted image compression package sent by the cloud platform;

[0122] The decryption submodule is used to decrypt the encrypted image compression package to obtain the decrypted image compression package;

[0123] The sending module 403 includes:

[0124] The first sending submodule is used to send the decrypted image compression package to the target edge node according to the configuration file.

[0125] Optionally, the acquisition module 402 includes:

[0126] Call submodule, used to call the configuration information reading interface;

[0127] A second sending submodule, used to send a configuration information reading instruction to the cloud platform through the configuration information reading interface, where the configuration information reading instruction carries an identifier of the node management platform;

[0128] The second receiving submodule is used to receive the configuration information of the target model sent by the cloud platform, and generate a configuration file of the target model based on the configuration information of the target model.

[0129] In an embodiment of the present application, the cloud platform compresses the model image of the target model to obtain an image compression package, sets a compression password during the compression of the model image, and stores the encrypted decompression password. After that, the cloud platform sends the image compression package to the node management platform, which is then sent to the target edge node by the node management platform. The target edge node then obtains the decompression password from the cloud platform, and then performs operations such as decompression of the image compression package according to the decompression password. The method provided in the embodiment of the present application saves data storage space, reduces the amount of data transmission and bandwidth occupancy, improves the efficiency of the model issuance process, and reduces the time of model issuance by compressing the model image. At the same time, the image compression package and the decompression password are not issued at the same time, which enhances the security of model issuance.

[0130] Figure 5 This is a structural diagram of another model sending device provided in an embodiment of the present application. The model sending device can be implemented by software, hardware or a combination of both to become part or all of the target edge node. Please refer to Figure 5 The device includes: a receiving module 501, a reading module 502 and a decompression module 503.

[0131] The receiving module 501 is used to receive the image compression package sent by the node management platform, where the image compression package is obtained by compressing the model image of the target model to be sent;

[0132] A reading module 502 is used to read a decompression password, where the decompression password is set during the compression process of the model image;

[0133] The decompression module 503 is used to decompress the image compression package according to the decompression password to obtain the target model.

[0134] Optionally, the reading module 502 includes:

[0135] Call submodule, used to call password reading interface;

[0136] A sending submodule, used to send a password acquisition instruction to the cloud platform through the password reading interface, where the password acquisition instruction carries the identifier of the target edge node;

[0137] The receiving submodule is used to receive the decompression password sent by the cloud platform.

[0138] In an embodiment of the present application, the cloud platform compresses the model image of the model to be issued to obtain an image compression package, sets a compression password during the compression of the model image, and stores the encrypted decompression password. After that, the cloud platform sends the image compression package to the node management platform, which is then sent to the corresponding edge node by the node management platform. The edge node then obtains the decompression password from the cloud platform, and then performs operations such as decompression of the image compression package according to the decompression password. This method saves data storage space, reduces data transmission volume and bandwidth occupancy, improves the efficiency of the model issuance process, and reduces the time of model issuance by compressing the model image. At the same time, the image compression package and the decompression password are not issued at the same time, which enhances the security of model issuance.

[0139] It should be noted that: the model sending device provided in the above embodiment only uses the division of the above functional modules as an example when sending the model. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the model sending device provided in the above embodiment and the model sending method embodiment belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0140] Figure 6 6 is a block diagram of a terminal device 600 provided in an embodiment of the present application. The terminal device 600 may be a portable mobile terminal device, such as a smart phone, a tablet computer, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer or a desktop computer. The terminal device 600 may also be referred to as a user device, a portable terminal device, a laptop terminal device, a desktop terminal device or other names.

[0141] Typically, the terminal device 600 includes: a processor 601 and a memory 602 .

[0142] The processor 601 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 601 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field Programmable Gate Array), and PLA (Programmable Logic Array). The processor 601 may also include a main processor and a coprocessor. The main processor is a processor for processing data in an awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in a standby state. In some embodiments, the processor 601 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 601 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0143] The memory 602 may include one or more computer-readable storage media, which may be non-transitory. The memory 602 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 602 is used to store at least one instruction, which is used to be executed by the processor 601 to implement the model delivery method provided in the method embodiment of the present application.

[0144] In some embodiments, the terminal device 600 may further optionally include: a peripheral device interface 603 and at least one peripheral device. The processor 601, the memory 602 and the peripheral device interface 603 may be connected via a bus or a signal line. Each peripheral device may be connected to the peripheral device interface 603 via a bus, a signal line or a circuit board. Specifically, the peripheral device includes: at least one of a radio frequency circuit 604, a display screen 605, a camera assembly 606, an audio circuit 607, a positioning assembly 608 and a power supply 609.

[0145] The peripheral device interface 603 may be used to connect at least one peripheral device related to I / O (Input / Output) to the processor 601 and the memory 602. In some embodiments, the processor 601, the memory 602, and the peripheral device interface 603 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 601, the memory 602, and the peripheral device interface 603 may be implemented on a separate chip or circuit board, which is not limited in this embodiment.

[0146] The radio frequency circuit 604 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 604 communicates with the communication network and other communication devices through electromagnetic signals. The radio frequency circuit 604 converts the electrical signal into an electromagnetic signal for transmission, or converts the received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 604 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, etc. The radio frequency circuit 604 can communicate with other terminal devices through at least one wireless communication protocol. The wireless communication protocol includes but is not limited to: the World Wide Web, a metropolitan area network, an intranet, various generations of mobile communication networks (2G, 3G, 4G and 5G), a wireless local area network and / or a WiFi (Wireless Fidelity) network. In some embodiments, the radio frequency circuit 604 may also include circuits related to NFC (Near Field Communication), which is not limited in the embodiments of the present application.

[0147] The display screen 605 is used to display the UI (User Interface). The UI may include graphics, text, icons, videos and any combination thereof. When the display screen 605 is a touch display screen, the display screen 605 also has the ability to collect touch signals on the surface or above the surface of the display screen 605. The touch signal can be input to the processor 601 as a control signal for processing. At this time, the display screen 605 can also be used to provide virtual buttons and / or virtual keyboards, also known as soft buttons and / or soft keyboards. In some embodiments, the display screen 605 can be one, and the front panel of the terminal device 600 is set; in other embodiments, the display screen 605 can be at least two, which are respectively set on different surfaces of the terminal device 600 or are folded; in some other embodiments, the display screen 605 can be a flexible display screen, which is set on the curved surface or folded surface of the terminal device 600. Even, the display screen 605 can also be set to a non-rectangular irregular shape, that is, a special-shaped screen. The display screen 605 can be made of materials such as LCD (Liquid Crystal Display), OLED (Organic Light-Emitting Diode, organic light-emitting diode).

[0148] The camera assembly 606 is used to capture images or videos. Optionally, the camera assembly 606 includes a front camera and a rear camera. Typically, the front camera is arranged on the front panel of the terminal device, and the rear camera is arranged on the back of the terminal device. In some embodiments, there are at least two rear cameras, which are any one of a main camera, a depth of field camera, a wide-angle camera, and a telephoto camera, so as to realize the fusion of the main camera and the depth of field camera to realize the background blur function, the fusion of the main camera and the wide-angle camera to realize panoramic shooting and VR (Virtual Reality) shooting function or other fusion shooting functions. In some embodiments, the camera assembly 606 may also include a flash. The flash can be a monochrome temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm light flash and a cold light flash, which can be used for light compensation at different color temperatures.

[0149] The audio circuit 607 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, and convert the sound waves into electrical signals and input them into the processor 601 for processing, or input them into the radio frequency circuit 604 to achieve voice communication. For the purpose of stereo acquisition or noise reduction, there may be multiple microphones, which are respectively arranged at different parts of the terminal device 600. The microphone may also be an array microphone or an omnidirectional acquisition microphone. The speaker is used to convert the electrical signal from the processor 601 or the radio frequency circuit 604 into sound waves. The speaker may be a traditional film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signal into sound waves audible to humans, but also convert the electrical signal into sound waves inaudible to humans for purposes such as ranging. In some embodiments, the audio circuit 607 may also include a headphone jack.

[0150] The positioning component 608 is used to locate the current geographical location of the terminal device 600 to implement navigation or LBS (Location Based Service). The positioning component 608 can be a positioning component of the GPS (Global Positioning System), Beidou system or Galileo system.

[0151] The power supply 609 is used to power various components in the terminal device 600. The power supply 609 can be an alternating current, a direct current, a disposable battery, or a rechargeable battery. When the power supply 609 includes a rechargeable battery, the rechargeable battery can be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery charged through a wired line, and a wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery can also be used to support fast charging technology.

[0152] Those skilled in the art will understand that Figure 6 The structure shown in the figure does not constitute a limitation on the terminal device 600, and may include more or less components than those shown in the figure, or combine certain components, or adopt a different component arrangement.

[0153] Figure 7 700 is a schematic diagram of the structure of a server provided in an embodiment of the present application. The server 700 includes a central processing unit (CPU) 701, a system memory 704 including a random access memory (RAM) 702 and a read-only memory (ROM) 703, and a system bus 705 connecting the system memory 704 and the central processing unit 701. The server 700 also includes a basic input / output system (I / O system) 706 that helps transmit information between various devices in the computer, and a large-capacity storage device 707 for storing an operating system 713, an application program 714, and other program modules 715.

[0154] The basic input / output system 706 includes a display 708 for displaying information and an input device 709 such as a mouse and a keyboard for user inputting information. The display 708 and the input device 709 are connected to the central processing unit 701 through an input / output controller 710 connected to the system bus 705. The basic input / output system 706 may also include an input / output controller 710 for receiving and processing inputs from a plurality of other devices such as a keyboard, a mouse, or an electronic stylus. Similarly, the input / output controller 710 also provides output to a display screen, a printer, or other types of output devices.

[0155] The mass storage device 707 is connected to the central processing unit 701 through a mass storage controller (not shown) connected to the system bus 705. The mass storage device 707 and its associated computer-readable media provide non-volatile storage for the server 700. That is, the mass storage device 707 may include a computer-readable medium (not shown) such as a hard disk or a CD-ROM drive.

[0156] Without loss of generality, computer readable media may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer readable instructions, data structures, program modules or other data. Computer storage media include RAM, ROM, EPROM, EEPROM, flash memory or other solid-state storage technologies, CD-ROM, DVD or other optical storage, cassettes, magnetic tapes, disk storage or other magnetic storage devices. Of course, those skilled in the art will appreciate that computer storage media are not limited to the above. The above-mentioned system memory 704 and mass storage device 707 can be collectively referred to as memory.

[0157] According to various embodiments of the present application, the server 700 can also be connected to a remote computer on the network through a network such as the Internet. That is, the server 700 can be connected to the network 712 through the network interface unit 711 connected to the system bus 705, or the network interface unit 711 can be used to connect to other types of networks or remote computer systems (not shown).

[0158] The above-mentioned memory also includes one or more programs, and the one or more programs are stored in the memory and configured to be executed by the CPU.

[0159] In some embodiments, a computer-readable storage medium is further provided, in which a computer program is stored, and when the computer program is executed by a processor, the steps of the model delivery method in the above embodiment are implemented. For example, the computer-readable storage medium can be a ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

[0160] It is worth noting that the computer-readable storage medium mentioned in the embodiments of the present application may be a non-volatile storage medium, in other words, a non-transitory storage medium.

[0161] It should be understood that all or part of the steps to implement the above embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. The computer instructions can be stored in the above-mentioned computer readable storage medium.

[0162] That is, in some embodiments, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute the steps of the model distribution method described above.

[0163] It should be understood that the "at least one" mentioned herein refers to one or more, and "multiple" refers to two or more. In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B; "and / or" in this article is only a description of the association relationship of the associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in order to facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, the words "first", "second" and the like are used to distinguish the same items or similar items with basically the same functions and effects. Those skilled in the art can understand that the words "first", "second" and the like do not limit the quantity and execution order, and the words "first", "second" and the like do not limit them to be necessarily different.

[0164] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals involved in the embodiments of the present application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0165] The above-mentioned embodiments are provided for the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A model distribution method, characterized in that: Applied to a cloud platform, the method includes: Obtain a model image of a target model to be issued, where the target model is a trained model; Compressing the model image to obtain an image compression package, and setting and storing a decompression password for the image compression package during the compression process of the model image; The image compression package is sent to a node management platform, and the node management platform sends the image compression package to a target edge node. The target edge node is used to decompress the image compression package using the decompression password to obtain the target model.

2. The method according to claim 1, characterized in that The sending the image compression package to the node management platform includes: Encrypting the image compression package; The encrypted image compression package is sent to the node management platform.

3. The method according to claim 1, characterized in that The method further comprises: Receiving a configuration information reading instruction sent by the node management platform through a configuration information reading interface, wherein the configuration information reading instruction carries an identifier of the node management platform; Verifying the identity of the node management platform based on the identifier of the node management platform; When the identity authentication of the node management platform is passed, the configuration information of the target model is sent to the node management platform.

4. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Receiving a password acquisition instruction sent by the target edge node through a password reading interface, wherein the password acquisition instruction carries an identifier of the target edge node; Verifying the identity of the target edge node based on the identifier of the target edge node; When the identity authentication of the target edge node is passed, the decompression password is sent to the target edge node.

5. A model distribution method, characterized in that: Applied to a node management platform, the method includes: Receiving a compressed image package sent by a cloud platform, wherein the compressed image package is obtained by compressing the model image of the target model to be sent; Obtaining a configuration file of the target model, wherein the configuration file indicates that the target model needs to be sent to a target edge node; According to the configuration file, the image compression package is sent to the target edge node, and the target edge node decompresses the image compression package using a decompression password to obtain the target model, where the decompression password is set during the compression process of the model image.

6. The method according to claim 5, characterized in that The receiving of the image compression package sent by the cloud platform includes: Receiving the encrypted image compression package sent by the cloud platform; Decrypting the encrypted image compression package to obtain the decrypted image compression package; The step of sending the image compression package to the target edge node according to the configuration file includes: According to the configuration file, the decrypted image compression package is sent to the target edge node.

7. The method according to claim 5 or 6, characterized in that The obtaining of the configuration file of the target model comprises: Call the configuration information reading interface; Sending a configuration information reading instruction to the cloud platform through the configuration information reading interface, wherein the configuration information reading instruction carries the identifier of the node management platform; Receive the configuration information of the target model sent by the cloud platform, and generate a configuration file of the target model based on the configuration information of the target model.

8. A model distribution method, characterized in that: Applied to a target edge node, the method comprises: Receive an image compression package sent by a node management platform, where the image compression package is obtained by compressing the model image of the target model to be sent; Reading a decompression password, where the decompression password is set during the compression process of the model image; The image compression package is decompressed according to the decompression password to obtain the target model.

9. The method according to claim 8, characterized in that The reading of the decompression password includes: Call the password reading interface; Sending a password acquisition instruction to the cloud platform through the password reading interface, wherein the password acquisition instruction carries the identifier of the target edge node; Receive the decompression password sent by the cloud platform.

10. A model distribution system, characterized in that: The system includes a cloud platform, a node management platform and a target edge node, the cloud platform is used to implement the steps of the method described in any one of claims 1 to 4, the node management platform is used to implement the steps of the method described in any one of claims 5 to 7, and the target edge node is used to implement the steps of the method described in claim 8 or 9.

Images