Method and device for testing interconnection and intercommunication between heterogeneous multi-party secure computing platforms

Abstract

The invention relates to the field of multi-party security computing, and discloses a test method and device for interconnection and intercommunication between heterogeneous multi-party security computing platforms, and the method comprises the steps: transmitting an instruction to a plurality of participant nodes provided with different security computing platforms through a control end by using a test script, the method comprises the following steps: firstly, indicating each node to load uniform environment configuration so as to ensure consistency during operation; a globally unique task identifier is then generated, and the node is instructed to run a secure compute master script to initiate a secure compute task. After the task is executed, based on the unique identifier, logs and state information of a whole link are aggregated and tracked across all heterogeneous platforms, and therefore effective testing and evaluation of the interconnection and interworking capacity are achieved. According to the method and the device, the participant nodes on different heterogeneous platforms are ensured to have a uniform environment basis, and the task initiation is associated with the task identifier, so that the full-link log and state information related to the test can be more conveniently aggregated and tracked based on the task identifier.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of multi-party secure computation, and in particular to a method and device for testing interconnection and intercommunication between heterogeneous multi-party secure computation platforms. BACKGROUND

[0002] With the increasing awareness of data security and privacy protection, multi-party secure computation technology, as a key path to achieve data "available but invisible", has been widely used in the fields of finance, telecommunications, medical treatment, etc. With the in-depth application, the technology platform of a single manufacturer has been difficult to meet the complex business ecological needs. Different institutions often build heterogeneous multi-party secure computation platforms based on different technical routes. Breaking the "platform island" and realizing the interconnection and intercommunication between heterogeneous platforms have become the inevitable requirement for releasing the value of data elements.

[0003] However, in such a complex distributed test environment involving multiple heterogeneous platforms and numerous participating nodes, the existing technology faces a prominent problem: the observability of the test process is extremely poor. When the test task is concurrently executed on multiple heterogeneous platforms and nodes, each platform and node will generate independent logs and state information. Due to the lack of effective correlation mechanism, these scattered information cannot establish a clear correspondence with a specific test task, making it difficult for the tester to completely track the execution path and state changes of a test task in the whole system. This problem makes performance bottleneck analysis, fault location and test result verification extremely difficult, seriously restricting the effective evaluation of the interconnection and intercommunication capability of heterogeneous platforms. SUMMARY

[0004] The purpose of the present application is to at least provide a method and device for testing interconnection and intercommunication between heterogeneous multi-party secure computation platforms, which can at least solve the technical problem that the tester cannot completely track the execution path and state changes of a test task in the whole system, and at least achieve the effect of improving the observability of the execution path and state changes of the test task in the whole system.

[0005] To solve the above technical problems, at least one embodiment of the present application provides a method for testing interconnection and intercommunication between heterogeneous multi-party secure computation platforms, which is executed by a test control end and includes: sending a first instruction to a plurality of predefined participant nodes through a preset test script, wherein the first instruction is used to instruct each participant node to load a local unified environment configuration file, a part of the plurality of participant nodes is installed with a first to-be-tested multi-party secure computation platform, another part of the plurality of participant nodes is installed with a second to-be-tested multi-party secure computation platform, and the first to-be-tested multi-party secure computation platform and the second to-be-tested multi-party secure computation platform are heterogeneous platforms; generating a globally unique task identifier; The second instruction is used to instruct each participant node to run a secure computing main script. After the secure computing task is initiated, based on the task identifier, full-link logs and state information related to the test are aggregated and tracked in the heterogeneous secure computing platforms corresponding to each participant node to complete the test.

[0006] At least one embodiment of the present application also provides a test device for interconnection and intercommunication between heterogeneous multi-party secure computing platforms, comprising: A first instruction sending module is configured to send a first instruction to a plurality of predefined participant nodes through a preset test script, wherein the first instruction is used to instruct each participant node to load a local unified environment configuration file, a part of the plurality of participant nodes is installed with a first to-be-tested multi-party secure computing platform, and another part is installed with a second to-be-tested multi-party secure computing platform, the first to-be-tested multi-party secure computing platform and the second to-be-tested multi-party secure computing platform are heterogeneous platforms. A task identifier generating module is configured to generate a globally unique task identifier. A second instruction sending module is configured to send a second instruction to each participant node through the test script, wherein the second instruction is used to instruct each participant node to run a secure computing main script, and the secure computing main script is configured to initiate a secure computing task to a to-be-tested secure computing platform to which each participant node belongs based on the unified environment configuration file and the task identifier. A test information aggregation module is configured to, after the secure computing task is initiated, aggregate and track full-link logs and state information related to the test in the heterogeneous secure computing platforms corresponding to each participant node based on the task identifier to complete the test.

[0007] At least one embodiment of the present application also provides an electronic device, comprising at least one processor and a memory in communication connection with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the above-mentioned test method for interconnection and intercommunication between heterogeneous multi-party secure computing platforms.

[0008] At least one embodiment of the present application also provides a computer readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the above-mentioned test method for interconnection and intercommunication between heterogeneous multi-party secure computing platforms.

[0009] The embodiments of the present application provide a test method and device for interconnection and intercommunication between heterogeneous multi-party secure computing platforms. A globally unique task identifier is generated during the test process, which provides a unified identification basis for subsequent tracking of the test task, so that all information related to the test task can be associated through the identifier. Through a preset test script, each participant node is first loaded with a local unified environment configuration file, and then instructed to run a secure computing main script. The secure computing main script is configured to initiate a secure computing task to the secure computing platform to be tested based on the unified environment configuration file and the task identifier. This step ensures that the participant nodes on different heterogeneous platforms have a unified environment basis when executing the test task, and the task initiation is associated with the task identifier. After the initiation of the secure computing task, based on the task identifier, the full-link logs and state information related to the test are aggregated and tracked in the corresponding heterogeneous secure computing platforms of the participant nodes. Due to the globally unique task identifier, the logs and state information scattered on different heterogeneous platforms and participant nodes can establish a clear corresponding relationship with a specific test task, so that the execution path and state change of a test task in the whole system can be completely tracked, the performance test is completed, and the problem of difficult tracking caused by poor observability in the test process is solved.

[0010] In some optional embodiments, the secure computing main script is further configured to perform the following operations: invoke an application program interface of the secure computing platform to be tested to create a task instance; set the task identifier as an associated identifier of the task instance; configure algorithm parameters of the secure computing task and specify a data set participating in the computation; submit and start the secure computing task.

[0011] In the embodiments, the operation of submitting and starting the secure computing task unifies the starting methods of tasks on different heterogeneous platforms. Although there may be differences between different platforms, through the unified invocation of the secure computing main script, the task starting process is more standardized and standardized, reducing the starting errors or inconsistency problems caused by platform differences, and improving the stability and reliability of the test.

[0012] In some optional embodiments, the number of participant nodes is greater than or equal to 8, and the size of all data sets participating in the secure computing task is above 1 billion records.

[0013] In this embodiment, the large-scale data set and multiple participant nodes put higher requirements on the computing power of the multi-party secure computing platform. The platform needs to complete complex computing tasks on a large amount of data within a limited time, while ensuring the accuracy and security of the computing results. Through this high-intensity test, the computing performance of the platform, including computing speed, throughput, etc., can be fully verified to ensure that the platform can meet the high-performance requirements in actual applications.

[0014] In some optional embodiments, the method further comprises: During the execution of the secure computing task, network traffic between the participant nodes is collected by a data packet capture tool; Based on the task identifier, the network traffic is filtered and analyzed to verify the encryption security of data transmission.

[0015] In this embodiment, the filtered network traffic is analyzed to check whether the data is encrypted by the correct encryption algorithm as expected during transmission. Based on the task identifier, the network traffic is filtered to accurately select data packets related to the specific test task, excluding irrelevant traffic interference, making the analysis work more focused and efficient.

[0016] In some optional embodiments, before executing the secure computing task, the method further comprises: Through the contract management interface provided by the first and second to-be-tested multi-party secure computing platforms, a bidirectional node registration and authentication process is performed to establish a trusted connection between the platforms.

[0017] In this embodiment, the bidirectional node registration and authentication process can strictly screen the nodes participating in the secure computing task. Only legally authenticated nodes can access the platform and participate in computing, effectively preventing the intrusion of malicious nodes and avoiding attacks such as data theft and tampering that may be launched by malicious nodes, thereby ensuring the security and privacy of data during transmission and processing.

[0018] In some optional embodiments, before executing the secure computing task, the method further comprises: Verify that the first to-be-tested multi-party secure computing platform can discover and access the metadata of the data set registered on the second to-be-tested multi-party secure computing platform, while ensuring that the original data content is not visible.

[0019] In this embodiment, through the discovery and access of metadata, the first platform can quickly understand the basic information of the second platform data set, such as data type, data size, data theme, etc. This allows the platform to accurately select and utilize appropriate data sets based on modeling needs, avoiding blind search and inefficient data access, thereby improving data utilization efficiency and reducing resource waste.

[0020] In some optional embodiments, the method further includes, prior to performing the secure computing task: Through the bidirectional approval interface between the first multi-party secure computing platform under test and the second multi-party secure computing platform under test, at least one of the approval processes, namely project-level approval and task-level approval, is executed.

[0021] In this embodiment, the two-way approval process establishes a strict access control mechanism for interactions between platforms. Both project-level and task-level approvals require review and confirmation from both platforms. This effectively prevents unauthorized project initiation or task execution, and avoids malicious users or unauthenticated entities from using the platform for illegal operations such as data theft or tampering with calculation results, thereby ensuring the security of the multi-party secure computing environment. Attached Figure Description

[0022] One or more embodiments are illustrated by way of example with reference to the accompanying drawings, and these illustrative descriptions do not constitute a limitation on the embodiments.

[0023] Figure 1 This is a flowchart of a testing method for interconnection between heterogeneous multi-party secure computing platforms provided in one embodiment of this application; Figure 2 This is a schematic diagram of a test system architecture for interconnection between heterogeneous multi-party secure computing platforms, provided in one embodiment of this application. Figure 3 This is a schematic diagram of a test device for interconnection between heterogeneous multi-party secure computing platforms provided in another embodiment of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the various embodiments of this application to help readers better understand this application. However, the technical solutions claimed in this application can be implemented even without these technical details and various changes and modifications based on the following embodiments. The division of the various embodiments below is for the convenience of description and should not constitute any limitation on the specific implementation of this application. The various embodiments can be combined with and referenced by each other without contradiction.

[0025] To address the technical problem mentioned above, where testers find it difficult to fully track the execution path and state changes of a single test task throughout the entire system, this invention proposes a testing method for interconnection between heterogeneous multi-party secure computing platforms. The implementation details of this method are described below. The following content is provided for ease of understanding and is not essential for implementing this solution.

[0026] Example 1: The interoperability testing method between heterogeneous multi-party secure computing platforms in this embodiment can be applied to the test control end, and the specific execution process can be as follows: Figure 1 As shown, it includes: Step 110: Send a first instruction to multiple predefined participating nodes through a preset test script. The first instruction is used to instruct each participating node to load a local unified environment configuration file. Some of the multiple participating nodes have a first multi-party secure computing platform under test installed, and another part has a second multi-party secure computing platform under test installed. The first multi-party secure computing platform under test and the second multi-party secure computing platform under test are heterogeneous platforms. Specifically, the test control terminal plays a dominant role in the entire test system, similar to the role of a commander. It is responsible for initiating, coordinating, and monitoring the entire test process of interconnection of heterogeneous multi-party secure computing platforms, ensuring that each participating node can perform operations accurately and orderly in accordance with the predetermined test plan.

[0027] Participating nodes refer to the various entities involved in the computation during the interoperability testing of heterogeneous multi-party secure computing platforms. These nodes can be computing resources such as servers, virtual machines, or containers. Before testing, it is necessary to define the network addresses, hardware configurations, and installed software environments of these nodes, and predefine them in the test configuration file so that the test scripts can communicate with them accurately. Some of the participating nodes may have a first multi-party secure computing platform under test installed, such as Ruilai's AI privacy protection platform, while others may have a second multi-party secure computing platform under test installed, such as Weishitong's multi-party secure computing platform.

[0028] A test script is a pre-written set of instructions, written in a specific programming language (such as Python or Shell), used to automate various operations during the testing process. In this step, the test script is responsible for communicating with multiple participating nodes and sending control commands.

[0029] The first instruction directs each participating node to load its local unified environment configuration file. This configuration file contains the environmental parameters required by all participating nodes when performing secure computing tasks, such as the version of the computing framework, the paths to dependent libraries, and network communication ports. By loading this configuration file, it ensures that all participating nodes run in the same environment, avoiding computational errors caused by environmental differences.

[0030] Step 120: Generate a globally unique task identifier; Specifically, a globally unique task identifier is a unique string or number used to uniquely identify this test task, serving as the basis for association and tracking throughout the testing process. It can be generated using specific algorithms, such as a combination of timestamps, random numbers, and node identifiers. For example, the current timestamp (accurate to milliseconds) can be combined with a randomly generated string and the unique identifier of the participating nodes (such as a MAC address) to generate the task identifier, ensuring its uniqueness.

[0031] Step 130: Send a second instruction to each of the participating nodes through the test script. The second instruction is used to instruct each participating node to run the secure computing main script. The secure computing main script is configured to initiate a secure computing task to the secure computing platform under test based on the unified environment configuration file and the task identifier. Specifically, the secure computing main script is configured to initiate secure computing tasks to the designated secure computing platform under test, based on a unified environment configuration file and a task identifier. The secure computing main script reads parameters from the unified environment configuration file, such as computing resource allocation and data input paths, and uses a task identifier to mark the current secure computing task for identification in subsequent log recording and status tracking.

[0032] Secure computing tasks can be "secure modeling," "secure statistics," "secure queries," "secure intersections," etc. The appropriate task should be chosen based on the specific circumstances; no specific restrictions are imposed here.

[0033] Step 140: After the secure computing task is initiated, based on the task identifier, aggregate and track the end-to-end logs and status information related to the test in the heterogeneous secure computing platforms corresponding to each of the participating nodes, so as to complete the test.

[0034] Specifically, during the execution of a secure computing task, each participating node generates a large amount of logs and status information, distributed across different nodes and heterogeneous platforms. Based on task identifiers, the test control terminal can collect this information from each node and aggregate it into a centralized location, such as the test server's database or log file. The aggregated information includes the start and end times of the computing task, intermediate results during the computation process, error messages, etc.

[0035] Task identifiers allow for the tracking of the entire process of a secure computing task. From task initiation, data transmission, computation execution, to result return, information at each stage can be accurately tracked and recorded. This helps testers understand the detailed execution flow of the task and identify potential performance bottlenecks and issues.

[0036] Aggregated and traced end-to-end logs and status information are crucial for interoperability testing. Testers can analyze this information to evaluate the interoperability metrics of heterogeneous multi-party secure computing platforms, such as computing speed, throughput, and resource utilization. By comparing these metrics with preset capability targets, testers can determine whether the platform meets performance requirements and provide data support for platform optimization and improvement.

[0037] In summary, the interoperability testing method between heterogeneous multi-party secure computing platforms provided in this embodiment generates a globally unique task identifier during the testing process. This identifier provides a unified basis for subsequent tracking of test tasks, allowing all information related to the test task to be associated through this identifier. Using a pre-set test script, each participating node first loads its local unified environment configuration file, and then instructs it to run the secure computing main script. The secure computing main script is configured to initiate a secure computing task to its assigned secure computing platform based on the unified environment configuration file and the task identifier. This step ensures that participating nodes on different heterogeneous platforms have a unified environmental foundation when executing test tasks, and that task initiation is associated with the task identifier. After the secure computing task is initiated, based on the task identifier, the entire chain of logs and status information related to this test is aggregated and tracked in the heterogeneous secure computing platforms corresponding to each participating node. With a globally unique task identifier, logs and status information scattered across various heterogeneous platforms and participating nodes can establish a clear correspondence with specific test tasks. This allows for the complete tracking of the execution path and status changes of a test task throughout the entire system, enabling performance testing and solving the problem of poor observability during the testing process.

[0038] In some optional embodiments, the secure computing main script is further configured to perform the following operations: call the application programming interface of the secure computing platform to which it belongs to create a task instance; set the task identifier as the associated identifier of the task instance; configure the algorithm parameters of the secure computing task and specify the dataset to be used for computing; submit and start the secure computing task.

[0039] Specifically, the secure computing main script can use the configuration loaded in sf-env.sh, such as PLATFORM_API_URL (platform API address) and AUTH_TOKEN (authentication token), to construct an authenticated HTTP request. This request is sent to the platform-specific API endpoint, such as POST / api / v1 / tasks. Upon receiving the request, the platform internally creates a new, unconfigured task instance and returns a platform-specific task ID (such as task_12345) as a response. This ID is different from the global task identifier and is only valid within the specific platform.

[0040] Next, the generated globally unique ID is "bound" to the task instance within the platform, that is, the global task identifier is set as a custom tag or metadata to the task instance. From then on, within this platform, all subsequent logs, status updates, and results of this task instance will carry this global tag.

[0041] Configure the algorithm parameters for the secure computation task and specify the dataset to be used for computation. This "shapes" the created task instance, defining "what" it will compute and "how" it will compute. Secure computation task types can include Privacy-Preserving Intersection (PSI), Federated Learning (HOMO_FL), or Secure Statistics (MPC), etc. When configuring the model type for Federated Learning, algorithm parameters can include the learning rate, number of iterations, and the PSI protocol type. Specify the data required for this task using the dataset ID or path. These data paths are usually pre-defined in sf-env.sh or passed as parameters in the secure computation main script.

[0042] After all configurations are complete, task execution is triggered. The secure computing main script sends the final start command to the platform, such as POST / api / v1 / tasks / task_12345 / start. Upon receiving the command, the platform performs a final dependency check and then schedules the task to the specified computing engine (such as wst-engine) for execution. The secure computing main script may subsequently enter a monitoring loop to periodically query the task status, or it may exit directly, with an external monitoring system tracking progress based on a global task identifier. This tests the interoperability, data exchange, and collaborative computing capabilities between heterogeneous platforms.

[0043] The innovation of the above series of steps, encapsulated in the main script of secure computing, lies not in the individual steps themselves, but in their serialization, automation, and seamless integration with global task identifiers and environment configuration systems. This achieves a "one-click deployment" effect: transforming a distributed secure computing task involving complex configuration into a standardized process that can be remotely, batch-, and automatically triggered by a test framework.

[0044] In this embodiment, the operation of submitting and starting a secure computing task standardizes the task startup method across different heterogeneous platforms. Although there may be differences between platforms, the unified invocation of the secure computing main script makes the task startup process more standardized and normalized, reducing startup errors or inconsistencies caused by platform differences and improving the stability and reliability of the test.

[0045] In some optional embodiments, the number of participating nodes is greater than or equal to 8, and the total size of the datasets participating in the secure computing task is greater than 1 billion records. For example, the dataset corresponding to each participating node may have more than 1.25 records.

[0046] This embodiment clarifies the scale benchmark of the test, defining specific indicators for "high performance" and "high reliability." Large-scale datasets and multiple participating nodes place higher demands on the computing power of the multi-party secure computing platform. The platform needs to complete complex computational tasks on large amounts of data within a limited time, while ensuring the accuracy and security of the computation results. Through this high-intensity testing, the platform's computing performance, including metrics such as computing speed and throughput, can be fully verified, ensuring that the platform can meet the high-performance requirements of practical applications.

[0047] In some optional embodiments, the method further includes: during the execution of the secure computing task, collecting network traffic between the participating nodes using a packet capture tool; filtering and analyzing the network traffic based on the task identifier to verify the encryption security of data transmission.

[0048] Specifically, in heterogeneous multi-party secure computing platforms, participating nodes may be distributed across different physical locations, connecting and communicating via a network. When collecting network traffic, packet capture tools need to be deployed in appropriate network locations. For example, port mirroring can be configured on the LAN switch where the participating nodes reside, copying the communication traffic between the participating nodes to another port. Then, a device with a packet capture tool installed can be connected to that port for collection. Commonly used packet capture tools include Wireshark, Tcpdump, and Snort. For instance, in Wireshark, display filters can be used to filter packets containing specific task identifiers. Detailed analysis of the filtered network traffic is then performed, including information such as the source / destination address, port number, protocol type, packet size, and transmission time. By analyzing this information, the communication patterns and data transmission between participating nodes can be understood, and it can be checked whether sensitive information is contained and whether this information is transmitted in encrypted form.

[0049] In this embodiment, analyzing the filtered network traffic allows us to check whether the data was encrypted using the correct encryption algorithm during transmission. Filtering network traffic based on task identifiers can accurately identify data packets related to a specific test task, eliminating interference from irrelevant traffic and making the analysis more focused and efficient.

[0050] In some optional embodiments, before performing the secure computing task, the method further includes: performing a bidirectional node registration and authentication process through the contract management interface provided by the first multi-party secure computing platform under test and the second multi-party secure computing platform under test, so as to establish a trusted connection between the platforms.

[0051] In this embodiment, the bidirectional node registration and authentication process can strictly screen nodes participating in secure computing tasks to verify the interoperability of the multi-party secure computing platform. Only nodes that have been legally authenticated can access the platform and participate in computing, effectively preventing intrusion by malicious nodes and avoiding potential data theft, tampering, and other attacks by malicious nodes, thereby ensuring the security and privacy of data during transmission and processing.

[0052] In some optional embodiments, before performing the secure computation task, the method further includes: verifying that the first multi-party secure computation platform under test can discover and access the metadata of the dataset registered on the second multi-party secure computation platform under test, while ensuring that the original data content is not visible.

[0053] In this embodiment, through metadata discovery and access, the first platform can quickly understand the basic information of the second platform's dataset, such as data type, data size, and data theme, to verify the dataset interoperability of the multi-party secure computing platform. This enables the platform to accurately select and utilize appropriate datasets according to modeling needs, avoiding blind searches and ineffective data access, thereby improving data utilization efficiency and reducing resource waste.

[0054] In some optional embodiments, before performing the secure computing task, the method further includes: performing at least one approval process, either project-level approval or task-level approval, through a bidirectional approval interface between the first multi-party secure computing platform under test and the second multi-party secure computing platform under test.

[0055] In this embodiment, the interoperability of projects and tasks within the multi-party secure computing platform is verified. A two-way approval process establishes a strict access control mechanism for interactions between the platforms. Both project-level and task-level approvals require review and confirmation from both platforms. This effectively prevents unauthorized project initiation or task execution, and avoids malicious users or unauthenticated entities from using the platform for illegal operations such as data theft or tampering with computation results, thereby ensuring the security of the multi-party secure computing environment.

[0056] In some optional embodiments, the interoperability analysis is performed based on the aggregated and tracked end-to-end logs and status information, including the following steps: Based on the global task identifier, aggregate end-to-end information from at least one channel, such as platform task engine logs, system monitoring data, network probe data, and platform audit logs; extract core dimension indicators from the end-to-end information, normalize the scores of the core dimension indicators, and use a weighted comprehensive evaluation model to obtain the interoperability index of the corresponding multi-party secure computing platform under test.

[0057] Specifically, the core dimension indicators are divided into three categories: functional indicators, security indicators, and performance indicators. Functional indicators are basic functional test items, including at least one of node interoperability, dataset interoperability, task interoperability, and project interoperability, with each functional test item corresponding to a test score; security indicators include at least one of communication security, identity authentication, and audit security; performance indicators include at least one of efficiency dimension indicators, reliability dimension indicators, resource efficiency dimension indicators, and collaboration quality dimension indicators.

[0058] Raw data related to core dimensional metrics can be obtained through platform task engine logs, system monitoring data, network probe data, and platform audit logs. All this raw data is cleaned, standardized, and associated with the same global task ID, then normalized to form a unified test fact base. Based on this test fact base, functional scores for functional-level metrics, security scores for security-level metrics, and performance scores for performance-level metrics are calculated. Finally, the interoperability index of the multi-party secure computing platform under test is calculated using the following formula: Interoperability capability index score = Functional score F* +Safety Score S* +Performance Score q*

[0059] in, + + =1.

[0060] Specifically, the functional score can be a weighted sum of the test scores corresponding to each functional test item; The formula for calculating the security score is as follows:

[0061] in, This represents the number of test items in the security test class. Let i be the total number of indicators for the i-th test item. and These represent the number of passes and the number of basic passes for the i-th test item, respectively. Let be the weight of the i-th test item.

[0062] Efficiency metrics may include at least one of the following: —Total task time: The number of seconds from task submission to final completion; —Total communication rounds: The number of interactions required to complete the entire task; Average round time: Total time / Number of communication rounds; —Communication cost per unit of data volume: (Total uplink data volume + Total downlink data volume) / Number of business data records; —Throughput: Total number of records processed / Total time elapsed.

[0063] Reliability metrics may include at least one of the following: —Task success rate: The percentage of times a task is successfully completed across multiple tests; —Automatic retry success rate: The percentage of cases where an error occurs and the system automatically retryes the attempt, ultimately succeeding. —Mean recovery time: The average time it takes for the system to return to normal operation after a non-fatal error occurs.

[0064] Resource efficiency metrics may include at least one of the following: —Resource Consumption Balance: Calculate the standard deviation of CPU / memory utilization across all nodes. The smaller the standard deviation, the more balanced the resource utilization, indicating no obvious "weak link." —Resource utilization rate: (∑Average utilization rate of all participating nodes) / Number of participating nodes. Under the premise of ensuring performance, a higher utilization rate indicates better resource efficiency; —Network resource cost: Total network traffic.

[0065] Collaborative quality metrics may include at least one of the following: —Computational consistency: The percentage of tests conducted to verify the consistency of multi-party computation results through verification mechanisms (such as hashing, zero-knowledge proofs); —Data alignment quality: In tasks such as PSI, the overlap rate of the data intersection IDs ultimately used for computation.

[0066] The weighted comprehensive evaluation model for the final performance score is as follows: Performance score q = Efficiency score * +Reliability Score* +Resource Efficiency Score* +Collaboration Quality Score* ; in, + + + = 1.

[0067] In some optional implementations, the weights of each core metric differ depending on the test application scenario. For example: In financial risk control testing scenarios: > > > (Reliability is paramount); In internet marketing testing scenarios: > > > (Efficiency and cost take priority).

[0068] In this embodiment, the subjective and vague concept of "interoperability" is transformed into a measurable, decomposable, traceable, and comparable quantitative scientific system. This is not only an innovation in testing methods but also in evaluation standards, providing a more objective and authoritative basis for decision-making in technology selection, contract acceptance, and product iteration.

[0069] Example 2: Based on the above embodiments, this embodiment provides an application example. The architecture diagram of the test system for interconnecting heterogeneous multi-party secure computing platforms in this embodiment is as follows: Figure 2 As shown. The test system consists of a test server and test laptops. The test server virtualizes eight participants ( Figure 2 The test involved deploying the heterogeneous MPC platforms (wst-21 to wst-24 and ruilai-25 to ruilai-28) on each participant's server. Specifically, the RealSecure privacy-preserving machine learning platform (heterogeneous platform 1) and the multi-party secure computation platform (heterogeneous platform 2) were deployed on each participant's server. Testing tools such as Xterminal and VMware ESXi were deployed on the test laptops. The deployment status of the eight participant nodes on the server was checked using Xterminal to confirm successful connection establishment between the eight nodes.

[0070] The test items are divided into two main categories: pre-construction interoperability functional items and comprehensive performance test items. Pre-construction interoperability functional items specifically include node connection testing, data sharing testing, project approval testing, and task execution testing. Comprehensive performance testing is conducted after the pre-construction interoperability functional items have passed the tests, and can be performed through the secure computing tasks described in the aforementioned embodiments.

[0071] The specific testing process includes the following: I. Pre-connectivity Function 1.1 Node Interoperability Test Through the contract management interfaces provided by the first and second multi-party secure computation platforms under test, a bidirectional node registration and authentication process is executed to establish a trusted connection between the platforms. Specifically, this includes the following steps: 1.1.1. Initiate contract approval through the RealSecure privacy-preserving machine learning platform - Contract Management - Add Contract, entering the multi-party secure computation platform node name "smpcp", authentication method "SHA256_RSA", and contract expiration time. View the approval process in the multi-party secure computation platform - Node Approval. If the multi-party secure computation platform receives the contract approval, the test is successful.

[0072] 1.1.2. Through the multi-party secure computation platform - node approval, select approval and click "Approve". Then, check the approval result in the RealSecure privacy-preserving machine learning platform - contract management. If the RealSecure privacy-preserving machine learning platform receives the contract approval result and the status shows "approved by the other party," the test is successful.

[0073] 1.1.3. Through RealSecure's privacy protection machine learning platform - contract management, select approval and click "cancel contract" to successfully cancel the contract. If the test result shows that the status is "cancellation was rejected by our party", then the test is passed.

[0074] 1.1.4. Initiate contract approval through the multi-party secure computation platform - node management - add heterogeneous node, entering the RealSecure privacy-preserving machine learning platform node ID, IP "10.131.191.31", and port "30007". View the approval process in RealSecure privacy-preserving machine learning platform - node management. If the test result shows successful contract approval and the status is "pending connection", indicating that the RealSecure privacy-preserving machine learning platform has received the contract approval, then the test is successful. 1.1.5. Through the RealSecure privacy-preserving machine learning platform - contract management, select the contract and click "Approve". Then, check the approval result in the multi-party secure computation platform - node management. If the actual test result is: approval successful, the status is "approved by this party", the multi-party secure computation party has received approval, and the connection request status is "accepted", then the test is successful. 1.1.6. View the multi-party secure computation platform dataset through the RealSecure Privacy-Preserving Machine Learning Platform - Collaborative Data. If the test result shows that the RealSecure Privacy-Preserving Machine Learning Platform can view the multi-party secure computation platform dataset after the node is connected, then the test is successful.

[0075] 1.2 Dataset Interoperability Test Verify that the first multi-party secure computation platform under test can discover and access the metadata of the dataset registered on the second multi-party secure computation platform under test, while ensuring that the original data content is not visible. Specifically: 1.2.1. Through the RealSecure privacy-preserving machine learning platform - My Data - Add Data, enter the dataset name "testunion", select the local dataset, upload, and submit. Successful dataset upload indicates a successful test. 1.2.2. Through the multi-party secure computing platform - My Datasets - Create New Dataset, enter the dataset name "union0613", select local dataset, upload and go online. If the dataset goes online successfully, the test is passed. 1.2.3. By using the RealSecure privacy-preserving machine learning platform - Collaboration Data - Add Collaboration Data, you can view the multi-party secure computation platform data. If you can view the datasets that have been launched on the multi-party secure computation platform, then the test is passed. 1.2.4. Use the multi-party secure computation platform - Data Square to view the RealSecure privacy-preserving machine learning platform data. If the RealSecure privacy-preserving machine learning platform dataset can be viewed, the test is successful.

[0076] 1.3 Project Interoperability Testing Through the bidirectional approval interface between the first and second multi-party secure computation platforms under test, at least one approval process, either project-level or task-level, is executed. Specific testing steps include the following: 1.3.1. Through the RealSecure privacy-preserving machine learning platform - Project List - Create a collaborative project, enter the project name "test-1", project type HOMO_FL, contract ID "2", and click "OK" to initiate project approval. Confirm whether the project approval was successfully initiated; 1.3.2. Through the multi-party secure computing platform - task approval, select "view approval details" to confirm whether you can view the approval details; 1.3.3. Through the multi-party secure computation platform - task approval, select approval and click "reject". Go to the RealSecure privacy protection machine learning platform - project list to view the approval result and confirm whether the project status is rejected by the other party. 1.3.4. Through the multi-party secure computation platform - Tasks - Create a new task, enter the task name "union-psi1", task type "Federated (FL)", task engine wst-engine, select the two-party dataset, configure the psi workflow, and click "Submit" to initiate the task workflow. Confirm that the actual test result is: the task workflow was successfully initiated and the project status is under approval; 1.3.5. View the acceptance process through RealSecure Privacy Protection Machine Learning Platform - Personal Center - Approval and click "Approval". Then view the approval result on the Multi-Party Secure Computing Platform and check if the status is: Pending Start, the participating party has approved the project. If the test results of steps 1.3.1 to 1.3.5 above are all passed, the project approval test is passed.

[0077] 1.4 Task Interoperability Test 1.4.1. Through the multi-party secure computing platform - Tasks, select the psi task and click "Start". Confirm whether you can successfully view the approval details; 1.4.2. Through the RealSecure Privacy Protection Machine Learning Platform - Personal Center - Approvals, view the task approval and click "Approval" to check the task progress on the Multi-Party Secure Computing Platform. Confirm whether you can view the PSI project execution progress; 1.4.3. Through the Multi-Party Secure Computation Platform - Tasks - Create New Task, enter the task name "union-Security Statistics", task type "MPC", task engine "wst", and task description "--". Select the two-party dataset, configure the security statistics process, and ensure the four arithmetic expressions are multiplications of the feature columns of the two-party datasets. Save the task and click "Start". To confirm successful execution, check the security statistics execution progress.

[0078] If all the test results in steps 1.4.1 to 1.4.3 above are passed, then the task execution test is passed.

[0079] II. Secure Computing Task Test After the aforementioned pre-construction interoperability tests are passed, the secure computing task test will then be conducted. This includes the following steps: 2.1. Using a pre-defined test script, a first instruction is sent to multiple predefined participating nodes. The first instruction is used to instruct each participating node to load a local unified environment configuration file. Some of the multiple participating nodes have a first multi-party secure computing platform under test installed, and another part has a second multi-party secure computing platform under test installed. The first multi-party secure computing platform under test and the second multi-party secure computing platform under test are heterogeneous platforms. 2.2 Generate a globally unique task identifier; 2.3. Using the test script, a second instruction is sent to each of the participating nodes, the second instruction being used to instruct each participating node to run the secure computing main script; wherein, the secure computing main script is configured to initiate a secure computing task to the secure computing platform under test to which it belongs, based on the unified environment configuration file and the task identifier; 2.4 After the secure computing task is initiated, based on the task identifier, the full-link logs and status information related to the test are aggregated and tracked in the heterogeneous secure computing platforms corresponding to each of the participating nodes to complete the test.

[0080] Specifically, a secure computing task is initiated to the Bafang service using the command `bsah 8p.sh "source / root / sf -env.sh&&bash / root / sf.sh'date +%Y%m%S%H%M%S%N'"`. The first command is: source / root / sf-env.sh. This command precisely executes the operation in step 2.1 of "instructing each participating node to load its local unified environment configuration file".

[0081] The command to generate a globally unique task identifier is: 'date +%Y%m%S%H%M%S%N', corresponding to the specific implementation of step 2.2 "Generate a globally unique task identifier". By directly executing the subshell command in the main command, a globally unique ID based on a nanosecond-level precision timestamp is generated in an extremely efficient way and passed as a parameter to subsequent scripts.

[0082] The second command is: `bash / root / sf.sh ...` which instructs each participating node to run the secure computation main script (i.e., `sf.sh`). Simultaneously, the task identifier generated in the previous step is passed as a parameter to the secure computation main script.

[0083] sf.sh corresponds to the main script for secure computing, which details the process of "initiating a secure computing task to the assigned secure computing platform under test" in step 2.3. It utilizes the environment configuration loaded by the first instruction and the unique task identifier passed in, and completes the creation, configuration, and startup of the task by calling platform APIs and other methods.

[0084] 8p.sh is the "preset test script" that corresponds to the initiation actions in steps 2.1 and 2.3. By integrating the above steps through the 8p.sh script, the operation of executing security modeling commands on a set of predefined hosts is realized.

[0085] In the secure computing task, network traffic between the participating nodes is collected using a packet capture tool; the network traffic is filtered and analyzed based on the task identifier to verify the encryption security of data transmission.

[0086] Example 3: Another embodiment of this application relates to a test apparatus for interconnection between heterogeneous multi-party secure computing platforms. The implementation details of this test apparatus are described below. The following details are for ease of understanding and are not essential for implementing this solution. A schematic diagram of the test apparatus for interconnection between heterogeneous multi-party secure computing platforms can be seen as follows: Figure 3 As shown, it includes a first instruction sending module 310, a task identifier generation module 320, a second instruction sending module 330, and a second instruction sending module 330.

[0087] The first instruction sending module 310 is used to send a first instruction to a predefined plurality of participating party nodes through a preset test script. The first instruction is used to instruct each participating party node to load a local unified environment configuration file. Among the plurality of participating party nodes, a portion of them are equipped with a first multi-party secure computing platform under test, and another portion are equipped with a second multi-party secure computing platform under test. The first multi-party secure computing platform under test and the second multi-party secure computing platform under test are heterogeneous platforms. Task identifier generation module 320 is used to generate globally unique task identifiers; The second instruction sending module 330 is used to send a second instruction to each of the participating nodes through the test script. The second instruction is used to instruct each participating node to run the secure computing main script. The secure computing main script is configured to initiate a secure computing task to the secure computing platform under test based on the unified environment configuration file and the task identifier. The test information aggregation module 340 is used to aggregate and track the end-to-end logs and status information related to the test in the heterogeneous secure computing platforms corresponding to each of the participating nodes after the secure computing task is initiated, based on the task identifier, so as to complete the test.

[0088] It is worth mentioning that all modules involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this application, this embodiment does not introduce units that are not closely related to solving the technical problems proposed in this application; however, this does not mean that other units are absent in this embodiment.

[0089] In some alternative embodiments, the apparatus further includes: The network traffic capture module is used to collect network traffic between the participating nodes using a packet capture tool during the execution of the secure computing task. The data transmission verification module is used to filter and analyze the network traffic based on the task identifier to verify the encryption security of data transmission.

[0090] In some alternative embodiments, the apparatus further includes: The node connection testing module is used to execute a bidirectional node registration and authentication process through the contract management interface provided by the first multi-party secure computing platform under test and the second multi-party secure computing platform under test, so as to establish a trusted connection between the platforms.

[0091] In some alternative embodiments, the apparatus further includes: The data sharing test module is used to verify that the first multi-party secure computing platform under test can discover and access the metadata of the dataset registered on the second multi-party secure computing platform under test, while ensuring that the original data content is not visible.

[0092] In some alternative embodiments, the apparatus further includes: The project approval testing module is used to execute at least one approval process, namely project-level approval and task-level approval, through the bidirectional approval interface between the first multi-party secure computing platform under test and the second multi-party secure computing platform under test.

[0093] Example 4: Another embodiment of this application relates to an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the test method for interconnection between heterogeneous multi-party secure computing platforms in the above embodiments.

[0094] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0095] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0096] Example 5: Another embodiment of this application relates to a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the method embodiments described above.

[0097] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0098] Those skilled in the art will understand that the above embodiments are specific embodiments for implementing this application, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of this application.

Claims

1. A test method for interconnection and interoperability between heterogeneous multi-party secure computing platforms, characterized in that, The method is executed by the test control terminal and includes: A first instruction is sent to multiple predefined participating nodes through a preset test script. The first instruction is used to instruct each participating node to load a local unified environment configuration file. Some of the multiple participating nodes have a first multi-party secure computing platform under test installed, and another part has a second multi-party secure computing platform under test installed. The first multi-party secure computing platform under test and the second multi-party secure computing platform under test are heterogeneous platforms. Generate a globally unique task identifier; The test script sends a second instruction to each of the participating nodes, which instructs each participating node to run the secure computing main script. The secure computing main script is configured to initiate a secure computing task to the secure computing platform under test based on the unified environment configuration file and the task identifier. After the secure computing task is initiated, based on the task identifier, the end-to-end logs and status information related to the test are aggregated and tracked in the heterogeneous secure computing platforms corresponding to each of the participating nodes to complete the test.

2. The test method for interconnection and interoperability between heterogeneous multi-party secure computing platforms according to claim 1, characterized in that, The secure computing main script is further configured to perform the following operations: Call the application programming interface of the security computing platform under test to create a task instance; Set the task identifier as the associated identifier of the task instance; Configure the algorithm parameters for the secure computing task and specify the dataset to be used in the computation; Submit and start the secure computing task.

3. The test method for interconnection and interoperability between heterogeneous multi-party secure computing platforms according to claim 1, characterized in that, The number of participating nodes is greater than or equal to 8, and the total size of all datasets participating in the secure computing task is more than 1 billion records.

4. The test method for interconnection and interoperability between heterogeneous multi-party secure computing platforms according to claim 1, characterized in that, The method further includes: During the execution of the secure computing task, network traffic between the participating nodes is collected using a packet capture tool; The network traffic is filtered and analyzed based on the task identifier to verify the encryption security of data transmission.

5. The test method for interconnection and interoperability between heterogeneous multi-party secure computing platforms according to claim 1, characterized in that, Before performing the secure computing task, the method further includes: Through the contract management interfaces provided by the first and second multi-party secure computation platforms under test, a bidirectional node registration and authentication process is executed to establish a trusted connection between the platforms.

6. The test method for interconnection and interoperability between heterogeneous multi-party secure computing platforms according to claim 1, characterized in that, Before performing the secure computing task, the method further includes: Verify that the first multi-party secure computation platform under test can discover and access the metadata of the dataset registered on the second multi-party secure computation platform under test, while ensuring that the original data content is not visible.

7. The test method for interconnection between heterogeneous multi-party secure computing platforms according to claim 1, characterized in that, Before performing the secure computing task, the method further includes: Through the bidirectional approval interface between the first multi-party secure computing platform under test and the second multi-party secure computing platform under test, at least one of the approval processes, namely project-level approval and task-level approval, is executed.

8. A testing device for interconnection and interoperability between heterogeneous multi-party secure computing platforms, characterized in that, include: The first instruction sending module is used to send a first instruction to a predefined number of participating nodes through a preset test script. The first instruction is used to instruct each participating node to load a local unified environment configuration file. Among the multiple participating nodes, a portion of them are equipped with a first multi-party secure computing platform under test, and another portion are equipped with a second multi-party secure computing platform under test. The first multi-party secure computing platform under test and the second multi-party secure computing platform under test are heterogeneous platforms. The task identifier generation module is used to generate globally unique task identifiers; The second instruction sending module is used to send a second instruction to each of the participating nodes through the test script. The second instruction is used to instruct each participating node to run the secure computing main script. The secure computing main script is configured to initiate a secure computing task to the secure computing platform under test based on the unified environment configuration file and the task identifier. The test information aggregation module is used to aggregate and track the end-to-end logs and status information related to the test in the heterogeneous secure computing platforms corresponding to each of the participating nodes after the secure computing task is initiated, based on the task identifier, so as to complete the test.

9. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the test method for interconnection between heterogeneous multi-party secure computing platforms as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the test method for interconnection between heterogeneous multi-party secure computing platforms as described in any one of claims 1 to 7.

Images