Safety multi-party computing method, device and equipment based on smart traffic and medium

Abstract

The invention relates to a safety multi-party computing method and device based on smart traffic, equipment and a medium. According to the scheme, the security weight of each participating node is determined through the data density index of each participating node, and the difference of the data contribution of each participating node can be balanced through the security weight of each participating node, so that the deviation caused by non-uniform data is counteracted; according to the scheme, the threshold parameter of the task node is adjusted through the data density index of the participating node and the abnormal mark information, so that the multi-party computing task can be still completed under the condition that part of participating node data is unstable; according to the scheme, each standardized data fragment is divided into the sub-batch data through the security weight, the exception mark information and the threshold parameter, and the task node of each sub-batch data is allocated, so that each sub-batch data is ensured to execute the security multi-party computing task through the enough number of task nodes with higher security weight and less exception conditions, and the security multi-party computing efficiency is improved. And the task completion rate and accuracy of the system in a high-dynamic traffic environment are improved.

Description

Technical Field

[0001] This application relates to the field of secure multi-party computation, and in particular to a secure multi-party computation method, apparatus, device, and medium based on intelligent transportation. Background Technology

[0002] Intelligent transportation is a new model of transportation development based on the combination of networks such as the Internet and the Internet of Things, with intelligent road networks, intelligent equipment, intelligent travel, and intelligent management as its key components. It has the basic characteristics of information connectivity, real-time monitoring, collaborative management, and integration of people and things.

[0003] Currently, in intelligent transportation scenarios, high-speed vehicle movement can lead to signal switching, short-term communication interruptions, and data delays during data collection. This can cause data imbalances during multi-party computation due to incomplete or highly fluctuating data collected by different participants, resulting in inaccurate results. For example, significant differences in the quantity and quality of traffic data collected from different regions and time periods can affect the accuracy of subsequent multi-party computations.

[0004] Therefore, how to improve the accuracy of multi-party computation of data is a problem that needs to be solved by those skilled in the art. Summary of the Invention

[0005] This application provides a secure multi-party computation method, apparatus, equipment, and medium based on intelligent transportation to improve the accuracy of data multi-party computation.

[0006] In a first aspect, this application provides a secure multi-party computation method based on intelligent transportation, wherein the secure multi-party computation method is applied to cloud nodes, and the secure multi-party computation method includes:

[0007] Receive target data in encrypted form sent by each participating node; the target data includes standardized data fragments, data density indices, and anomaly marker information;

[0008] The security weight of each participating node is determined based on its data density index.

[0009] Threshold parameters are determined using the data density indices and anomaly marker information of each participating node; the threshold parameters are the threshold parameters of the task nodes participating in the secure multi-party computation task.

[0010] Based on the security weights, anomaly marking information, and threshold parameters of each participating node, each standardized data shard is divided into at least one sub-batch of data according to its spatiotemporal characteristics, and the task nodes for each sub-batch of data are determined.

[0011] Send the corresponding secure multi-party computation task to the task node of each sub-batch of data, and calculate the global computation result based on the local aggregation result of each sub-batch of data.

[0012] Optionally, determining the security weight of each participating node based on its data density index includes:

[0013] Determine the data density metrics for each participating node; the data density metrics include: data quality metrics and data quantity metrics;

[0014] A comprehensive index is calculated based on the data quality and data quantity indicators of each participating node.

[0015] The security weight of each participating node is calculated using the comprehensive index of each participating node.

[0016] Optionally, threshold parameters are determined using the data density metrics and anomaly marker information of each participating node, including:

[0017] The global contribution is determined using the data density index of each participating node;

[0018] The percentage of anomalies is determined by using the anomaly marking information of each participating node;

[0019] The threshold parameters are adjusted based on the global contribution and anomaly percentage of each participating node.

[0020] Optionally, the step of dividing each standardized data shard into at least one sub-batch of data according to its spatiotemporal characteristics based on the security weights of each participating node, anomaly marking information, and the threshold parameter, and determining the task node for each sub-batch of data, includes:

[0021] Each standardized data segment is divided into at least one batch of data according to time.

[0022] Divide each batch of data into at least one group based on geographical location;

[0023] Initial participating nodes are determined from the participating nodes corresponding to each group of data; the initial participating nodes are determined using the security weight and anomaly marking information of each participating node;

[0024] Determine whether the number of initial participating nodes in each group of data exceeds the threshold parameter;

[0025] If yes, then each group of data will be used as the data for each sub-batch, and the initial participating nodes of each group of data will be used as the task nodes of each sub-batch of data; if no, then regions will be merged according to geographical location, at least one group of data will be re-divided, and the step of determining the initial participating nodes from the participating nodes corresponding to each group of data will continue.

[0026] Secondly, this application provides a secure multi-party computation method based on intelligent transportation, wherein the secure multi-party computation method is applied to participating nodes, and the secure multi-party computation method includes:

[0027] The collected raw traffic data is preprocessed to generate standardized data fragments;

[0028] Determine data density indices using standardized data fragments;

[0029] Perform anomaly detection on each standardized data fragment and generate anomaly marker information;

[0030] Based on the standardized data sharding, the data density index, and the anomaly marker information, dense target data is generated and sent to the cloud node; the cloud node is used to execute the above-described secure multi-party computation method based on the cloud node.

[0031] If a secure multi-party computation task is received from the cloud node, the secure multi-party computation task is executed, and the local aggregation results of each sub-batch data are sent to the cloud node so that the cloud node can calculate the global computation result based on each local aggregation result.

[0032] Optionally, the step of performing anomaly detection on each standardized data shard and generating anomaly marker information includes:

[0033] Detect outliers in each standardized data segment;

[0034] The abnormal data is marked to generate abnormal marking information;

[0035] Interpolation compensation is performed on the abnormal data in each standardized data segment.

[0036] Optionally, performing the secure multi-party computation task includes:

[0037] Determine the sub-batch data to be processed corresponding to the secure multi-party computation task;

[0038] Based on the sub-batch data to be processed and the security weight of the current node, the secure multi-party computation task is executed to generate a local aggregation result.

[0039] Thirdly, this application provides a secure multi-party computing device based on intelligent transportation, wherein the secure multi-party computing device is applied to a cloud node, and the secure multi-party computing device includes:

[0040] The receiving module is used to receive target data in encrypted form sent by each participating node; the target data includes standardized data fragments, data density indicators, and anomaly marker information;

[0041] The first determination module is used to determine the security weight of each participating node based on the data density index of each participating node.

[0042] The second determining module is used to determine the threshold parameter by utilizing the data density index and anomaly marking information of each participating node; the threshold parameter is the threshold parameter of the task node participating in the secure multi-party computation task.

[0043] The partitioning module is used to divide each standardized data shard into at least one sub-batch of data according to the spatiotemporal characteristics based on the security weight of each participating node, the anomaly marking information and the threshold parameter, and to determine the task node of each sub-batch of data.

[0044] The first sending module is used to send the corresponding secure multi-party computation task to the task nodes of each sub-batch of data.

[0045] The calculation module is used to calculate the global calculation result based on the local aggregation results of each sub-batch data.

[0046] Fourthly, this application also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0047] Memory, used to store computer programs;

[0048] When a processor executes a program stored in memory, it implements the steps of the above-described secure multi-party computation method.

[0049] Fifthly, this application also provides a computer storage medium storing computer-executable instructions for performing the steps of the secure multi-party computation method described above.

[0050] Compared with the prior art, the technical solutions provided in this application have the following advantages: This application discloses a safe multi-party computation method, device, equipment, and medium based on intelligent transportation. In the safe multi-party computation method provided in this solution, the safety weight of each participating node can be determined by the data density index of each participating node. By using the safety weight of each participating node, the differences in data contribution of each participating node can be balanced, thereby offsetting the deviation caused by uneven data. This solution can also adjust the threshold parameters of task nodes by using the data density index and anomaly marking information of participating nodes, thereby ensuring that the multi-party computation task can still be completed even when the data of some participating nodes is unstable. This solution can also divide each standardized data into sub-batch data by using safety weight, anomaly marking information, and threshold parameters, and assign task nodes to each sub-batch data, thereby ensuring that each sub-batch data performs the safe multi-party computation task through a sufficient number of task nodes with high safety weight and few anomalies, thereby improving the task completion rate and accuracy of the system in a highly dynamic traffic environment. Attached Figure Description

[0051] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0054] Figure 1 A schematic diagram of a secure multi-party computation method based on intelligent transportation is provided for an embodiment of this application;

[0055] Figure 2 A schematic diagram of a secure multi-party computing device based on intelligent transportation is provided in an embodiment of this application.

[0056] Figure 3 This is a schematic diagram of an electronic device structure provided in an embodiment of this application. Detailed Implementation

[0057] In existing solutions, addressing data imbalance primarily relies on external engineering scheduling, such as data sharding, load balancing, and fault-tolerant retransmission. While these methods are generally applicable across different scenarios, they lack sufficiently sophisticated compensation mechanisms for the specific needs of intelligent transportation. For instance, they fail to effectively compensate for data imbalances caused by high-speed vehicle movement or significant regional density variations. Furthermore, these compensation mechanisms are not adaptively controlled within the MPC (Secure Multi-party Computation) algorithm. Traditional secure multi-party computation protocols typically assume stable and balanced input data. Therefore, in intelligent transportation scenarios, issues such as high-speed vehicle movement and uneven data distribution can lead to inaccurate computation results during secure multi-party computation.

[0058] Therefore, in this embodiment, a secure multi-party computation method, apparatus, device, and medium based on intelligent transportation are provided to improve the accuracy of data multi-party computation.

[0059] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0060] The following disclosure provides numerous different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the scope of the invention. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed.

[0061] See Figure 1 This is a schematic diagram of a secure multi-party computation method based on intelligent transportation, provided in an embodiment of this application. The secure multi-party computation method is applied to cloud nodes and specifically includes the following steps:

[0062] S101, Receive target data in encrypted form sent by each participating node; the target data includes standardized data fragments, data density indicators and anomaly marking information;

[0063] In this application, participating nodes refer to all data providers that actually participate in secure multi-party computation, including: vehicle terminals with autonomous data collection capabilities, roadside units (RSUs) deployed in fixed areas, and edge server nodes (such as regional control nodes). These nodes are usually computation participants in the MPC protocol, so this application refers to each computation participant as a participating node.

[0064] Each participating node in this application needs to collect raw traffic data and perform preprocessing, index calculation, and anomaly marking on the raw traffic data. When collecting raw traffic data, the vehicle terminal mainly collects location information, speed, acceleration, and driving behavior data, and uses local caching, breakpoint resume, and real-time interpolation compensation to cope with signal switching and short-term interruptions caused by high-speed movement; the roadside unit mainly collects road conditions, traffic flow, and environmental data in fixed areas to supplement the vehicle terminal's shortcomings in low-density areas.

[0065] After each participating node collects raw traffic data, preprocessing operations such as time synchronization, noise filtering, and format standardization need to be performed locally at each node to generate standardized data fragments. The preprocessing operations are explained below: 1. Time Synchronization: Introducing a unified time reference for participating nodes such as vehicle terminals and road test units, correcting the timestamps of each data collection to a unified timeline, supporting microsecond-level synchronization accuracy; 2. Noise Filtering: Applying moving average, bidirectional exponential smoothing, or median filtering methods to the raw traffic data sequence to filter out short-term jumps or pulses introduced by sensor or communication interference; 3. Format Standardization: Converting heterogeneous data from each terminal (such as sensor ID (Identity document, unique code), unit, and format) into a unified data structure format, for example, fixed fields such as timestamp, data type, value, and flag bits, to meet the needs of subsequent secure multi-party computation processing.

[0066] After generating standardized data fragments, this application requires real-time anomaly detection of data interruptions or delays caused by high-speed movement, utilizing vehicle motion status, communication quality information, and data upload latency. Abnormal data segments are then marked, generating anomaly marker bits for the standardized data fragments, and anomaly marking information is generated based on these marker bits. This anomaly marking information includes the location of the abnormal data segment, the number of times the abnormal data segment occurs, etc. In this application, abnormal data segments can be those exhibiting continuous packet loss, communication interruption, or abnormally increased sampling intervals. Upon the occurrence of such abnormal data segments, they are detected and recorded in real-time by participating nodes. In this application, the anomaly marking operation is performed locally on the participating nodes, and the anomaly marking information is sent as additional metadata along with the standardized data fragments to the cloud node.

[0067] It should be noted that during the real-time data acquisition process, if the vehicle terminal detects abnormal data segments with continuous packet loss or communication interruption, this application records the interval as an "abnormal window" and marks it with an abnormal flag bit. Furthermore, this application can also use interpolation to compensate for this interval, while the abnormal flag information is still recorded and reported with the data for subsequent abnormal weighting, threshold adjustment, or data removal strategies in MPC. The interpolated data can be used to improve computational continuity, but it will not cover up the abnormal behavior of the original data. The "abnormal flag" and "data compensation" in this application are parallel and complementary.

[0068] Furthermore, in this application, after each participating node generates standardized data shards, it is also necessary to calculate data density indicators based on the standardized data shards. These data density indicators may include data quality indicators and data quantity indicators, etc.

[0069] Through the above process, each participating node will generate standardized data shards, data density indicators, and anomaly marking information locally. It will also generate target data in a secret state through a secret sharing mechanism, and upload the target data to the cloud node. The cloud node will then aggregate the target data from all participating nodes in the secret sharing environment and use the target data to complete the subsequent determination of security weights, threshold parameters, etc., for use in weighted aggregation and scheduling decisions during subsequent multi-party computation.

[0070] S102. Determine the security weight of each participating node based on the data density index of each participating node;

[0071] In this application, after each participating node completes data preprocessing locally, it calculates its own data density indicators, such as data quality indicators and data quantity indicators. Then, each participating node uploads the calculated data density indicators to the cloud node after secret sharing encoding. The cloud node receives the data density indicators sent by each participating node and executes a weighting function in the secret sharing environment to uniformly calculate the security weights of all participating nodes. It should be noted that in this application, each participating node only uploads the encrypted form of its local data density indicators and does not need to exchange plaintext information with each other.

[0072] The security weight in this application is closely related to the data density index. If the data density index of a participating node is large, then the security weight of that participating node is also relatively large. The difference in the security weight of each participating node can reflect the degree of data contribution of each participating node. Therefore, this application introduces security weight in secure multi-party computation, which can balance the differences in data contribution of each participating node in secret sharing computation, thereby offsetting the deviation caused by the difference in the amount of data in the region.

[0073] S103. Determine the threshold parameter using the data density index and anomaly marker information of each participating node; this threshold parameter is the threshold parameter for the task node participating in the secure multi-party computation task.

[0074] In this application, after obtaining the security weights of each participating node, it is also necessary to dynamically adjust the threshold parameters in the secret sharing protocol based on data density indicators and anomaly marker information. These threshold parameters are the task node threshold parameters for the secure multi-party computation task; that is, they control the minimum number of task nodes' encrypted data required to complete the computation in each round of secure multi-party computation. This application, through adaptive threshold adjustment, can enhance system fault tolerance and ensure uninterrupted multi-party computation by dynamically lowering the threshold parameters in scenarios such as high-speed vehicle movement and network fluctuations, where individual participating nodes may experience data loss or communication interruptions. Conversely, when the overall data quality of the participating parties is stable, the threshold parameters can be increased to enhance protocol security.

[0075] For example, if there are currently 10 task nodes performing multi-party computation, and the original threshold parameter is 7, meaning the original system requires data from at least 7 participating nodes to complete the computation, but in this round of computation, data uploaded by 3 participating nodes is marked as abnormal by the system, such as incomplete data or unstable signals, then to prevent computation interruption, the system will automatically lower the threshold parameter to 6. This means that as long as data from 6 participating nodes is valid, this round of computation can be completed by these 6 nodes. In this way, even if some participating nodes experience abnormalities due to high-speed movement or network instability, the computation of the entire system can still continue without failure.

[0076] S104. Based on the security weight, anomaly marking information and threshold parameters of each participating node, divide each standardized data shard into at least one sub-batch of data according to its spatiotemporal characteristics, and determine the task node for each sub-batch of data.

[0077] In this application, after the cloud node calculates the security weights and threshold parameters of each participating node through the above process, it can combine anomaly marking information to divide the data into multiple sub-batch data according to time and spatial characteristics. This application divides the data according to time and spatial characteristics, ensuring that the sub-batch data are similar in time and space. Through security weights, anomaly marking information, and threshold parameters, several task nodes with high data quality and few anomalies that meet the threshold parameter requirements can be selected, thereby ensuring that the data quality used for secure multi-party computation also meets the standards, ensuring that subsequent multi-party computation can proceed smoothly.

[0078] Furthermore, cloud nodes can monitor the computational load and network status of each participating node in real time, and dynamically adjust task allocation using anomaly marker information, security weights, and threshold parameters to determine the task nodes participating in task computation. For example, the anomaly marker information of each participating terminal indicates that a vehicle terminal is experiencing data instability due to high-speed movement. In this case, some tasks of the original vehicle terminal can be transferred to the roadside or cloud, and anomaly verification and secure retransmission mechanisms can be triggered to ensure computational continuity and result accuracy.

[0079] S105. Send the corresponding secure multi-party computation task to the task node of each sub-batch data, and calculate the global computation result based on the local aggregation result of each sub-batch data.

[0080] After determining the data for each sub-batch and its corresponding task nodes, this application allows for localized secure multi-party computation (MPC) to generate localized aggregation results, which are then aggregated to form the global computation result. The localized aggregation result in this application is the result of a single round of secure MPC completed by multiple task nodes for a specific sub-batch of data (e.g., a specific area within a specific time period). This secure MPC task can calculate factors such as average vehicle speed, congestion score, and safety risk factors for a road segment. This process involves task nodes providing dense data, which is then organized by the MPC system of the cloud nodes to complete the relevant calculations. By decomposing the global computation task into localized computation and global aggregation, this application reduces the load on each task node and effectively addresses the issue of data imbalance across different regions.

[0081] Furthermore, to enhance data security, this application can package MPC local aggregation results, global calculation results, security weights, threshold parameters, and fault-tolerant processing proof data (proof information generated through Merkle trees, multi-signature technology, or zero-knowledge proofs) into batch result packages and upload them to the blockchain. The authenticity and integrity of the results are verified through smart contracts, thereby achieving the immutability and full auditing of the data calculation process.

[0082] Cloud nodes can obtain verified global computation results from the blockchain and distribute them to participating nodes, which then update their local states accordingly. If a participating node detects inaccurate global computation results, a callback or retransmission mechanism can be automatically triggered. The cloud node then retrieves the target data from the participating nodes again, further adjusting its internal security weighting and adaptive threshold strategies to provide a basis for improvement for the next round of computation.

[0083] In summary, the secure multi-party computation method provided in this application can determine the security weight of each participating node through its data density index. This security weight balances the differences in data contributions among participating nodes, thus offsetting biases caused by uneven data distribution. Furthermore, this scheme can adjust the threshold parameters of task nodes using the data density index and anomaly marking information of participating nodes, ensuring that multi-party computation tasks can still be completed even when some participating nodes have unstable data. Additionally, this scheme can divide standardized data into sub-batch data segments using security weights, anomaly marking information, and threshold parameters, and assign task nodes to each sub-batch data segment. This ensures that each sub-batch data segment executes secure multi-party computation tasks through a sufficient number of task nodes with high security weights and fewer anomalies, improving the system's task completion rate and accuracy in highly dynamic traffic environments.

[0084] In another embodiment of this application, the process of determining the security weight of each participating node based on the data density index of each participating node specifically includes the following steps: determining the data density index of each participating node; the data density index includes: data quality index and data quantity index; calculating a comprehensive index based on the data quality index and data quantity index of each participating node; and calculating the security weight of each participating node using the comprehensive index of each participating node.

[0085] In this application, the data density metrics for each participating node include: data quality metrics and data quantity metrics, which can be obtained through Q. i This represents the data quality metric for the i-th participating node, expressed as d. i The data quantity metric for the i-th participating node is represented by the data quality metric Q. i This represents the average validity and completeness of the data collected by participating node i. The evaluation dimensions include: data continuity (uninterrupted sampling rate), sampling interval stability (time interval variance), missing rate (percentage of null values), and communication signal strength level. After standardizing these multidimensional features, data quality indicators can be determined through a custom weighted average calculation method; the data quantity indicator d... i This refers to the number of data entries collected and successfully uploaded by participating node i within one computation cycle (e.g., 5 seconds / 10 seconds).

[0086] When calculating the security weight of each participating node, this application first needs to calculate a comprehensive index using data quality indicators and data quantity indicators. This application defines a calculation function f(Q) i d i The comprehensive index can be calculated using the function: f(Q) i d i )=Q i ×d iIn other words, the comprehensive index is determined by multiplying the data quality index and the data quantity index. The larger the values ​​of the data quality index and the data quantity index, the larger the calculated security weight. This calculation function can also be f(Q). i ,d i ) = log(1 + Q i ×d i This is not specifically limited to this. This application is approved by w i The security weight of the i-th participating node is defined as follows:

[0087]

[0088] In summary, this application can dynamically calculate the security weight of each participating node in a confidential sharing environment by using the data quality and data quantity indicators of each participating node. This security weight can be used in operations such as determining threshold parameters, dividing sub-batch data, and allocating task nodes, thereby balancing the contribution of data from different regions and offsetting the bias caused by uneven data distribution.

[0089] In another embodiment of this application, the process of determining the threshold parameter using the data density index and anomaly marking information of each participating node specifically includes: determining the global contribution using the data density index of each participating node; determining the anomaly percentage using the anomaly marking information of each participating node; and adjusting the threshold parameter based on the global contribution and anomaly percentage of each participating node.

[0090] In this application, the cloud node can aggregate the data density metrics uploaded by each participating node and determine the global contribution of each participating node based on these metrics. The calculation formula for the global contribution can be customized and is not specifically limited here, as long as it satisfies the rule that: the larger the data density metric value, the larger the global contribution value; and the smaller the data density metric value, the smaller the global contribution value. The cloud node can also aggregate the anomaly marking information uploaded by each participating node and determine the anomaly percentage of each participating node based on this information. The calculation formula for the anomaly percentage can be customized and is not specifically limited here, as long as it satisfies the rule that: the more anomalies recorded in the anomaly marking information, the larger the anomaly percentage value; and the fewer anomalies recorded in the anomaly marking information, the smaller the anomaly percentage value.

[0091] Furthermore, after determining the global contribution and anomaly percentage of each participating node, the threshold parameter can be adjusted based on these parameters. For example, if a large number of participating nodes have a high anomaly percentage, the threshold parameter value is dynamically lowered; if the anomaly percentage of each participating node is low, the threshold parameter value is dynamically increased. If the global contribution of each node is relatively high, the threshold parameter value is dynamically increased; if only some nodes have a relatively high global contribution, the threshold parameter value is dynamically decreased.

[0092] In summary, this application can ensure that multi-party computation tasks can still be completed even when the data of some participants is unstable by dynamically adjusting the threshold parameters through data density indicators and anomaly marking information, thereby enhancing the fault tolerance of the system.

[0093] In another embodiment of this application, the process of dividing each standardized data shard into at least one sub-batch of data according to its spatiotemporal characteristics based on the security weights of each participating node, anomaly marking information, and the threshold parameters, and determining the task nodes for each sub-batch of data, includes:

[0094] Each standardized data segment is divided into at least one batch of data according to time.

[0095] Divide each batch of data into at least one group based on geographical location;

[0096] Initial participating nodes are determined from the participating nodes corresponding to each group of data; the initial participating nodes are determined using the security weight and anomaly marking information of each participating node;

[0097] Determine whether the number of initial participating nodes in each group of data exceeds the threshold parameter;

[0098] If yes, then each group of data will be used as the data for each sub-batch, and the initial participating nodes of each group of data will be used as the task nodes of each sub-batch of data; if no, then regions will be merged according to geographical location, at least one group of data will be re-divided, and the step of determining the initial participating nodes from the participating nodes corresponding to each group of data will continue.

[0099] In this application, the security weights of each participating node allow us to understand the data quality and quantity of each node. Threshold parameters determine the minimum number of task nodes required for each round of secure multi-party computation. Standardized data sharding determines the collection time and location of each data entry. Therefore, when dividing at least one batch of data by time, this application segments the data according to the collection time of each data entry within the standardized data shards, such as dividing data into batches every 30 seconds or every minute. When dividing at least one group of data by geographical location, this application analyzes the data based on its collection location, such as dividing adjacent road segments or areas into groups. Then, within each group, the participating nodes uploading the group data are identified, and preliminary participating nodes are selected based on their security weights and anomaly flags. These preliminary participating nodes are those with high security weights and few anomaly flags; these are considered high-quality nodes. This method allows for the dynamic adjustment of initial participating nodes based on security weights and anomaly markers when communication or computing resources become unstable due to high-speed movement of some vehicle terminals. This enables the automatic transfer of some tasks to other high-quality nodes via the protocol, thereby balancing the contributions of each participating node's data, ensuring the continuous execution of multi-party computation and the accuracy of the overall results.

[0100] Furthermore, this application also needs to detect whether the number of initially selected participating nodes meets the threshold parameter. If the number of initially selected participating nodes exceeds the threshold parameter, each group of data is directly used as the data of each sub-batch, and the initially selected participating nodes of each group of data are used as the task nodes of each sub-batch of data. The more task nodes there are, the more detailed the secure multi-party computation task is divided, thereby reducing the computational pressure on each task node. If the number of initially selected participating nodes is less than the threshold parameter, it means that the currently selected initially selected participating nodes are insufficient to complete the secure multi-party computation task. At this time, regions can be merged according to geographical location to expand the region range, increase the redundancy coefficient, re-divide at least one group of data, re-select new initially selected participating nodes, and continue to execute the step of determining the initially selected participating nodes from the participating nodes corresponding to each group of data.

[0101] In another embodiment provided in this application, when each task node executes the secure multi-party computation task, it first needs to determine the sub-batch data to be processed corresponding to the secure multi-party computation task. Then, based on the sub-batch data to be processed and the security weight of the current node, the secure multi-party computation task is executed to generate a local aggregation result. For example, this application divides each standardized data into multiple sub-batch data: B1, B2, ..., B m Let m represent the total number of sub-batch data. For the j-th sub-batch data, its local aggregation result is:

[0102]

[0103] Among them, w i x represents the security weight of the i-th task node in the j-th sub-batch data. i This represents the standardized data fragment of the i-th task node within the j-th sub-batch data. Then, all local aggregation results are safely summarized to obtain the global computation result:

[0104]

[0105] It should be noted that the cloud node in this application serves as the scheduling center, monitoring the computational load and network status of each participating node in real time within the secure MPC protocol. After dividing the data into sub-batch data and determining task nodes through security weights, anomaly markers, and threshold parameters, this application dynamically adjusts the task allocation for each node to ensure a balanced contribution of data from different regions. For example, task nodes with high security weights and low anomaly rates are prioritized for core computational tasks. For nodes with high anomaly rates, the system can adjust their task roles, allowing them to participate only in redundant or fault-tolerant computations. If necessary, their task allocation can be suspended to avoid affecting computational accuracy and the continuity of protocol execution. This dynamic scheduling mechanism ensures a stable and controllable node composition within each sub-batch, improving the system's task completion rate in highly dynamic traffic environments.

[0106] Furthermore, each task node is embedded with an anomaly verification module. This anomaly verification module can be collaboratively executed by the task node group (including cloud, roadside, and vehicle) participating in MPC during the operation of the MPC protocol. It can automatically remove low-quality or missing data based on anomaly marking information or use redundant data for error compensation, and trigger a secure retransmission mechanism when necessary.

[0107] In summary, this application proposes an adaptive data balancing mechanism originating from within the MPC algorithm. Utilizing mechanisms such as security weighting, adaptive threshold adjustment, hierarchical computation, and dynamic task scheduling, this application provides an internal adaptive compensation scheme specifically addressing the problems of unstable data collection, computational deviations, and communication interruptions caused by high-speed vehicle movement and data imbalance in intelligent transportation. This achieves dynamic balancing and compensation of the data contributions of each participant, thereby ensuring that multi-party computation tasks can be completed continuously, efficiently, and accurately in intelligent transportation scenarios, while protecting the privacy of all parties' data.

[0108] In another embodiment of this application, a secure multi-party computation method based on intelligent transportation is disclosed. This secure multi-party computation method is applied to participating nodes and includes the following steps:

[0109] The collected raw traffic data is preprocessed to generate standardized data fragments;

[0110] Determine data density indices using standardized data fragments;

[0111] Perform anomaly detection on each standardized data fragment and generate anomaly marker information;

[0112] Based on the standardized data sharding, the data density index, and the anomaly marker information, dense target data is generated and sent to the cloud node; the cloud node is used to execute the secure multi-party computation method described in any of the above method embodiments;

[0113] If a secure multi-party computation task is received from the cloud node, the secure multi-party computation task is executed, and the local aggregation results of each sub-batch data are sent to the cloud node so that the cloud node can calculate the global computation result based on each local aggregation result.

[0114] In another embodiment of this application, anomaly detection is performed on each standardized data segment to generate anomaly marking information, including: detecting abnormal data in each standardized data segment; marking the abnormal data to generate anomaly marking information; and interpolating and compensating the abnormal data in each standardized data segment.

[0115] It should be noted that the secure multi-party computation method applied to participating nodes described in the embodiments of this application has been described in detail in any of the above method embodiments, and will not be elaborated here.

[0116] See Figure 2 , Figure 2 This application provides a schematic diagram of a secure multi-party computation device structure based on intelligent transportation. The device is applied to a cloud node and specifically includes:

[0117] The receiving module 11 is used to receive target data in encrypted form sent by each participating node; the target data includes standardized data fragments, data density indicators and anomaly marking information;

[0118] The first determining module 12 is used to determine the security weight of each participating node based on the data density index of each participating node.

[0119] The second determining module 13 is used to determine the threshold parameter by using the data density index and anomaly marking information of each participating node; the threshold parameter is the threshold parameter of the task node participating in the secure multi-party computation task.

[0120] The partitioning module 14 is used to divide each standardized data shard into at least one sub-batch of data according to the spatiotemporal characteristics based on the security weight of each participating node, the anomaly marking information and the threshold parameter, and to determine the task node of each sub-batch of data.

[0121] The first sending module 15 is used to send the corresponding secure multi-party computation task to the task node of each sub-batch data;

[0122] The calculation module 16 is used to calculate the global calculation result based on the local aggregation results of each sub-batch data.

[0123] In another embodiment of this application, the first determining module includes:

[0124] The first determining unit is used to determine the data density index of each participating node; the data density index includes: data quality index and data quantity index.

[0125] The first calculation unit is used to calculate the comprehensive index based on the data quality index and data quantity index of each participating node;

[0126] The second calculation unit is used to calculate the security weight of each participating node using the comprehensive index of each participating node.

[0127] In another embodiment of this application, the second determining module includes:

[0128] The second determining unit is used to determine the global contribution using the data density index of each participating node;

[0129] The third determining unit is used to determine the percentage of anomalies by utilizing the anomaly marking information of each participating node;

[0130] The adjustment unit is used to adjust the threshold parameters based on the global contribution and anomaly rate of each participating node.

[0131] In another embodiment of this application, the partitioning module includes:

[0132] The first partitioning unit is used to divide each standardized data fragment into at least one batch of data according to time.

[0133] The second partitioning unit is used to divide each batch of data into at least one group of data according to geographical location;

[0134] The fourth determining unit is used to determine the initial participating nodes from the participating nodes corresponding to each group of data; the initial participating nodes are determined using the security weight and anomaly marking information of each participating node;

[0135] The judgment unit is used to determine whether the number of initial selection participating nodes of each group of data exceeds the threshold parameter; if so, each group of data is used as each sub-batch of data, and the initial selection participating nodes of each group of data are used as the task nodes of each sub-batch of data; if not, the second division unit is triggered to merge regions according to geographical location, re-divide at least one group of data, and trigger the fourth determination unit.

[0136] In another embodiment of this application, a secure multi-party computation device based on intelligent transportation is also disclosed. This device is applied to participating nodes and includes:

[0137] The preprocessing module is used to preprocess the collected raw traffic data and generate standardized data fragments;

[0138] The third determination module is used to determine the data density index using each standardized data fragment;

[0139] The detection module is used to perform anomaly detection on each standardized data fragment and generate anomaly marker information;

[0140] The second sending module is used to generate target data in a dense state according to the standardized data fragmentation, the data density index and the anomaly marking information, and send it to the cloud node; the cloud node is used to execute the above-mentioned secure multi-party computation method based on the cloud node.

[0141] The execution module is used to receive the secure multi-party computation task issued by the cloud node and execute the secure multi-party computation task;

[0142] The third sending module is used to send the local aggregation results of each sub-batch data to the cloud node, so that the cloud node can calculate the global calculation result based on each local aggregation result.

[0143] In another embodiment of this application, the detection module includes:

[0144] The detection unit is used to detect abnormal data in each standardized data segment;

[0145] A marking unit is used to mark the abnormal data and generate abnormal marking information;

[0146] The compensation unit is used to interpolate and compensate for abnormal data in each standardized data segment.

[0147] In another embodiment of this application, the execution module includes:

[0148] The fifth determining unit is used to determine the sub-batch data to be processed corresponding to the secure multi-party computation task;

[0149] The execution unit is used to execute the secure multi-party computation task based on the sub-batch data to be processed and the security weight of the current node, and generate a local aggregation result.

[0150] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0151] See Figure 3 , Figure 3 A schematic diagram of an electronic device structure provided in this application embodiment includes a processor 21, a communication interface 22, a memory 23 and a communication bus 24, wherein the processor 21, the communication interface 22 and the memory 23 communicate with each other through the communication bus 24;

[0152] Memory 23 is used to store computer programs;

[0153] When the processor 21 executes the program stored in the memory 23, it implements the steps of the secure multi-party computation method described in any of the above method embodiments, which will not be repeated here.

[0154] Wherein, if the electronic device is a cloud node, the electronic device executes the above-mentioned secure multi-party computation method based on cloud nodes, and the cloud node can be a server; if the electronic device is a participating node, when the electronic device executes the above-mentioned secure multi-party computation method based on participating nodes, the participating node can be a vehicle terminal, road test unit, etc.

[0155] The communication bus mentioned in the above terminal can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0156] The communication interface is used for communication between the aforementioned terminal and other devices.

[0157] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0158] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0159] In another exemplary embodiment, a computer storage medium is also provided, wherein the program instructions, when executed by a processor, implement the steps of the secure multi-party computation method described in any of the above method embodiments. The storage medium may include various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0160] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments, and will not be repeated here.

[0161] It should be understood that the terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “described” as used herein may also include the plural forms. The terms “comprising,” “including,” “containing,” and “having” are inclusive and therefore indicate the presence of the stated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not construed as requiring them to be performed in a particular order described or illustrated unless the order of performance is explicitly indicated. It should also be understood that additional or alternative steps may be used.

[0162] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A secure multi-party computation method based on intelligent transportation, characterized in that, The secure multi-party computation method is applied to cloud nodes, and the secure multi-party computation method includes: Receive target data in encrypted form sent by each participating node; the target data includes standardized data fragments, data density indices, and anomaly marker information; The security weight of each participating node is determined based on its data density index. Threshold parameters are determined using the data density indices and anomaly marker information of each participating node; the threshold parameters are the threshold parameters of the task nodes participating in the secure multi-party computation task. Based on the security weights, anomaly marking information, and threshold parameters of each participating node, each standardized data shard is divided into at least one sub-batch of data according to its spatiotemporal characteristics, and the task nodes for each sub-batch of data are determined. Send the corresponding secure multi-party computation task to the task node of each sub-batch of data, and calculate the global computation result based on the local aggregation result of each sub-batch of data.

2. The secure multi-party computation method according to claim 1, characterized in that, The process of determining the security weight of each participating node based on its data density index includes: Determine the data density metrics for each participating node; the data density metrics include: data quality metrics and data quantity metrics; A comprehensive index is calculated based on the data quality and data quantity indicators of each participating node. The security weight of each participating node is calculated using the comprehensive index of each participating node.

3. The secure multi-party computation method according to claim 1, characterized in that, Using the data density metrics and anomaly marker information of each participating node, threshold parameters are determined, including: The global contribution is determined using the data density index of each participating node; The percentage of anomalies is determined by using the anomaly marking information of each participating node; The threshold parameters are adjusted based on the global contribution and anomaly percentage of each participating node.

4. The secure multi-party computation method according to claim 1, characterized in that, The step of dividing each standardized data shard into at least one sub-batch of data according to its spatiotemporal characteristics based on the security weights of each participating node, anomaly marking information, and the threshold parameters, and determining the task nodes for each sub-batch of data, includes: Each standardized data segment is divided into at least one batch of data according to time. Divide each batch of data into at least one group based on geographical location; Initial participating nodes are determined from the participating nodes corresponding to each group of data; the initial participating nodes are determined using the security weight and anomaly marking information of each participating node; Determine whether the number of initial participating nodes in each group of data exceeds the threshold parameter; If yes, then each group of data will be used as the data for each sub-batch, and the initial participating nodes of each group of data will be used as the task nodes of each sub-batch of data; if no, then regions will be merged according to geographical location, at least one group of data will be re-divided, and the step of determining the initial participating nodes from the participating nodes corresponding to each group of data will continue.

5. A method for secure multi-party computation based on intelligent transportation, characterized in that, The secure multi-party computation method is applied to participating nodes, and the secure multi-party computation method includes: The collected raw traffic data is preprocessed to generate standardized data fragments; Determine data density indices using standardized data fragments; Perform anomaly detection on each standardized data fragment and generate anomaly marker information; Based on the standardized data sharding, the data density index, and the anomaly marking information, dense target data is generated and sent to the cloud node; the cloud node is used to execute the secure multi-party computation method according to any one of claims 1 to 4. If a secure multi-party computation task is received from the cloud node, the secure multi-party computation task is executed, and the local aggregation results of each sub-batch data are sent to the cloud node so that the cloud node can calculate the global computation result based on each local aggregation result.

6. The secure multi-party computation method according to claim 5, characterized in that, The process of performing anomaly detection on each standardized data fragment and generating anomaly marker information includes: Detect outliers in each standardized data segment; The abnormal data is marked to generate abnormal marking information; Interpolation compensation is performed on the abnormal data in each standardized data segment.

7. The secure multi-party computation method according to claim 5, characterized in that, The execution of the secure multi-party computation task includes: Determine the sub-batch data to be processed corresponding to the secure multi-party computation task; Based on the sub-batch data to be processed and the security weight of the current node, the secure multi-party computation task is executed to generate a local aggregation result.

8. A secure multi-party computing device based on intelligent transportation, characterized in that, The secure multi-party computation device is applied to a cloud node, and the secure multi-party computation device includes: The receiving module is used to receive target data in encrypted form sent by each participating node; the target data includes standardized data fragments, data density indicators, and anomaly marker information; The first determination module is used to determine the security weight of each participating node based on the data density index of each participating node. The second determining module is used to determine the threshold parameter by utilizing the data density index and anomaly marking information of each participating node; the threshold parameter is the threshold parameter of the task node participating in the secure multi-party computation task. The partitioning module is used to divide each standardized data shard into at least one sub-batch of data according to the spatiotemporal characteristics based on the security weight of each participating node, the anomaly marking information and the threshold parameter, and to determine the task node of each sub-batch of data. The first sending module is used to send the corresponding secure multi-party computation task to the task nodes of each sub-batch of data. The calculation module is used to calculate the global calculation result based on the local aggregation results of each sub-batch data.

9. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the steps of the secure multi-party computation method according to any one of claims 1 to 8.

10. A computer storage medium, characterized in that, The computer storage medium stores computer-executable instructions for performing the steps of the secure multi-party computation method according to any one of claims 1 to 8 of this application.

Images