Multi-agent interaction security verification method and device, equipment, storage medium and program product

Abstract

This application relates to a secure verification method, apparatus, device, storage medium, and program product for multi-agent interaction. The method includes: acquiring state change events of multiple agents for each interaction stage of a target interaction scenario; each state change event includes at least one of a state change type, a change timestamp, and a post-change state identifier; determining target verification rules and target credential generation rules based on the state change events; performing secure verification on the agents participating in the interaction based on the target verification rules; and, if the security verification passes, acquiring the context parameters of the sending node, fusing the context parameters according to the target credential generation rules to obtain the interaction credential for the interaction stage; the interaction credential is used to characterize the verifiability of the operation during the interaction stage. By loading corresponding verification rules and credential generation rules based on the agent states and performing targeted verification on the agents participating in the interaction based on the verification rules, data security is improved.

Description

Technical Field

[0001] This application relates to the field of intelligent communication technology, and in particular to a secure verification method, apparatus, device, storage medium, and program product for multi-agent interaction. Background Technology

[0002] With the rapid development of artificial intelligence technology, multi-agent systems have been widely applied in various fields such as industrial manufacturing, intelligent transportation, financial services, and smart cities. These systems can significantly improve task execution efficiency, resource utilization, and decision-making speed through the collaboration and interaction of multiple agents. For example, in intelligent manufacturing, multiple agents can collaboratively optimize production processes and improve production efficiency. These applications not only enhance the intelligence of the systems but also drive the digital transformation and intelligent upgrading of industries.

[0003] However, in the process of multi-agent collaboration, data interaction is a core link and faces severe challenges. Since multiple agents need to frequently exchange a large amount of sensitive information during the interaction process, how to ensure data security and privacy protection has become an urgent problem to be solved. Summary of the Invention

[0004] Therefore, it is necessary to provide a secure verification method, apparatus, device, storage medium, and program product for multi-agent interaction that can improve the security of multi-agent data interaction processes, addressing the aforementioned technical problems.

[0005] Firstly, this application provides a security verification method for multi-agent interaction, including:

[0006] For each interaction stage of the target interaction scenario, multiple agent state change events are acquired. Each state change event includes at least one of the following: state change type, change timestamp, and post-change state identifier.

[0007] Determine the target verification rules and target credential generation rules based on the state change events;

[0008] Security verification is performed on the intelligent agents participating in the interaction based on the target verification rules;

[0009] If the security verification is successful, the context parameters of the sending node are obtained, and the context parameters are fused according to the target credential generation rules to obtain the interaction credential for the interaction phase; the interaction credential is used to characterize the verifiability of the operation in the interaction phase.

[0010] In one embodiment, security verification of the agents participating in the interaction is performed based on target verification rules, including:

[0011] Acquire the agent's attribute information and interaction risk level. The attribute information includes at least one of the agent's role, data sensitivity level, and the participant's historical credit score.

[0012] The attribute information is processed based on the target validation rules to obtain the first validation method;

[0013] The interaction risk level is processed based on the target verification rules to obtain the second verification method;

[0014] The agents participating in the interaction are verified according to the first verification method and / or the second verification method.

[0015] In one embodiment, security verification of the intelligent agent participating in the interaction is performed according to a first verification method, including:

[0016] When the first verification method is the high verification strength method, a signature verification request is sent to the sender node, receiver node and supervisor node respectively, and a signature verification response is received from the sender node, receiver node and supervisor node. When multiple signature verification responses are all passed, the security verification is determined to be successful.

[0017] When the first verification method is the low verification strength method, a signature verification request is sent to the sender node and the receiver node, and a signature verification response is received from the sender node and the receiver node. If multiple signature verification responses are all successful, the security verification is determined to be successful.

[0018] In one embodiment, security verification of the intelligent agent participating in the interaction is performed according to a second verification method, including:

[0019] When the second verification method is a high-risk operation verification method, a strict verification process is adopted to perform security verification on the intelligent agent;

[0020] When the second verification method is a low-risk operation verification method, a rapid verification process is used to perform security verification on the intelligent agent.

[0021] In one embodiment, the context parameters are fused according to the target credential generation rule to obtain the interaction credential, including:

[0022] Obtain the first verification parameter of the current interaction stage and the second verification parameter of the previous stage;

[0023] According to the target credential generation rules, the first verification parameter and the second verification parameter are bound together by a hash chain to generate a staged credential for the current interaction stage.

[0024] The phased credentials and context parameters are solidified to generate interaction credentials; the context parameters include the algorithm version number, network environment parameters, and a summary of the negotiation terms.

[0025] In one embodiment, determining the target verification rule and the target credential generation rule based on the state change event includes:

[0026] Based on the mapping relationship between status identifiers and rules, the target verification rules and target credential generation rules corresponding to the changed status identifiers are determined.

[0027] Secondly, this application also provides a security verification device for multi-agent interaction, comprising:

[0028] The acquisition module is used to acquire state change events of multiple agents for each interaction stage of the target interaction scenario. The state change event includes at least one of the following: state change type, change timestamp, and post-change state identifier.

[0029] The determination module is used to determine the target verification rules and target credential generation rules based on the state change event;

[0030] The verification module is used to perform security verification on the intelligent agents participating in the interaction based on the target verification rules;

[0031] The generation module is used to obtain the context parameters of the sending node after the security verification is passed, and to perform fusion processing on the context parameters according to the target credential generation rules to obtain the interaction credential for the interaction stage; the interaction credential is used to characterize the verifiability of the operation in the interaction stage.

[0032] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0033] For each interaction stage of the target interaction scenario, multiple agent state change events are acquired. Each state change event includes at least one of the following: state change type, change timestamp, and post-change state identifier.

[0034] Determine the target verification rules and target credential generation rules based on the state change events;

[0035] Security verification is performed on the intelligent agents participating in the interaction based on the target verification rules;

[0036] If the security verification is successful, the context parameters of the sending node are obtained, and the context parameters are fused according to the target credential generation rules to obtain the interaction credential for the interaction phase; the interaction credential is used to characterize the verifiability of the operation in the interaction phase.

[0037] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0038] For each interaction stage of the target interaction scenario, multiple agent state change events are acquired. Each state change event includes at least one of the following: state change type, change timestamp, and post-change state identifier.

[0039] Determine the target verification rules and target credential generation rules based on the state change events;

[0040] Security verification is performed on the intelligent agents participating in the interaction based on the target verification rules;

[0041] If the security verification is successful, the context parameters of the sending node are obtained, and the context parameters are fused according to the target credential generation rules to obtain the interaction credential for the interaction phase; the interaction credential is used to characterize the verifiability of the operation in the interaction phase.

[0042] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0043] For each interaction stage of the target interaction scenario, multiple agent state change events are acquired. Each state change event includes at least one of the following: state change type, change timestamp, and post-change state identifier.

[0044] Determine the target verification rules and target credential generation rules based on the state change events;

[0045] Security verification is performed on the intelligent agents participating in the interaction based on the target verification rules;

[0046] If the security verification is successful, the context parameters of the sending node are obtained, and the context parameters are fused according to the target credential generation rules to obtain the interaction credential for the interaction phase; the interaction credential is used to characterize the verifiability of the operation in the interaction phase.

[0047] The aforementioned security verification method, apparatus, device, storage medium, and program product for multi-agent interaction acquires state change events of multiple agents for each interaction stage of the target interaction scenario. Each state change event includes at least one of a state change type, a change timestamp, and a post-change state identifier. Based on the state change events, target verification rules and target credential generation rules are determined. Security verification is performed on the participating agents based on the target verification rules. If the security verification passes, the context parameters of the sending node are acquired, and the context parameters are fused according to the target credential generation rules to obtain the interaction credential for the interaction stage. The interaction credential is used to characterize the verifiability of the operation during the interaction stage. By real-time monitoring of the agent states and correspondingly loading the verification rules and credential generation rules, dynamic rule adaptation is achieved. Targeted verification of participating agents based on the verification rules improves data security during multi-agent interaction. Simultaneously, by combining the interaction context parameters to generate cross-stage chain-linked immutable credentials, traceability of data during the interaction process is achieved. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a block diagram of a multi-agent state management system in one embodiment;

[0050] Figure 2 This is a flowchart illustrating a security verification method for multi-agent interaction in one embodiment;

[0051] Figure 3 This is a schematic diagram of a process for performing security verification on intelligent agents participating in interaction based on target verification rules in one embodiment;

[0052] Figure 4 This is a flowchart illustrating the process of fusing context parameters according to the target credential generation rules to obtain an interactive credential in one embodiment.

[0053] Figure 5 This is a flowchart illustrating the process of determining the target verification rule and the target credential generation rule based on a state change event in one embodiment.

[0054] Figure 6 This is a flowchart illustrating a security verification method for multi-agent interaction in another embodiment;

[0055] Figure 7 This is a structural block diagram of a security verification device for multi-agent interaction in one embodiment;

[0056] Figure 8 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0058] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0059] With the rapid development of artificial intelligence technology, multi-agent systems have been widely applied in various fields such as industrial manufacturing, intelligent transportation, financial services, and smart cities. These systems can significantly improve task execution efficiency, resource utilization, and decision-making speed through the collaboration and interaction of multiple agents. For example, in intelligent manufacturing, multiple agents can collaboratively optimize production processes and improve production efficiency. These applications not only enhance the intelligence of the systems but also drive the digital transformation and intelligent upgrading of industries. However, data interaction, as a core link in multi-agent collaboration, faces severe challenges. Because multiple agents need to frequently exchange large amounts of sensitive information during the interaction process, ensuring data security and privacy protection has become an urgent problem to be solved.

[0060] In view of the above-mentioned technical problems, this application provides a security verification method for multi-agent interaction that can improve the security of multi-agent data interaction. The following embodiments will specifically describe the security verification method for multi-agent interaction.

[0061] The security verification method for multi-agent interaction provided in this application can be applied to, for example... Figure 1In the illustrated intelligent agent state management system, processing device 102 communicates with multiple intelligent agents 104, and these agents 104 communicate with each other. In the data interaction scenario, each intelligent agent 104 can act as a data sender, data receiver / user, coordinator, or supervisor. Processing device 102 performs security verification on the intelligent agents in the interaction scenario. In specific applications, multiple intelligent agents interact with each other in the application scenario. During each interaction phase, the processing device monitors the state change events of each intelligent agent in real time and performs security verification on the data sender based on the attribute information of the intelligent agent in the current interaction phase. Processing device 102 can be a server, which can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0062] In one exemplary embodiment, such as Figure 2 As shown, a security verification method for multi-agent interaction is provided. This embodiment illustrates the application of this method to a processing device. In this embodiment, the method includes:

[0063] S201: For each interaction stage of the target interaction scenario, acquire state change events of multiple agents.

[0064] The state change event includes at least one of the following: state change type, change timestamp, and post-change state identifier. The target interaction scenario can be any one of the following: data request and transmission, data negotiation, sensitive data transaction, sensitive command issuance, routine information synchronization, derived data processing, credit assessment, and state change. The interaction stages are different for each scenario. For example, the data request and transmission interaction scenario includes a request stage and a transmission stage, while the sensitive data transaction interaction scenario includes a request stage, a negotiation stage, a transmission stage, and a derived data processing stage.

[0065] In the embodiments of this application, the agent state bus is communicatively connected to the permission database and the credit scoring system, and captures agent state change events in real time through the agent state bus. Specifically, the state bus registers listening filters with the permission database and the credit scoring system, specifying the types and scope of changes to be monitored. For example, permission change monitoring focuses on types such as "authorized role change" and "permission level change"; credit score change monitoring focuses on events such as "30-day performance rate update" and "credit rating increase / decrease". When a permission change operation occurs in the permission database, for example, agent A changes from "ordinary role" to "authorized role", the database triggers a change log record, the state bus captures the change record, and extracts the change object identifier (agent A), the permission value before the change, the permission value after the change, and the change timestamp. When a credit score update operation occurs in the credit scoring system, for example, agent B's 30-day performance rate decreases from 85% to 55%, the state bus captures the change record, and extracts the change object identifier (agent B), the credit score before the change, the credit score after the change, and the change timestamp.

[0066] S202, determine the target verification rules and target credential generation rules based on the state change event.

[0067] In the embodiments of this application, in response to a state change event, the processing device obtains the corresponding rule identifier list, loads the corresponding lightweight rule script from the code repository based on the rule identifier list, and parses and instantiates the lightweight rule script through the rule engine to generate an executable rule instance. For example, when node A transitions from a pending authorization state to a data provision state, the high-priority transmission protocol rule module corresponding to the data provision state, as well as the target verification rule and target credential generation rule, are loaded to perform security verification on the participating agent A.

[0068] S203, performs security verification on the intelligent agents participating in the interaction based on the target verification rules.

[0069] In the embodiments of this application, for each participating agent, the processing device parses the information possessed by the agent to obtain the agent's attribute information and / or subsequent interaction risk level. The attribute information of the agent can be processed using target verification rules to obtain a corresponding verification method, and the agent is then security-verified based on this verification method. Alternatively, the interaction risk level of the agent can be processed using target verification rules to obtain a corresponding verification method, and the agent is then security-verified based on this verification method. Furthermore, the attribute information and interaction risk level of the agent can be processed using target verification rules to obtain multiple types of verification methods, and the agent is then security-verified based on these multiple verification methods.

[0070] S204. If the security verification is successful, obtain the context parameters of the sending node, and perform fusion processing on the context parameters according to the target credential generation rules to obtain the interaction credential for the interaction phase.

[0071] The interaction credential is used to characterize the verifiability of the operation during the interaction phase. The context parameters include the algorithm version number, network environment parameters, and a summary of the negotiation terms.

[0072] In the embodiments of this application, if the security verification is passed, it indicates that the intelligent agent participating in the interaction is risk-free. The processing device obtains the first verification parameter of the current interaction stage and the second verification parameter of the previous stage; binds the first verification parameter and the second verification parameter to a hash chain according to the target credential generation rule to generate a staged credential for the current interaction stage; and solidifies the staged credential with the context parameters to generate an interaction credential.

[0073] The aforementioned security verification method for multi-agent interaction acquires state change events of multiple agents for each interaction stage of the target interaction scenario. Each state change event includes at least one of the following: state change type, change timestamp, and post-change state identifier. Based on the state change events, target verification rules and target credential generation rules are determined. Security verification is performed on the participating agents based on the target verification rules. If the security verification passes, the context parameters of the sending node are acquired, and these parameters are fused according to the target credential generation rules to obtain the interaction credential for the interaction stage. The interaction credential is used to characterize the verifiability of the operation during the interaction stage. By real-time monitoring of agent states and correspondingly loading the corresponding verification and credential generation rules, dynamic rule adaptation is achieved. Targeted verification of participating agents based on the verification rules improves data security during multi-agent interaction. Simultaneously, by combining the interaction context parameters to generate cross-stage chain-linked immutable credentials, traceability of data during the interaction process is achieved.

[0074] In one exemplary embodiment, such as Figure 3 As shown, security verification of participating agents based on target verification rules includes:

[0075] S301, obtain the agent's attribute information and interaction risk level.

[0076] Attribute information includes at least one of the following: agent role, data sensitivity level, and participant's historical credit score;

[0077] In the embodiments of this application, after obtaining the target verification rules, the processing device parses the intelligent agents participating in the interaction to obtain the attribute information and interaction risk level of each intelligent agent.

[0078] S302, The attribute information is processed based on the target verification rules to obtain the first verification method.

[0079] The first verification method can include joint verification by three parties or basic verification, or it can be a verification method for a number of other nodes, depending on the actual scenario requirements.

[0080] In the embodiments of this application, for attribute information, the processing device can select the mapping relationship between roles and verification methods from the target verification rules, match the intelligent agent roles and the mapping relationship between roles and verification methods, and determine the verification method corresponding to the successfully matched role as the first verification method. For example, if the intelligent agent role is an authorized role, the first verification method is determined to be three-party joint verification; if the intelligent agent role is an unauthorized role, the first verification method is determined to be basic verification.

[0081] In some embodiments, the processing device can select a mapping relationship between sensitivity levels and verification methods from the target verification rules, match the data sensitivity levels and the mapping relationship between sensitivity levels and verification methods, and determine the verification method corresponding to the successfully matched sensitivity level as the first verification method. For example, if the data sensitivity level is high-level sensitive data, the first verification method is determined to be three-party joint verification; if the data sensitivity level is low-level sensitive data, the first verification method is determined to be basic verification.

[0082] In some embodiments, the processing device can select a mapping relationship between credit scores and verification methods from the target verification rules, match the participant's historical credit scores and the mapping relationship between credit scores and verification methods, and determine the verification method corresponding to the successfully matched credit score as the first verification method. For example, if the participant's historical credit score is a high credit rating, the first verification method is determined to be joint verification by three parties; if the participant's historical credit score is a low credit rating, the first verification method is determined to be basic verification.

[0083] In some embodiments, the processing device can determine the weight coefficients of each attribute information according to a preset weight mapping table, and then perform a weighted summation of each attribute information and its corresponding weight coefficients to obtain a comprehensive score. Finally, based on the score interval into which the comprehensive score falls, the verification intensity requirement is determined, where the score interval and the verification intensity requirement have a unique mapping relationship. A first verification method is then determined based on the verification intensity requirement. For example, when the verification intensity requirement is high, the first verification method is determined to be three-party joint verification; when the verification intensity requirement is high, the first verification method is determined to be basic verification.

[0084] In some embodiments, the processing device can simultaneously match the agent's attribute information with various preset mapping relationships in the target verification rules to obtain multiple verification methods, and select the verification method with the most verification nodes from among the multiple verification methods as the first verification method. For example, if multiple verification methods simultaneously include three-party joint verification and basic verification, the three-party joint verification is determined as the first verification method.

[0085] S303, based on the target verification rules, processes the interaction risk level to obtain the second verification method.

[0086] In the embodiments of this application, for each agent participating in the interaction, the processing device extracts the mapping relationship between risk level and verification method from the target verification rules, matches the agent's interaction risk level with the mapping relationship between risk level and verification method, and determines the verification method corresponding to the successfully matched risk level as the second verification method. For example, when the interaction risk level is high risk, the second verification method is determined to be a strict verification process; when the interaction risk level is low risk, the second verification method is determined to be a fast verification process.

[0087] S304, Verify the intelligent agent participating in the interaction according to the first verification method and / or the second verification method.

[0088] In embodiments of this application, after determining the first verification method and the second verification method, the computer device may verify the intelligent agent participating in the interaction based solely on the first verification method. Alternatively, it may verify the intelligent agent participating in the interaction based solely on the second verification method. In some embodiments, it may also verify the intelligent agent participating in the interaction based on both the first and second verification methods.

[0089] Targeted verification of participating agents based on verification rules improves data security during multi-agent interactions.

[0090] In an exemplary embodiment, security verification of the intelligent agent participating in the interaction is performed according to a first verification method, including:

[0091] When the first verification method is the high verification strength method, signature verification requests are sent to the sender node, receiver node and supervisor node respectively, and signature verification responses are received from the sender node, receiver node and supervisor node. When multiple signature verification responses are all passed, the security verification is determined to be successful.

[0092] In the embodiments of this application, the sending node, receiving node, and supervisory node correspond to intelligent agents that perform data sending, data receiving, and data supervision in the interaction scenario, respectively. When the first verification method is a high-strength verification method, characterizing this interaction phase is crucial, and the data transmission process is complex. This situation requires distributed multi-verification. Therefore, signature verification requests are sent to the sending node, receiving node, and supervisory node respectively, and signature verification responses are received from the sending node, receiving node, and supervisory node. Security verification is considered successful when multiple signature verification responses are all successful.

[0093] When the first verification method is the low verification strength method, a signature verification request is sent to the sender node and the receiver node, and a signature verification response is received from the sender node and the receiver node. If multiple signature verification responses are all successful, the security verification is determined to be successful.

[0094] In the embodiments of this application, when the first verification method is a low-strength verification method, it indicates that the interaction stage is a conventional data processing scenario, requiring only simple verification by the sending node and the receiving node. Therefore, a signature verification request is sent to the sending node and the receiving node, and signature verification responses are received from the sending node and the receiving node. If multiple signature verification responses are all successful, the security verification is determined to be successful.

[0095] In an exemplary embodiment, security verification of the intelligent agent participating in the interaction is performed according to a second verification method, including:

[0096] When the second verification method is a high-risk operation verification method, a strict verification process is adopted to perform security verification on the intelligent agent.

[0097] High-risk operations may include cross-border data transfers and the issuance of sensitive instructions.

[0098] In the embodiments of this application, for each intelligent agent participating in the interaction, when the second verification method is a high-risk operation verification method, the processing device adopts a strict verification process to perform security verification on the intelligent agent. The strict verification process can be to add the signature of an authoritative institution or to extend the verification time window.

[0099] When the second verification method is a low-risk operation verification method, a rapid verification process is used to perform security verification on the intelligent agent.

[0100] Among them, low-risk operations can involve synchronizing non-sensitive information.

[0101] In the embodiments of this application, for each intelligent agent participating in the interaction, when the second verification method is a low-risk operation verification method, the processing device can use a fast verification process to perform security verification on the intelligent agent. The fast verification process can be a simplified signature process or a process to shorten the response time.

[0102] In one exemplary embodiment, such as Figure 4 As shown, the context parameters are fused according to the target credential generation rules to obtain the interactive credential, including:

[0103] S401, obtain the first verification parameter of the current interaction stage and the second verification parameter of the previous stage.

[0104] The first and second verification parameters can both include a hash value, a timestamp, and the identity of the participant.

[0105] In the embodiments of this application, the processing device can obtain the first verification parameter of the current interaction stage and the second verification parameter of the preceding stage of the sending node. It can also obtain the first verification parameter of the current interaction stage and the second verification parameter of the preceding stage for each agent participating in the interaction in that interaction stage.

[0106] S402, bind the first verification parameter and the second verification parameter to a hash chain according to the target credential generation rules to generate a staged credential for the current interaction stage.

[0107] In the embodiments of this application, the processing device obtains the end timestamp of the current interaction phase as the associated time point and extracts the identity list of all smart agents participating in the current interaction phase. The first verification parameter, the second verification parameter, the current timestamp, and the participant identity list are combined according to a preset association concatenation format to generate an associated data string. A hash operation is performed on the associated data string to generate a chained associated hash value. This chained associated hash value simultaneously binds the current phase credential, the preceding credential, and the participant information, forming an indivisible chain structure. The chained associated hash value, together with the current phase identifier, the preceding phase identifier, and the associated timestamp, is encapsulated into block data and written into the current segment storage block of the corresponding session in the blockchain ledger, generating a phased credential for the current interaction phase.

[0108] S403 solidifies the staged credentials and context parameters to generate interactive credentials.

[0109] In the embodiments of this application, the processing device can first serialize the context parameters according to a predefined data structure, converting them into structured data objects to ensure that each field has a clear key-value correspondence. The serialized context parameters are then forcibly bound to the phased credential, generating a composite data structure containing a credential hash value, a context parameter object, and a binding timestamp. Using a system private key or a designated signing key, the composite data structure is digitally signed to generate a signature value, ensuring the integrity and non-repudiation of the binding relationship. The signed composite data is then written to a distributed ledger, a trusted timestamp server, or other tamper-proof storage medium via a blockchain transaction or trusted timestamp service. After successful writing, a solidification completion credential is generated. This credential includes at least a storage address index, a write timestamp, and a solidified data hash value. During subsequent auditing, the hash credential and its solidified context parameters can be completely read from the tamper-proof storage medium using this index.

[0110] In one exemplary embodiment, such as Figure 5 As shown, the target verification rules and target credential generation rules are determined based on the state change event, including:

[0111] S202', based on the mapping relationship between status identifiers and rules, determines the target verification rule and target credential generation rule corresponding to the changed status identifier.

[0112] In the embodiments of this application, the processing device uses the changed status identifier as an index key to perform an exact match query from a pre-stored status-rule mapping table to obtain a list of rule module identifiers corresponding to the changed status identifier. Each record in the status-rule mapping table contains at least a status identifier field, a verification rule module identifier field, and a credential generation rule module identifier field. The obtained list of rule module identifiers is parsed to separate the target verification rule module identifier and the target credential generation rule module identifier. When there are multiple verification rule module identifiers in the list, they are sorted according to a priority field to determine the loading order. Based on the target verification rule module identifier and the target credential generation rule module identifier, the corresponding verification rule script file and credential generation rule script file are located from the rule code repository, respectively. The rule code repository establishes an index directory based on the rule module identifier; the rule engine reads and loads the located verification rule script file and credential generation rule script file into memory space, and the loading operation includes script syntax verification and dependency resolution; the rule engine parses and instantiates the loaded verification rule script and credential generation rule script to generate executable target verification rule instance and target credential generation rule instance, and the instantiation operation includes compiling the condition judgment logic in the script into executable conditional branch statements; the mapping relationship between the changed status identifier and the target verification rule instance and target credential generation rule instance is written into the status rule cache area. When the same changed status identifier appears again, the corresponding rule instance is directly read from the status rule cache area to skip the mapping table lookup and script loading operation.

[0113] In addition to the methods of all the above embodiments, a security verification method for multi-agent interaction is also provided, such as... Figure 6 As shown, the method includes:

[0114] S601, for each interaction stage of the target interaction scenario, acquire multiple state change events of intelligent agents, and the state change event includes at least one of state change type, change timestamp and post-change state identifier;

[0115] S602, Based on the mapping relationship between status identifiers and rules, determine the target verification rule and target credential generation rule corresponding to the changed status identifier;

[0116] S603, Obtain the attribute information and interaction risk level of the intelligent agent. The attribute information includes at least one of the intelligent agent role, data sensitivity level and historical credit score of the participants.

[0117] S604, The attribute information is processed based on the target verification rules to obtain the first verification method;

[0118] S605, based on the target verification rules, the interaction risk level is processed to obtain the second verification method;

[0119] S606, Verify the intelligent agent participating in the interaction according to the first verification method and / or the second verification method;

[0120] S607, if the security verification is successful, obtain the context parameters of the sending node; the context parameters include the algorithm version number, network environment parameters, and a summary of the negotiation terms;

[0121] S608, obtain the first verification parameter of the current interaction stage and the second verification parameter of the previous stage;

[0122] S609, bind the first verification parameter and the second verification parameter to a hash chain according to the target credential generation rule to generate a staged credential for the current interaction stage;

[0123] S610, solidify the staged credentials and context parameters to generate interaction credentials; interaction credentials are used to characterize the verifiability of operations in the interaction stage.

[0124] Each of the above steps has been described in the foregoing embodiments. For details, please refer to the foregoing content. They will not be repeated here.

[0125] The following examples illustrate the above method.

[0126] The target interaction scenario is a transaction scenario, including the request phase, negotiation phase, and transaction phase. Specifically:

[0127] 1. Request Phase

[0128] Input: init∈D (initial request)

[0129] State transition:

[0130] ;

[0131] Where fcid:D→H is the content identifier function (hash operation|request target type|time stamp|random number).

[0132] B-signature authentication:

[0133]

[0134] That is, there are identity and role labels for B (such as "coordinator", "auditor" etc.).

[0135] During the request phase, agent B, as a participant in the transaction interaction scenario, needs to undergo security verification. This is just an example; in real-world scenarios, verification of other agents is also required, but this verification process is omitted here.

[0136] 2. Negotiation Phase

[0137] Terms generated:

[0138] ;

[0139] ;

[0140] (gterms: Terms generation function, containing context parameters for the current interaction).

[0141] Consistency Protocol:

[0142] ;

[0143] Voucher generation:

[0144] ;

[0145] For the credential function (P: credential space, containing a summary of the negotiation terms, a list of participating agents and their roles, a timestamp of consensus reached, and a hash value of the preceding request phase (associated via f_cid(D))).

[0146] Must meet: .

[0147] 3. Transaction Phase

[0148] Atomic execution:

[0149] ;

[0150] Payment and service functions:

[0151] ;

[0152] Atomicity rule:

[0153] .

[0154] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0155] Based on the same inventive concept, this application also provides a security verification device for multi-agent interaction to implement the security verification method for multi-agent interaction described above. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations in one or more embodiments of the security verification device for multi-agent interaction provided below can be found in the limitations of the security verification method for multi-agent interaction described above, and will not be repeated here.

[0156] In one exemplary embodiment, such as Figure 7 As shown, a security verification device for multi-agent interaction is provided, comprising: an acquisition module 71, a determination module 72, a verification module 73, and a generation module 74, wherein:

[0157] The acquisition module 71 is used to acquire state change events of multiple intelligent agents for each interaction stage of the target interaction scenario. The state change event includes at least one of state change type, change timestamp and post-change state identifier.

[0158] The determination module 72 is used to determine the target verification rule and the target credential generation rule based on the state change event;

[0159] Verification module 73 is used to perform security verification on the intelligent agents participating in the interaction based on the target verification rules;

[0160] The generation module 74 is used to obtain the context parameters of the sending node when the security verification is passed, and to perform fusion processing on the context parameters according to the target credential generation rule to obtain the interaction credential for the interaction stage; the interaction credential is used to characterize the verifiability of the operation in the interaction stage.

[0161] In an exemplary embodiment, the verification module 73 described above is specifically used for:

[0162] Obtain the attribute information and interaction risk level of the intelligent agent, wherein the attribute information includes at least one of the intelligent agent role, data sensitivity level, and historical credit score of the participant;

[0163] The attribute information is processed based on the target verification rules to obtain the first verification method;

[0164] The interaction risk level is processed based on the target verification rules to obtain a second verification method;

[0165] The intelligent agent participating in the interaction is verified according to the first verification method and / or the second verification method.

[0166] In an exemplary embodiment, the verification module 73 described above is specifically used for:

[0167] When the first verification method is a high verification strength method, a signature verification request is sent to the sender node, the receiver node and the supervisor node respectively, and a signature verification response is received from the sender node, the receiver node and the supervisor node. When multiple signature verification responses are all passed, the security verification is determined to be successful.

[0168] When the first verification method is a low verification strength method, a signature verification request is sent to the sending node and the receiving node, and a signature verification response is received from the sending node and the receiving node. When multiple signature verification responses are all passed, the security verification is determined to be successful.

[0169] In an exemplary embodiment, the verification module 73 described above is specifically used for:

[0170] When the second verification method is a high-risk operation verification method, a strict verification process is used to perform security verification on the intelligent agent;

[0171] When the second verification method is a low-risk operation verification method, a fast verification process is used to perform security verification on the intelligent agent.

[0172] In an exemplary embodiment, the generation module 74 described above is specifically used for:

[0173] Obtain the first verification parameter of the current interaction stage and the second verification parameter of the previous stage;

[0174] According to the target credential generation rule, the first verification parameter and the second verification parameter are bound together by a hash chain to generate the staged credential for the current interaction stage;

[0175] The phased credentials and the context parameters are solidified to generate the interaction credentials; the context parameters include the algorithm version number, network environment parameters, and a summary of the negotiation terms.

[0176] In an exemplary embodiment, the determining module 72 described above is specifically used for:

[0177] Based on the mapping relationship between status identifiers and rules, the target verification rule and target credential generation rule corresponding to the changed status identifier are determined.

[0178] Each module in the aforementioned multi-agent interaction security verification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0179] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores authentication rules and credential generation rules for intelligent agents. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a secure authentication method for multi-agent interaction.

[0180] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0181] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0182] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0183] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0184] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0185] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0186] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A secure verification method for multi-agent interaction, characterized in that, The method includes: For each interaction stage of the target interaction scenario, multiple state change events of intelligent agents are obtained. The state change event includes at least one of state change type, change timestamp and post-change state identifier. The target verification rules and target credential generation rules are determined based on the state change events; Security verification is performed on the intelligent agents participating in the interaction based on the aforementioned target verification rules; If the security verification is successful, the context parameters of the sending node are obtained, and the context parameters are fused according to the target credential generation rule to obtain the interaction credential for the interaction stage; the interaction credential is used to characterize the verifiability of the operation in the interaction stage.

2. The method according to claim 1, characterized in that, The security verification of the participating agents based on the target verification rule includes: Obtain the attribute information and interaction risk level of the intelligent agent, wherein the attribute information includes at least one of the intelligent agent role, data sensitivity level, and historical credit score of the participant; The attribute information is processed based on the target verification rules to obtain the first verification method; The interaction risk level is processed based on the target verification rules to obtain a second verification method; The intelligent agent participating in the interaction is verified according to the first verification method and / or the second verification method.

3. The method according to claim 2, characterized in that, The security verification of the participating intelligent agent according to the first verification method includes: When the first verification method is a high verification strength method, a signature verification request is sent to the sender node, the receiver node and the supervisor node respectively, and a signature verification response is received from the sender node, the receiver node and the supervisor node. When multiple signature verification responses are all passed, the security verification is determined to be successful. When the first verification method is a low verification strength method, a signature verification request is sent to the sending node and the receiving node, and a signature verification response is received from the sending node and the receiving node. When multiple signature verification responses are all passed, the security verification is determined to be successful.

4. The method according to claim 2, characterized in that, The security verification of the participating intelligent agent according to the second verification method includes: When the second verification method is a high-risk operation verification method, a strict verification process is used to perform security verification on the intelligent agent; When the second verification method is a low-risk operation verification method, a fast verification process is used to perform security verification on the intelligent agent.

5. The method according to claim 1, characterized in that, The step of fusing the context parameters according to the target credential generation rule to obtain the interaction credential includes: Obtain the first verification parameter of the current interaction stage and the second verification parameter of the previous stage; According to the target credential generation rule, the first verification parameter and the second verification parameter are bound together by a hash chain to generate the staged credential for the current interaction stage; The phased credentials and the context parameters are solidified to generate the interaction credentials; the context parameters include the algorithm version number, network environment parameters, and a summary of the negotiation terms.

6. The method according to any one of claims 1-5, characterized in that, The step of determining the target verification rule and the target credential generation rule based on the state change event includes: Based on the mapping relationship between status identifiers and rules, the target verification rule and target credential generation rule corresponding to the changed status identifier are determined.

7. A security verification device for multi-agent interaction, characterized in that, The device includes: The acquisition module is used to acquire state change events of multiple intelligent agents for each interaction stage of the target interaction scenario. The state change event includes at least one of state change type, change timestamp and post-change state identifier. The determination module is used to determine the target verification rule and the target credential generation rule based on the state change event; The verification module is used to perform security verification on the intelligent agents participating in the interaction based on the target verification rules; The generation module is used to obtain the context parameters of the sending node when the security verification is passed, and to perform fusion processing on the context parameters according to the target credential generation rule to obtain the interaction credential for the interaction stage; the interaction credential is used to characterize the verifiability of the operation in the interaction stage.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images