Unmanned aerial vehicle network reputation enhancement message authentication method and system supporting cross-regional reputation reevaluation

Abstract

The invention discloses an unmanned aerial vehicle network reputation enhancement message authentication method and system supporting cross-regional reputation reevaluation, and the method comprises the steps: carrying out the management and maintenance of a secret key through a trusted center, carrying out the evaluation of the reputation values of unmanned aerial vehicles in a local region and a cross-regional region, and stopping the transmission of the secret key for the unmanned aerial vehicle with the reputation value not meeting the requirements; the edge server verifies the message signature in the unmanned aerial vehicle network, judges the authenticity of the unmanned aerial vehicle-ground message content, and feeds back to the trusted center; and the unmanned aerial vehicle regularly feeds back the surrounding information of the unmanned aerial vehicle with low authenticity to the edge server. According to the invention, a unified unmanned aerial vehicle reputation evaluation method is established among areas with different message interaction conditions; the message authentication condition is incorporated into reputation evaluation, and a reputation evaluation system is perfected; in the message authentication stage, the edge server carries out message authentication in a unified manner, so that the resource overhead of the unmanned aerial vehicle is reduced, and accurate evaluation and message authentication of the reputation of the unmanned aerial vehicle across regions and the region are finally realized.

Description

A Reputation Enhancement Message Authentication Method and System for Unmanned Aerial Vehicle Networks Supporting Cross-Regional Reputation Reassessment Technical Field

[0001] This invention relates to reputation mechanisms and message authentication technology, specifically to a message authentication method and system for enhancing the reputation of unmanned aerial vehicle (UAV) networks that supports cross-regional reputation reassessment. Background Technology

[0002] Unmanned aerial vehicle (UAV) networks have attracted much attention due to their flexible deployment, wide coverage, and high mobility. They are increasingly widely used in both military and civilian fields, bringing great convenience but also presenting security challenges. In UAV networks, inter-UAV messages and UAV-to-ground messages are the two most important message types. However, the open environment makes these two types of messages vulnerable to attacks from both internal (e.g., sending false messages) and external (e.g., eavesdropping, tampering, and replay). Currently, encryption schemes incorporating reputation assessment are considered an effective solution. However, traditional schemes do not consider the reputation changes of UAVs across regions. When a UAV moves from one region to another, its initial reputation value, established based on a limited amount of interaction, may not accurately reflect its reputation level in a more active environment with greater interaction. To more accurately identify malicious entities with low reputation values ​​and ensure the fairness of reputation assessment, the original reputation value must be reassessed in conjunction with the specific circumstances of the new region to determine whether its reputation level meets the requirements. Therefore, this invention studies a more practical UAV network reputation enhancement message authentication scheme that supports cross-domain reputation reassessment.

[0003] In drone networks, cryptographic techniques can effectively defend against attacks such as eavesdropping, tampering, and replay attacks by external attackers. Reputation-based methods can assess the reputation of entities within the network and detect malicious entities. Combining these two approaches can revoke the keys of entities with low reputation values, preventing them from participating in information exchange. However, this only guarantees a one-way relationship between reputation and cryptography; it cannot achieve a close integration between the two, as message signing results cannot affect reputation levels. Furthermore, in scenarios involving cross-regional drone interaction, it is impossible to accurately assess the reputation level of drones entering new areas.

[0004] Many scholars are now focusing on research combining reputation mechanisms and cryptographic techniques. For example, they are exploring methods to assess the reputation of drone entities within a single region, methods to assess the reputation of drone entities based on multiple metrics, and various signature technologies. However, these existing solutions suffer from several problems: they neglect the reputation assessment of drone entities across regions; they fail to tightly integrate reputation mechanisms and cryptographic techniques into a two-way feedback loop; and the computational overhead and communication efficiency of some existing solutions are not ideal, making them unsuitable for direct application in drone systems.

[0005] Chinese patent application CN103338194A discloses a cross-security domain access control system and method based on reputation assessment. The reputation value of an entity within a region is evaluated by scores from other entities, which is highly susceptible to positive / negative review attacks, potentially resulting in excessively high / low reputation values ​​and ultimately inaccurate results. Furthermore, cross-region reputation values ​​are calculated through weighted multiplication, failing to consider the imbalance in data sources for reputation assessment. For example, if an entity moves from region A to region B, assuming all entities in region B have a high average interaction volume, and if the entity also has a large number of interaction records in region A, then the reputation value obtained in region A based on these records more accurately reflects its true reputation value in region B. However, if the entity has only a few interaction records in region A, then the reputation value obtained based on these records does not accurately reflect its reputation value in region B.

[0006] Chinese patent application CN114125728A discloses a lightweight, privacy-preserving trust assessment method and system for drone networks, but it does not consider the reputation assessment mechanism for drones across regions, and only assesses the reputation of drones within a region. Summary of the Invention

[0007] Purpose of the invention: The purpose of this invention is to address the shortcomings of existing technologies and provide a message authentication method and system for enhancing the reputation of drone networks that supports cross-regional reputation reassessment. This method utilizes edge servers to verify message signatures within the system, reducing the resource overhead of drones while providing data for subsequent reputation assessments. Furthermore, it establishes a unified standard for cross-regional drone reputation assessments.

[0008] Technical solution: The present invention provides a method for enhancing the reputation of unmanned aerial vehicle (UAV) networks by message authentication, which supports cross-regional reputation reassessment, comprising the following steps:

[0009] Step S1, signing and authenticating drone messages, the specific method is as follows:

[0010] Step S1.1: System initialization, i.e., generating and publishing system parameters: elliptic curves. System public key and private key Symmetric encryption keys used for communication between drones Symmetric encryption keys for drone-edge servers Safe hash function Key seed used for signing on edge servers ;

[0011] Next, drones are registered, and each drone has a real identity. The drone applies for registration with the Trusted Center, which calculates the drone's pseudonym. and signing private key ;

[0012] Step S1.2: The drone encrypts and signs the message, that is, the drone generates a random point. For the two main types of messages in the network: firstly, messages between drones, a symmetric encryption key is calculated using a random point and the private key. Finally, ciphertext is generated. and signature Secondly, there's the drone-to-ground messaging, which can be directly used... Encryption is performed to obtain ciphertext, and the signature is the same as above;

[0013] Step S1.3: The edge server performs signature verification and records the verification result: After receiving the signature from the drone, the edge server uses the sender's identity and system public parameters. The validity of the signatures of the two types of messages mentioned above is verified, and the verification results are saved for subsequent reputation evaluation.

[0014] For drone-to-ground messages, the edge server verifies them, and decryption can be performed directly after successful verification.

[0015] For messages between drones, the edge server needs to generate two lists. and This stores the hash values ​​of illegal and legal messages separately, and combines the two lists. use Generated key Sign it;

[0016] Step S1.4: Sign the drone verification result and decrypt the message: The drone first uses the previous round The generated key right Verification is performed. If the verification is successful, the signature of the message to be decrypted is checked to see if it is in the valid list. If it is, a decryption key is generated based on the sender's identity and a random point, and the original message is decrypted.

[0017] Step S2, drone reputation assessment within the area, the specific method is as follows:

[0018] Step S2.1, Reputation value evaluation for message authentication: When performing message signature authentication, the edge server retains the number of invalid signatures for each drone sending two types of messages. and Calculate the reputation value of the final message authentication. ;

[0019] Step S2.2, Message Authenticity Judgment: This mainly involves calculating the authenticity of the feedback messages from the drone-edge server, for the set of feedback messages. The edge server first calculates the set of trust levels. Then assess the probability of the event occurring above the credibility level. Simultaneously, based on feedback from other drones, a penalty factor is selected. ;

[0020] Step S2.3, Message content reputation assessment: based on the probability density function of the beta distribution and the penalty factor. Calculate the reputation value of the drone's message content ;

[0021] Step S2.4, Calculation of final reputation value: The final reputation value of the drone in the area is calculated based on the reputation value of the message content and the reputation value of message authentication. ;

[0022] Step S3, cross-regional drone reputation assessment, the specific method is as follows:

[0023] Step S3.1: Edge server calculates the interaction status of drones in the region: The edge server calculates the average value of the two types of messages sent by all drones in the region. To ensure reliability, outliers need to be removed.

[0024] Step S3.2: The Trusted Center queries the edge server in the original region for the number of message interactions between the drone and the original region. ;

[0025] Finally, the reputation value of the cross-regional drone in the new region was calculated. .

[0026] This invention incorporates cryptographic signature verification of messages into regional reputation assessment and also designs a cross-regional drone reputation assessment method. Furthermore, this invention primarily targets two main types of messages in drone networks: drone-to-drone messages and drone-to-ground messages (drone-to-ground messages are messages sent by drones to edge servers), designing different message encryption keys to ultimately obtain the drone reputation value within the region. And the reputation value of cross-regional drones in new regions .

[0027] Furthermore, the detailed process of system initialization and drone registration in step S1.1 is as follows:

[0028] First, select an elliptic curve. and its generators Then, the system private key, public key, and global encryption key are generated. , , as well as ;

[0029] Define a hash function , , , , Then, the edge server signing key is generated. ;

[0030] Next, for those with For drones, the trusted center selects random numbers. Calculate the public key and kana and private key Then calculate Stored on the drone.

[0031] Furthermore, the detailed process of the drone encrypting and signing the message in step S1.2 is as follows:

[0032] First, select a random point on the elliptic curve. ,in ;

[0033] Then, for messages between drones, the drones need to use a global encryption key. Calculation for messages Encrypted symmetric key Then encrypt the message and the random number respectively. , For messages perceived by the drone-ground (edge ​​server), use directly. Encrypt , ;

[0034] Next, the drone generates a signature for the corresponding message and calculates... , ,in It's a timestamp; the final signature and message combination is... ,in ;

[0035] These are random points on an encrypted elliptic curve. It is a public key. It is the message signature; These correspond to kana. It's a timestamp.

[0036] Furthermore, the detailed method for step S1.3, verifying the signature and signing the verification result, is as follows:

[0037] First, the edge server restores the random points. And calculate two hash functions. , ;

[0038] Next, the edge server verifies... The validity of a signature is determined by whether both sides of the equation are true, and batch authentication is performed simultaneously. ,

[0039] in , It is a small random integer that can resist attacks;

[0040] For all messages with valid signatures sent to the edge server, the edge server directly decrypts them. ,

[0041] For messages between drones, the edge server generates two lists after verification. and It stores the hash values ​​of messages with invalid and valid signatures respectively, and combines them. and for Generate signature ;

[0042] ,

[0043] in The timestamp used to generate the signature. Generated through a hash chain. .

[0044] Furthermore, the detailed method for signing and decrypting the UAV verification result in step S1.4 is as follows:

[0045] The drones, based on the previous round calculate Verify the equation Check if it is true; if true, accept the verification result.

[0046] The next step, for the message we want to find, is to first calculate its hash value. And search in both lists if and only if the above hash value is only in When the message signature appears in the signature, it proves that the corresponding message signature has been verified as valid, and only then will the drone receive the message.

[0047] The random points also need to be restored first. Calculate the symmetric encryption key Finally, the original message was obtained through decryption. .

[0048] Furthermore, the specific content of step S2 is as follows:

[0049] First, the edge server counts the signature verification results of two types of messages from drones (messages between drones and drone-to-ground messages), and calculates the number of invalid signatures sent by the drones. ;

[0050] Next, feedback from other drones was checked, and a penalty factor was determined based on the number of messages containing false information sent by the target drone. , ;

[0051] Then the edge server bases its response on the message set. ,in Indicates drone Regarding the event The feedback, the set of calculated credibility , ,in Indicates drone Distance Event Distance from the location of the incident and It is a variable that controls the rate;

[0052] Next, the edge server calculates the event based on trustworthiness. probability of occurrence If the value is greater than the threshold, the event occurs; otherwise, it is considered not to have occurred.

[0053] Next, the reputation value of the drone's message content is calculated. , and These represent the accumulation of correct and false information, respectively.

[0054] Finally, the Trust Center calculates the drone's final reputation score. ,in It is a historical reputation value. These are the corresponding weights. .

[0055] Furthermore, the specific method for cross-regional drone reputation assessment in step S3 is as follows:

[0056] The edge server first uses the absolute median difference algorithm to filter out outliers and calculates the average value of all drone interactions within the region. ,

[0057] Then the Trust Center queries the edge server in the original region of the cross-regional drone for its interaction volume. Next, the activity level of the cross-domain drone in the new area is calculated. ,

[0058] Finally, the reputation value of the cross-domain drone in the new area is calculated. .

[0059] This invention also discloses a drone network reputation enhancement message authentication system supporting cross-regional reputation reassessment, comprising a trusted center, an edge server, drones, and a base station. The trusted center initializes the system, generates drone keys, and assesses drone reputation; the generated keys are sent to the drones via a secure channel. The base station is responsible for expanding the communication range and ensuring stable and timely transmission. The edge server verifies the signatures of both internal drone messages and drone-to-ground messages within the network, saving the verification results for subsequent reputation assessments. It also verifies the message content returned by the drones to facilitate the calculation of their reputation values. Drones perform tasks, interact with other drones, and transmit ground (edge ​​server) perception messages. The trusted center performs a unified assessment of the reputation values ​​of drones across different regions, more accurately identifying malicious entities. The trusted center periodically assesses the reputation values ​​of drones, stopping the distribution of global encryption keys to drones whose reputation values ​​do not meet the requirements.

[0060] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0061] (1) The solution of the present invention can establish a unified credit level reassessment mechanism for drones in different regions within the credit mechanism;

[0062] (2) This invention closely integrates reputation mechanism and cryptographic technology to achieve two-way feedback, thereby enhancing the reputation assessment mechanism and the ability to identify malicious entities.

[0063] (3) The present invention uses an edge server to uniformly verify the signature of the message, which reduces the overhead on the drone. Attached Figure Description

[0064] Figure 1 is a schematic diagram of the system structure of the present invention;

[0065] Figure 2 is a flowchart of the message authentication process of the present invention;

[0066] Figure 3 is a flowchart of the credit assessment process in this invention;

[0067] Figure 4 is a comparative experimental diagram of the technical solution in Example 1;

[0068] Figure 5 is a comparative experimental diagram of the technical solution in Example 2;

[0069] Figure 6 is a comparative experimental diagram of the technical solution in Example 3;

[0070] Figure 7 is a comparison diagram of the experimental results of the technical solution in Example 4. Detailed Implementation

[0071] The technical solution of the present invention will be described in detail below, but the scope of protection of the present invention is not limited to the embodiments described.

[0072] As shown in Figure 1, the UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment of the present invention includes the following steps:

[0073] Step S1, signing and authenticating drone messages, the specific method is as follows:

[0074] Step S1.1: System initialization, i.e., generating and publishing system parameters: elliptic curves. System public key and private key Symmetric encryption keys used for communication between drones Symmetric encryption keys for drone-edge servers Safe hash function Key seed used for signing on edge servers ;

[0075] Next, drones are registered, and each drone has a real identity. The drone applies for registration with the Trusted Center, which calculates the drone's pseudonym. and signing private key ;

[0076] Step S1.2: The drone encrypts and signs the message, that is, the drone generates a random point. For two types of messages in the network: firstly, messages between drones, a symmetric encryption key is calculated using a random point and the private key. Finally, ciphertext is generated. and signature Secondly, there's the drone-to-ground messaging, which can be directly used... Encryption yields ciphertext;

[0077] Step S1.3: The edge server performs signature verification and records the verification result: After receiving the signature from the drone, the edge server uses the sender's identity and system public parameters. Verify the validity of the signatures for both types of messages;

[0078] Step S1.4: Sign the drone verification result and decrypt the message: The drone first uses the previous round The generated key right Verification is performed. If the verification is successful, the signature of the message to be decrypted is checked to see if it is in the valid list. If it is, a decryption key is generated based on the sender's identity and a random point, and the original message is decrypted.

[0079] Step S2, drone reputation assessment within the area, the specific method is as follows:

[0080] Step S2.1: Message authentication reputation value assessment. When performing message signature authentication, the edge server retains the number of invalid signatures for each drone sending two types of messages. and Calculate the reputation value of the final message authentication. ;

[0081] Step S2.2: Message authenticity judgment. Calculate the authenticity of the feedback messages from the drone-edge server for the set of feedback messages. The edge server first calculates the set of trust levels. Then assess the probability of the event occurring above the credibility level. Simultaneously, based on feedback from other drones, a penalty factor is selected. ;

[0082] Step S2.3: Evaluation of the message content's reputation value, based on the probability density function of the beta distribution and the penalty factor. Calculate the reputation value of the drone's message content ;

[0083] Step S2.4: Calculate the final reputation value. Based on the reputation value of the message content and the reputation value of message authentication, calculate the final reputation value of the drone within the area. ;

[0084] Step S3, cross-regional drone reputation assessment, the specific method is as follows:

[0085] Step S3.1: Edge server calculates the interaction status of drones in the region: The edge server calculates the average value of the two types of messages sent by all drones in the region. To ensure reliability, outliers need to be removed.

[0086] Step S3.2: The Trusted Center queries the edge server in the original region for the number of message interactions between the drone and the original region. ;

[0087] Finally, the reputation value of the cross-regional drone in the new region was calculated. .

[0088] The system used for the aforementioned UAV network reputation enhancement authentication method includes the following participants: a trusted center, an edge server, UAVs, and a base station. The trusted center initializes the system, generates UAV keys, and evaluates UAV reputation. The generated keys are sent to the UAVs via a secure channel. The base station acts as a communication bridge, responsible for expanding the communication range and ensuring communication quality. The edge server verifies the signatures of both internal UAV messages and UAV-to-ground messages within the network, saving the verification results for subsequent reputation evaluation. It also verifies the content of messages returned by the UAVs to facilitate the calculation of their reputation values. UAVs perform tasks, interact with other UAVs, and transmit ground (edge ​​server) perception messages. The trusted center conducts a unified evaluation of the reputation values ​​of UAVs operating across different regions, more accurately identifying malicious entities. The trusted center periodically evaluates the reputation values ​​of UAVs, and stops distributing global encryption keys to any UAVs whose reputation values ​​do not meet the requirements.

[0089] As shown in Figure 2, the message authentication scheme in this embodiment is described as follows: Before sending a message, the drone encrypts the data and signs it using a private key to ensure the integrity and trustworthiness of the message source. The edge server verifies the signatures of all messages in the network, including messages between drones and drone-to-ground messages, and performs corresponding operations according to the message category: For messages between drones, the server signs the verification result and sends it back to the drone, which then verifies it and chooses to decrypt it; for drone-to-ground messages, the server can directly decrypt them after signature verification. The specific details are as follows:

[0090] First, the edge server counts the signature verification results of the two types of messages sent by the drone and calculates the number of invalid signatures sent by the drone. ;

[0091] Next, feedback from other drones was checked, and a penalty factor was determined based on the number of messages containing false information sent by the target drone. , ;

[0092] Then, the edge server bases its response on the set of messages. The set for calculating credibility , ,in Indicates drone Distance Event Distance from the location of the incident and It is a variable that controls the rate; Indicates drone Regarding the event Feedback;

[0093] Edge server computing is based on trust, events probability of occurrence If the value is greater than the threshold, the event occurs; otherwise, it is considered not to have occurred.

[0094] And calculate the reputation value of the drone's message content. , and These represent the accumulation of correct and false information, respectively.

[0095] Finally, the Trust Center calculates the drone's final reputation score. ,in It is a historical reputation value. These are the corresponding weights. .

[0096] As shown in Figure 3, the reputation assessment scheme in this embodiment is described as follows: For drones operating across regions, the edge server first calculates the average interaction volume in its region; the trusted center then queries the drone's interaction records in the original region and calculates its reputation value in the new region based on these two sets of data. For drones operating within a region, the edge server first assesses the credibility of its message content, and then the trusted center calculates the reputation value of the message content and message authentication based on the server's feedback. Finally, the trusted center integrates the obtained multi-dimensional reputation indicators to calculate the final reputation value of the target drone and determines whether it meets the key distribution conditions based on a preset threshold. If the reputation value exceeds the threshold, the system continues to distribute keys to it; otherwise, the key distribution process is terminated. The specific method is as follows:

[0097] The edge server first uses the absolute median difference algorithm to filter out outliers and calculates the average value of all drone interactions within the region. ,

[0098] Then the Trust Center queries the edge server in the original region of the cross-regional drone for its interaction volume. Next, the activity level of the cross-domain drone in the new area is calculated. ,

[0099] Finally, the reputation value of the cross-domain drone in the new area is calculated. .

[0100] For the evaluation of message authentication and reputation mechanisms, the implementation examples used the C++ PBC cryptography library and MATLAB to simulate the operation of the scheme on a machine configured with Windows 11 and an Intel Core i7-10710U.

[0101] Example 1: Example 1 tested the time required for signing and verification under different schemes. Figure 4(a) compares the time required for signing and verifying a single message under different schemes. Because the technical solution of this invention does not involve paired operations and has a reasonable design, the technical solution of this invention can simultaneously reduce the computational overhead of signing and verification. Figure 4(b) tests the overhead of batch verification on a drone entity. Because in the technical solution of this invention, the signature verification is entirely completed by the edge server, the overhead of signature verification on the drone can be greatly reduced.

[0102] Example 2: Example 2 tested the impact of different behaviors on the reputation level of a drone. In the test shown in Figure 5(a), the weights of the different dimensions of the final reputation value were fixed at... , , For a single plane, sending genuine messages and having a higher historical reputation score results in a higher reputation score in the evaluation, and vice versa; the upper plane, due to fewer illegally signed messages, achieves a higher reputation score than the lower plane. In the test shown in Figure 5(b), the scheme implements dynamic adjustment of weights. This dynamic adjustment allows the system to slow down the rate at which the reputation score of drones with higher historical reputation scores increases, thus increasing the difficulty for them to obtain a higher reputation score. Simultaneously, when a drone with a high message content reputation score sends an authentication failure message, the system accelerates the rate at which its reputation score decreases. This dynamic adjustment controls the upper and lower limits of reputation scores, effectively increasing the cost of malicious behavior and enhancing the system's ability to detect and punish malicious entities.

[0103] Example 3: Example 3 tested the reputation evaluation mechanism of the technical solution of the present invention and the comparative solution. For fairness, the reputation value range was specified to be... The interval, and the two schemes in The reputation value is initialized to 0.7 at any given moment; in each round, there are... Individual entities participate in the feedback, and the penalty factor .from arrive The entity sends each round Message; from arrive The entity sends each round Message. Specifically, in arrive period, The message was malicious; arrive period, The message was malicious; and arrive as well as arrive During this period, the proportion of malicious messages was... The result is shown in Figure 6. to During the study, the proposed scheme effectively identified malicious entities. In contrast, the comparative scheme RSMA was more susceptible to benign message masking, leading to failure in malicious entity detection. to During this period, the reputation score of the proposed solution declined faster than that of the comparison solution, indicating that it was more sensitive to malicious behavior. From arrive The proposed solution maintained a more stable reputation value, while the reputation value of the comparison solution fluctuated more significantly and may even exceed the preset range.

[0104] Example 4: Example 4 tested the scheme's evaluation results for cross-regional drone reputation. As shown in Figure 7(a), when the drone's historical reputation value was set to... At that time, the new reputation value of the target area increases with the ratio. It gradually increases with the increase of [something], as mentioned above. This represents the ratio of drone interaction volume to the local average interaction volume. This indicates that a higher level of previous interaction provides reliable confirmation of its historical reputation value. Furthermore, as the reputation threshold increases, the reputation value of drones across regions also increases accordingly, thereby enhancing the fairness and consistency of cross-regional reputation assessment. As shown in Figure 7(b), for drones with different historical reputation values, the proposed assessment method ensures that their assessed reputation value in the new region will not exceed their original historical reputation value, thus maintaining the reputation boundary. In summary, this invention performs reputation assessment through edge servers and a trusted center, and the evaluations of other drones do not directly affect the reputation value of the target drone. Simultaneously, a unified cross-regional reputation value reassessment method is designed based on the interaction volume in different regions, which can incorporate the "cryptographic verification result of message signature" as an indicator into the reputation assessment of drones within the region, better identifying malicious drones.

Claims

1. A message authentication method for enhancing the reputation of unmanned aerial vehicle (UAV) networks that supports cross-regional reputation reassessment, characterized in that, Includes the following steps: Step S1, signing and authentication of drone messages, specifically as follows: Step S1.1, system initialization generates and publishes system parameters: elliptic curve System public key and private key Symmetric encryption keys used for communication between drones Symmetric encryption keys for drone-edge servers Safe hash function Key seed used for signing on edge servers Next, drones are registered, and each drone has a real identity. The drone applies for registration with the Trusted Center, which calculates the drone's pseudonym. and signing private key ； Step S1.2: The drone encrypts and signs the message, that is, the drone generates a random point. For the two types of messages in the network: messages between drones and drone-to-ground messages; step S1.3, the edge server performs signature verification and records the signature verification result: after receiving the drone's signature, the edge server uses the sender's identity and system public parameters Verify the validity of the signatures for both messages; Step S1.4: The drone verifies the signature of the message and decrypts it: The drone first uses the previous round The generated key right Verification is performed. After successful verification, the signature of the message to be decrypted is checked to see if it is in the valid list. If it is, a decryption key is generated based on the sender's identity and a random point, and the original message is decrypted. Step S2: Reputation assessment of drones in the area. The specific method is as follows: Step S2.1: Reputation value assessment for message authentication. When performing message signature authentication, the edge server retains the number of invalid signatures for each drone sending the two types of messages. and Calculate the reputation value for final message authentication. ； Step S2.2: Message authenticity judgment. Calculate the authenticity of the feedback messages from the drone-edge server for the set of feedback messages. The edge server first calculates the set of trust levels. Then assess the probability of the event occurring above the credibility level. Simultaneously, based on feedback from other drones, a penalty factor is selected. ； Step S2.3: Evaluation of the message content's reputation value, based on the probability density function of the beta distribution and the penalty factor. Calculate the reputation value of the drone's message content ； Step S2.4: Calculate the final reputation value. Based on the reputation value of the message content and the reputation value of message authentication, calculate the final reputation value of the drone within the area. Step S3, Cross-regional drone reputation assessment, specifically: Step S3.1, Edge server calculates regional drone interaction: The edge server calculates the average of the two types of messages sent by all drones in the region. To ensure reliability, outliers need to be removed; Step S3.2: The Trusted Center queries the edge server in the original region for the number of message interactions between the drone and the original region. Finally, the reputation value of the cross-regional drone in the new region was calculated. 。 2. The UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment according to claim 1, characterized in that, The detailed process of system initialization and drone registration in step S1.1 is as follows: First, select an elliptic curve. and its generators It generates a system private key, a public key, and a global encryption key. ， ， as well as Define a hash function ， ， ， ， Then, the edge server signing key is generated. Next, for those with For drones, the trusted center selects random numbers. Calculate the public key and kana and private key Finally, calculate Stored on the drone.

3. The UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment according to claim 1, characterized in that, Step S1.2 The detailed process of the drone encrypting and signing the message is as follows: First, a random point on the elliptic curve is selected. ,in ； Then, for messages between drones, the drones use a global encryption key. Calculation for messages Encryption symmetric key Then encrypt the message and the random number respectively. ， For drone-to-ground perception messages, use directly Encrypt ， Next, the drone generates a signature for the corresponding message and calculates... ， ,in It's a timestamp; the final signature and message combination is... ,in ； These are random points on an encrypted elliptic curve. It is a public key. It is the message signature; These correspond to the kana. It's a timestamp.

4. The UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment according to claim 1, characterized in that, The detailed method for step S1.3, verifying the signature and the signature verification result, is as follows: First, the edge server restores the random point. And calculate two hash functions. ， Next, the edge server verifies... The validity of a signature is determined by whether both sides of the equation are true, and batch authentication is performed simultaneously. The expression is as follows: ,in ， It is a small random integer that can resist attacks; for all messages with valid signatures sent to the edge server, the edge server directly decrypts them. For messages between drones, the edge server generates two lists after verification. and It stores the hash values ​​of messages with invalid and valid signatures respectively, and combines them. and for Generate signature ； ,in The timestamp used to generate the signature. Generated through hash chains 。 5. The UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment according to claim 1, characterized in that, The detailed method for signing and decrypting the drone verification result in step S1.4 is as follows: the drone, based on the previous round... calculate Verify the equation If the statement is true, accept the verification result; then, for the message you want to find, first calculate its hash value. And search in both lists if and only if the above hash value is only in When the message signature appears in the signature, it proves that the corresponding message signature has been verified as valid, and only then will the drone receive the message. First, restore the random points. Calculate the symmetric encryption key Finally, the original message was obtained through decryption. 。 6. The UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment according to claim 1, characterized in that, The specific content of step S2 is as follows: First, the edge server counts the signature verification results of the two types of messages from the drone and calculates the number of illegal signatures sent by the drone. Next, feedback from other drones is queried, and a penalty factor is determined based on the number of messages containing false information sent by the target drone. ， ； Then, the edge server bases its response on the set of messages. The set for calculating credibility ， ,in Indicates drone Distance Event Distance from the location of the incident and It is a variable that controls the rate; Indicates drone Regarding the event Feedback; Edge server computing is based on trust, events probability of occurrence If the value is greater than the threshold, the event occurs; otherwise, it is considered not to have occurred. The reputation value of the drone's message content is then calculated. ， and These represent the accumulation of true and false information, respectively; finally, the Trust Center calculates the drone's final reputation score. ,in It is a historical reputation value. These are the corresponding weights. 。 7. The UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment according to claim 1, characterized in that, The specific method for cross-regional drone reputation assessment in step S3 is as follows: The edge server first uses the absolute median difference algorithm to filter out outliers and calculates the average value of all drone interactions within the region. Then the Trust Center queries the edge server in the original region of the cross-regional drone for its interaction volume. Next, the activity level of the cross-domain drone in the new area is calculated. Finally, the reputation value of the cross-domain drone in the new area is calculated. 。 8. A system for implementing the UAV network reputation enhancement message authentication method supporting cross-regional reputation reassessment as described in any one of claims 1 to 7, characterized in that, The system includes a trusted center, an edge server, drones, and a base station. The trusted center initializes the system, generates drone keys, and evaluates drone reputation. The generated keys are sent to the drones through a secure channel. The base station is responsible for expanding the communication range and ensuring stable and timely transmission. The edge server verifies the signatures of both internal drone messages and drone-to-ground messages in the network and saves the verification results for subsequent reputation evaluation. It also verifies the message content fed back by the drones to facilitate the calculation of the drones' reputation values. The drone performs tasks, interacts with other drones, and transmits perception messages back to the edge server. The Trust Center assesses the reputation of drones across different regions; the Trust Center periodically evaluates the reputation of drones and stops distributing global encryption keys to any drones whose reputation does not meet the requirements.

Images