Flexible and intelligent group key negotiation method under unmanned aerial vehicle self-organizing network

Abstract

The invention discloses a flexible and intelligent group key negotiation method under an unmanned aerial vehicle ad hoc network, which is applied to each unmanned aerial vehicle of an unmanned aerial vehicle group, the unmanned aerial vehicle has a secret share, the secret share is a univariate polynomial determined based on two threshold values, and the two threshold values are alternately updated during continuous group key updating. When the group key is updated at the current time, generating a new threshold value of a threshold value which is not updated when the group key is updated at the last time by adopting a threshold value decision model; according to the new threshold value, the updated threshold value when the group key is updated last time and the updated secret share when the group key is updated last time, the updated secret share when the group key is updated at the current time is determined; and according to the private key of the unmanned aerial vehicle and the updated secret share of the unmanned aerial vehicle when the group key is updated at the current time, determining secret data of the unmanned aerial vehicle, broadcasting and receiving secret data of other unmanned aerial vehicles, and calculating the group key at the current time according to the secret data. According to the invention, the threshold value can be changed flexibly, and the group key can be updated.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicles, and in particular relates to a flexible and intelligent group key negotiation method under a self-organizing network of unmanned aerial vehicles. Background Art

[0002] At present, a lot of research has been done on group key agreement. In traditional group key agreement protocols, all participants usually need to participate in the group key agreement process at the same time, which is often impractical in practical applications, especially in UANETs where the communication link is unstable. In recent years, some researchers have proposed group key agreement protocols that can overcome this shortcoming. Some works focus on asynchronous group key agreement, where one member initially calculates the group key and then distributes key materials to help other responders calculate the group key. However, this requires the help of a central server to intelligently relay user messages to ensure sufficiently low communication complexity. However, these protocols always rely on a central server to deliver messages, which is not suitable for decentralized UANETs. Other works are based on threshold secret sharing, however, these solutions cannot be directly applied to group key agreement in UANETs. On the one hand, the threshold variation range in these schemes is limited and predetermined, which is not suitable for dynamic UANETs. On the other hand, in these schemes, group members cannot update their local secret shares according to the latest threshold, and the final recovered secret remains unchanged. This makes the group key difficult to change and does not meet the needs of continuous group key update in UANETs.

[0003] That is to say, although some works have proposed threshold group key agreement schemes using secret sharing, the threshold change range in these schemes is limited and determined in advance, and group members cannot update the secret share and secret, which makes the group key difficult to change. This is not applicable to UANET where link conditions and members change dynamically. Summary of the invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides a flexible and intelligent group key negotiation method in a UAV self-organizing network.

[0005] The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0006] The present invention provides a flexible and intelligent group key negotiation method under a drone self-organizing network, which is applied to an i-th drone in a drone self-organizing network including n drones, where i ranges from 1 to n; the i-th drone has a public key, a private key and a secret share, the secret share is a univariate polynomial determined based on two thresholds, and is updated as the thresholds are updated, and the two thresholds are alternately updated when performing consecutive group key updates, and the method comprises:

[0007] When performing the current group key update, a new threshold value of the threshold value to be updated is generated using the trained threshold decision model; the threshold value to be updated is the threshold value that was not updated when the group key was last updated;

[0008] Based on the updated secret share of the n drones when the group key was last updated, the new threshold and the number of the j-th drone, generate the first value and the second value of the j-th drone and encrypt the first value and the second value of the j-th drone and send them to the j-th drone, and receive the first value and the second value sent by the j-th drone; j is 1 to n and j is not equal to i; wherein the updated secret shares of the n drones when the group key was last updated together constitute the bivariate polynomial used for sharing of the drone self-organizing network last time, and the two degrees of the bivariate polynomial used for sharing last time are the threshold value that was not updated when the group key was last updated and the threshold value that was updated when the group key was last updated, respectively;

[0009] Determine, according to the first value and the second value of the self and the threshold value updated when the group key was last updated, the updated secret share of the self when the group key is currently updated;

[0010] Determine its own secret data according to its own private key and its own updated secret share when the group key is updated for the last time, encrypt it and send it to the j-th UAV, and receive the secret data sent by the j-th UAV;

[0011] When the amount of received secret data exceeds the value of the new threshold minus 1, a hash algorithm is used to calculate the current group key according to the received secret data, and the current group key update is completed.

[0012] Compared with the prior art, the present invention has the following beneficial effects:

[0013] Through the method of the present invention, the threshold can be flexibly changed by switching the dimensions of a bivariate polynomial, and the freshness of each group key can be ensured by updating the secret share. A trained network model can also be used to guide the UAV self-organizing network to reasonably select the threshold during the group key negotiation process, thereby making the group key updateable and making the group key negotiation more efficient, secure and robust.

[0014] The present invention will be further described in detail below with reference to the accompanying drawings and specific implementation methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1It is a flow chart of a flexible and intelligent group key negotiation method in a drone self-organizing network provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The present invention is further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0017] The embodiment of the present invention provides a flexible and intelligent group key negotiation method in a drone self-organizing network, which is applied to each drone in a drone self-organizing network. Exemplarily, the drone self-organizing network is a UANET composed of n drones, U = {u1, u2, ..., u n}, each drone has a fixed identifier, for example, u i We assume that the entire UANET is a connected network. i and u j A communication link can be established within a suitable distance ij , and mark each other as neighbor nodes. Each drone can transmit messages to any other drone in UANET through multi-hop communication links. The network topology of UANET can be viewed as an undirected graph in, Represents the vertex set of each drone ui, ε={e1,e2,...,e m} represents the edge set of the communication link between adjacent drones. The following will take the application of this method to the i-th drone ui (i ranges from 1 to n) in the drone self-organizing network as an example to specifically introduce the present invention.

[0018] Here, the i-th UAV has a public key, a private key and a secret share. The secret share is a univariate polynomial determined based on two thresholds and is updated as the thresholds are updated. The two thresholds are updated alternately when the group key is updated consecutively. The i-th UAV has an initial secret share. The initial secret share, public key and private key are all generated by the ground control station (GCS) for the i-th UAV; the initial secret share of the i-th UAV is generated by the GCS based on the initial bivariate polynomial for sharing of the UAV self-organizing network, the preset secret value s and the number i; the two variables of the initial bivariate polynomial for sharing are x and y, and the two degrees are the initial threshold u and the initial threshold v, respectively. The variable corresponding to the threshold u is x, and the variable corresponding to the threshold v is y.

[0019] When initializing the UAV self-organizing network, GCS generates a large prime number q and an elliptic curve E(F) with a base point P according to the preset security parameter 1λ. p). Then, GCS randomly selects a secure hash function H: in, is a finite field of order q, {0,1} * represents a bit string of any length consisting of 0s and 1s, * Asterisks represent arbitrary lengths. For the i-th drone u i , GCS randomly selects an integer And calculate x i ·P, and transmit x through a secure channel i Assign the private key of drone ui to drone ui, and the public key of drone ui to i P is announced to all drones in the group. In addition, GCS randomly selects an integer As a secret, generate the initial bivariate polynomial for sharing of the drone self-organizing network with the initial threshold u and the initial threshold v respectively. Among them, a ij ∈GF(p) and a 0,0 =s, GF(p) is a finite field. Then, according to F(x,y), GCS divides the secret s into n secret shares and distributes these n secret shares to n drones in the drone self-organizing network as the initial secret shares of each of the n drones, where the initial secret share of drone ui is expressed as {sh i =f i (y)=F(i,y)}. Specifically, when GCS calculates the secret share of drone ui, it substitutes the number i of drone ui as the value of variable x into F(x,y), and then a univariate polynomial with y as the variable can be obtained. The univariate polynomial with y as the variable is the initial secret share of drone ui.

[0020] Figure 1 is a flow chart of a flexible and intelligent group key negotiation method in a drone self-organizing network provided by an embodiment of the present invention. The method is applied to the i-th drone mentioned above, such as Figure 1 As shown, the method includes:

[0021] S101. When performing the current group key update, a trained threshold decision model is used to generate a new threshold for the threshold to be updated; the threshold to be updated is the threshold that was not updated when the group key was last updated.

[0022] It should be noted that the new threshold generated by the trained threshold decision model can be any integer.

[0023] S102. Generate the first value and the second value of itself and the j-th drone based on the secret share updated when the group key was last updated, the new threshold and the number of the j-th drone, encrypt the first value and the second value of the j-th drone and send them to the j-th drone, and receive the first value and the second value sent by the j-th drone; the value of j is 1 to n and j is not equal to i; wherein the secret shares updated when the group key of the n drones was last updated together constitute the bivariate polynomial used for sharing of the drone self-organizing network last time, and the two degrees of the bivariate polynomial used for sharing last time are the threshold that was not updated when the group key was last updated and the threshold that was updated when the group key was last updated, respectively.

[0024] It should be noted that the jth drone, namely drone u j , is every drone except drone ui in the drone self-organizing network.

[0025] S103: Determine the updated secret share of the current group key update according to the first value and the second value of the current group key and the threshold updated when the group key was last updated.

[0026] S104. Determine its own secret data based on its own private key and its own updated secret share when the group key is updated for the last time, encrypt it and send it to the j-th drone, and receive the secret data sent by the j-th drone.

[0027] S105. When the amount of received secret data exceeds the value of the new threshold minus 1, a hash algorithm is used to calculate the current group key according to the received secret data, and the current group key update is completed.

[0028] In some embodiments, the above S101 can be implemented by the following steps:

[0029] S1011. When performing the current group key update, obtain the current network connectivity status matrix of the drone self-organizing network; the network connectivity status matrix is ​​used to characterize the connection status of the communication link between drones.

[0030] For example, when a group key update is required at time t1, UAV i can obtain the network connectivity state matrix of the UAV self-organizing network at time t1 and obtain a network connectivity state matrix M.

[0031] S1012: Convert the current network connectivity state matrix into a grayscale image.

[0032] Here, the network connectivity state matrix M is a matrix containing multiple elements, and each element is composed of a floating point number in the range [0,1]. In order to map these values ​​to the pixel range (0 to 255) of the grayscale image, for the element m at the (i,j) position in the network connectivity state matrix M ij ,calculate The calculated value is used as the pixel value of the (i, j) position in the converted grayscale image Image. Since the matrix M is a single channel, the obtained Image is a grayscale image.

[0033] S1013. Input the converted grayscale image into a trained threshold decision model, and the trained threshold decision model outputs an optimal threshold corresponding to the grayscale image, and the optimal threshold is used as the new threshold to be updated; wherein the trained threshold decision model is obtained by training a convolutional neural network CNN using a data set, and the data set is obtained by collecting data from a self-organizing network of a drone used for data collection, and the data set contains multiple different network connectivity state matrices, and each network connectivity state matrix has a threshold as a label; the threshold corresponding to each network connectivity state matrix is ​​the maximum threshold among the multiple thresholds corresponding to the network connectivity state matrix, and the key negotiation success rate of the drone self-organizing network used for data collection at each threshold among the multiple thresholds is greater than the preset success rate.

[0034] Currently, there are several datasets on drone scenarios, but these datasets mainly focus on tasks such as target recognition and tracking, trajectory prediction, etc., and do not specifically address group key negotiation and threshold selection issues. Therefore, we constructed a dataset D through a large number of tests. Each sample in dataset D is i' ={M i' ,Label} is composed of a network connection state matrix M i' and a label Label (i.e., the optimal threshold for group key negotiation). Specifically, the NS3 platform can be used for data acquisition. Given a UAV self-organizing network with a group size of n, the topology is randomly set and the neighboring UAVs ui and u j A random packet loss rate is set between the two to generate a UANET instance and its corresponding network connection state matrix M i' In practical applications, many studies have been conducted to explore how to obtain the topology of UANET and the connection probability between drones, so the present invention will not be repeated here. Then, the method proposed in the present invention is run in this UANET configuration, and the success rate ssuc of group key negotiation under different thresholds t is recorded, where It should be noted that when The success rate of group key negotiation is the highest when ssuc>90%, because this requires that each group member receives the least number of messages from other members. However, lowering the threshold will damage the security of group key negotiation, so a trade-off needs to be made between security and efficiency. In order to simplify this process, the present invention sets a preset success rate value for the success rate of group key negotiation, requiring ssuc>90%. This means that the present invention selects the largest threshold among all thresholds that meet the success rate requirements as the label of the network connection state matrix. Of course, this preset success rate value can be adjusted as needed, and when the preset success rate value changes, the label of the connection state matrix will change.

[0035] After obtaining the above data set, it is necessary to use the above conversion principle to convert each network connection state matrix in the data set into a grayscale image, and use the label of the network connection state matrix as the label of the converted grayscale image. After that, the converted grayscale image with label is used to train the convolutional neural network CNN. In the specific training process, the loss is calculated through forward propagation, and then the model parameters are adjusted through backward propagation. The loss function uses cross entropy loss (CEL), which is the standard loss function for multi-classification tasks. After multiple training rounds, the model can accurately associate grayscale images with thresholds.

[0036] During training or testing, the convolutional neural network CNN first extracts the features of the link connectivity matrix of the current UANET from the input grayscale image through the convolution layer and the pooling layer, where each convolution kernel slides on the image, calculates the weighted sum with the local pixel area, and then performs a nonlinear transformation through an activation function (such as ReLU). The specific convolution process is expressed as: Among them, σ is the activation function, m is the size of the convolution kernel, and w j1 is the j1th weight of the convolution kernel, b is the bias, x i1+j1 is the (i1+j1)th element of the input grayscale image, h i1is the output of the convolution operation. Usually, a CNN contains multiple convolutional layers. As the network depth increases, CNN can not only capture simple and direct features (such as the connectivity of adjacent drones in UANET), but also capture complex features of the input data (such as the overall link connectivity status of UANET). After the convolution layer, the pooling layer is used to reduce the spatial dimension of the feature map, further compress the features extracted by the convolution layer, retain the key features, and finally output several feature maps to complete the feature extraction. After the feature extraction is completed, the output feature map is multi-dimensional. In order to be able to be input to the fully connected layer, these multi-dimensional feature maps need to be flattened into a one-dimensional vector. The flattened vector is passed to multiple fully connected layers connected in series. Each fully connected layer multiplies each feature in the input vector with a set of weights and adds a bias term to perform a linear transformation. After processing multiple fully connected layers, nonlinearity is introduced by adding an activation function (such as ReLU), so that the model can fit more complex nonlinear relationships. After processing multiple fully connected layers and activation functions, a vector representing the result of feature combination is obtained, which is finally used for decision making (i.e. threshold selection). After that, the obtained vector is input into the output layer, which makes a specific decision based on the input vector and outputs an optimal threshold value corresponding to the input grayscale image.

[0037] In some embodiments, the above S102 is implemented by the following steps:

[0038] S1021. Use its own number i and the number j of the j-th drone as two values ​​of the variable, and substitute them into its own updated secret share when the group key was updated last time, to obtain the first value of the j-th drone and its own first value.

[0039] For example, when the current group key is updated for the first time, the bivariate polynomial used for sharing last time is F(x, y), and the secret share updated by drone ui when the group key is updated last time is the initial secret share f of drone ui i (y), and the threshold to be updated is the initial threshold v, the threshold updated when the group key was last updated is the initial threshold u, and the new threshold is a new threshold v', then drone ui will use its own number i as the value of variable y to enter f i (y), we can get the polynomial f i Point f on (y) i (i) The f i (i) is the first value of the drone ui, and the drone ui inserts the number j as the value of the variable y into f i (y), we can get the polynomial f i Point f on (y) i (j), the f i(j) is the drone u j The first value of .

[0040] Exemplarily, when the current time is the second group key update, the bivariate polynomial used for sharing last time is the bivariate polynomial used for sharing of the drone self-organizing network obtained after the first group key update, and the secret share of drone ui after its own update when the last group key was updated is the secret share of drone ui after its own update when the group key was updated for the first time, and the threshold to be updated is the initial threshold u, the threshold updated when the group key was updated last time is the threshold v' updated for the first time, and the new threshold is a new threshold u', and the calculation of the subsequent first and second values ​​is the same.

[0041] S1022. Randomly generate a first univariate polynomial, where the degree of the first univariate polynomial is a new threshold.

[0042] Exemplarily, when the current group key is updated for the first time, the first univariate polynomial is a polynomial g of degree v' with y as the variable and i as the value of the variable x. i (y).

[0043] Exemplarily, when the current group key update is the second time, the first univariate polynomial is a polynomial g of degree u' with x as the variable and i as the value of the variable y. i (x).

[0044] S1023. Use the number i of the own drone and the number j of the j-th drone as two values ​​of the variable, and substitute them into the first univariate polynomial to obtain the second value of the j-th drone and the second value of the own drone.

[0045] For example, when the current group key is updated for the first time, the drone ui uses its own number i as the value of the variable y and inserts it into g i (y), we can get the polynomial g i Point g on (y) i (i) The g i (i) is the second value of drone ui, and drone ui takes the number j as the value of variable y and puts it into g i (y), we can get the polynomial g i Point g on (y) i (j), the g i (j) is the drone u j The second value of .

[0046] When the current time is the second group key update, the second data is calculated in the same way.

[0047] S1024. Use its own private key and public key to encrypt the first value and the second value of the j-th drone and send them to the j-th drone. At the same time, receive the first value and the second value sent by the j-th drone.

[0048] Specifically, the encryption function MBE() in the multi-message multi-receiver encryption algorithm can be called to encrypt the drone u using the private key and public key of the drone ui. j The first value and the second value are encrypted to generate a broadcast message, and the broadcast message is broadcast to the drone self-organizing network, so that the drone u j After receiving the broadcast message, the broadcast message will be decrypted to obtain the drone ui as drone u j The first value and the second value are calculated.

[0049] In some embodiments, the above S103 is implemented by the following steps:

[0050] S1031. Construct a second univariate polynomial by Lagrange interpolation method according to the first value calculated by itself and the first value received from the j-th UAV.

[0051] Exemplarily, when the current group key update is the first time, the second univariate polynomial is a univariate polynomial F(x,i) calculated by inserting the number i as the variable y into F(x,y), F(x,i)=f i (x).

[0052] Exemplarily, when the current group key update is the second time, the second univariate polynomial is a univariate polynomial calculated by substituting the number i as the variable x into the bivariate polynomial for sharing of the drone self-organizing network obtained after the first group key update.

[0053] S1032. Sum the second value calculated by itself and the second value received from the j-th UAV to obtain the sum value, and randomly generate a third univariate polynomial, wherein the degree of the third univariate polynomial is the threshold updated when the group key was last updated, and the variables in the third univariate polynomial are the same as the second univariate polynomial, and when the value of the variable is 0, the value calculated by the third univariate polynomial is the sum value.

[0054] Exemplarily, when the current group key update is the first time, the third univariate polynomial is a polynomial g with degree u and x as the variable and i as the value of the variable y. i (x); In addition, the drone ui calculates its own second value g i (i) and from the drone u jThe second value g of the drone ui received at j (i) The expression of the sum value obtained after the joint summation is: and,

[0055] Exemplarily, when the current time is the second group key update, the third univariate polynomial is a polynomial with degree u', y as a variable and i as the value of the variable x.

[0056] S1033. Add the second univariate polynomial to the third univariate polynomial to obtain the updated secret share when the group key is updated for the current time; wherein, the third univariate polynomials of the n drones together constitute a new bivariate polynomial; the new bivariate polynomial and the bivariate polynomial used for sharing last time together constitute the bivariate polynomial used for sharing this time.

[0057] For example, when the current group key is updated for the first time, the updated secret share of drone ui when the group key is updated for the current time is f i '(x), and f i '(x)=f i (x)+g i (x), the new bivariate polynomial is denoted by G(x,y), and g i (x)=G(x,j), the current bivariate polynomial used for sharing is expressed as F'(x,y), and F'(x,y)=F(x,y)+G(x,y), the two degrees of F'(x,y) are the initial threshold u and the new threshold v' respectively.

[0058] Exemplarily, when the current group key update is the second time, the new bivariate polynomial is expressed as G'(x,y), the current bivariate polynomial used for sharing is expressed as F"(x,y), and F"(x,y)=F'(x,y)+G'(x,y), and the two degrees of F"(x,y) are the new threshold u' and threshold v' respectively.

[0059] In some embodiments, the above S104 is implemented by the following steps:

[0060] S1041. Substitute 0 as a variable into the secret share updated by itself when the group key was updated last time, obtain the share value, and use its own private key and share value as its own secret data.

[0061] For example, when the drone ui updates the group key for the current time, the updated secret share is f i '(x), 0 is substituted into f as the value of variable x. i '(x) can be used to calculate the share value fi '(0), then, the private key x of the drone ui i and share value f i '(0) together as the secret data of drone ui (x i ,f i '(0)).

[0062] S1042. Use its own private key and public key to encrypt its own secret data and send it to the j-th drone, and receive the secret data sent by the j-th drone.

[0063] Specifically, the encryption function MBE() in the multi-message multi-recipient encryption algorithm can be called to encrypt the secret data (x i ,f i '(0)) is encrypted and a broadcast message is generated, which is then broadcast to the drone self-organizing network. At the same time, the receiving drone u j The secret data sent (x j ,f j '(0)).

[0064] In some embodiments, the above S105 is implemented by the following steps:

[0065] S1051. When the amount of received secret data exceeds the value of the new threshold minus 1, determine a first sum value according to the received share value and its own share value.

[0066] S1052: When the amount of received secret data exceeds the value of the new threshold minus 1, determine a second sum value according to the received private key and its own private key.

[0067] S1053. Use a hash function to perform a hash calculation on the product of the first sum value and the second sum value to obtain the current group key, and complete the current group key update; wherein the hash function is generated by the ground control station.

[0068] For example, when the current group key is updated for the first time, when the secret data received by the drone ui exceeds v'-1, the current group key is calculated using the following formula:

[0069]

[0070] Among them, GK is the current group key, H(.) is the hash function, and f i ′(0) is the share value of drone ui, x i is the private key of the drone ui, x j For drone u j The private key of is the first summation value mentioned above, is the second summation value mentioned above. Obviously, in the above process, a threshold of group key negotiation is changed from v to v', and each drone can also locally update its own secret share, and v' is an arbitrary integer, so the present invention can flexibly change the threshold, and in addition, by simultaneously updating the shared secret and secret share of each drone, the present invention realizes continuous group key negotiation.

[0071] The present invention improves the existing bivariate polynomial secret sharing, thereby allowing flexible threshold adjustment, and designs a secret and share update algorithm, so that the key contribution and group key of the drone group members can be dynamically updated. On this basis, the present invention also designs a variable threshold group key negotiation method. In addition, the present invention also uses supervised learning to enable the drone to explore the mapping relationship between the UANET link condition and the optimal threshold through offline training, thereby realizing intelligent threshold selection. Through the method of the present invention, the flexible change of the threshold can be realized by switching the dimension of the bivariate polynomial, and the freshness of each group key can be ensured by updating the secret share. It is also possible to guide the drone self-organizing network to reasonably select the threshold in the group key negotiation process through a trained network model, so that the group key can be updated, and the negotiation of the group key is made more efficient, secure and robust.

[0072] It should be noted that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0073] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification.

[0074] In the specification, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality of components. Certain measures are recorded in different embodiments, but this does not mean that these measures cannot be combined to produce good effects.

[0075] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A flexible and intelligent group key negotiation method in a drone self-organizing network, characterized in that: The method is applied to an i-th drone in a drone self-organizing network including n drones, where i ranges from 1 to n; the i-th drone has a public key, a private key and a secret share, the secret share is a univariate polynomial determined based on two thresholds and is updated as the thresholds are updated, and the two thresholds are updated alternately when performing consecutive group key updates, and the method comprises: When performing the current group key update, a new threshold value of the threshold value to be updated is generated using the trained threshold decision model; the threshold value to be updated is the threshold value that was not updated when the group key was last updated; Based on the updated secret share of the n drones when the group key was last updated, the new threshold and the number of the j-th drone, generate the first value and the second value of the j-th drone and encrypt the first value and the second value of the j-th drone and send them to the j-th drone, and receive the first value and the second value sent by the j-th drone; j is 1 to n and j is not equal to i; wherein the updated secret shares of the n drones when the group key was last updated together constitute the bivariate polynomial used for sharing of the drone self-organizing network last time, and the two degrees of the bivariate polynomial used for sharing last time are the threshold value that was not updated when the group key was last updated and the threshold value that was updated when the group key was last updated, respectively; Determine, according to the first value and the second value of the self and the threshold value updated when the group key was last updated, the updated secret share of the self when the group key is currently updated; Determine its own secret data according to its own private key and its own updated secret share when the group key is updated for the last time, encrypt it and send it to the j-th UAV, and receive the secret data sent by the j-th UAV; When the amount of received secret data exceeds the value of the new threshold minus 1, a hash algorithm is used to calculate the current group key according to the received secret data, and the current group key update is completed.

2. According to claim 1, a flexible and intelligent group key negotiation method in a drone self-organizing network is characterized in that: When performing the current group key update, a new threshold to be updated is generated by using the trained threshold decision model, including: When performing the current group key update, obtaining the current network connectivity state matrix of the drone self-organizing network; the network connectivity state matrix is ​​used to characterize the connection state of the communication link between the drones; Converting the current network connectivity state matrix into a grayscale image; Inputting the converted grayscale image into the trained threshold decision model, the trained threshold decision model outputs an optimal threshold corresponding to the grayscale image, and using the optimal threshold as a new threshold of the threshold to be updated; Among them, the trained threshold decision model is obtained by training a convolutional neural network using a data set, and the data set is obtained by collecting data from a drone self-organizing network used for data collection. The data set contains multiple different network connectivity state matrices, and each network connectivity state matrix has a threshold as a label; the threshold corresponding to each network connectivity state matrix is ​​the maximum threshold among the multiple thresholds corresponding to the network connectivity state matrix, and at each threshold among the multiple thresholds, the key negotiation success rate of the drone self-organizing network used for data collection is greater than the preset success rate.

3. According to claim 1, a flexible and intelligent group key negotiation method in a drone self-organizing network is characterized in that: The i-th UAV has an initial secret share, and the initial secret share, the public key and the private key are all generated by the ground control station for the i-th UAV; the initial secret share of the i-th UAV is generated by the ground control station according to the initial bivariate polynomial for sharing of the UAV self-organizing network, a preset secret value s and a number i; the two variables of the initial bivariate polynomial for sharing are x and y, and the two degrees are the initial threshold u and the initial threshold v, respectively, the variable corresponding to the threshold u is x, and the variable corresponding to the threshold v is y.

4. The flexible and intelligent group key negotiation method in a drone self-organizing network according to claim 3 is characterized in that: The method generates a first value and a second value of itself and the j-th drone based on the secret share updated by itself when the group key was last updated, the new threshold and the number of the j-th drone, encrypts the first value and the second value of the j-th drone and sends them to the j-th drone, and receives the first value and the second value sent by the j-th drone, including: The number i of the own drone and the number j of the j-th drone are used as two values ​​of the variable, and are substituted into the updated secret share of the own drone when the group key was last updated, to obtain the first value of the j-th drone and the first value of the own drone; randomly generating a first univariate polynomial, where the degree of the first univariate polynomial is the new threshold; The number i of the own drone and the number j of the j-th drone are used as two values ​​of the variable, and are substituted into the first univariate polynomial to obtain the second value of the j-th drone and the second value of the own drone; The first value and the second value of the j-th UAV are encrypted using its own private key and public key and then sent to the j-th UAV. At the same time, the first value and the second value sent by the j-th UAV are received.

5. A flexible and intelligent group key negotiation method in a drone self-organizing network according to claim 4, characterized in that: When the current time is the first group key update, the bivariate polynomial used for sharing last time is the initial bivariate polynomial used for sharing, the initial bivariate polynomial used for sharing is expressed as F(x, y), the secret share of the i-th drone after it is updated when the group key is last updated is the initial secret share of the i-th drone, the threshold to be updated is the initial threshold v, the threshold updated when the group key is last updated is the initial threshold u, the new threshold is a new threshold v', and the initial secret share of the i-th drone is a univariate polynomial F(i, y) calculated by substituting the number i as the variable x into F(x, y), and F(i, y) is expressed as f i (y), the first univariate polynomial is a polynomial of degree v' with y as the variable and i as the value of the variable x, and the first univariate polynomial is represented by g i (y).

6. The flexible and intelligent group key negotiation method in a drone self-organizing network according to claim 3, characterized in that: The determining, based on the first value and the second value of the self and the threshold value updated when the group key was last updated, the updated secret share of the self when the group key is currently updated includes: constructing a second univariate polynomial by Lagrange interpolation method according to the first value calculated by itself and the first value received from the j-th drone; summing up the second value calculated by itself and the second value received from the j-th UAV to obtain a sum value, and randomly generating a third univariate polynomial, wherein the degree of the third univariate polynomial is the threshold value updated when the group key was last updated, and the variables in the third univariate polynomial are the same as those in the second univariate polynomial, and when the value of the variable is 0, the value calculated by the third univariate polynomial is the sum value; Adding the second univariate polynomial to the third univariate polynomial to obtain the updated secret share when the group key is updated for the current time; Among them, the third univariate polynomials of the n drones together constitute a new bivariate polynomial; the new bivariate polynomial and the bivariate polynomial used for sharing last time together constitute the bivariate polynomial used for sharing this time.

7. A flexible and intelligent group key negotiation method in a drone self-organizing network according to claim 6, characterized in that: When the current time is the first group key update, the bivariate polynomial used for sharing last time is the initial bivariate polynomial used for sharing, the initial bivariate polynomial used for sharing is expressed as F(x, y), the secret share of the i-th drone after it is updated when the group key is last updated is the initial secret share of the i-th drone, the threshold to be updated is the initial threshold v, the threshold updated when the group key is last updated is the initial threshold u, the new threshold is a new threshold v', and the initial secret share of the i-th drone is a univariate polynomial F(i, y) calculated by substituting the number i as the variable x into F(x, y), and F(i, y) is expressed as f i (y), the second univariate polynomial is a univariate polynomial F(x,i) calculated by inserting the number i as the variable y into F(x,y), and F(x,i) is represented by f i (x), the third univariate polynomial is a polynomial of degree u with x as variable and i as value of variable y, the third univariate polynomial is represented by g i (x), g i (x) is represented as G(x,j), the new bivariate polynomial is represented as G(x,y), the current bivariate polynomial used for sharing is represented as F'(x,y), and F'(x,y)=F(x,y)+G(x,y), the two degrees of F'(x,y) are the initial threshold u and the new threshold v' respectively.

8. The flexible and intelligent group key negotiation method in a drone self-organizing network according to claim 3, characterized in that: The determining of the secret data of the user according to the private key of the user and the secret share updated by the user when the group key is updated for the current time, encrypting the secret data and sending the encrypted data to the j-th drone, and receiving the secret data sent by the j-th drone, comprises: Substitute 0 as a variable into the secret share updated by the user when the group key is updated for the current time to obtain a share value, and use the user's private key and the share value as the user's secret data; Use its own private key and public key to encrypt its own secret data and send it to the j-th drone, and receive the secret data sent by the j-th drone.

9. The flexible and intelligent group key negotiation method in a drone self-organizing network according to claim 3, characterized in that: The secret data of each drone includes the private key of the drone and the share value of the drone, and the share value of the drone is determined according to the secret share of the drone after it is updated when the group key is updated for the current time; When the amount of the received secret data exceeds the value of the new threshold minus 1, a hash algorithm is used to calculate the current group key according to the received secret data to complete the current group key update, including: When the amount of received secret data exceeds the value of the new threshold minus 1, determining a first sum value according to the received share value and its own share value; When the amount of received secret data exceeds the value of the new threshold minus 1, determining a second sum value according to the received private key and its own private key; A hash function is used to perform a hash calculation on the product of the first sum value and the second sum value to obtain the current group key, thereby completing the current group key update; wherein the hash function is generated by the ground control station.

10. A flexible and intelligent group key negotiation method in a drone self-organizing network according to claim 9, characterized in that: When the current time is the first group key update, the bivariate polynomial used for sharing last time is the initial bivariate polynomial used for sharing, the initial bivariate polynomial used for sharing is expressed as F(x, y), the secret share of the i-th drone after it is updated when the group key is last updated is the initial secret share of the i-th drone, the threshold to be updated is the initial threshold v, the threshold updated when the group key is last updated is the initial threshold u, the new threshold is a new threshold v', and the initial secret share of the i-th drone is a univariate polynomial F(i, y) calculated by substituting the number i as the variable x into F(x, y), the expression of the current group key is: Wherein, GK is the current group key, H(.) is the hash function, and f i ′(0) is the share value of the i-th drone, x i is the private key of the i-th drone, x j is the private key of the j-th drone.

Images