A method and apparatus for resilient consistency optimization under an agent-based honeycomb architecture

Abstract

This invention relates to the field of intelligent agent cellular architecture technology applications, specifically providing a method and apparatus for elastic consistency optimization under intelligent agent cellular architecture. This method is applied to distributed network communication and intelligent cloud platform cellular architecture, and includes the following steps: S1, network attack modeling; erroneous data injection network attack model: x i (k+1)=f i (x i (k)); x i (k) represents the information sent by agent i to neighboring agents, f i (x i (k) represents any update method; the agent honeycomb architecture introduces a consensus mechanism and implements a hierarchical topology; S2, agent composition assumes that these protected nodes constitute the dominant connected subset of the network; S3, the convergence of the resilient consensus optimization method is verified. Compared with the prior art, this invention can solve the problem of agents in a honeycomb architecture reaching consensus in an unreliable communication network.

Description

Technical Field

[0001] This invention relates to the field of intelligent agent honeycomb architecture technology, specifically providing a method and apparatus for elastic consistency optimization under intelligent agent honeycomb architecture. Background Technology

[0002] The agent honeycomb architecture pattern utilizes a large number of large-model inference agents that collaborate to solve a problem, with each agent contributing from its unique perspective. The resulting outcome demonstrates a collective intelligence that surpasses the capabilities of any single AI entity, particularly in handling complex datasets and scenarios.

[0003] While the honeycomb architecture pattern of agents has advantages in terms of flexibility, adaptability and compatibility of swarm intelligence, it is challenging to solve the problem of how agents under the honeycomb architecture can reach consensus in unreliable communication networks. Summary of the Invention

[0004] This invention addresses the shortcomings of the prior art by providing a highly practical method for optimizing elastic consistency under an intelligent agent honeycomb architecture.

[0005] A further technical objective of this invention is to provide a reasonably designed, safe, and applicable elastic consistency optimization device under an intelligent agent honeycomb architecture.

[0006] The technical solution adopted by this invention to solve its technical problem is:

[0007] A resilient consistency optimization method for an agent-based honeycomb architecture, applicable to distributed network communication and intelligent cloud platform honeycomb architecture, comprises the following steps:

[0008] S1. Network attack modeling;

[0009] Error data injection network attack model:

[0010] x i (k+1)=f i (x i (k));

[0011] Where, x i (k) represents the information sent by agent i to neighboring agents, f i (x i (k) represents any update method;

[0012] The intelligent agent honeycomb architecture introduces a consensus mechanism and implements a layered topology.

[0013] S2. Agent Composition: Assuming these protected nodes constitute the dominant connected subset of the network, the resilient consistency optimization method is as follows:

[0014] The input is the set of ordinal numbers of the protected nodes in the network, represented as T = {v i The ordinal set of other nodes is represented as O = {v | i = 1, 2, ..., n1}. i The attack node ordinal set is represented as A = {v | i = n1 + 1, ..., n0} (n0 = n1 + n2). i , |i=n0+1,...,n} and their corresponding state information;

[0015] S3. Convergence verification of the elastic consistency optimization method.

[0016] Furthermore, in step S1, the consensus mechanism involves a virtual leader layer with one or more leaders. Leaders do not communicate with each other, but only send information to nodes in the first layer. This is achieved by dividing n nodes into m layers, labeled from 1 to m from top to bottom. Nodes in the first layer only send information to the second layer, and nodes in the i-th layer send information to at least one node in the (i+1)-th layer. Information transmission between nodes in the i-th layer can be bidirectional. Assuming each layer has r... i If there are n nodes, then r1 + r2 + K + r m =n.

[0017] Furthermore, the communication network of the agent cellular architecture uses an undirected graph G = (V, E, A), where each node i of the agent has a scalar state at time k, represented as... The states of all nodes in the system are represented by a vector x = [x1, x2, k, x...]. n ] T express;

[0018] If a subset G of graph G = (V, E, A) is G d Each one that does not belong to G d A node has at least one neighbor belonging to G. d And belongs to G d All nodes form a connected graph, G d It is a dominant connected subset of G = (V, E, A).

[0019] Furthermore, specifically including:

[0020] S2.1, Each node v i Receive the states of the neighbors to form a set

[0021] S2.2, Node v i By identifying the state information of trusted nodes and combining it with the state information of its own nodes, the minimum threshold is selected. and the maximum threshold

[0022] S2.3, Judgment set S i element x in (k) j Does (k) satisfy...? Form a set

[0023] S2.4, each node v i The update law is:

[0024]

[0025] x j (k)∈R i (k);

[0026] S2.5, Repeat steps S2.1-S2.4 until |x i (k)-x j (k)|<ε,v j ∈T i .

[0027] Furthermore, in step S3, each protected node or ordinary node v i The minimum threshold is determined based on the information of the protected nodes in the neighborhood and the node's own information. and maximum threshold Set R is obtained through threshold filtering. i (k), node v i In the k-th iteration, according to R i The sources of agents in (k) categorize states into three types: those from a set of trusted nodes. The state x of its own node i (k), and a set from ordinary nodes or attacking nodes.

[0028] Furthermore, according to formula (1), it can be rewritten as:

[0029]

[0030] in, The information can be represented by a minimum threshold and a maximum threshold, i.e., there exists 0 < ρ. j <1, satisfying the following equation

[0031] Furthermore, formula (2) can be rewritten as follows:

[0032]

[0033] remember but

[0034]

[0035] Among them, M 11 M represents the interaction between trusted nodes. 21 M represents the role of a trusted node in relation to a normal node. 22 To represent the role of its own node, the consensus mechanism divides the nodes of the communication network into three layers: the first layer consists of trusted nodes, the second layer consists of ordinary nodes, and the third layer consists of attacking nodes.

[0036] Furthermore, for ease of analysis and recording If we represent the state of the trusted node and its own state, then...

[0037] It is not difficult to see that M(k) is a row random matrix. The elements of M(k) are non-negative, and the non-zero elements have a lower bound τ. Furthermore, τ = 1 / (d M +1), d M For R i The maximum value of (k) base, i.e.

[0038] Furthermore, the elastic consistency of the formula mechanism is analyzed:

[0039]

[0040] By the properties of row random matrices, we have:

[0041] in, It is a random vector, independent of t, in which the states of the nodes will reach a consensus and converge to t.

[0042] A resilient consistency optimization device under an agent honeycomb architecture includes: at least one memory and at least one processor;

[0043] The at least one memory is used to store a machine-readable program;

[0044] The at least one processor is used to call the machine-readable program to execute a resilient consistency optimization method under an agent honeycomb architecture.

[0045] Compared with existing technologies, the resilient consistency optimization method and apparatus under the intelligent agent honeycomb architecture of the present invention have the following outstanding advantages:

[0046] This invention's honeycomb architecture pattern utilizes a large number of agents based on large model inference. These agents collaborate to solve a problem, with each agent contributing from its unique perspective. The resulting synergistic outcome demonstrates collective intelligence, surpassing the capabilities of any single AI entity, particularly in handling complex datasets and large-scale scenarios. It also addresses the challenge of achieving consensus among agents in a honeycomb architecture within unreliable communication networks. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Appendix Figure 1 This is a schematic diagram of the agent honeycomb structure in a resilient consistency optimization method under an agent honeycomb architecture.

[0049] Appendix Figure 2 This is a schematic diagram of the consensus mechanism under the honeycomb architecture in a resilient consistency optimization method under an intelligent agent honeycomb architecture. Detailed Implementation

[0050] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to specific embodiments. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] The following is a preferred embodiment:

[0052] like Figure 1-2 As shown in this embodiment, a resilient consistency optimization method under an agent honeycomb architecture has the following steps:

[0053] S1. Network attack modeling;

[0054] False data injection (FDI) attacks are easy to execute and possess strong stealth and disruptive capabilities, posing a potential threat to consensus-based decision-making among multiple agents in a cellular architecture. Attackers, knowing the current system configuration, can carefully manipulate sensor measurement data or controller input signals to mislead the system's state estimation, causing the system's state to deviate from stability.

[0055] Consider the following error data injection network attack model:

[0056] x i (k+1)=f i (x i (k))

[0057] Where, x i (k) represents the information sent by agent i to neighboring agents, f i (x i (k) represents any update method.

[0058] The consensus mechanism of a honeycomb architecture for intelligent agents has significant implications. Consider a hierarchical topology. For example... Figure 2 As shown, the virtual leader layer has one or more leaders. Leaders do not communicate with each other and only send information to nodes in the first layer. The n nodes are divided into m layers, labeled from 1 to m from top to bottom. Nodes in the first layer only send information to the second layer, and nodes in the i-th layer send information to at least one node in the (i+1)-th layer. Information transmission between nodes in the i-th layer can be bidirectional. Assume each layer has r... i If there are n nodes, then r1 + r2 + K + r m =n.

[0059] Considering the impact of spoofed data injection on an agent-based cellular architecture, a consensus mechanism model with unidirectional information transmission is constructed by adjusting the original communication topology. The hierarchical topology reduces the impact of attacking nodes on the network, improving its topological robustness. The communication network of the agent-based cellular architecture is represented by an undirected graph G = (V, E, A), where each node i of an agent has a scalar state at time k, denoted as... The states of all nodes in the system are represented by a vector x = [x1, x2, k, x...]. n ] T express.

[0060] If a graph is represented by G = (V, E, A), its subset G d Each of the numbers not belonging to G d A node has at least one neighbor belonging to G. d And belongs to G d All nodes form a connected graph, G d It is a dominant connected subset of G = (V, E, A).

[0061] S2. Agent Composition: Assuming these protected nodes constitute the dominant connected subset of the network, the resilient consistency optimization method is as follows:

[0062] In the event of a fake data injection attack, setting up trusted nodes can prevent intrusion by the attacking nodes. Protective measures include: ① typically increasing the firewall security level of some nodes and performing redundant backups of node resources to ensure that nodes receive authentic and trustworthy information; ② using digital signatures and data encryption technology between nodes to prevent the authentic information sent by these nodes from being tampered with.

[0063] Assuming these protected nodes constitute the dominant connected subset of the network, the consistency optimization method is as follows:

[0064] The input to this method is the ordinal set of protected nodes in the network, represented as T = {v i The ordinal set of other nodes is represented as O = {v | i = 1, 2, ..., n1}. i The attack node ordinal set is represented as A = {v | i = n1 + 1, ..., n0} (n0 = n1 + n2). i , |i=n0+1,...,n} and their corresponding state information.

[0065] Specifically, it includes:

[0066] S2.1, Each node v i Receive the states of the neighbors to form a set

[0067] S2.2, Node v i By identifying the state information of trusted nodes and combining it with the state information of its own nodes, the minimum threshold is selected. and the maximum threshold

[0068] S2.3, Judgment set S i element x in (k) j Does (k) satisfy...? Form a set

[0069] S2.4, each node v i The update law is:

[0070]

[0071] x j (k)∈R i (k);

[0072] S2.5, Repeat steps S2.1-S2.4 until |x i (k)-x j (k)|<ε,v j ∈T i .

[0073] S3. Convergence verification of the elastic consistency optimization method.

[0074] Each protected node or ordinary node v i The minimum threshold is determined based on the information of the protected nodes in the neighborhood and the node's own information. and maximum threshold Set R is obtained through threshold filtering. i (k). Node v i In the k-th iteration, according to R i The sources of agents in (k) categorize states into three types: those from a set of trusted nodes. The state x of its own node i (k), and a set from ordinary nodes or attacking nodes.

[0075] Rewritten according to formula (1):

[0076]

[0077] in, The information can be represented by a minimum threshold and a maximum threshold, i.e., there exists 0 < ρ. j <1, satisfying the following equation

[0078] Formula (2) can be further rewritten as:

[0079]

[0080] remember but

[0081]

[0082] Among them, M 11 M represents the interaction between trusted nodes. 21 M represents the role of a trusted node in relation to a normal node. 22 To represent the role of its own node, the consensus mechanism divides the nodes of the communication network into three layers: the first layer consists of trusted nodes, the second layer consists of ordinary nodes, and the third layer consists of attacking nodes.

[0083] For ease of analysis and recording If we represent the state of the trusted node and its own state, then...

[0084] It is not difficult to see that M(k) is a row random matrix. The elements of M(k) are non-negative, and the non-zero elements have a lower bound τ. Furthermore, τ = 1 / (d M +1), d M For R i The maximum value of (k) base, i.e.

[0085] Furthermore, the elastic consistency of the formula mechanism is analyzed:

[0086]

[0087] By the properties of row random matrices, we have:

[0088] in, It is a random vector, independent of t, in which the states of the nodes will reach a consensus and converge to t in the above topology reconstruction method.

[0089] Based on the above method, a resilient consistency optimization device under an agent honeycomb architecture in this embodiment includes: at least one memory and at least one processor;

[0090] At least one memory for storing machine-readable programs;

[0091] At least one processor is used to invoke the machine-readable program to execute a resilient consistency optimization method under an agent honeycomb architecture.

[0092] The above-described specific embodiments are merely specific examples of the present invention. The patent protection scope of the present invention includes, but is not limited to, the above-described specific embodiments. Any technical solution that conforms to the above-described specific embodiments of the present invention and any appropriate changes or substitutions made by those skilled in the art should fall within the patent protection scope of the present invention.

[0093] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for resilient consistency optimization under an agent honeycomb architecture, characterized in that, This method, applied to distributed network communication and intelligent cloud platform cellular architecture, comprises the following steps: S1. Network attack modeling; Error data injection network attack model: x i (k+1)=f i (x i (k)); Where, x i (k) represents the information sent by agent i to neighboring agents, f i (x i (k) represents any update method; The intelligent agent honeycomb architecture introduces a consensus mechanism and implements a layered topology. The consensus mechanism involves a virtual leader layer with one or more leaders who do not communicate with each other but only send information to nodes in the first layer. This is achieved by dividing n nodes into m layers, labeled 1 to m from top to bottom. Nodes in the first layer only send information to the second layer, and nodes in the i-th layer send information to at least one node in the (i+1)-th layer. Information transmission between nodes in the i-th layer is bidirectional. Assume each layer has r nodes. i If there are n nodes, then r1 + r2 + K + r m =n; S2. Agent Composition: Assuming these protected nodes constitute the dominant connected subset of the network, the resilient consistency optimization method is as follows: The input is the set of ordinal numbers of the protected nodes in the network, represented as T = {v i The ordinal set of other nodes is represented as O = {v | i = 1, 2, ..., n1}. i The attack node ordinal set is represented as A = {v}, |i = n1+1, ..., n0} (n0 = n1+n2). i , |i=n0+1,...,n} and their corresponding state information; The communication network of an agent-based honeycomb architecture uses an undirected graph G = (V, E, A). Each node i of an agent has a scalar state at time k, denoted as... The states of all nodes in the system are represented by a vector x = [x1, x2, k, x...]. n ] T express; If a subset G of graph G = (V, E, A) is G d Each one that does not belong to G d A node has at least one neighbor belonging to G. d And belongs to G d All nodes form a connected graph, G d It is a dominant connected subset of G = (V, E, A); Specifically, it includes: S2.1, Each node v i Receive the states of the neighbors to form a set S2.2, Node v i By identifying the state information of trusted nodes and combining it with the state information of its own nodes, the minimum threshold is selected. and the maximum threshold S2.3, Judgment set S i element x in (k) j Does (k) satisfy...? Form a set S2.4, each node v i The update law is: S2.5, Repeat steps S2.1-S2.4 until |x i (k)-x j (k)|<ε,v j ∈T i ; S3. Convergence verification of the elastic consistency optimization method.

2. The elastic consistency optimization method under an agent honeycomb architecture according to claim 1, characterized in that, In step S3, each protected node or ordinary node v i The minimum threshold is determined based on the information of the protected nodes in the neighborhood and the node's own information. and maximum threshold Set R is obtained through threshold filtering. i (k), node v i In the k-th iteration, according to R i The sources of agents in (k) categorize states into three types: those from a set of trusted nodes. The state x of its own node i (k), and a set from ordinary nodes or attacking nodes.

3. The elastic consistency optimization method under an agent honeycomb architecture according to claim 2, characterized in that, Rewritten according to formula (1): in, The information can be represented by a minimum threshold and a maximum threshold, i.e., there exists 0 < ρ. j <1, satisfying the following equation 4. The elastic consistency optimization method under an agent honeycomb architecture according to claim 3, characterized in that, Formula (2) can be further rewritten as: remember but in, M 11 This represents the interaction between trusted nodes. M 21 This indicates the role of trusted nodes in relation to ordinary nodes. M 22 To represent the role of its own node, the consensus mechanism divides the nodes of the communication network into three layers: the first layer consists of trusted nodes, the second layer consists of ordinary nodes, and the third layer consists of attacking nodes.

5. The elastic consistency optimization method under an agent honeycomb architecture according to claim 4, characterized in that, If we represent the state of the trusted node and its own state, then... M(k) is a row random matrix. The elements of M(k) are non-negative, and the non-zero elements have a lower bound τ, where τ = 1 / (d M +1), d M For R i The maximum value of (k) base, i.e.

6. The elastic consistency optimization method under an agent honeycomb architecture according to claim 5, characterized in that, Analyze the elastic consistency of the formula mechanism: By the properties of row random matrices, we have: in, It is a random vector, independent of t, in which the states of the nodes reach a consensus and converge to t in the topology reconstruction method.

7. A resilient consistency optimization device under an intelligent agent honeycomb architecture, characterized in that, include: At least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is configured to invoke the machine-readable program to execute the method according to any one of claims 1 to 6.

Images