A fuzzy control protocol method for intelligent communication networks under DOS attacks

Abstract

This paper considers the challenges posed by DoS attacks in communication network systems and establishes a state-space model of intelligent communication network systems using a T-S fuzzy positive multi-agent system. Using T-S fuzzy rules to approximate linear systems, and employing Lyapunov functions and matrix decomposition techniques, a consistent control method based on a distributed PID controller is proposed. This method effectively prevents failures and other issues in intelligent communication network systems, regardless of DoS attacks. This modeling approach fully considers the positivity and nonlinearity inherent in actual communication network systems, and based on this, a fuzzy control protocol for intelligent communication networks under DoS attacks is designed.

Description

Technical Field

[0001] The present invention belongs to the field of automation technology and modern control, and relates to the modeling of a TS fuzzy positive multi-agent system and the consistency problem of an intelligent communication network system based on a PID control protocol under a DOS attack. Background Art

[0002] The openness, interactivity, and distributed nature of the internet enable people to realize their demands for information sharing, openness, flexibility, and speed. The internet has created an ideal space for information sharing, communication, and services. The rapid development and widespread application of internet technology has provided a powerful impetus for human development. Daily communication online has become an integral part of people's lives and work. However, as the frequency and volume of data transmitted over networks increase, data communication security issues are becoming increasingly serious. Once a security incident occurs in a communication network, it not only hinders communication between thousands of people but also leads to unforeseen losses of social and economic value. Therefore, the security of communication networks is of paramount importance. Only by comprehensively ensuring their security and reliability can the value of communication networks be maximized, thereby promoting my country's economic and social development.

[0003] The security of communication networks has become a major issue in the field of automatic control. Intelligent communication network systems are a type of control technology that will significantly reduce network attacks, reduce the occurrence of security incidents, and enhance the anti-interference capabilities of communication networks. Communication network security refers to measures implemented through hardware, software, operating systems, and other protective measures to protect communication data from infringement. Communication network security primarily refers to the security of communication network data. Generally, network security issues occur when a communication network system is attacked by an external source (such as a DOS attack), resulting in the leakage, theft, modification, and deletion of communication data, or when the continuity of data communication is affected, resulting in transmission interruptions, data packet loss, and the like. Multi-agent consistency refers to the state of each agent in a multi-agent system being ultimately consistent through a suitable control law. The controller used in this invention, which has a proportional-integral-differential control law, is called a PID controller. A PID controller (proportional-integral-differential controller) adjusts the deviation of the entire control system by setting the proportional unit P, integral unit I, and differential unit D, so that the actual value of the controlled variable is consistent with the predetermined value required by the process. The P part of the PID controller can reduce the steady-state error of the system, thereby improving the control accuracy of the system. The I part can improve the steady-state performance of the system. The D part can improve the dynamic performance of the system. The PID controller has been used in the research of multi-agent systems. In the communication network system, the data packets transmitted by each communication subnet are regarded as agents to form a communication topology for information exchange. Figure 1 The smart communication network system shown takes five communication subnets as an example. The connection between the communication subnets represents that the two communication subnets transmit data packets and can communicate with each other. The absence of a connection represents that there is no communication information exchange between the two subnets.

[0004] In a communication network system, the number of data packets transmitted by a communication subnet is always non-negative. Communication network security control primarily involves adjusting the structure of the communication subnet to maintain a relatively stable transmission structure between itself and surrounding communication subnets, changing the packet transmission rate as needed, improving anti-interference capabilities, and ensuring normal communication. Most current communication network systems utilize wireless communication technologies, such as LTE and 5G, as well as local area networks (LANs) such as WiFi, to achieve communication. When encountering external interference or malicious attacks, the security of the communication network system is difficult to ensure. Therefore, it is particularly important to leverage intelligent network communication systems to ensure normal communication even under external DoS attacks. The present invention utilizes a positive system to construct a communication network system, utilizing a system with non-negative characteristics for modeling, resulting in a more accurate model. In an intelligent communication network system, the frequency of DoS attacks on different subnets during the same period varies, resulting in different structural changes in the subnets and different packet transmission speeds. This presents significant nonlinear characteristics when modeling. In principle, the nonlinear process of packet flow changes can be described using a nonlinear positive system. However, nonlinear systems are difficult to handle, and even if a control method for a nonlinear general network system is designed, it is difficult to implement. The present invention uses the TS fuzzy model to approximate the nonlinear system into a system with linear characteristics, which is more convenient to handle and the designed related control method is easier to implement. The intelligent communication network system is composed of multiple subnets and a large number of data packets. It is more appropriate to regard the data packets transmitted by the subnet as multi-agents to establish a positive multi-agent system model. The use of PID controllers can improve the steady-state performance and dynamic performance of the system to control the subnet structure and the speed of transmitting data packets. Finally, based on the transmission data packet flow information, a fuzzy intelligent control protocol is constructed to achieve the normal transmission of data packets by all subnets, maintain consistent operation, and improve the ability to resist DOS attacks.

[0005] Based on the above analysis, this paper uses a positive TS fuzzy multi-agent system to establish a mathematical model for the operation of communication network data packets and proposes a consistency method based on a distributed PID controller. This consistency control method can improve the steady-state and dynamic performance of intelligent communication systems, ensuring system consistency and smooth communication. Summary of the Invention

[0006] This paper establishes a positive TS fuzzy multi-agent system model for the data packet operation process of the intelligent communication network system and proposes a consistency method based on a distributed PID controller. The specific technical solution is as follows:

[0007] A fuzzy control protocol method for intelligent communication networks under DOS attacks, comprising the following steps:

[0008] Step 1: Establish a state space model of the number of data packets in the intelligent communication network data transmission process;

[0009] Step 2: Establish a distributed PID control protocol for the intelligent communication network system;

[0010] Step 3: Design the proportion of distributed PID control protocol;

[0011] Step 4: Design the number of DOS attacks N(k0,k);

[0012] Step 5: Design the distributed PID control protocol and PID gain matrix of the communication network system according to the two situations with and without DOS attacks;

[0013] Step 6: Establish conditions for the smooth operation of the intelligent communication network system;

[0014] Step 7: Construct a positive verification process for the intelligent communication network system;

[0015] Step 8: Construct the consistency verification process of the intelligent communication network system.

[0016] Furthermore, the specific method of step 1 is: analyzing the dynamic process of data transmission in the communication network and collecting model data, and establishing a system state space model in the following form:

[0017]

[0018]

[0019] Among them, x i (k)∈R n is the data packet transmitted by the intelligent communication network, n represents the number of sub-networks in the intelligent communication network; y i (k)∈R q It indicates the number of data packets received by the data terminal, and q indicates the number of output sensors measured;

[0020] represents the control input for the next operating state of the ith subnet at time k, p represents the number of measurement input sensors; h r (θ(k)) represents the usage of the network in the communication system, A∈R n×n , B∈R n×p , C∈R q×n is the system matrix, R n , R q , R p , R n×n , R n×p , R q×n , representing n-dimensional vector, q-dimensional vector, p-dimensional vector, n×n-dimensional, n×p-dimensional, and q×n-dimensional matrices respectively.

[0021] Furthermore, the distributed PID control protocol in step 2 is constructed as follows:

[0022]

[0023] in, and These are the gain matrices of the PID control protocol to be designed; is a matrix related to the communication topology between agents. If the i-th agent can communicate with the j-th agent, then otherwise, Its dimension is related to the number of agents in the multi-agent system.

[0024] Furthermore, in step 3, e i (k) and Δy i (k) are the integral and differential parts of the distributed PID control protocol, where the integral part e i (k) = y i (k-1)+(1-α)e i (k-1), differential part Δy i (k) = y i (k)-y i (k-1), α is a tuning parameter and α>0.

[0025] Furthermore, in step 4, N(k0,k) satisfies the following conditions:

[0026]

[0027] Furthermore, step 5 includes the following steps:

[0028] Step 5.1 When there is no DOS attack in the communication network system, that is, k∈Θ(k0,k):

[0029] PID gain matrix:

[0030]

[0031]

[0032]

[0033]

[0034] Distributed PID control protocol:

[0035]

[0036] Step 5.2 When there is a DOS attack in the communication network system, that is, k∈Γ(k0,k):

[0037] PID gain matrix:

[0038]

[0039]

[0040]

[0041]

[0042] Distributed PID control protocol:

[0043]

[0044] Furthermore, the method for constructing the conditions for the stable operation of the intelligent communication network system in step 6 is as follows:

[0045] Design constant λ>1,0<α≤1, μ2>1, Vector Sum vector

[0046] Make

[0047]

[0048]

[0049]

[0050]

[0051]

[0052]

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060] Where 1-[D] ii =∑ i≠j [D] ij , then, under the distributed PID control protocol in step 2, the intelligent communication network system achieves positivity and consistency, and the gain matrix is

[0061]

[0062]

[0063] Among them 1 p represents a p-dimensional column vector whose elements are all 1, represents a p-dimensional column vector whose lth element is 1 and the rest of the elements are 0, and the DOS attack duration satisfies:

[0064]

[0065] Furthermore, the construction form of step 7, the positive verification process of the intelligent communication network system is as follows:

[0066] Step 7.1 Based on the state space model of the intelligent communication network established in step 1, the PID control protocol constructed in step 2, the integral part of the PID control protocol constructed in step 3, and the two cases divided according to whether there is a DOS attack in step 5, we can get

[0067] When there is no DOS attack k∈Θ(k0,k),

[0068]

[0069]

[0070] When there is a DOS attack k∈Γ(k0,k),

[0071]

[0072] Among them, I M and I q are the identity matrices of M×M and q×q dimensions respectively, is a Kronecker product operator, and

[0073]

[0074]

[0075] Step 7.2 Definition Combining steps 6 and 7.1, we can get the value when there is no DOS attack.

[0076]

[0077] in,

[0078]

[0079]

[0080]

[0081]

[0082]

[0083] When there is a DOS attack

[0084]

[0085] in,

[0086]

[0087]

[0088]

[0089]

[0090]

[0091] Let the matrix and The diagonal and off-diagonal matrices of are:

[0092]

[0093]

[0094]

[0095]

[0096] Step 7.3: Based on the first five conditions in step 6, design the gain matrix and and

[0097] get:

[0098]

[0099]

[0100] therefore The intelligent communication network system is positive in both cases.

[0101] Furthermore, the consistency verification process of the intelligent communication network system in step 8 is as follows:

[0102] Step 8.1 Select the linear copositive Lyapunov function based on whether there is a DOS attack

[0103] When k∈[k 2f-2 ,k 2f-1 )hour,

[0104]

[0105] When k∈[k 2f-1 ,k 2f ),

[0106]

[0107] in,

[0108]

[0109]

[0110]

[0111] Step 8.2 Construction The difference of k∈[k 2f-2 ,k 2f-1 )hour,

[0112]

[0113] When k∈[k 2f-1 ,k 2f )hour,

[0114]

[0115] Step 8.3 combines steps 8.1 and 8.2 to obtain 2f-2 ,k 2f-1 ),

[0116]

[0117] in

[0118]

[0119] When k∈[k 2f-1 ,k 2f )hour,

[0120]

[0121] in

[0122]

[0123] Step 8.4 From the conditions in step 6, we can get: when k∈[k 2f-2 ,k 2f-1 ),

[0124]

[0125]

[0126]

[0127]

[0128] Further we get:

[0129]

[0130]

[0131]

[0132] When k∈[k 2f-1 ,k 2f )hour,

[0133]

[0134]

[0135]

[0136]

[0137] Further we get:

[0138]

[0139]

[0140]

[0141] 8.5 According to the conclusions of steps 8.1 and 8.4, the conditions are met in both cases.

[0142] Therefore, the smart communication network system is consistent, that is, each subnet in the smart communication network system can operate normally with the surrounding subnets regardless of whether there is a DOS attack.

[0143] The beneficial effects of the present invention are as follows:

[0144] The method of the present invention first uses a TS fuzzy positive multi-agent system to establish a state-space model of the communication network system. Using a communication topology diagram and graph theory, information exchange between data packets transmitted across different subnets is achieved. Using the constructed linear copositive Lyapunov function and matrix decomposition techniques, a distributed PID control protocol is designed. Its positivity and consistency are analyzed, achieving a relatively stable motion state between data packets transmitted across each subnet in the communication network system. BRIEF DESCRIPTION OF THE DRAWINGS

[0145] Figure 1 It is a process diagram of a communication network system to which the present invention is applied;

[0146] Figure 2 This is a schematic diagram of the intelligent communication network system of the present invention. DETAILED DESCRIPTION

[0147] The purpose of this invention is to study the consistency control problem of data packets transmitted in subnets in intelligent communication network systems by using a distributed PID control protocol, and to provide a distributed PID control protocol method for intelligent communication network systems based on TS fuzzy positive multi-agent system modeling.

[0148] like Figure 2 As shown, the fuzzy control protocol method of the intelligent communication network under DoS attack of the present invention includes the following steps:

[0149] Step 1: Establish a state space model of the number of data packets in the intelligent communication network data transmission process. The specific method is: analyze the dynamic process of communication network data transmission and collect model data to establish a system state space model in the following form:

[0150]

[0151]

[0152] Among them, x i (k)∈R n is the data packet transmitted by the intelligent communication network, n represents the number of sub-networks in the intelligent communication network; y i (k)∈R q It indicates the number of data packets received by the data terminal, and q indicates the number of output sensors measured; represents the control input for the next operating state of the ith subnet at time k, p represents the number of measurement input sensors; h r (θ(k)) represents the usage of the network in the communication system, A∈R n×n , B∈R n×p , C∈R q×n is the system matrix, R n, R q , R p , R n×n , R n×p , R q×n , representing n-dimensional vector, q-dimensional vector, p-dimensional vector, n×n-dimensional, n×p-dimensional, and q×n-dimensional matrices respectively.

[0153] Step 2: Establish a distributed PID control protocol for the intelligent communication network system. Its construction form is as follows:

[0154]

[0155] in, and These are the gain matrices of the PID control protocol to be designed; is a matrix related to the communication topology between agents. If the i-th agent can communicate with the j-th agent, then otherwise, Its dimension is related to the number of agents in the multi-agent system.

[0156] Step 3: Design the proportional, integral, and differential parts of the distributed PID control protocol. i (k) and Δy i (k) are the integral and differential parts of the distributed PID control protocol, where the integral part e i (k) = y i (k-1)+(1-α)e i (k-1), differential part Δy i (k) = y i (k)-y i (k-1), α is a small tuning parameter and α>0.

[0157] Step 4: Design the number of DOS attacks N(k0,k) to meet the following conditions:

[0158]

[0159] Step 5: According to the actual situation, there are two situations: with DOS attack and without DOS attack, and the distributed PID control protocol and PID gain matrix of the communication network system are designed respectively.

[0160] Step 5.1 When there is no DOS attack in the communication network system, that is, k∈Θ(k0,k):

[0161] PID gain matrix:

[0162]

[0163]

[0164]

[0165]

[0166] Distributed PID control protocol:

[0167]

[0168] Step 5.2 When there is a DOS attack in the communication network system, that is, k∈Γ(k0,k):

[0169] PID gain matrix:

[0170]

[0171]

[0172]

[0173]

[0174] Distributed PID control protocol:

[0175]

[0176] Step 6: The conditions for the stable operation of the intelligent communication network system are constructed as follows:

[0177] Design constant λ>1,0<α≤1, μ2>1, Vector Sum vector

[0178] Make

[0179]

[0180]

[0181]

[0182]

[0183]

[0184]

[0185]

[0186]

[0187]

[0188]

[0189]

[0190]

[0191]

[0192] in Then, under the distributed PID control protocol in step 2, the intelligent communication network system achieves positivity and consistency, and the gain matrix is

[0193]

[0194]

[0195] Among them 1 p represents a p-dimensional column vector whose elements are all 1, represents a p-dimensional column vector whose lth element is 1 and the rest of the elements are 0, and the DOS attack duration satisfies:

[0196]

[0197] Step 7: The positive verification process of the smart communication network system is constructed as follows:

[0198] Step 7.1 Based on the state space model of the intelligent communication network established in step 1, the PID control protocol constructed in step 2, the integral part of the PID control protocol constructed in step 3, and the two cases divided according to whether there is a DOS attack in step 5, we can get

[0199] When there is no DOS attack k∈Θ(k0,k),

[0200]

[0201]

[0202] When there is a DOS attack k∈Γ(k0,k),

[0203]

[0204] Among them, I M and I q are the identity matrices of M×M and q×q dimensions respectively, is a Kronecker product operator, and

[0205]

[0206]

[0207] Step 7.2 Definition Combining steps 6 and 7.1, we can get the value when there is no DOS attack.

[0208]

[0209] in,

[0210]

[0211]

[0212]

[0213]

[0214]

[0215] When there is a DOS attack

[0216]

[0217] in,

[0218]

[0219]

[0220]

[0221]

[0222]

[0223] Let the matrix and The diagonal and off-diagonal matrices of are:

[0224]

[0225]

[0226]

[0227]

[0228] Step 7.3: Based on the first five conditions in step 6, design the gain matrix and and

[0229] You can get:

[0230]

[0231]

[0232] therefore Therefore, the intelligent communication network system is positive in both cases.

[0233] Step 8: The consistency verification process of the smart communication network system is as follows:

[0234] Step 8.1 Select the linear copositive Lyapunov function based on whether there is a DOS attack

[0235] When k∈[k 2f-2 ,k 2f-1 )hour,

[0236]

[0237] When k∈[k 2f-1 ,k 2f ),

[0238]

[0239] in,

[0240]

[0241]

[0242]

[0243] Step 8.2 Construction The difference,

[0244] When k∈[k 2f-2 ,k 2f-1 )hour,

[0245]

[0246] When k∈[k 2f-1 ,k 2f )hour,

[0247]

[0248] Step 8.3 Combining steps 8.1 and 8.2, we can get 2f-2 ,k 2f-1 ),

[0249]

[0250] in

[0251]

[0252] When k∈[k 2f-1 ,k 2f )hour,

[0253]

[0254] in

[0255]

[0256] Step 8.4 From the conditions in step 6, we can get: when k∈[k 2f-2 ,k 2f-1 ),

[0257]

[0258]

[0259]

[0260]

[0261] Further we get:

[0262]

[0263]

[0264]

[0265] When k∈[k 2f-1 ,k 2f )hour,

[0266]

[0267]

[0268]

[0269]

[0270] Further we get:

[0271]

[0272]

[0273]

[0274] Step 8.5 Based on the conclusions of steps 8.1 and 8.4, the condition ΔV(X(k))<0 is satisfied in both cases.

[0275] Therefore, the smart communication network system is consistent, that is, each subnet in the smart communication network system can operate normally with the surrounding subnets regardless of whether there is a DOS attack.

Claims

1. A fuzzy control protocol method for intelligent communication networks under DOS attacks, characterized in that The steps are as follows: Step 1: Establish a state space model of the number of data packets in the intelligent communication network data transmission process; Step 2: Establish a distributed PID control protocol for the intelligent communication network system; Step 3: Design the proportion of distributed PID control protocol; Step 4: Design the number of DOS attacks N(k0,k); Step 5: Design the distributed PID control protocol and PID gain matrix of the communication network system according to the two situations with and without DOS attacks; Step 6: Establish conditions for the smooth operation of the intelligent communication network system; Step 7: Construct a positive verification process for the intelligent communication network system; Step 8: Construct the consistency verification process of the intelligent communication network system; The specific method of step 1 is: analyze the dynamic process of data transmission in the communication network and collect model data to establish the system state space model in the following form: Among them, x i (k)∈R n is the data packet transmitted by the intelligent communication network, n represents the number of sub-networks in the intelligent communication network; y i (k)∈R q It indicates the number of data packets received by the data terminal, and q indicates the number of output sensors measured; represents the control input for the next operating state of the ith subnet at time k, p represents the number of measurement input sensors; h r (θ(k)) represents the usage of the network in the communication system, A∈R n×n , B∈R n×p , C∈R q×n is the system matrix, R n , R q , R p , R n×n , R n×p , R q×n , respectively represent n-dimensional vector, q-dimensional vector, p-dimensional vector, n×n-dimensional, n×p-dimensional, and q×n-dimensional matrices; The distributed PID control protocol in step 2 is constructed as follows: in, and These are the gain matrices of the PID control protocol to be designed; is a matrix related to the communication topology between agents. If the i-th agent can communicate with the j-th agent, then otherwise, Its dimension is related to the number of agents in the multi-agent system; Step 3 i (k) and Δy i (k) are the integral and differential parts of the distributed PID control protocol, where the integral part e i (k) = y i (k-1)+(1-α)e i (k-1), differential part Δy i (k) = y i (k)-y i (k-1), α is a tuning parameter and α>0; Step 4, N(k0,k) satisfies the following conditions: Step 5 includes the following steps: Step 5.1 When there is no DOS attack in the communication network system, that is, k∈Θ(k0,k): PID gain matrix: Distributed PID control protocol: Step 5.2 When there is a DOS attack in the communication network system, that is, k∈Γ(k0,k): PID gain matrix: Distributed PID control protocol:

2. The fuzzy control protocol method for intelligent communication networks under DOS attacks according to claim 1 is characterized in that: The method for constructing the conditions for the stable operation of the intelligent communication network system in step 6 is as follows: Design constant λ>1,0<α≤1, μ2>1, Vector Sum vector Make Where 1-[D] ii =∑ i≠j [D] ij , then, under the distributed PID control protocol in step 2, the intelligent communication network system achieves positivity and consistency, and the gain matrix is Among them 1 p represents a p-dimensional column vector whose elements are all 1, represents a p-dimensional column vector whose lth element is 1 and the rest of the elements are 0, and the DOS attack duration satisfies:

3. The fuzzy control protocol method for intelligent communication networks under DOS attacks according to claim 2 is characterized in that: Step 7: The construction form of the positive verification process of the intelligent communication network system is as follows: Step 7.1 Based on the state space model of the intelligent communication network established in step 1, the PID control protocol constructed in step 2, the integral part of the PID control protocol constructed in step 3, and the two cases divided according to whether there is a DOS attack in step 5, we can get When there is no DOS attack k∈Θ(k0,k), When there is a DOS attack k∈Γ(k0,k), Among them, I M and I q are the identity matrices of M×M and q×q dimensions respectively, is a Kronecker product operator, and Step 7.2 Definition Combining steps 6 and 7.1, we can get the value when there is no DOS attack. in, When there is a DOS attack in, Let the matrix and The diagonal and off-diagonal matrices of are: Step 7.3: Based on the first five conditions in step 6, design the gain matrix and and get: therefore The intelligent communication network system is positive in both cases.

4. The fuzzy control protocol method for intelligent communication networks under DOS attacks according to claim 3 is characterized in that: The consistency verification process of the intelligent communication network system in step 8 is as follows: Step 8.1 Select the linear copositive Lyapunov function based on whether there is a DOS attack When k∈[k 2f-2 ,k 2f-1 )hour, When k∈[k 2f-1 ,k 2f ), in, Step 8.2 Construction The difference, When k∈[k 2f-2 ,k 2f-1 )hour, When k∈[k 2f-1 ,k 2f )hour, Step 8.3 combines steps 8.1 and 8.2 to obtain When k∈[k 2f-2 ,k 2f-1 ), in When k∈[k 2f-1 ,k 2f )hour, in Step 8.4 From the conditions in step 6, we can get: when k∈[k 2f-2 ,k 2f-1 ), Further we get: When k∈[k 2f-1 ,k 2f )hour, Further we get: Step 8.5 Based on the conclusions of Step 8.1 and Step 8.4, the conditions are met in both cases.

Images