Switch load balancing method based on fuzzy logic control

Abstract

The invention discloses a switch load balancing method based on fuzzy logic control, and the method comprises the following steps: S1, monitoring a network state in real time through quantum sensing equipment, and obtaining related parameters; s2, transmitting the parameters to a fuzzy logic controller for processing; s3, generating a load distribution strategy according to a predefined rule; s4, transmitting the preliminary strategy to each switch node in the block chain network, and recording related data; s5, verifying the data by using a distributed consensus mechanism of the block chain; s6, dynamically evaluating a node trust value based on an enhanced fuzzy trust model, and selecting nodes meeting requirements to participate in a load balancing game decision; s7, carrying out cooperative calculation on the selected nodes by utilizing a game theory cooperation mechanism, and generating a globally optimized load distribution strategy; and S8, optimizing a load distribution strategy through a quantum genetic algorithm, and performing automatic execution through an intelligent contract. According to the method, the advantages of fuzzy logic, the block chain and quantum computing are combined, and the intelligence and safety of load balancing are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of network communication, and particularly relates to a switch load balancing method based on fuzzy logic control. BACKGROUND

[0002] With the rapid development of information technology, network load balancing has become a key means to ensure network stability and efficiency. In a highly interconnected network environment, uneven distribution of data traffic can cause some nodes to run overload, resulting in network bottlenecks and degradation of service quality, and even cause network paralysis. The application of load balancing technology can ensure the effective use of resources and the stable operation of services, and has become an indispensable part of modern networks.

[0003] Currently, load balancing technology mainly relies on static and dynamic methods. Static load balancing method usually distributes traffic to each node according to a fixed proportion based on pre-set rules. This method is simple to implement, but cannot adapt to dynamic changes in network environment, and is prone to low resource utilization and uneven load distribution. Dynamic load balancing method adjusts the traffic distribution strategy dynamically according to the current traffic situation by real-time monitoring of network state, and has higher adaptability and flexibility. However, the existing dynamic load balancing technology also has certain limitations.

[0004] Firstly, the existing dynamic load balancing method often relies on simple algorithms and models in data processing and decision-making process, and cannot effectively handle complex network states and multi-dimensional parameters. For example, traditional load balancing algorithms usually only consider single factors such as bandwidth utilization or node response time, without considering multiple influencing factors, which limits the accuracy and effectiveness of load distribution decisions.

[0005] Secondly, the existing technology lacks intelligent means in the generation and adjustment of load balancing strategy, and is difficult to quickly respond to changes in network state. Most load balancing methods rely on pre-set rules or simple feedback mechanisms, and cannot make intelligent analysis and decisions based on real-time monitoring of complex network states. Especially in the case of high load or burst traffic, traditional methods are difficult to quickly adjust the load distribution strategy, which can easily lead to network congestion and degradation of service quality.

[0006] In addition, the existing load balancing technology also has deficiencies in the security and credibility of data. With the increase of network attacks and malicious behaviors, it is difficult to guarantee the integrity of data and the credibility of node behavior in the process of load distribution, only relying on traditional distributed systems and centralized management methods. For example, malicious nodes may affect load balancing decisions by falsifying load information or tampering with data, thereby disrupting the normal operation of the network.

[0007] Finally, the prior art also has the problem of insufficient evaluation of node credibility. In the process of load balancing, the selection of nodes not only needs to consider their current load and performance, but also needs to evaluate their historical behavior and game strategy. However, traditional methods often ignore this, which may introduce unreliable or malicious nodes in the node selection process, affecting the security and stability of the overall network.

[0008] Therefore, how to provide a switch load balancing method based on fuzzy logic control is a problem that those skilled in the art need to solve. SUMMARY

[0009] One object of the present application is to provide a switch load balancing method based on fuzzy logic control. The present application combines fuzzy logic, blockchain consensus mechanism and quantum genetic algorithm, and generates and optimizes load distribution strategy through quantum perception device real-time monitoring network state, realizing efficient and secure load balancing. This method makes full use of the processing power of fuzzy logic, the security of blockchain and the optimization efficiency of quantum genetic algorithm, and has the advantages of high intelligence, strong adaptability and high security.

[0010] According to the switch load balancing method based on fuzzy logic control, the method comprises the following steps:

[0011] S1, real-time monitoring network state through quantum perception device, acquiring network parameters including bandwidth utilization rate, delay, jitter, etc., and transmitting parameters to fuzzy logic controller;

[0012] S2, fuzzy processing of received network parameters in fuzzy logic controller, generating corresponding fuzzy set, and processing fuzzy set according to predefined fuzzy inference rule to generate fuzzy output of load distribution;

[0013] S3, defuzzification processing of fuzzy output to generate preliminary load distribution strategy;

[0014] S4, transmitting preliminary load distribution strategy to each switch node in blockchain network, and recording historical behavior data and current game strategy of each switch node;

[0015] S5, verifying historical behavior data and current game strategy of each switch node through distributed consensus mechanism of blockchain;

[0016] S6, based on enhanced fuzzy trust model, dynamically evaluating trust value of each switch node, and selecting nodes with trust value meeting requirements to participate in load balancing game decision;

[0017] S7, using game theory collaboration mechanism to perform collaborative calculation on selected switch nodes to generate globally optimized load distribution strategy;

[0018] S8, optimizing the global load distribution strategy through the quantum genetic algorithm to generate a final global optimal load distribution scheme, and automatically executing the strategy through a smart contract on the blockchain.

[0019] Optionally, the S2 specifically includes:

[0020] S21, receiving network parameters obtained through quantum sensing devices, including but not limited to bandwidth utilization, delay and jitter;

[0021] S22, preprocessing the received network parameters, removing noise and normalizing data to form a set of normalized network parameters, the normalized bandwidth utilization being B n , the normalized delay being D n , and the normalized jitter being J n ;

[0022] S23, in the fuzzy logic controller, the normalized network parameters are fuzzified according to the predefined fuzzy set membership function to generate a fuzzy set, the fuzzy membership degree of bandwidth utilization being the fuzzy membership degree of delay being and the fuzzy membership degree of jitter being

[0023] S24, according to the predefined fuzzy inference rule, the fuzzy set is input into the fuzzy inference engine for processing, the fuzzy inference rule being defined as: if B n is and D n is and J n is , then the output is μ L , wherein μ L represents the fuzzy output of load distribution.

[0024] S25, using the fuzzy inference engine to perform rule matching and reasoning on the input fuzzy set to generate the corresponding fuzzy output μ L :

[0025]

[0026] wherein j represents an imaginary unit, represents the fuzzy membership degree of bandwidth utilization B i in the ith rule, θ B represents the phase angle of bandwidth utilization, θ D represents the phase angle of delay, θ J represents the phase angle of jitter, and w i is the weight of each rule, and N is the number of rules participating in the calculation.

[0027] Optionally, the S3 specifically comprises:

[0028] S31, the fuzzy output μ generated by the fuzzy inference engine is converted into a membership function μ(x), where x represents the specific value of the load allocation strategy; L and transmitted to the de-fuzzification module for de-fuzzification processing;

[0029] S32, the de-fuzzification method based on complex logic and entropy optimization is used to convert the fuzzy output μ into a membership function μ(x), where x represents the specific value of the load allocation strategy; L L (x), where x represents the specific value of the load allocation strategy;

[0030] S33, the fuzzy entropy H(μ L ) is calculated to evaluate the uncertainty of the fuzzy output:

[0031] H(μ L ) = -∫μ L (x)·log e (μ L (x))dx;

[0032] where μ L (x) represents the membership function of the load allocation strategy;

[0033] S34, the fuzzy entropy is adjusted using the complex logic optimization method, and the complex optimization function C(L) of the load allocation strategy is defined, with the optimization goal being to minimize the fuzzy entropy and the load deviation:

[0034] C(L) = argmin L [H(μ L )+λ·∫|x-L|·μ L (x)dx];

[0035] where λ is the balance coefficient, x represents the specific value of the load allocation strategy, L is the optimized load allocation strategy, |x-L| is the difference between the specific value x of the load allocation strategy and the optimized load allocation strategy L, and represents the load deviation;

[0036] S35, the optimized load allocation strategy L is stored in the intermediate data cache area of the system for subsequent processing steps;

[0037] S36, the optimized load allocation strategy L is subjected to adaptive quantization processing, and the dynamic quantization levels {Q1, Q2, …, Q n} of the load allocation strategy are set, where n is the number of quantization levels;

[0038] S37, the multi-objective optimization method is used to map the optimized load allocation strategy to the optimal quantization level Q i , and generate the quantized load allocation strategy L q ​:

[0039]

[0040] wherein a is a weight coefficient, Q i denotes a predefined quantization level, denotes a membership function of quantization level, m is the total number of quantization levels, |L-Q i | is the difference between the optimized load allocation strategy L and the quantization level Q i . i . k . 2 . i . k is the squared difference between the quantization levels Q

[0041] Optionally, the S4 specifically comprises:

[0042] S41, transmitting a preliminary load allocation strategy L q to each switch node in the blockchain network, each switch node receiving and storing the preliminary load allocation strategy;

[0043] S42, recording historical behavior data H and current game strategy G of each switch node, wherein the historical behavior data H = {h1, h2, …, hn}, n is the number of historical records, the current game strategy G = {g1, g2, …, gm}, m is the number of current game strategies; n m .

[0044] S43, writing the preliminary load allocation strategy and its related data into the blockchain;

[0045] S44, encrypting and verifying the historical behavior data and the current game strategy of each switch node by using quantum Fourier transform and chaotic mapping encryption method:

[0046]

[0047] wherein, is the quantum Fourier transform item of historical data, is the quantum Fourier transform item of game strategy, ψ i and respectively represent the quantum state of historical data and game strategy, λ k is a chaotic mapping parameter, N and M are the lengths of historical data and game strategy, is the chaotic mapping item of game strategy, QFT(H, G) is the encrypted verification data, h i is the i-th historical behavior data, g i ​The i-th game strategy data;

[0048] S45, verifying and consensus of the encrypted data through a quantum chain consensus mechanism:

[0049]

[0050] Where QFT i is a quantum Fourier transform and a single data item encrypted by a chaotic mapping, V(QFT) represents a verification value, λ k is a chaotic mapping parameter, θ is an integral variable, and η is a Fourier transform parameter, is a complex exponential representation of the phase angle θ i . is a Fourier transform integral of the encrypted data, is an integral of the chaotic mapping term, is a weighted sum of the Fourier transform parameter η, is a complexity verification integral of the encrypted data.

[0051] S46, selecting a trusted node according to the verification value V(QFT), and the selection process is based on the quantum Fourier transform and chaotic mapping verification results and historical behavior data of the node;

[0052] S47, recording the selection result of the trusted node on the quantum chain.

[0053] Optionally, the S6 specifically includes:

[0054] S61, dynamically evaluating the trust value T i of each switch node based on an enhanced fuzzy trust model;

[0055] S62, collecting historical behavior data H i and current game strategy G i of each switch node, wherein the historical behavior data H i ={h i1 ,h i2 ,…,h in} and the current game strategy G i ={g i1 ,g i2 ,…,g im};

[0056] S63, inputting the historical behavior data and the current game strategy into the enhanced fuzzy trust model:

[0057]

[0058] Where μ T (H i ,G i) is a trust degree function calculated by the enhanced fuzzy trust model ij is the historical behavior data of node i in the jth record, g ij is the game strategy data of node i in the jth record, θ H , θ G and θ j respectively represent the historical data, game strategy and weighted phase angle, j is the imaginary unit, is the phase angle weight of the historical data, is the phase angle weight of the game strategy;

[0059] S64, generate a preliminary trust value by combining the membership degrees of the historical data and the game strategy through fuzzy logic inference rules, and the inference rules are as follows:

[0060]

[0061] wherein, is the preliminary trust value, μ H (h i1 ) is the membership function value of the historical data h i1 , μ G (g i1 ) is the membership function value of the game strategy g i1 .

[0062] S65, dynamically adjust the preliminary trust value , and the adjusted trust value T i :

[0063]

[0064] wherein, T i is the trust value of the switch node, μ H (h ij ) is the membership function of the historical data h ij , μ G (g ij ) is the membership function of the game strategy g ij , λ and α are adjustment parameters, p is the number of records participating in the adjustment, H max is the maximum value of the historical data, G max is the maximum value of the game strategy, is the weighted sum of the normalized historical data and game strategy, is the quantized trust degree.

[0065] S66, record the adjusted trust value T i in the blockchain network.

[0066] Optionally, the S7 specifically comprises:

[0067] S71, selecting a trust value T i The qualified switch nodes participate in the load balancing game decision;

[0068] S72, constructing a game matrix G according to the selected switch nodes:

[0069]

[0070] Wherein, T i and T j represent the trust values of nodes i and j respectively, G ij represents the load allocation benefit between nodes i and j, θ ij represents the phase difference between nodes i and j, μ H (h ik ) and μ G (g ik ) represent the membership functions of historical data and game strategy respectively;

[0071] S73, using the game theory collaboration mechanism to solve the game matrix, and the optimization goal is to minimize the overall load fluctuation of the system:

[0072]

[0073] Wherein, is the overall load fluctuation function, ω ij represents the frequency difference between nodes i and j;

[0074] S74, generating a globally optimized load allocation strategy L * according to the results of the optimization algorithm:

[0075]

[0076] Wherein, represents the optimized load allocation coefficient, γ is an adjustment parameter, Δ i j represents the load difference between nodes i and j, μ L (x) represents the membership function of the load allocation strategy;

[0077] S75, distributing the globally optimized load allocation strategy L * to each switch node and recording on the blockchain.

[0078] Optionally, the S8 specifically comprises:

[0079] S81, inputting the globally optimized load allocation strategy L * to a quantum genetic algorithm optimization module;

[0080] S82, initialize a population with quantum superposition states, the population size is N, and each individual P i is represented as a linear combination of quantum bits;

[0081] S83, perform quantum gate operations on each individual in the population, including quantum mutation and quantum crossover;

[0082] S84, calculate the fitness value F(P i ) of each individual:

[0083]

[0084] where L ij is the jth load allocation value in the individual P i , γ j is an adjustment parameter, is a membership function of the load allocation value, μ H (h ik ) and μ G (g ik ) are membership functions;

[0085] S85, select individuals with higher fitness values for quantum measurement and quantum genetic operations to generate a new population, and repeat the quantum evolution process until the termination condition is met;

[0086] S86, according to the optimized population, select the individual with the highest fitness value as the final load allocation strategy, and input it to the smart contract module in the blockchain network to automatically execute the load allocation strategy.

[0087] The beneficial effects of the present application are:

[0088] (1) The present application combines fuzzy logic control, blockchain consensus mechanism and quantum genetic algorithm, significantly improves the intelligence and security of network load balancing. Especially in the generation and optimization of load allocation strategy, the present application overcomes the shortcomings of traditional methods, can dynamically adapt to complex and changeable network environment, and realizes efficient load allocation.

[0089] (2) The present application introduces quantum sensing devices and enhanced fuzzy trust model, realizes real-time monitoring of network state and dynamic evaluation of node trust degree, improves the accuracy and reliability of load balancing decision. At the same time, by using the distributed consensus mechanism of blockchain, the integrity and credibility of data are ensured, and the security threats of malicious attacks and data tampering are effectively dealt with.

[0090] (3) The present application provides an efficient, safe and intelligent network load balancing solution by comprehensively using multiple advanced technologies. The scheme generates a preliminary load distribution strategy through a fuzzy logic controller, and further optimizes it through a quantum genetic algorithm, ensuring the optimality of load distribution. The introduction of the blockchain consensus mechanism not only enhances data security, but also improves the transparency and fairness of the load balancing strategy, thereby achieving efficient utilization of network resources and improvement of service quality.

[0091] (4) By using the method of combining complex logic and entropy optimization, the present application improves the accuracy of the load distribution strategy during the de-fuzzification process. Through complex fitness function and phase angle calculation, the optimization effect of load distribution is further improved. The application of quantum gate operation and quantum evolution process makes the load balancing algorithm have higher calculation efficiency and optimization ability when dealing with large-scale data and complex network environment.

[0092] (5) The present application provides a comprehensive solution, from data monitoring, processing, transmission to the final load distribution strategy optimization and execution, each step is carefully designed and optimized to ensure the overall performance of the system. Through multi-dimensional optimization and intelligent processing, the present application has made significant improvement in the accuracy, real-time performance and security of load balancing, and is suitable for load balancing requirements in various complex network environments. BRIEF DESCRIPTION OF DRAWINGS

[0093] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, which together with the embodiments of the present application, serve to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0094] Figure 1 A flowchart of a switch load balancing method based on fuzzy logic control is proposed for the present application;

[0095] Figure 2 A flowchart of dynamically evaluating node trust value based on an enhanced fuzzy trust model is provided;

[0096] Figure 3 A schematic diagram of optimizing global load distribution strategy through quantum genetic algorithm is provided;

[0097] Figure 4 A schematic diagram of verifying data and recording historical behavior data through blockchain consensus mechanism is provided. DETAILED DESCRIPTION

[0098] The present application will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams that only illustrate the basic structure of the present application in a schematic manner, and therefore only show the components related to the present application.

[0099] ReferenceFigures 1-4 A switch load balancing method based on fuzzy logic control, comprising the following steps:

[0100] S1, real-time monitoring the network state through a quantum sensing device, obtaining network parameters including bandwidth utilization, delay, jitter, etc., and transmitting the parameters to a fuzzy logic controller;

[0101] S2, fuzzy processing the received network parameters in the fuzzy logic controller, generating corresponding fuzzy sets, and processing the fuzzy sets according to predefined fuzzy inference rules to generate fuzzy output of load distribution;

[0102] S3, defuzzification processing the fuzzy output to generate a preliminary load distribution strategy;

[0103] S4, transmitting the preliminary load distribution strategy to each switch node in the blockchain network, and recording the historical behavior data and current game strategy of each switch node;

[0104] S5, verifying the historical behavior data and current game strategy of each switch node through the distributed consensus mechanism of the blockchain;

[0105] S6, based on the enhanced fuzzy trust model, dynamically evaluating the trust value of each switch node, and selecting nodes with trust values meeting the requirements to participate in the load balancing game decision;

[0106] S7, using the game theory collaboration mechanism to perform collaborative calculation on the selected switch nodes to generate a globally optimized load distribution strategy;

[0107] S8, optimizing the global load distribution strategy through a quantum genetic algorithm to generate a final globally optimal load distribution scheme, and automatically executing the strategy through a smart contract on the blockchain.

[0108] In the embodiment, S2 specifically includes:

[0109] S21, receiving network parameters obtained by monitoring through a quantum sensing device, including but not limited to bandwidth utilization, delay, and jitter;

[0110] S22, preprocessing the received network parameters to remove noise and normalize the data, forming a normalized network parameter set, where the normalized bandwidth utilization is B n , the normalized delay is D n , and the normalized jitter is J n ;

[0111] S23, in the fuzzy logic controller, according to the predefined fuzzy set membership function, fuzzy processing the normalized network parameters to generate fuzzy sets, and the fuzzy membership degree of bandwidth utilization is The fuzzy membership degree of delay is The fuzzy membership degree of jitter is

[0112] S24, according to the pre-defined fuzzy inference rule, the fuzzy set input fuzzy inference engine for processing, the fuzzy inference rule is defined as: if B n is and D n is and J n is then the output is μ L , wherein μ L represents the fuzzy output of load distribution;

[0113] S25, using fuzzy inference engine for rule matching and reasoning of input fuzzy set, generate corresponding fuzzy output μ L :

[0114]

[0115] wherein j represents the imaginary unit, represents the fuzzy membership degree of bandwidth usage B i in the ith rule, θ B represents the phase angle of bandwidth usage, θ D represents the phase angle of delay, θ J represents the phase angle of jitter, w i is the weight of each rule, and N is the number of rules participating in the calculation.

[0116] In the embodiment, the S3 specifically comprises:

[0117] S31, the fuzzy output μ L generated by the fuzzy inference engine is transmitted to the defuzzification module for defuzzification processing;

[0118] S32, using the defuzzification method based on complex logic and entropy optimization combination, the fuzzy output μ L is converted into membership function μ L (x), wherein x represents the specific value of load distribution strategy;

[0119] S33, calculate the fuzzy entropy H(μ L ) to evaluate the uncertainty of fuzzy output:

[0120] H(μ L ) = -∫ μ L (x)·log e (μ L (x))dx;

[0121] wherein μL (x) is a membership function representing the load distribution strategy;

[0122] S34, adjust the fuzzy entropy by using a complex logic optimization method, define a complex optimization function C(L) of the load distribution strategy, and the optimization goal is to minimize the fuzzy entropy and the load deviation:

[0123] C(L) = argmin L [H(μ L )+λ·∫|x-L|·μ L (x)dx];

[0124] Wherein, λ is a balance coefficient, x represents a specific value of the load distribution strategy, L is the optimized load distribution strategy, |x-L| is the difference between the specific value x of the load distribution strategy and the optimized load distribution strategy L, and represents the load deviation;

[0125] S35, store the optimized load distribution strategy L in the intermediate data cache area of the system for subsequent processing steps;

[0126] S36, perform adaptive quantization processing on the optimized load distribution strategy L, and set the dynamic quantization levels {Q1, Q2, …, Q n} of the load distribution strategy, wherein n is the number of quantization levels;

[0127] S37, map the optimized load distribution strategy to the optimal quantization level Q i by using a multi-objective optimization method, and generate the quantized load distribution strategy L q :

[0128]

[0129] Wherein, α is a weight coefficient, Q i represents a predefined quantization level, represents a membership function of the quantization level, m is the total number of quantization levels, |L-Q i | is the difference between the optimized load distribution strategy L and the quantization level Q i , |Q i -Q k | 2 is the squared difference between the quantization level Q i and Q k , which is used to measure the distance between the quantization levels.

[0130] In this embodiment, the S4 specifically comprises:

[0131] S41, store the preliminary load distribution strategy L qTransmitted to each switch node in the blockchain network, each switch node receives and stores the preliminary load distribution strategy;

[0132] S42, record the historical behavior data H and current game strategy G of each switch node, where the historical behavior data H = {h1, h2, ..., h n}, n is the number of historical records, the current game strategy G={g1,g2,…,g m}, m is the number of current game strategies;

[0133] S43. Writing the preliminary load distribution strategy and related data into the blockchain;

[0134] S44. Use quantum Fourier transform and chaos map encryption methods to encrypt and verify the historical behavior data and current game strategy of each switch node:

[0135]

[0136] in, is the quantum Fourier transform term of the historical data, is the quantum Fourier transform term of the game strategy, ψ i and Represent the quantum states of historical data and game strategies, respectively, k is the chaotic mapping parameter, N and M are the lengths of historical data and game strategy respectively, is the chaotic mapping term of the game strategy, QFT(H,G) is the encrypted verification data, h i is the i-th historical behavior data, g i is the i-th game strategy data;

[0137] S45. Verify and reach consensus on the encrypted data through the Qtum consensus mechanism:

[0138]

[0139] Among them, QFT i is a single data item encrypted by quantum Fourier transform and chaotic mapping, V(QFT) represents the verification value, λ k is the chaotic mapping parameter, θ is the integral variable, η is the Fourier transform parameter, is the phase angle θ i The complex exponential representation of is the Fourier transform integral of the encrypted data, is the integral of the chaotic mapping term, is the weighted sum of the Fourier transform parameters η, Verify the complexity of the encrypted data;

[0140] S46. Select a trusted node based on the verification value V(QFT), where the selection process is based on the node's quantum Fourier transform and chaos map verification results and historical behavior data;

[0141] S47. Record the selection result of the trusted node on the quantum chain.

[0142] In this embodiment, S6 specifically includes:

[0143] S61. Based on the enhanced fuzzy trust model, dynamically evaluate the trust value T of each switch node i ;

[0144] S62: Collect historical behavior data H of each switch node i and the current game strategy G i , where historical behavior data H i ={h i1 ,h i2 ,…,h in}, current game strategy G i ={g i1 ,g i2 ,…,g im};

[0145] S63. Input historical behavior data and current game strategy into the enhanced fuzzy trust model:

[0146]

[0147] Among them, μ T (H i ,G i ) is the trust function calculated by the enhanced fuzzy trust model, h ij is the historical behavior data of node i in the jth record, g ij is the game strategy data of node i in the jth record, θ H ,θ G and θ j Represent historical data, game strategy and weighted phase angle respectively, j is an imaginary unit, is the phase angle weight of historical data, is the phase angle weight of the game strategy;

[0148] S64. Generate a preliminary trust value by combining historical data and the membership of the game strategy through fuzzy logic reasoning rules. The reasoning rules are as follows:

[0149]

[0150] in, is the initial trust value, μ H (h i1is the membership function value of historical data h i1 . G is the membership function value of game strategy g i1 . i1 .

[0151] S65, dynamically adjusts the preliminary trust value T . i

[0152]

[0153] wherein T i is the trust value of the switch node, μ H (h ij ) is the membership function of historical data h ij , μ G (g ij ) is the membership function of game strategy g ij , λ and α are adjustment parameters, p is the number of records participating in adjustment, H max is the maximum value of historical data, G max is the maximum value of game strategy, T is the normalized weighted sum of historical data and game strategy, and T is the quantified trust degree.

[0154] S66, records the adjusted trust value T i in the blockchain network.

[0155] In the embodiment, the S7 specifically comprises:

[0156] S71, selects the switch node whose trust value T i meets the requirements to participate in the load balancing game decision;

[0157] S72, constructs a game matrix G according to the selected switch node:

[0158]

[0159] wherein T i and T j respectively represent the trust values of nodes i and j, G ij represents the load distribution benefit between nodes i and j, θ ij represents the phase difference between nodes i and j, μ H (h ik ) and μ G (g ik ) respectively represent the membership functions of historical data and game strategy;

[0160] ​S73, solve the game matrix by using the game theory collaboration mechanism, and the optimization target is to minimize the overall load fluctuation of the system:

[0161]

[0162] wherein, is an overall load fluctuation function, ω ij represents the frequency difference between node i and node j;

[0163] S74, according to the result of the optimization algorithm, generate the globally optimized load distribution strategy L * :

[0164]

[0165] wherein, represents the optimized load distribution coefficient, γ is an adjustment parameter, Δ i j represents the load difference between node i and node j, μ L (x) represents the membership function of the load distribution strategy;

[0166] S75, distribute the globally optimized load distribution strategy L * to each switch node, and record it on the blockchain.

[0167] In the embodiment, the S8 specifically comprises:

[0168] S81, input the globally optimized load distribution strategy L * to the quantum genetic algorithm optimization module;

[0169] S82, initialize the population by using the quantum superposition state, the population size is N, and each individual P i is represented as a linear combination of quantum bits;

[0170] S83, perform quantum gate operations on each individual in the population, including quantum mutation and quantum crossover;

[0171] S84, calculate the fitness value F(P i ) of each individual:

[0172]

[0173] wherein, L ij is the jth load distribution value in the individual P i , γ j is an adjustment parameter, μ L *(L ij ) is the membership function of the load distribution value, μ H (h ik ) and μG (g ik ) is a membership function;

[0174] S85, select individuals with higher fitness values for quantum measurement and quantum genetic operation, generate a new population, repeat the quantum evolution process until the termination condition is met;

[0175] S86, according to the optimized population, select the individual with the highest fitness value as the final load distribution strategy, and input it into the smart contract module in the blockchain network to automatically execute the load distribution strategy.

[0176] Embodiment 1:

[0177] In order to verify the feasibility and effectiveness of the present application, the present application is applied to the network management system of a large Internet data center. The data center is interconnected by multiple high-performance switches, and internally arranged with thousands of servers, bearing a large amount of network traffic and computing tasks. With the continuous expansion of business scale, the data center is facing problems such as network bandwidth bottleneck, increased delay and data packet loss, especially during peak traffic periods, network congestion causes some servers to be overloaded, while other servers are idle, and network efficiency decreases significantly. To solve these problems, the data center decides to introduce the method and system of the present application to achieve more efficient network load balancing and resource optimization.

[0178] In the implementation of the method of the present application, first of all, quantum perception devices are deployed in the data center for real-time monitoring of the network status of each switch. The quantum perception device can accurately collect key network parameters including bandwidth utilization, delay, jitter, and pass these parameters to the fuzzy logic controller.

[0179] The fuzzy logic controller receives and processes these network parameters to generate a fuzzy set. Then, according to the predefined fuzzy inference rules, the preliminary load distribution strategy is obtained. This strategy is then transmitted to each switch node in the blockchain network, and the distributed ledger of the blockchain records the historical behavior data and current game strategy of each node.

[0180] Through the distributed consensus mechanism of the blockchain, each node verifies the strategy data to ensure the authenticity and consistency of the data. Next, the system uses the enhanced fuzzy trust model to dynamically evaluate the trust value of each switch node, and selects high-trust nodes to participate in the subsequent game decision.

[0181] Under the support of game theory collaboration mechanism, each switch node generates a globally optimized load distribution strategy through collaborative computation. To further optimize the strategy, the system uses quantum genetic algorithm for final optimization, and automatically executes the optimized strategy on the blockchain network through smart contracts. With the dynamic changes of network state, the system can trigger real-time monitoring and optimization process, ensuring efficient use of network resources and stability.

[0182] To evaluate the actual effect of the method, we conducted a one-day comparative test in a data center network. The test covered network performance during peak and off-peak hours, focusing on network bandwidth usage, latency, data packet loss rate, server load balancing, and node trustworthiness.

[0183] Table 1: Comparison of network performance indicators before and after the application of the method

[0184]

[0185] From the above test data, it can be seen that after using the method, the network performance of the data center has been significantly improved. During peak hours, network bandwidth usage decreased by 18 percentage points, network latency decreased by 35 milliseconds, and data packet loss rate decreased to less than 1%. In addition, the server load balancing degree increased from 65% in the traditional method to 92%, showing the significant effect of the invention in optimizing resource utilization and improving system stability. The node trustworthiness score also increased from 80 to 95, indicating that the system performs well in data reliability and security. Overall, the method effectively solves the problem of uneven resource allocation and network performance bottleneck in high-load operation of data centers, providing strong technical support for network management in similar environments.

[0186] The above is only the preferred embodiment of the present application, but the protection scope of the present application is not limited to this. Any skilled person in the art can make equivalent replacements or changes within the technical scope disclosed by the present application according to the technical solution and inventive concept of the present application, which should be covered within the protection scope of the present application.

Claims

1. A switch load balancing method based on fuzzy logic control, characterized in that: The steps include: S1. Use quantum sensing devices to monitor network status in real time, obtain network parameters including bandwidth utilization, delay, jitter, etc., and pass the parameters to the fuzzy logic controller; S2, fuzzifying the received network parameters in the fuzzy logic controller to generate corresponding fuzzy sets, and processing the fuzzy sets according to predefined fuzzy inference rules to generate fuzzy outputs for load distribution; S3, defuzzify the fuzzy output and generate a preliminary load distribution strategy; S4. Transmit the preliminary load distribution strategy to each switch node in the blockchain network, and record the historical behavior data and current game strategy of each switch node; S5. Verify the historical behavior data and current game strategy of each switch node through the distributed consensus mechanism of the blockchain; S6. Based on the enhanced fuzzy trust model, dynamically evaluate the trust value of each switch node and select nodes with trust values ​​that meet the requirements to participate in the load balancing game decision; S7. Using the game theory collaboration mechanism to perform collaborative calculations on the selected switch nodes to generate a globally optimized load distribution strategy; S8. Optimize the global load distribution strategy through quantum genetic algorithm to generate the final global optimal load distribution plan, and automatically execute the strategy through the smart contract on the blockchain.

2. A switch load balancing method based on fuzzy logic control according to claim 1, characterized in that: The S2 specifically includes: S21. Receive network parameters obtained through monitoring by the quantum sensing device, including but not limited to bandwidth usage, latency, and jitter; S22, pre-process the received network parameters, remove noise and normalize the data to form a standardized network parameter set. The bandwidth utilization rate after normalization is B n , after delay normalization, it is D n , after jitter normalization, it is J n ; S23. In the fuzzy logic controller, the normalized network parameters are fuzzified according to the predefined fuzzy set membership function to generate a fuzzy set. The fuzzy membership of the bandwidth utilization rate is μ Bn , the fuzzy membership of the delay is μ Dn , the blur membership of the jitter is μ Jn ; S24, according to the predefined fuzzy inference rules, the fuzzy set is input into the fuzzy inference engine for processing, and the fuzzy inference rules are defined as: if B n is μ Bn and D n is μ Dn And J n is μ Jn , then the output is μ L , where μ L Fuzzy output representing load distribution; S25, use the fuzzy inference engine to match and infer the input fuzzy set rules and generate the corresponding fuzzy output μ L : Where j represents the imaginary unit, Indicates the bandwidth usage B in the i-th rule i The fuzzy membership degree, θ B The phase angle representing bandwidth utilization, θ D The phase angle of the delay, θ J Indicates the phase angle of jitter, w i is the weight of each rule, and N is the number of rules involved in the calculation.

3. The switch load balancing method based on fuzzy logic control according to claim 1, characterized in that: The S3 specifically includes: S31, the fuzzy output μ generated by the fuzzy inference engine L Pass it to the defuzzification module for defuzzification processing; S32, using a defuzzification method based on complex logic and entropy optimization, the fuzzy output μ L Converted to membership function μ L (x), where x represents the specific value of the load distribution strategy; S33, calculate the fuzzy entropy H(μ L ) Evaluate the uncertainty of the fuzzy output: H(μ L )=-∫μ L (x)·log e (μ L (x))dx; Among them, μ L (x) represents the membership function of the load distribution strategy; S34. Use the complex logic optimization method to adjust the fuzzy entropy and define the complex optimization function C(L) of the load distribution strategy. The optimization goal is to minimize the fuzzy entropy and load deviation: C(L)=argmin L [H(μ L )+λ·∫|xL|·μ L (x)dx]; Where λ is the balancing coefficient, x represents the specific value of the load distribution strategy, L is the optimized load distribution strategy, and |xL| is the difference between the specific value x of the load distribution strategy and the optimized load distribution strategy L, indicating the load deviation. S35. Store the optimized load distribution strategy L in the intermediate data buffer area of ​​the system for use in subsequent processing steps; S36, perform adaptive quantization processing on the optimized load distribution strategy L, and set the dynamic quantization level {Q1, Q2, ..., Q n }, where n is the number of quantization levels; S37, using multi-objective optimization method to map the optimized load distribution strategy to the optimal quantization level Q i , generate the quantized load distribution strategy L q : Among them, α is the weight coefficient, Q i represents a predefined quantization level, represents the membership function of the quantization level, m is the total number of quantization levels, |LQ i | is the optimized load distribution strategy L and quantization level Q i The difference between |Q i -Q k | 2 is the quantization level Q i and Q k The squared difference between , which measures the distance between quantization levels.

4. The switch load balancing method based on fuzzy logic control according to claim 1, characterized in that: The S4 specifically includes: S41, the initial load distribution strategy L q Transmitted to each switch node in the blockchain network, each switch node receives and stores the preliminary load distribution strategy; S42, record the historical behavior data H and current game strategy G of each switch node, where the historical behavior data H = {h1, h2, ..., h n }, n is the number of historical records, the current game strategy G={g1,g2,…,g m }, m is the number of current game strategies; S43. Writing the preliminary load distribution strategy and related data into the blockchain; S44. Use quantum Fourier transform and chaos map encryption methods to encrypt and verify the historical behavior data and current game strategy of each switch node: in, is the quantum Fourier transform term of the historical data, is the quantum Fourier transform term of the game strategy, ψ i and Represent the quantum states of historical data and game strategies, respectively, k is the chaotic mapping parameter, N and M are the lengths of historical data and game strategy respectively, is the chaotic mapping term of the game strategy, QFT(H,G) is the encrypted verification data, h i is the i-th historical behavior data, g i is the i-th game strategy data; S45. Verify and reach consensus on the encrypted data through the Qtum consensus mechanism: Among them, QFT i is a single data item encrypted by quantum Fourier transform and chaos map, represents the verification value, λ k is the chaotic mapping parameter, θ is the integral variable, η is the Fourier transform parameter, e -jθi is the phase angle θ i The complex exponential representation of is the Fourier transform integral of the encrypted data, is the integral of the chaotic mapping term, is the weighted sum of the Fourier transform parameters η, Verify the complexity of the encrypted data; S46. Select a trusted node based on the verification value V(QFT), where the selection process is based on the node's quantum Fourier transform and chaos map verification results and historical behavior data; S47. Record the selection result of the trusted node on the quantum chain.

5. The switch load balancing method based on fuzzy logic control according to claim 1, characterized in that: The S6 specifically includes: S61. Based on the enhanced fuzzy trust model, dynamically evaluate the trust value T of each switch node i ; S62: Collect historical behavior data H of each switch node i and the current game strategy G i , where historical behavior data H i ={h i1 ,h i2 ,…,h in }, current game strategy G i ={g i1 ,g i2 ,…,g im }; S63. Input historical behavior data and current game strategy into the enhanced fuzzy trust model: Among them, μ T (H i ,G i ) is the trust function calculated by the enhanced fuzzy trust model, h ij is the historical behavior data of node i in the jth record, g ij is the game strategy data of node i in the jth record, θ H ,θ G and θ j Represent historical data, game strategy and weighted phase angle respectively, j is an imaginary unit, is the phase angle weight of historical data, is the phase angle weight of the game strategy; S64. Generate a preliminary trust value by combining historical data and the membership of the game strategy through fuzzy logic reasoning rules. The reasoning rules are as follows: in, is the initial trust value, μ H (h i1 ) is the historical data h i1 The membership function value, μ G (g i1 ) is the game strategy g i1 The membership function value of . S65. Based on the initial trust value Dynamic adjustment is performed, and the adjusted trust value T i : Among them, T i is the trust value of the switch node, μ H (h ij ) is the historical data h ij The membership function, μ G (g ij ) is the game strategy g ij The membership function, λ and α are adjustment parameters, p is the number of records involved in the adjustment, H max is the maximum value of historical data, G max is the maximum value of the game strategy, is the weighted sum of normalized historical data and game strategies, is the quantified trust. S66, the adjusted trust value T i Recorded in the blockchain network.

6. A switch load balancing method based on fuzzy logic control according to claim 1, characterized in that: The S7 specifically includes: S71. Select trust value T i Switch nodes that meet the requirements participate in the load balancing game decision-making; S72. Construct a game matrix G based on the selected switch nodes: Among them, T i and T j Represent the trust values ​​of node i and node j respectively, G ij represents the load distribution benefit between node i and node j, θ ij represents the phase difference between node i and node j, μ H (h ik ) and μ G (g ik ) represent the membership functions of historical data and game strategies respectively; S73. Use the game theory collaboration mechanism to solve the game matrix, and the optimization goal is to minimize the overall load fluctuation of the system: in, is the overall load fluctuation function, ω ij represents the frequency difference between node i and node j; S74. Generate a globally optimized load distribution strategy L based on the results of the optimization algorithm. * : in, represents the optimized load distribution coefficient, γ is the adjustment parameter, Δ ij represents the load difference between node i and node j, μ L (x) represents the membership function of the load distribution strategy; S75, the load distribution strategy L after global optimization * Distributed to each switch node and recorded on the blockchain.

7. A switch load balancing method based on fuzzy logic control according to claim 1, characterized in that: The S8 specifically includes: S81, the load distribution strategy L after global optimization * Input to the quantum genetic algorithm optimization module; S82. Use quantum superposition to initialize the population. The population size is N. Each individual P i Represented as a linear combination of qubits; S83. Perform quantum gate operations on each individual in the population, including quantum mutation and quantum crossover; S84, calculate the fitness value F(P i ): Among them, L ij For individual P i The jth load distribution value in ,γ j To adjust the parameters, is the membership function of the load distribution value, μ H (h ik ) and μ G (g ik ) is the membership function; S85. Select individuals with higher fitness values ​​for quantum measurement and quantum genetic operations to generate a new population, and repeat the quantum evolution process until the termination condition is met. S86. Based on the optimized population, select the individual with the highest fitness value as the final load distribution strategy, input it into the smart contract module in the blockchain network, and automatically execute the load distribution strategy.

Images