A method for selecting access to the network of magnetotelluric acquisition terminals based on comprehensive evaluation theory

Abstract

The present invention discloses a method for selecting and accessing a magnetotelluric acquisition terminal network based on comprehensive evaluation theory, belonging to the field of geophysical exploration. The method comprises the following steps: first, selecting candidate networks and their attributes and attribute values. Then, using the analytic hierarchy process (AHP) to calculate the subjective weight of each candidate network. Subsequently, preprocessing the original attribute values ​​of the candidate networks and using the improved CRITIC method to calculate the objective weight of each candidate network. Based on the subjective and objective weights of the candidate networks, a comprehensive weight is calculated using game theory. Finally, the TOPSIS method is used to calculate the scores of the candidate networks, thereby selecting the optimal network. The present invention utilizes multi-attribute decision making and a comprehensive subjective and objective weighting approach to model the magnetotelluric acquisition terminal access network selection problem. The method can effectively reflect the performance of wireless communication networks and ensure that the data transmission process of the magnetotelluric acquisition terminal has the characteristics of low delay rate, low switching rate, high reliability, high speed, and high stability.

Description

Technical Field

[0001] The present invention belongs to the field of geophysical exploration, and in particular relates to a method for selecting and accessing a magnetotelluric acquisition terminal network based on comprehensive evaluation theory. Background Art

[0002] Currently, data communication networks are transitioning to a wide-area, open, and shared internet, becoming an essential component of geophysical exploration. Currently, magnetotelluric acquisition terminals in the geophysical field typically utilize a variety of communication technologies, generally including wired and wireless communication technologies. However, wired communication systems are gradually being replaced by wireless communication due to their high construction costs, demanding operating environments, and poor communication flexibility. Wireless communication, with its inherent flexibility, low investment costs, and wide range of service terminal support, has gradually become a leader in data communication networks. However, due to the relatively dispersed distribution of magnetotelluric acquisition terminals and the varying reliability requirements of different acquisition locations, a single wireless data transmission network cannot meet the diverse needs of magnetotelluric acquisition terminals. Therefore, how to select network access for magnetotelluric acquisition terminals has become a critical issue in geophysical exploration.

[0003] Not only do magnetotelluric acquisition terminals have different requirements for data communication networks, but each wireless data communication technology also has its own advantages and disadvantages. Therefore, how to utilize the advantages of various wireless data communication networks to achieve complementary advantages and disadvantages, and comprehensively select networks based on the performance of wireless data communication networks and the needs of magnetotelluric acquisition terminals, while ensuring data transmission efficiency and stability, choosing the most appropriate wireless communication network access will be crucial to improving the performance of geophysical exploration magnetotelluric acquisition and transmission systems. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a network selection access method for magnetotelluric acquisition terminals based on comprehensive evaluation theory, so as to improve the problems of high transmission delay rate and high network switching rate of magnetotelluric acquisition systems.

[0005] The present invention is achieved in this way. Compared with the prior art, the present invention has the following beneficial effects:

[0006] The network selection access method for magnetotelluric acquisition terminals constructed by the method of the present invention can well reflect the performance of the wireless communication network and can ensure that the data transmission process of the acquisition terminal has the characteristics of low switching rate, low delay rate, high reliability, high speed and high stability.

[0007] The method of the present invention adopts the hierarchical analysis method and the improved CRITIC method to calculate the subjective and objective weights of candidate network attributes, which improves the reliability of network selection decision-making from a subjective perspective, reflects the importance that decision-makers attach to different indicators, and significantly reduces the sensitivity problem caused by differences in candidate network attributes from an objective perspective.

[0008] The method of the present invention obtains comprehensive weights through game theory, realizes the unification of subjective and objective weights, makes the determination of weights more comprehensive and specific, and well reflects the comprehensive relationship between the demand of the magnetotelluric acquisition terminal for access network and network attributes.

[0009] The beneficial effects of using the TOPSIS method to obtain the ranking results among candidate networks are: it can make full use of the information of the original data, accurately and specifically reflect the gaps between the candidate networks, and remove the restrictions on the selection of network attributes.

[0010] This network selection and access method combines the aforementioned methods, ensuring both the rationality of determining subjective and objective weights and the overall weight, and the scientific and acceptable nature of the ranking results. Simulation results demonstrate that, compared to existing network selection and access methods, this method can more effectively reduce network handoff rates and delays, while improving transmission speed, reliability, and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a schematic diagram of network access for a magnetotelluric acquisition terminal provided by an embodiment of the present invention;

[0012] Figure 2 It is a hierarchical model diagram of the hierarchical analysis method provided by an embodiment of the present invention;

[0013] Figure 3 This is a flow chart of a candidate network selection and access algorithm provided by an embodiment of the present invention;

[0014] Figure 4 This is a line chart comparing average switching rates of different network access selection algorithms provided by an embodiment of the present invention;

[0015] Figure 5 This is a line chart comparing average delay rates of different network access selection algorithms provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] A method for selecting and accessing a magnetotelluric acquisition terminal network based on comprehensive evaluation theory is characterized by the following steps:

[0018] Step 1: Obtain candidate networks at the acquisition terminal.

[0019] Magnetotelluric acquisition terminals are mostly distributed in the wild. In the wild, many wireless communication networks cannot cover them, so the number of available wireless communication networks is extremely small. Therefore, the present invention selects wireless communication networks that can perform well in the wild, including 5G, WiMAX, LTE-A and WLAN, as candidate networks for the magnetotelluric acquisition terminals to access.

[0020] Step 2: Determine the attributes of the candidate network and measure the attribute values ​​of the candidate network attributes.

[0021] In order to select the wireless communication network of the acquisition terminal from a more comprehensive and objective perspective, while reducing the influence of subjective bias and improving the scientificity and timeliness of network selection, a multi-attribute decision-making method is used for network selection, so it is necessary to select multiple attributes of the candidate network. The attribute selection of the candidate network in the present invention is determined based on the acquisition service requirements, including bandwidth, rate, security index, delay, packet loss rate and cost index. The original attribute value b of the corresponding candidate network attribute is the same as the original attribute value b of the candidate network attribute. mn Obtained by measurement method.

[0022] Wherein, m=4 is the number of candidate networks in the embodiment, n=6 is the number of candidate network attributes in the embodiment, and b mn Represents the original attribute value obtained by measuring the nth attribute in the mth candidate network.

[0023] Step 3: Use the analytic hierarchy process to determine the subjective weights of network attributes.

[0024] The analytic hierarchy process combines qualitative and quantitative methods. Starting from the essence of the evaluation problem, the evaluator determines the subjective weights of the candidate network attributes in accordance with the thinking mode of comparison, decomposition, judgment and synthesis, thereby making complex logical problems systematic and mathematical, making them easier to solve.

[0025] Step 3.1: Build a hierarchical analysis model, see Figure 3 As shown in the figure, based on the requirements of the acquisition terminal for wireless communication network access selection, the AHP model for wireless communication network access selection is divided into three layers: the target layer, the criterion layer, and the solution layer. The solution layer represents the candidate networks for network selection, the criterion layer represents the corresponding network attributes that affect wireless communication network selection, and the target layer selects the optimal communication network as the decision target.

[0026] Step 3.2: Construct a decision matrix. The decision matrix is ​​a matrix that expresses the relative importance of each indicator in each layer relative to the indicators in the upper layer. In this embodiment, the influence of the six attributes of the candidate network on the target optimal communication network is compared by pairwise comparison to construct a decision matrix A = (a ij ) n×n ,Right now

[0027] In the formula, use a ij Indicates c i Relative to c j The importance of c i Represents row attributes, c j Represents the column attribute, and the value of n is the total number of candidate network attributes. Therefore, in this embodiment, n=6, and the corresponding conditions must also be met, namely

[0028] The above formula shows a ij The value characteristics and a ji with a ij The value relationship of .

[0029] a in the decision matrix ij The value of is expressed using a nine-level scale method, which has a total of 9 values. The larger the value, the greater the c i Relative to c j The more important it is, as shown in Table 1 below.

[0030] Table 1

[0031]

[0032] Step 3.3: Preprocess the decision matrix to obtain a standardized decision matrix Right now

[0033]

[0034] Step 3.4: Use the eigenvalue method to calculate the subjective weight. First calculate The maximum eigenvalue λ max and the largest eigenvector x, i.e.

[0035]

[0036] Where x is a non-zero vector. After obtaining the maximum eigenvector x, it is necessary to normalize the maximum eigenvector x, that is,

[0037]

[0038] Thus we get W1=[w 11 ,w 12 ,…,w16 ] T Subjective weight matrix, where w 11 ,w 12 ,w 13 ,w 14 ,w 15 ,w 16 They represent the subjective weight values ​​of bandwidth, rate, security index, delay, packet loss rate and cost index respectively.

[0039] Step 3.5: Conduct consistency test to eliminate the misleading weight results caused by subjective judgment and prove that the subjective weight results are reasonable and acceptable. The test index used in consistency test is CI, that is,

[0040]

[0041] After obtaining the CI, it needs to be corrected to ensure that the consistency test results are reliable. The corrected formula is:

[0042]

[0043] Where RI represents the average random consistency index, and its value is shown in Table 2. After obtaining the results, the attribute weight results are judged according to the value of CR. If CR is less than 0.1, it means that the weight is reasonable, otherwise the weight needs to be recalculated.

[0044] Table 2

[0045]

[0046] The decision matrix of the embodiment and the corresponding calculation results are shown in Table 3 below. It can be seen that in the embodiment, CR=0.058, indicating that the subjective weight is reasonable and acceptable.

[0047] Table 3

[0048]

[0049]

[0050] Therefore, the subjective weight matrix of candidate network attributes W1 = [w 11 ,w 12 ,…,w 16 ] T The value is reasonable and acceptable, and the subjective weight result is W1 = [0.2109, 0.1014, 0.2109, 0.1218, 0.3218, 0.0332] T

[0051] Step 4: Use the improved CRITIC method to determine the objective weights of network attributes.

[0052] like Figure 2 As shown, this wireless communication network selection method uses an improved CRITIC method to determine the objective weights of network attributes, which requires unifying the dimensions of network attribute values. The improved CRITIC method considers both the contrast strength and the conflict between different network attributes. It uses a different standard deviation coefficient than the traditional CRITIC method to represent the contrast strength and the absolute value of the correlation coefficient to measure the conflict between different network attributes. This overcomes the effects of different orders of magnitude and negative correlation coefficients on the objective weight results, making it more suitable for calculating the objective weights of network attributes than the entropy weight method and the traditional CRITIC method.

[0053] Step 4.1: Construct the decision matrix C = (c ij ) m×n ,The data in the decision matrix should be the attribute values ​​after preprocessing the original attribute values. This is because the dimensions of various network attributes may not be the same, so some singular samples of network attributes will have an adverse effect on the network selection algorithm. Therefore, it is necessary to preprocess the attribute values ​​of the network attributes so that the preprocessed data is limited to a certain range, thereby eliminating the impact.

[0054] For positive attributes, including bandwidth, rate, and security index, they need to be preprocessed using the maximum value method:

[0055]

[0056] For inverse attributes, including delay, packet loss rate and cost index, they need to be preprocessed using the minimum method:

[0057]

[0058] The pre-processed network attribute values ​​constitute the decision matrix C = (c ij ) m×n ,Right now

[0059]

[0060] Wherein, m=4 is the number of candidate networks in the embodiment, n=6 is the number of candidate network attributes in the embodiment, and c mn Represents the attribute value of the nth attribute in the mth candidate network after preprocessing,

[0061] Step 4.2: Construct the comparative strength expression of network attributes. The traditional CRITIC method uses the standard deviation to express it. The improved CRITIC method in this embodiment uses the standard deviation coefficient to express it, that is,

[0062]

[0063] Where, δ j Represents the standard deviation of the j-th network attribute, Sdc j represents the standard deviation coefficient of the j-th network attribute, Represents the average value of the jth network attribute. Due to the differences in the order of magnitude of each network attribute, the result obtained by using the standard deviation has a large deviation, so the standard deviation coefficient needs to be used to overcome the deviation. The larger the standard deviation coefficient, the stronger the evaluation strength of the attribute, and the more weight it should be assigned.

[0064] Step 4.3: Construct the conflict expression of network attributes in the form of correlation coefficient, that is,

[0065]

[0066] Where r ij represents the correlation coefficient between the i-th network attribute and the j-th network attribute, M j Represents the conflict between different network attributes. Improved CRITIC method to calculate M j When using |r ij |Not r ij , thus solving the impact of negative correlation coefficient on calculation conflict.

[0067] Step 4.4: Construct an expression for the amount of information, i.e.

[0068] N j =Sdc j ×M j ,j=1,2,…,n

[0069] Step 4.5: Construct the final expression of the objective weight of network attributes obtained by the improved CRITIC method, that is,

[0070]

[0071] W2=[w 21 ,w 22 ,…,w 26 ] T

[0072] Where w 21 ,w 22 ,w 23 ,w 24 ,w 25 ,w 26 They represent the objective weights of bandwidth, rate, security index, latency, packet loss rate, and cost index respectively.

[0073] For this embodiment, the objective weight of network attributes obtained by the improved CRITIC method is:

[0074] W2=[0.2930,0.1685,0.1332,0.1332,0.1351,0.1370] T

[0075] Step 5: Use game theory to determine the comprehensive weight of network attributes.

[0076] The determination of weights will directly affect whether the network selection scheme is reasonable, and there are differences and functional relationships between different network attributes. Therefore, integrating and unifying the subjective and objective weights to form a scientific and comprehensive comprehensive weight is the key to network selection. Current research uses a simple weighted addition method or exponential multiplication to perform comprehensive weighting, which is not scientific and standardized. Therefore, Figure 2 As shown, in the embodiment, the game theory is used to perform comprehensive weighting, and the comprehensive weight is obtained by calculating the difference between the optimal weight and the subjective and objective weights and minimizing it.

[0077] Step 5.1: Construct a vector weight set, i.e.

[0078]

[0079] Where W k represents the weight vector involved in comprehensive weighting, k is the number of weight results, L represents the number of weighting methods used, in the embodiment, L = 2, respectively, the hierarchical analysis method and the improved CRITIC method, z k is the combined weight coefficient, W com Represents the comprehensive weight of network attributes.

[0080] Step 5.2: Construct a comprehensive weight assignment model based on game theory, namely

[0081]

[0082] Where, It's W k The transposed matrix of .

[0083] Step 5.3: Construct the minimum deviation condition, that is

[0084]

[0085] Where, It's W i The transposed matrix of , i=1,2,…,L.

[0086] Step 5.4: Calculate the combination weight coefficient z according to the above formula k ,Right now

[0087]

[0088] Step 5.5: Normalize the combined weight coefficient to obtain the final combined weight coefficient, that is,

[0089]

[0090] Where, satisfy and

[0091] Step 5.6: Combine the final combined weight coefficient with the desired subjective and objective weights to obtain the comprehensive weight W com ,Right now

[0092]

[0093] Where, Represents the comprehensive weight of bandwidth, rate, security index, delay, packet loss rate and cost index, and satisfies

[0094] The combined weight coefficients and comprehensive weights of network attributes for this example are shown in Table 4 below. The results are as follows:

[0095] Table 4

[0096]

[0097] Step 6: Use TOPSIS to rank the candidate networks.

[0098] The TOPSIS method ranks candidate networks by determining their proximity to the optimal and worst access networks. This method makes the most efficient use of raw data and has no strict attribute restrictions. Compared to the SAW and MEW methods, it is well suited for solutions like the embodiment, which have multiple evaluation targets. By calculating the distances between candidate networks and the optimal and worst networks, it can select both the optimal candidate network whose attribute values ​​are closest to the optimal value and the worst candidate network whose attribute values ​​are closest to the worst value. This allows the selection of the optimal candidate network that is closest to the optimal access network and farthest from the worst access network.

[0099] Step 6.1: The comprehensive weight W obtained in step 5 com Multiplying it with the pre-processed C, we get the weighted decision matrix P = (p ij ) m×n Right now

[0100]

[0101] Where p ijRepresents the attribute value of the j-th network attribute of the i-th candidate network after weighted processing.

[0102] Step 6.2: Calculate the positive and negative ideal values ​​as a reference for the candidate network attribute values, i.e.

[0103]

[0104]

[0105] Where, represents the optimal value of attribute j in the candidate network, Represents the worst value of attribute j in the candidate network. Step 6.3: After obtaining the positive and negative ideal values, it is necessary to calculate the Euclidean distance between each candidate network and the best network and the worst network, that is,

[0106]

[0107] Where, Represents the Euclidean distance between each candidate network and the best network, where Sd i - represents the Euclidean distance between each network and the worst network.

[0108] Step 6.4: Calculate the relative closeness η between each candidate network and the ideal access network i , thus obtaining the ranking result of the candidate network, that is,

[0109]

[0110] Where η i The larger the value, the closer the candidate network is to the optimal network, and the better the candidate network is. i The results of ranking the candidate networks are shown in Table 5 below:

[0111] Table 5

[0112]

[0113] As can be seen from Table 5, WLAN is the best network choice, followed by WiMAX, and then LTE-A and 5G.

[0114] like Figure 1 As shown, a simulation design is performed using the scenario in the figure, in which the positions of the magnetotelluric acquisition terminals are randomly generated, and the number gradually increases from 0 to 120. The performance of the algorithm in the present invention is compared with that of the networks selected by the hierarchical analysis method and the improved CRITIC method, respectively, to prove the superiority and effectiveness of the algorithm in the present invention.

[0115] like Figure 4As shown in the figure, the average switching rates of different network access selection algorithms are given. It can be seen from the figure that the average switching rates of the three algorithms fluctuate, but the algorithm adopted by the present invention has a lower flat switching rate than the other two algorithms. This is because the other two algorithms only consider a single weight and do not comprehensively consider the correlation between various network attributes, which leads to a high average switching rate.

[0116] like Figure 5 As shown in the figure, the average delay rates of different network access selection algorithms are given. It can be seen from the figure that the average delay rates of the three algorithms are all zero when the number of users is less than 50. However, as the number of users continues to increase, the average delay rates of the three algorithms gradually increase. However, the algorithm of the present invention still has the lowest average delay rate compared with the other two algorithms. This is because the algorithm adopted by the present invention comprehensively considers the collection business requirements and attribute characteristics. Even when the number of users is 120, the average delay rate is still reduced by 26.7% and 18.5% compared with the hierarchical analysis method and the improved CRITIC method.

[0117] In summary, it can be seen that the algorithm in the present invention can effectively reduce the average switching rate and the average delay rate, thereby improving network performance, making the optimal network selection access decision, and improving the communication performance of the magnetotelluric acquisition terminal.

[0118] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for selecting and accessing a magnetotelluric acquisition terminal network based on comprehensive evaluation theory, characterized in that: The method comprises the following steps: Step 1: Obtain the wireless communication network at the location of the magnetotelluric acquisition terminal, which includes 5G, WiMAX, LTE-A and WLAN, and use them as candidate networks; Step 2: Based on the service requirements of the acquisition terminal for the wireless communication network, select representative candidate network attributes, including bandwidth, rate, security index, latency, packet loss rate, and cost index, and measure the original attribute value of each network attribute; Step 3: Use the AHP method to determine the subjective weights of candidate network attributes. The AHP model for network access selection is divided into three levels: the solution level, the criteria level, and the target level. The solution level represents the candidate networks for network selection, the criteria level represents the corresponding network attributes, and the target level represents the optimal network selection solution. Step 4: Use the maximum method and the minimum method to preprocess the original attribute values ​​of the candidate network attributes, and use the improved CRITIC method to determine the objective weights of the candidate network attributes; Step 4 specifically includes: Step 4.1: Construct the decision matrix C = (c ij ) m×n ,The data in the decision matrix are the attribute values ​​after preprocessing the original attribute values. For positive attributes, including bandwidth, rate and security index, the maximum value method is used for preprocessing: For the inverse attributes, including delay, packet loss rate and cost index, the minimum value method is used to preprocess them: Where m = 4, is the number of candidate networks, n = 6, is the number of candidate network attributes, c mn represents the attribute value of the nth attribute in the mth candidate network after preprocessing, b mn represents the original attribute value of the nth attribute in the mth candidate network, The pre-processed network attribute values ​​constitute the decision matrix C = (c ij ) m×n : Where m = 4, is the number of candidate networks, n = 6, is the number of candidate network attributes, c mn Represents the attribute value of the nth attribute in the mth candidate network after preprocessing; Step 4.2: Construct an expression for the comparative strength of network attributes, expressed in the form of standard deviation coefficients: Where, δ j Represents the standard deviation of the j-th network attribute, Sdc j represents the standard deviation coefficient of the j-th network attribute, represents the average value of the j-th network attribute; Step 4.3: Construct the conflict expression of network attributes in the form of correlation coefficient: Where r ij represents the correlation coefficient between the i-th network attribute and the j-th network attribute, M j Represents the conflict between different network attributes; Step 4.4: Construct an expression for the amount of information: N j =Sdc j ×M j ,j=1,2,…,n; Step 4.5: Construct the final expression of the objective weight of network attributes obtained by the improved CRITIC method: W2=[in 21 ,In 22 ,…,In 26 ] T Where w 21 ,w 22 ,w 23 ,w 24 ,w 25 ,w 26 Represent the objective weights of bandwidth, rate, security index, latency, packet loss rate, and cost index respectively; Step 5: Obtain the comprehensive weight by calculating the difference between the subjective and objective weights compared to the optimal weight and minimizing it; Step 6: Sort the candidate networks using the TOPSIS method to obtain the optimal access network selection.

2. The method for selecting and accessing a magnetotelluric acquisition terminal network based on comprehensive evaluation theory according to claim 1, characterized in that: The step 3 specifically includes: Step 3.1: Establish the analytic hierarchy process model; Step 3.2: Construct a decision matrix to express the relative importance of each indicator in each layer compared with the upper layer indicators in the form of a matrix, compare the impact of the six attributes of the candidate network on the target optimal access network, and construct the decision matrix A = (a ij ) n×n , expressed as: In the formula, use a ij Indicates c i Relative to c j The importance of c i Represents row attributes, c j represents the column attribute, n=6 is the total number of candidate network attributes, a ij Requirements: Step 3.3: Preprocess the decision matrix to obtain a standardized decision matrix Right now Step 3.4: Use the eigenvalue method to calculate the subjective weight. First calculate The maximum eigenvalue λ max And the maximum eigenvector x, the calculation formula is: Where x is a non-zero vector. After obtaining the maximum eigenvector x, the maximum eigenvector x is normalized and expressed as: Get the subjective weight matrix W1=[w 11 ,w 12 ,…,w 16 ] T , where w 11 ,w 12 ,w 13 ,w 14 ,w 15 ,w 16 Represent the subjective weight values ​​of bandwidth, rate, security index, delay, packet loss rate and cost index respectively; Step 3.5: Perform consistency test. The test indicator used for consistency test is CI: After obtaining the CI, it is corrected and the corrected formula is: Where RI represents the average random consistency index. The attribute weight result is judged according to the value of CR. If CR is less than 0.1, it means that the weight is reasonable. Otherwise, the weight needs to be recalculated.

3. The method for selecting and accessing a magnetotelluric acquisition terminal network based on comprehensive evaluation theory according to claim 1, characterized in that: Step 5 specifically includes: Step 5.1: Construct vector weight set: Where W k represents the weight vector involved in comprehensive weighting, k is the number of weight results, L represents the number of weighting methods used, z k is the combined weight coefficient, W com Represents the comprehensive weight of network attributes; Step 5.2: Construct a comprehensive weight assignment model based on game theory: Where W j T It's W k The transposed matrix of Step 5.3: Construct the minimum deviation condition: Where W i T It's W i The transposed matrix of , i = 1, 2, ..., L; Step 5.4: Calculate the combined weight coefficient z according to step 5.3 k : Step 5.5: Normalize the combined weight coefficients to obtain the final combined weight coefficients: Where, satisfy and Step 5.6: Combine the final combined weight coefficient with the required subjective and objective weights to obtain the comprehensive weight W com : Represents the comprehensive weight of bandwidth, rate, security index, delay, packet loss rate and cost index, and satisfies 4. The method for selecting and accessing a magnetotelluric acquisition terminal network based on comprehensive evaluation theory according to claim 3, characterized in that: Step 6 specifically includes: Step 6.1: The comprehensive weight W obtained in step 5 com Multiplying it with the pre-processed C, we get the weighted decision matrix P = (p ij ) m×n : Where p ij Represents the attribute value of the jth network attribute of the i-th candidate network after weighted processing; Step 6.2: Calculate positive and negative ideal values ​​as references for candidate network attribute values: Where, represents the optimal value of attribute j in the candidate network, represents the worst value of attribute j in the candidate network; Step 6.3: After obtaining the positive and negative ideal values, calculate the Euclidean distance between each candidate network and the best network and the worst network: Where, represents the Euclidean distance between each candidate network and the best network, where represents the Euclidean distance between each network and the worst network; Step 6.4: Calculate the relative closeness between each candidate network and the ideal access network to obtain the ranking result of the candidate networks. The expression of relative closeness is:

Images