RIS mode selection method based on grey correlation clustering method

Abstract

The invention relates to the technical field of data analysis, in particular to an RIS mode selection method based on a grey correlation clustering method. The method comprises the following steps: S1, acquiring communication data of a user by a base station, and dividing the communication data of the user into two types of data: continuously valued numerical data and discretely valued type data; s2, establishing a gray possibility degree mapping matrix for the type data of discrete values, and directly mapping discrete variables to final pattern classification; s3, establishing a gray possibility degree function for the numerical value class data with continuous values, firstly mapping continuous variables to connotation classification, and then mapping the continuous variables to final pattern classification; s4, setting weights, and performing overall statistics on prediction scores of various modes through the weights; and S5, selecting an optimal decision for the RIS mode according to the sorting of the prediction scores. According to the method, the calculation complexity is reduced while accurate capturing of the user data is guaranteed, and the method can be more suitable for real-time application under a large model.

Description

Technical Field

[0001] The invention relates to the technical field of data analysis, and in particular to a RIS mode selection method based on a grey relational clustering method. Background Art

[0002] Reconfigurable Intelligent Surface (RIS) is an emerging wireless communication technology, and its reflection mode is one of the core characteristics of this technology. Reconfigurable intelligent surfaces are usually composed of a large number of programmable reflection units, which can dynamically adjust their phase and amplitude to achieve intelligent reconstruction of the propagation channel. When the wireless signal reaches the RIS surface, these reflection units will reflect, refract or scatter the signal by adjusting the phase and amplitude according to the preset algorithm or real-time feedback to achieve the purpose of optimizing the signal propagation environment.

[0003] In terms of application, the reflection mode of RIS shows great potential in many fields: In complex indoor environments (such as office buildings, shopping malls, etc.), RIS can improve signal coverage and reduce blind spots. In large-scale outdoor environments (such as city streets, stadiums, etc.), RIS can optimize signal propagation paths and enhance signal strength and stability. In future 5G and 6G networks, RIS can be used for beamforming, interference management, and spectrum efficiency improvement to meet higher bandwidth and connection requirements. RIS can also provide a more stable and efficient communication environment for IoT devices, especially in equipment-intensive environments such as smart cities and industrial automation.

[0004] At present, most RIS mode selection technologies are based on engineers' experience and preset rules or on simple data statistics and machine learning methods to analyze user data to select modes. Simple mode selection methods based on experience generally lack flexibility and are difficult to adapt to rapidly changing environments and diverse user needs. In addition, their simple models often fail to capture complex network characteristics, which greatly limits their accuracy. As for the method of using machine learning algorithms for mode selection, its computational complexity is high and it is difficult to achieve real-time application under large models. Summary of the invention

[0005] The present invention provides a RIS mode selection method based on a grey relational clustering method, aiming to solve the technical problems existing in most current RIS mode selection technologies.

[0006] The present invention provides a RIS mode selection method based on a grey relational clustering method, comprising the following steps:

[0007] S1. The base station obtains the user's communication data and divides the user's communication data into two types of data: continuous value data and discrete value type data;

[0008] S2. For the discrete value type data, a grey possibility mapping matrix is ​​established to cluster the discrete variables and directly map the discrete variables to the final pattern classification;

[0009] S3. For numerical data with continuous values, a grey possibility function is established to cluster continuous variables, and the continuous variables are first mapped to connotation classification and then mapped to the final pattern classification;

[0010] S4. Set weights and make overall statistics on the prediction scores of various modes by weights;

[0011] S5. Select the optimal decision for the RIS model based on the ranking of the prediction scores.

[0012] As a further improvement of the present invention, in step S1, the numerical data of continuous value includes distance value, SNR value, angle value; the type data of discrete value includes service priority value.

[0013] As a further improvement of the present invention, the step S2 specifically includes:

[0014] Shared numerical data, define the rth Continuous variable χ r Its connotation category Grey possibility function for:

[0015]

[0016] in, is the gray value interval of the kth class, for different variables χ r , the number of categories Different, and different k corresponding to each of Different; y is the value of χ, y∈[a,b] is the mathematical representation of y in the interval a~b, y is not in this interval; when χ r When the value of is different, χ r The probability of taking different categories is also different.

[0017] As a further improvement of the present invention, in step S3, a gray possibility function is established for the numerical data of continuous values, and the continuous variables are clustered, and the continuous variables are first mapped to the connotation classification, which specifically includes:

[0018] Shared type data, define the dth Discrete variable χ d For Category The gray possibility mapping matrix is:

[0019]

[0020] The matrix size is N rows. Column, that is There are N discrete values ​​that can be assigned to in the categories; in the matrix, the nth (n=1,…N) row and the kth (k=1,2,…,K χ ) column elements mean that when χ d Pick When X d The probability of belonging to the kth class, the value of the element is less than 1, It represents the discrete value variable X d For Category The gray possibility.

[0021] As a further improvement of the present invention, in step S3, after the continuous variable is mapped to the connotation classification, it is mapped to the final pattern classification, which specifically includes:

[0022] Define the rth Continuous variable X r Connotation category The two-step possible degree mapping matrix is:

[0023]

[0024] The size of this matrix is ​​K A OK Column, K A is the final target category number, is the variable χ r The number of connotation categories; the ξ-th (ξ=1,2,…,K A ) The meaning of the column elements is that when the variable χ r When it belongs to the kth connotation classification, the probability of belonging to the ξth final target category, the value of the element is less than 1, It means the rth Continuous variable X r Connotation category The second step possibility.

[0025] As a further improvement of the present invention, in step S4, the specific process of setting the weight includes:

[0026] Shared Numerical data, Type data, for the rth Continuous variables, whose weights are defined as For the dth Discrete value variable, define its weight as ω d , the initial calculation formula is as follows:

[0027]

[0028] As a further improvement of the present invention, in step S4, the prediction scores of various modes are generally counted by weight, specifically including:

[0029] For the mth (m=1,…,K C ) class model, predicting the score γ selected by the current user i m Calculated by the following formula:

[0030]

[0031] in For variable X d The value is of

[0032] As a further improvement of the present invention, step S5 specifically includes:

[0033] If the total number of users is n, the final decision score is calculated as follows:

[0034]

[0035] Again m The values ​​are sorted in descending order, and the corresponding index values ​​are This index sequence is the ordering of the RIS modes predicted to be most suitable for the communication needs at this time; The corresponding mode is the predicted optimal mode, and the RIS mode will be automatically set to this optimal mode.

[0036] The beneficial effects of the present invention are as follows: the method can obtain and quantify user data, divide user data into continuous data and discrete data, use the grey possibility function and mapping matrix to realize clustering of continuous variables and discrete variables, directly map discrete variables to final mode classification, map continuous variables to connotation classification first, and then to final mode classification, and then perform overall statistics on the prediction scores of various modes by adaptively setting weights, and finally realize the decision of RIS mode. While ensuring accurate capture of user data, the method reduces the complexity of calculation and is more adaptable to real-time applications under large machine learning models. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of the RIS mode selection method based on the grey relational clustering method of the present invention;

[0038] Figure 2 It is a schematic diagram of a model for selecting the reflection mode of the reconfigurable intelligent surface by using a grey correlation clustering method in the present invention. DETAILED DESCRIPTION

[0039] 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 accompanying drawings and embodiments.

[0040] like Figure 1 As shown, the present invention provides a RIS mode selection method based on the grey relational clustering method, RIS refers to a reconfigurable intelligent surface (RIS), and the RIS mode refers to the reflection mode of the reconfigurable intelligent surface, which specifically includes the following steps:

[0041] S1. The base station obtains the user's communication data and divides the user's communication data into two types of data: numerical data with continuous values ​​and type data with discrete values.

[0042] When transmitting data to a user, the base station can first obtain the following communication data: distance value, SNR value, angle value, service priority value, etc. Among them, the continuous value type data includes distance value, SNR value, angle value, etc.; the discrete value type data includes service priority value, etc.

[0043] S2. For the type data with discrete values, a grey possibility mapping matrix is ​​established (the category setting corresponds to the RIS model), discrete variables are clustered, and the discrete variables are directly mapped to the final model classification.

[0044] Shared numerical data, define the rth Continuous variable X r Its connotation category Grey possibility function for:

[0045]

[0046] in, is the gray value interval of the kth class, which is artificially defined. r , the number of categories Different, and different k corresponding to each of Different; y is the value of χ, y∈[a,b] is the mathematical representation of y in the interval a~b, y is not in this interval. The meaning of this function is that when χ r When the value of is different, the possibility of taking different categories is also different.

[0047] S3. For numerical data with continuous values, a grey possibility function is established (the category setting is not limited), and continuous variables are clustered. The continuous variables are first mapped to the connotation classification and then mapped to the final pattern classification.

[0048] For example, the continuous variable distance between the user and the base station is mapped to its corresponding connotation classification such as far, medium, and near, and then the weight is set to map it to the final mode classification, which corresponds to the RIS reflection mode such as 0°, 45°, 90°, etc.

[0049] Shared type data, define the dth Discrete variable χ d For Category The gray possibility mapping matrix is:

[0050]

[0051] The matrix size is N rows. Column, that is There are N discrete values ​​that can be assigned to in the categories; in the matrix, the nth (n=1,…N) row and the kth (k=1,2,…,K χ ) column elements mean that when χ d Pick When d The probability of belonging to the kth category. The values ​​of these elements are required to be less than 1 and can be given artificially based on experience. represents the discrete value variable χ d For Category If the statistical method is used, the calculation formula is:

[0052]

[0053] To unify the subsequent calculations, the total number of categories of all discrete value variables during algorithm execution Take the total number of modes K C .

[0054] Two-step cluster mapping of continuous variables This step helps map the connotation classification of continuous variables to the final target category (pattern category). The above gray possibility function only maps the variable to its connotation classification, not the final pattern classification. To get the final result, you need to do another mapping.

[0055] Define the rth Continuous variable χ r Connotation category The two-step possible degree mapping matrix is:

[0056]

[0057] The size of this matrix is ​​K A OK Column, K A is the final target category number, is the variable χ r The number of connotation categories; the ξ-th (ξ=1,2,…,K A ) The meaning of the column elements is that when the variable χ r When belonging to the kth connotation classification, the probability of belonging to the ξth final target category, the value of these elements is required to be less than 1, which can be given artificially based on experience, It means the rth Continuous variable χ r Connotation category The two-step possibility of . If the statistical method is used, the calculation formula is:

[0058]

[0059] During the algorithm execution, another K A The value is the total number of modes K C .

[0060] S4. Set weights and use the weights to make overall statistics on the prediction scores of each type of mode.

[0061] The specific process of setting weights includes: Continuous variables, whose weights are defined as For the dth Discrete value variable, define its weight as ω d , the initial calculation formula is as follows:

[0062]

[0063] Grey clustering coefficient calculation,This step obtains the prediction scores of various modes selected by the current user i data.

[0064] For the mth (m=1,…,K C ) class model, predicting the score Υ selected by the current user i m Calculated by the following formula:

[0065]

[0066] in is the variable χ d The value is of

[0067] S5. Select the optimal decision for the RIS model based on the ranking of the prediction scores.

[0068] Final decision execution: If the total number of users is n, the final decision score is calculated as follows:

[0069]

[0070] Then m The values ​​are sorted in descending order, and the corresponding index values ​​are This index sequence is the RIS mode ordering predicted by the algorithm to be most suitable for the communication needs at this time; The corresponding mode is the predicted optimal mode, that is, the "current most suitable RIS mode". The RIS mode will be automatically set to the optimal mode so that the reflection link of the base station can cover as many users as possible and users with higher demand priority.

[0071] Most existing models use FPGA to control RIS to select reflection mode, such as 90°, 0°, 45°, etc. Figure 2 As shown, the present invention selects the reflection mode of the reconfigurable intelligent surface (RIS) by the grey association clustering method. After the selection by the grey association clustering algorithm, the blue mode in the figure is selected as the optimal RIS mode in the current state, so that the reflection link can cover as many users as possible and users with higher demand priority. The selection method of the present invention is highly flexible and can adapt to the problems of rapidly changing environments and diversified user needs. While ensuring accurate capture of user data, it reduces the complexity of calculation and is more adaptable to real-time applications under large machine learning models.

[0072] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the scope of protection of the present invention.

Claims

1. The RIS mode selection method based on grey relational clustering method is characterized by: The following steps are involved: S1. The base station obtains the user's communication data and divides the user's communication data into two types of data: continuous value data and discrete value type data; S2. For the discrete value type data, a grey possibility mapping matrix is ​​established to cluster the discrete variables and directly map the discrete variables to the final pattern classification; S3. For numerical data with continuous values, a grey possibility function is established to cluster continuous variables, and the continuous variables are first mapped to connotation classification and then mapped to the final pattern classification; S4. Set weights and make overall statistics on the prediction scores of various modes by weights; S5. Select the optimal decision for the RIS model based on the ranking of the prediction scores.

2. The RIS mode selection method based on the grey relational clustering method according to claim 1 is characterized in that: In step S1, the continuously-valued numerical data includes distance values, SNR values, and angle values; The discrete value type data includes the service priority value.

3. The RIS mode selection method based on the grey relational clustering method according to claim 1 is characterized in that: The step S2 specifically includes: Shared numerical data, define the rth Continuous variable χ r Its connotation category Grey possibility function for: in, is the gray value interval of the kth class, for different variables χ r , the number of categories Different, and different k corresponding to each of Different; y is the value of χ, y∈[a,b] is the mathematical representation of y in the interval a~b, y is not in this interval; when χ r When the value of is different, χ r The probability of taking different categories is also different.

4. The RIS mode selection method based on the grey relational clustering method according to claim 1 is characterized in that: In step S3, a grey possibility function is established for the numerical data of continuous values, and the continuous variables are clustered, and the continuous variables are first mapped to the connotation classification, which specifically includes: Shared type data, define the dth Discrete variable χ d For Category The gray possibility mapping matrix is: The matrix size is N rows. Column, that is There are N discrete values ​​that can be assigned to in the categories; in the matrix, the nth (n=1,…N) row and the kth (k=1,2,…,K χ ) column elements mean that when χ d Pick When d The probability of belonging to the kth class, the value of the element is less than 1, represents the discrete value variable χ d For Category The gray possibility.

5. The RIS mode selection method based on the grey relational clustering method according to claim 4 is characterized in that: In step S3, the continuous variable is mapped to the connotation classification and then mapped to the final pattern classification, which specifically includes: Define the rth Continuous variable χ r Connotation category The two-step possible degree mapping matrix is: The size of this matrix is ​​K A OK Column, K A is the final target category number, is the variable χ r The number of connotation categories; the ξ-th (ξ=1,2,…,K A ) The meaning of the column elements is that when the variable χ r When it belongs to the kth connotation classification, the probability of belonging to the ξth final target category, the value of the element is less than 1, It means the rth Continuous variable χ r Connotation category The second step possibility.

6. The RIS mode selection method based on the grey relational clustering method according to claim 1 is characterized in that: In step S4, the specific process of setting the weight includes: Shared Numerical data, Type data, for the rth Continuous variables, whose weights are defined as For the dth Discrete value variable, define its weight as ω d , the initial calculation formula is as follows:

7. The RIS mode selection method based on the grey relational clustering method according to claim 6 is characterized in that: In step S4, the prediction scores of various modes are generally counted by weight, specifically including: For the mth (m=1,…,K C ) class model, predicting the score Υ selected by the current user i m Calculated by the following formula: in is the variable χ d The value is of 8. The RIS mode selection method based on the grey relational clustering method according to claim 7 is characterized in that: The step S5 specifically includes: If the total number of users is n, the final decision score is calculated as follows: Then m The values ​​are sorted in descending order, and the corresponding index values ​​are This index sequence is the ordering of the RIS modes predicted to be most suitable for the communication needs at this time; The corresponding mode is the predicted optimal mode, and the RIS mode will be automatically set to this optimal mode.

Images