A 5G base station service energy consumption prediction method considering temperature and humidity factors

Abstract

The present invention discloses a 5G base station business energy consumption prediction method considering temperature and humidity factors. On the basis of obtaining the temperature and humidity business energy consumption information under the operating state of the 5G base station, by mapping the 5G base station temperature and humidity business energy consumption three-dimensional spatial points and fitting the 5G base station temperature and humidity business energy consumption two-dimensional feature space prediction reference straight line, a 5G base station business energy consumption prediction value plane is constructed according to the criterion of minimum cumulative average prediction error, so as to realize the prediction of the business energy consumption value of the 5G base station with a given temperature and humidity, reduce the computational complexity of the 5G base station business energy consumption prediction method and improve the prediction accuracy, and realize the lightweight computing effect of the 5G base station business energy consumption prediction.

Description

Technical Field

[0001] The present invention relates to the field of 5G base station energy consumption calculation technology, and in particular to a 5G base station service energy consumption prediction method considering temperature and humidity factors. Background Art

[0002] The fifth generation mobile communication technology (5G) is a new generation of broadband mobile communication technology with the characteristics of high speed, low latency and large connection. It is the network infrastructure for realizing the interconnection of people, machines and things. In the operation of 5G network, the base station equipment room in the entire 5G communication network is the main power consumption part of the entire network, accounting for about 60% of the energy consumption of the entire network, of which 5G base station equipment consumes about 50% of the power, and the radio frequency equipment energy consumption accounts for more than 90% of the energy consumption of the entire base station; the research on energy-saving technology of 5G base station is particularly important. The research on energy consumption prediction of 5G base station business is the basis of the research on energy-saving technology of 5G base station, but on the other hand, environmental dynamic factors (such as temperature and humidity, etc.) will also make the energy consumption of 5G base station business show the characteristics of dynamic fluctuation, which further increases the computational complexity of energy consumption prediction of 5G base station business.

[0003] Data show that the existing technical routes for predicting the energy consumption of communication base station services that take dynamic factors into consideration mostly use multiple regression methods. The multiple regression method can obtain more appropriate prediction results for the energy consumption prediction of 5G base station services that take temperature and humidity factors into consideration. In addition to causing high computational complexity problems, the multiple regression method considers temperature and humidity as independent factors, and does not consider the joint effect of temperature and humidity on the communication business volume of 5G base stations, which affects the prediction accuracy.

[0004] Chinese patent document CN112566226A discloses a "5G base station intelligent energy-saving method". It includes: S1, distinguishing the specific characteristics of the wireless base station and determining the initial energy-saving configuration; S2, predicting the energy-saving parameter threshold through a second-order smoothing prediction algorithm; S3, an energy-saving parameter adjustment mechanism based on real-time KPI monitoring. The advantage of the present invention is that it effectively overcomes the problems of rigid application mode, poor flexibility, slow response time, poor energy-saving effect, and inability to effectively combine with user perception and KPI of traditional energy-saving means. The system interface is compiled in C++ language to process the massive cell-level historical performance data of the existing network, and the data is screened and classified according to the energy-saving effect. The cells suitable for the energy-saving strategy are selected, and the traffic development trend of the applicable cells is predicted through a second-order smoothing prediction algorithm to obtain the energy-saving time window of the sub-cell. Then, the energy-saving parameters are dynamically adjusted according to the changes in the cell load through an energy-saving parameter adjustment mechanism based on real-time KPI monitoring, so as to achieve the best balance between the cell energy-saving effect and user perception. The above technical scheme has the technical limitation of not considering the joint effect of temperature and humidity factors on the prediction of base station energy consumption. Summary of the invention

[0005] The present invention mainly solves the technical problem that the original technical solution does not consider the technical limitations of the joint effects of temperature and humidity factors on base station energy consumption prediction, and provides a 5G base station business energy consumption prediction method considering temperature and humidity factors. On the basis of obtaining the temperature and humidity business energy consumption information under the operating state of the 5G base station, by mapping the 5G base station temperature and humidity business energy consumption three-dimensional spatial points and fitting the 5G base station temperature and humidity business energy consumption two-dimensional feature space prediction reference line, a 5G base station business energy consumption prediction value plane is constructed according to the minimum cumulative average prediction error criterion, and the business energy consumption value of the 5G base station is predicted for a given temperature and humidity, which reduces the computational complexity of the 5G base station business energy consumption prediction method and improves the prediction accuracy, thereby achieving a lightweight computing effect of 5G base station business energy consumption prediction.

[0006] The above technical problem of the present invention is mainly solved by the following technical solution: The present invention comprises the following steps:

[0007] S1 obtains the temperature and humidity service energy consumption information of the 5G base station in operation state;

[0008] S2 maps the three-dimensional spatial points of 5G base station temperature and humidity service energy consumption;

[0009] S3 fits the two-dimensional feature space prediction benchmark line of the service energy consumption of 5G base station temperature and humidity;

[0010] S4 constructs the 5G base station service energy consumption prediction value plane according to the minimum cumulative average prediction error criterion;

[0011] S5 predicts the service energy consumption value of 5G base station given temperature and humidity.

[0012] Preferably, in step S1, for a given 5G base station with service energy consumption to be predicted, the statistical period is set to D, the D value is 60 days to 360 days, the collection period is T, the T value is 1 minute to 5 minutes, and the total number of collected data in the statistical period is calculated as N according to the formula N=D*24*60 / T; for No.i (i∈N) collection periods, the temperature average value x, humidity average value y and service energy consumption average value p in the current collection period are obtained to form a 5G base station service energy consumption parameter triplet, expressed as (x i ,y i ,p i The unit of the average temperature x is degree, the unit of the average humidity y is %, and the unit of the average service power consumption p is W.

[0013] As a preferred embodiment, in step S2, temperature, humidity and 5G base station service energy consumption are used as x-axis, y-axis and z-axis to form a three-dimensional space of service energy consumption, and each No.i triple (x i ,yi ,p i ) is mapped into a spatial point in the three-dimensional space of business energy consumption. Subsequently, the spatial points in the three-dimensional space of business energy consumption are applied to the projection to achieve a prediction benchmark straight line for the two-dimensional feature space of business energy consumption fitting the temperature and humidity of the 5G base station.

[0014] Preferably, the step S3 specifically includes:

[0015] S3.1 For each No.i triple (x i ,y i ,p i ) is projected vertically onto the XoZ two-dimensional plane where the x-axis and the z-axis intersect, and each No.i triple (x i ,y i ,p i ) is mapped into a point in the XoZ two-dimensional plane, marked as a temperature business energy consumption tuple (x i ,p i );

[0016] S3.2 If multiple temperature service energy consumption pairs (x i ,p i ) has multiple p values ​​for the same x. i ,p i ) is used to perform a unique operation, using a temperature business energy consumption tuple Replace the above multiple temperature business energy consumption tuples (x i ,p i ),in is the p in the above multiple temperature energy consumption binary groups i The mathematical expectation of XoZ is such that every identical x i The value corresponds to only one energy consumption value p i , after the unique marking operation, the temperature service energy consumption tuple is

[0017] S3.3 Temperature business energy consumption binary The corresponding points in the XoZ two-dimensional plane are used as the data source, and the least squares method is used to calculate the benchmark characteristic straight line p=k for 5G base station temperature service energy consumption prediction. x x+b x , where k x and b x are the slope and intercept of the above straight line equation respectively;

[0018] S3.4 Repeat step S3.3 to calculate and obtain the 5G base station humidity service energy consumption prediction benchmark characteristic straight line p = k y y+b y , where k yand b y are the slope and intercept of the above straight line equation respectively.

[0019] Preferably, the step S4 specifically includes:

[0020] S4.1 The two characteristic lines p = k x x+b x and p = k y y+b y Move ΔZ up and down along the Z axis in the XoZ two-dimensional plane and the YoZ two-dimensional plane x and ΔZ y To achieve intersection, the two intersecting straight lines are used as the two sides of the parallelogram, and the 5G base station service energy consumption prediction value plane s is formed by constructing a parallelogram plane in three-dimensional space; where ΔZ x and ΔZ y The value is determined in accordance with S4.2.

[0021] S4.2 Mark each No.i triple (x i ,y i ,p i ) The vertical distance from the corresponding three-dimensional space point to the energy consumption prediction value plane s is d i , according to the formula Calculate the cumulative average prediction error w; take the cumulative average prediction error w as the minimum optimization goal, use the particle swarm optimization method to obtain the parameter ΔZ x and ΔZ y The best value.

[0022] Preferably, after the temperature and humidity of the 5G base station whose business energy consumption is to be predicted are given, step S5 calculates the business energy consumption value of the 5G base station business energy consumption prediction value plane s corresponding to the temperature and humidity, that is, the Z-axis value corresponding to the three-dimensional plane, as the business energy consumption prediction value of the 5G base station.

[0023] The beneficial effects of the present invention are as follows: by mapping the three-dimensional spatial points of the temperature and humidity business energy consumption of the 5G base station and fitting the two-dimensional feature space prediction reference straight line of the business energy consumption of the temperature and humidity of the 5G base station, a 5G base station business energy consumption prediction value plane is constructed according to the criterion of minimum cumulative average prediction error, so as to predict the business energy consumption value of the 5G base station with a given temperature and humidity, reduce the computational complexity of the 5G base station business energy consumption prediction method and improve the prediction accuracy, and achieve a lightweight computing effect of the 5G base station business energy consumption prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0025] The technical solution of the present invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0026] Embodiment: A 5G base station service energy consumption prediction method considering temperature and humidity factors in this embodiment is as follows: Figure 1 As shown, the following steps are included:

[0027] S1 obtains the temperature and humidity service energy consumption information of the 5G base station under operating status.

[0028] For a given 5G base station with service energy consumption to be predicted, let the statistical period be D (D takes a value of 60 days to 360 days), the collection period be T (take a value of 1 minute to 5 minutes), and the total number of collected data in the statistical period be calculated as N according to the formula N=D*24*60 / T; for No.i (i∈N) collection periods, obtain the average temperature x (in degrees), the average humidity y (in %) and the average service energy consumption p (in W) in the current collection period, and form a 5G base station service energy consumption parameter triple (hereinafter referred to as triple) expressed as (x i ,y i ,p i ).

[0029] S2 maps the three-dimensional spatial points of 5G base station temperature, humidity and service energy consumption.

[0030] The temperature, humidity and 5G base station service energy consumption are respectively used as the x-axis, y-axis and z-axis to form a three-dimensional space of service energy consumption. i ,y i ,p i ) is mapped into a spatial point in the three-dimensional space of business energy consumption. Subsequently, the spatial points in the three-dimensional space of business energy consumption are applied to the projection to achieve a prediction benchmark straight line for the two-dimensional feature space of business energy consumption fitting the temperature and humidity of the 5G base station.

[0031] S3 fits the service energy consumption two-dimensional feature space prediction benchmark line of 5G base station temperature and humidity, including:

[0032] S3.1 For each No.i triple (x i ,y i ,p i ) is projected vertically onto the XoZ two-dimensional plane where the x-axis and the z-axis intersect, and each No.i triple (x i ,y i ,p i ) is mapped into a point in the XoZ two-dimensional plane, marked as a temperature business energy consumption tuple (x i ,p i );

[0033] S3.2 If multiple temperature service energy consumption pairs (xi ,p i ) has multiple p values ​​for the same x. i ,p i ) is used to perform a unique operation, using a temperature business energy consumption tuple Replace the above multiple temperature business energy consumption tuples (x i ,p i ),in is the p in the above multiple temperature energy consumption binary groups i The mathematical expectation of XoZ is such that every identical x i The value corresponds to only one energy consumption value p i , after the unique marking operation, the temperature service energy consumption tuple is

[0034] S3.3 Temperature business energy consumption binary The corresponding points in the XoZ two-dimensional plane are used as the data source, and the least squares method is used to calculate the benchmark characteristic straight line p=k for 5G base station temperature service energy consumption prediction. x x+b x , where k x and b x are the slope and intercept of the above straight line equation respectively;

[0035] S3.4 Repeat step S3.3 to calculate and obtain the 5G base station humidity service energy consumption prediction benchmark characteristic straight line p = k y y+b y , where k y and b y are the slope and intercept of the above straight line equation respectively.

[0036] S4 constructs the 5G base station service energy consumption prediction value plane according to the minimum cumulative average prediction error criterion, specifically including:

[0037] S4.1 The two characteristic lines p = k x x+b x and p = k y y+b y Move ΔZ up and down along the Z axis in the XoZ two-dimensional plane and the YoZ two-dimensional plane x and ΔZ y The intersection is achieved, and the two intersecting straight lines are used as the two sides of the parallelogram. The 5G base station service energy consumption prediction value plane s is formed by constructing a parallelogram plane in three-dimensional space;

[0038] S4.2 Mark each No.i triple (x i ,y i ,p i) The vertical distance from the corresponding three-dimensional space point to the energy consumption prediction value plane s is d i , according to the formula Calculate the cumulative average prediction error w; take the cumulative average prediction error w as the minimum optimization goal, use the particle swarm optimization method to obtain the parameter ΔZ x and ΔZ y The best value.

[0039] S5 predicts the service energy consumption value of the 5G base station given the temperature and humidity. After the temperature and humidity of the 5G base station to be predicted are given, the service energy consumption value of the 5G base station service energy consumption prediction value plane s corresponding to the temperature and humidity is calculated, that is, the Z-axis value corresponding to the three-dimensional plane, as the service energy consumption prediction value of the 5G base station.

[0040] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

[0041] Although this article uses more terms such as temperature and humidity service energy consumption and reference straight line, it does not exclude the possibility of using other terms. These terms are used only to more conveniently describe and explain the essence of the present invention; interpreting them as any additional restrictions is contrary to the spirit of the present invention.

Claims

1. A 5G base station service energy consumption prediction method considering temperature and humidity factors, characterized in that: The following steps are involved: S1 obtains the temperature and humidity service energy consumption information of the 5G base station in operation state; S2 maps the three-dimensional spatial points of 5G base station temperature and humidity service energy consumption; S3 fits the two-dimensional feature space prediction benchmark line of the service energy consumption of 5G base station temperature and humidity, and converts each No.i triple (x i ,y i ,p i ) is projected vertically onto the XoZ two-dimensional plane to form the temperature business energy consumption tuple (x i ,p i ), as the data source, obtain the temperature prediction characteristic straight line p = k x x+b x And humidity prediction characteristic straight line p = k y y+b y ; S4 constructs the 5G base station service energy consumption prediction value plane according to the minimum cumulative average prediction error criterion; S5 predicts the service energy consumption value of the 5G base station given the temperature and humidity, and moves the two characteristic straight lines to intersect and construct a prediction value plane; Calculate the prediction error of each No.i triplet to the energy consumption prediction value plane, minimize the average prediction error, obtain the optimal value of the characteristic straight line movement, and the Z-axis value corresponding to the three-dimensional plane is the service energy consumption prediction value of the 5G base station.

2. According to a method for predicting 5G base station service energy consumption considering temperature and humidity factors according to claim 1, it is characterized in that: In step S1, for a given 5G base station with service energy consumption to be predicted, the statistical period is set to D, where D is 60 days to 360 days, the collection period is T, where T is 1 minute to 5 minutes, and the total number of collected data in the statistical period is calculated as N according to the formula N=D*24*60 / T; for No.i (i∈N) collection periods, the average temperature x, the average humidity y and the average service energy consumption p in the current collection period are obtained to form a 5G base station service energy consumption parameter triplet, expressed as (x i ,y i ,p i ).

3. According to a method for predicting 5G base station service energy consumption considering temperature and humidity factors according to claim 2, it is characterized in that: In step S2, temperature, humidity and 5G base station service energy consumption are used as x-axis, y-axis and z-axis to form a three-dimensional space of service energy consumption. i ,y i ,p i ) is mapped into a spatial point in the three-dimensional space of business energy consumption.

4. A 5G base station service energy consumption prediction method considering temperature and humidity factors according to claim 2 or 3, characterized in that: The step S3 specifically includes: S3.1 For each No.i triple (x i ,y i ,p i ) is projected vertically onto the XoZ two-dimensional plane where the x-axis and the z-axis intersect, and each No.i triple (x i ,y i ,p i ) is mapped into a point in the XoZ two-dimensional plane, marked as a temperature business energy consumption tuple (x i ,p i ); S3.2 If multiple temperature service energy consumption pairs (x i ,p i ) has multiple p values ​​for the same x. For the above binary pair (x i ,p i ) is used to perform a unique operation, using a temperature business energy consumption tuple Replace the above multiple temperature business energy consumption tuples (x i ,p i ),in is the p in the above multiple temperature service energy consumption tuples i The mathematical expectation of XoZ is such that every identical x i The value corresponds to only one energy consumption value p i , after the unique marking operation, the temperature service energy consumption tuple is S3.3 Temperature business energy consumption binary The corresponding points in the XoZ two-dimensional plane are used as the data source, and the least squares method is used to calculate the benchmark characteristic straight line p=k for 5G base station temperature service energy consumption prediction. x x+b x , where k x and b x are the slope and intercept of the above straight line equation respectively; S3.4 Repeat step S3.3 to calculate and obtain the 5G base station humidity service energy consumption prediction benchmark characteristic straight line p = k y y+b y , where k y and b y are the slope and intercept of the above straight line equation respectively.

5. According to a method for predicting 5G base station service energy consumption considering temperature and humidity factors according to claim 1, it is characterized in that: The step S4 specifically includes: S4.1 The two characteristic lines p = k x x+b x and p = k y y+b y Move ΔZ up and down along the Z axis in the XoZ two-dimensional plane and the YoZ two-dimensional plane x and ΔZ y The intersection is achieved, and the two intersecting straight lines are used as the two sides of the parallelogram. The 5G base station service energy consumption prediction value plane s is formed by constructing a parallelogram plane in three-dimensional space; S4.2 Mark each No.i triple (x i ,y i ,p i ) The vertical distance from the corresponding three-dimensional space point to the energy consumption prediction value plane s is d i , according to the formula Calculate the cumulative average prediction error w; take the cumulative average prediction error w as the minimum optimization goal, use the particle swarm optimization method to obtain the parameter ΔZ x and ΔZ y The best value.

6. A 5G base station service energy consumption prediction method considering temperature and humidity factors according to claim 1, characterized in that: After the temperature and humidity of the 5G base station whose business energy consumption is to be predicted are given, step S5 calculates the business energy consumption value of the 5G base station business energy consumption prediction value plane s corresponding to the temperature and humidity, that is, the Z-axis value corresponding to the three-dimensional plane, as the business energy consumption prediction value of the 5G base station.

Images