5G network antenna weight optimization method and device and computer program product

Abstract

The invention discloses a 5G network antenna weight optimization method and device and a computer program product, and relates to the technical field of 5G networks. The 5G network antenna weight optimization method comprises the following steps: acquiring terminal data through measurement report data; decoupling a network into a plurality of sub-network slices based on the terminal data, wherein the sub-network slices comprise a plurality of cells; establishing a single-cell beam weight library according to different coverage scenes; constructing a multi-cell beam weight library based on the single-cell beam weight library; and performing weight searching on the multi-cell beam weight library by using a fusion algorithm to obtain an optimal weight group, the fusion algorithm being an algorithm combining a population algorithm and a genetic algorithm.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of 5G network, and particularly relates to a 5G network antenna weight optimization method and device and a computer program product. BACKGROUND

[0002] The fifth generation mobile communication technology (5G) is the latest generation of cellular mobile communication technology, which has the advantages of high data, high speed, low delay, low consumption, etc. It will become a new generation of general-purpose technology after steam engine, power technology and Internet technology. Accelerating the construction of new basic settings such as 5G network and data center, and actively enriching the application scenarios of 5G technology are the top priority of 5G construction.

[0003] As a key technology of 5G, large-scale antenna technology can effectively improve the network deep coverage performance. It uses large-scale array antenna to enhance the spatial dimension analysis ability and use efficiency.

[0004] However, in the related art, the commonly used method for 5G beam weight optimization has poor global search ability and is easy to fall into a local optimal solution. SUMMARY

[0005] Embodiments of the present application provide a 5G network antenna weight optimization method, device and computer program product to at least solve the problem that the 5G beam weight optimization in the related art is easy to fall into a local optimal solution.

[0006] In a first aspect, embodiments of the present application provide a 5G network antenna weight optimization method, comprising: obtaining terminal data through measurement report data; decoupling a network into a plurality of sub-network slices based on the terminal data, wherein the sub-network slices include a plurality of cells; establishing a single-cell beam weight library according to different coverage scenarios; constructing a multi-cell beam weight library based on the single-cell beam weight library; searching for weights in the multi-cell beam weight library by using a fusion algorithm to obtain an optimal weight group, wherein the fusion algorithm is an algorithm combining a population algorithm and a genetic algorithm.

[0007] In a second aspect, embodiments of the present application provide a 5G network antenna weight optimization device, comprising: an obtaining module configured to obtain terminal data through measurement report data; a decoupling module configured to decouple a network into a plurality of sub-network slices based on the terminal data, wherein the sub-network slices include a plurality of cells; The establishing module is configured to establish a single-cell beam weight value library according to different coverage scenarios; The constructing module is configured to construct a multi-cell beam weight value library based on the single-cell beam weight value library; The optimizing module is configured to search for optimal weights from the multi-cell beam weight value library by using a fusion algorithm, the fusion algorithm being an algorithm combining a population algorithm and a genetic algorithm.

[0008] In a third aspect, an electronic device is provided, which includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, and the program or instructions, when executed by the processor, implement the steps of the method according to the first aspect.

[0009] In a fourth aspect, a computer program product is provided, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions executable by a computer to implement the steps of the method according to the first aspect.

[0010] In the embodiments of the present application, first, terminal data is acquired through measurement report data, second, a network is decoupled into multiple sub-network slices based on the terminal data, the sub-network slices including multiple cells, then, a single-cell beam weight value library is established according to different coverage scenarios, a multi-cell beam weight value library is constructed based on the single-cell beam weight value library, finally, a fusion algorithm is used to search for optimal weights from the multi-cell beam weight value library, the fusion algorithm being an algorithm combining a population algorithm and a genetic algorithm. In the embodiments of the present application, a multi-cell beam weight value library is constructed from a single-cell beam weight value library, and a population algorithm and a genetic algorithm are combined to search for weights from the multi-cell beam weight value library, the population algorithm generates a well-distributed initial population, and the genetic algorithm is used to solve the problem to reduce the possibility of the algorithm falling into a local optimum, so that the global optimal solution of the problem is more easily searched, and the algorithm converges faster and the efficiency is improved. BRIEF DESCRIPTION OF DRAWINGS

[0011] The accompanying drawings, which are included to provide a further understanding of the present application, form a part of the present application and illustrate the illustrative embodiments of the present application and together with the description serve to explain the present application. In the drawings: Figure 1 is a flowchart of the 5G network antenna weight optimization method provided by the embodiments of the present application; Figure 2 is a flowchart of relationship formula management provided by the embodiments of the present application; Figure 3 is a schematic diagram of a 5G network antenna weight optimization device provided by the embodiments of the present application; Figure 4This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0013] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0014] The following is in conjunction with the appendix Figures 1 to 4 This application provides a detailed description of a 5G network antenna weight optimization method, apparatus, and computer program product through specific embodiments and application scenarios.

[0015] like Figure 1 The diagram shown is a flowchart of a 5G network antenna weight optimization method provided in an embodiment of this application. Figure 1 As shown, the 5G network antenna weight optimization method may include the contents shown in S101 to S105.

[0016] In S101, terminal data is obtained through measurement report data.

[0017] The Measurement Report (MR) data includes the direction of arrival (DOA) information for users in one serving cell and multiple neighboring cells.

[0018] In S102, the network is decoupled into multiple subnet slices based on terminal data, and each subnet slice includes multiple cells.

[0019] In this embodiment, the network is decoupled into multiple independent subnet slices, each of which is regarded as an independent network, and the beam adjustment does not affect each other.

[0020] In S103, a single-cell beam weight library is established according to different coverage scenarios.

[0021] The beam weight library is a set of key parameters used in 5G networks to optimize base station coverage and signal quality. It includes azimuth, downtilt, horizontal beam width (BeamWidthH), and vertical beam width (BeamWidthV), which are used to control the antenna waveform and achieve beamforming.

[0022] In S104, a multi-cell beam weight library is constructed based on the single-cell beam weight library.

[0023] In S105, a fusion algorithm is used to search for the optimal weight set in the multi-cell beam weight library. The fusion algorithm is a combination of population algorithm and genetic algorithm.

[0024] This embodiment uses a combination of population algorithm and genetic algorithm to search for the optimal weight set for multiple cells, thereby improving the network coverage of multiple cells.

[0025] In this embodiment, terminal data is first acquired through measurement report data. Then, based on the terminal data, the network is decoupled into multiple sub-network slices, each containing multiple cells. Next, a single-cell beam weight library is established according to different coverage scenarios. Based on this single-cell beam weight library, a multi-cell beam weight library is constructed. Finally, a fusion algorithm is used to search for the optimal weights in the multi-cell beam weight library. The fusion algorithm combines population dynamics and genetic algorithms. This embodiment constructs a multi-cell beam weight library from the single-cell beam weight library and then uses a combination of population dynamics and genetic algorithms to search for the optimal weights. The population dynamics algorithm generates a well-distributed initial population, while the genetic algorithm solves the problem, reducing the possibility of the algorithm getting trapped in local optima and making it easier to find the global optimum. This also accelerates algorithm convergence and improves efficiency.

[0026] In one possible implementation of this application, after establishing a single-cell beam weight library according to different coverage scenarios, the 5G network antenna weight optimization method may further include: using a particle swarm optimization algorithm to search the single-cell beam weight library to obtain the optimal weight group for a single cell.

[0027] In this embodiment, an improved particle swarm optimization algorithm is used to search for the optimal weight set for a single cell, thereby improving its 5G network coverage.

[0028] In one possible implementation of this application, the particle swarm optimization algorithm is used to search the single-cell beam weight library to obtain the optimal weight set for the single cell. This may include: initializing the number of particles in the swarm, randomizing the position and movement speed of the beam weight set; evaluating the fitness value of each particle according to the fitness function; updating the movement speed of each particle using a chaotic operator as a random number; updating the position of the beam weight set based on the current position of the particle's weight set and the historical optimal position of the weight set; and ending the update when a termination condition is met, thereby obtaining the optimal weight set for the single cell. The termination condition includes at least one of the following: reaching the maximum number of iterations, or the increment of the fitness value being less than the increment threshold.

[0029] Because the distribution of users in a single cell exhibits regular changes within a week, cell coverage cannot be achieved using a single antenna weight. Instead, an improved particle swarm optimization algorithm is needed to optimize antenna weights. After optimization, the antenna weights are used to shape the coverage, enabling differentiated coverage during the day and night to better match cell coverage with service requirements.

[0030] In one example, the improved particle swarm optimization algorithm process is as follows: Let the position of the i-th particle be denoted as The position of each particle can be viewed as a set of feasible solutions to an optimization problem, and each particle calculates its own fitness value f using a fitness function. This is used to determine the quality of the position, and the velocity of the i-th particle is expressed as... The velocity parameter determines the distance and direction of particle flight, and can be adjusted based on the particle's own and its companions' flight experience; The best position that the i-th particle experienced during the iterative optimization process is: = ( ), i=1,2,…,N The best position the particle swarm has ever reached is: = ( ), i=1,2,…,N In the iterative optimization process of the particle swarm, the velocity update formula for the i-th particle is: (k+1)=w (k)+ ( - (k))+ ( - (k)) The formula for updating the position of the i-th particle is: = + (k+1) Where k is the number of iterations; (k) represents the velocity parameter of the i-th particle in the k-th iteration. , For the minimum speed, This represents the maximum speed. Let be the position parameter of the i-th particle in the k-th iteration; , As an acceleration factor; , Let w be a random number on (0,1), and w be the inertia weight. The number of particles can be 150.

[0031] The weights are optimized by simultaneously and linearly decreasing both the inertia weight and the acceleration factor. The linear decrease in the inertia weight refers to its maximum value at the beginning of the iteration. It decreases linearly to the minimum value through continuous iteration. Its adjustment formula is:

[0032] In the formula, k is the number of iterations, which can take the value 200; This represents the maximum number of iterations. The maximum inertia weight can be 0.9. This is the minimum inertia weight, and its value can be 0.4. The linear decrease in the acceleration factor refers to its maximum value at the beginning of the iteration. It decreases linearly to the minimum value through continuous iteration. Its adjustment formula is:

[0033] In the formula, This is the maximum acceleration factor, and its value can be 2.5; This is the minimum acceleration factor, which can take a value of 0.5.

[0034] Introducing the chaotic operator L to replace random numbers and This allows the particle swarm to retain considerable diversity in the later stages of the search. The chaotic operator L is generated by mapping in the (0,1) interval, and its calculation formula is as follows:

[0035] In the formula, L (0,1);k=0,1,2,…,N; These are system parameters.

[0036] when At that time, the system is in a chaotic state; The logical mapping at this time is called a full mapping. At this time, the system has ergodicity, and its probability density function is independent of the initial value. The chaotic operator L is:

[0037] In the iterative optimization process of the particle swarm, the velocity update formula for the i-th particle is:

[0038] The selection of the fitness function directly affects the optimization result. Using the ITAE (time multiplied by the integral of absolute error) performance index as the fitness function, the control system constructs the following fitness evaluation function: dt+

[0039] The fitness function consists of four parts: the first part is the ITAE performance index; the second part is the difference between the output at the current time and the previous time; the third part is the square of the controller output; and the fourth part is the rise time of the response. The weights are selected as follows: = Therefore, the fitness function is: = dt+

[0040] In one example, updating the beam weight set position based on the particle's current weight set position and its historical best weight set position can include: for each particle, comparing the fitness value corresponding to the particle's current weight set position with the fitness value corresponding to the particle's historical best weight set position; if the fitness value corresponding to the current weight set position is greater than the fitness value corresponding to the historical best weight set position, updating the current weight set position to the particle's optimal weight set position; for each particle, comparing the particle's current fitness value with the fitness value corresponding to the population's optimal weight set position; if the particle's current fitness value is greater than the fitness value corresponding to the population's optimal weight set position, updating the particle's current fitness value to the population's optimal weight set position.

[0041] In this embodiment, the particle swarm optimization algorithm for optimizing beam weight groups in a single cell includes the following steps: Step 1: Particle swarm initialization, initializing the number of particles in the swarm, and randomizing the beam weight group position and movement speed; Step 2: Fitness evaluation, evaluating the fitness value of each particle according to the fitness function; Step 3: Determining the position of the individual optimal weight group, for each particle, comparing the fitness value corresponding to its current weight group position with the fitness value corresponding to its historical optimal weight group position. If the current fitness value is higher, then the current weight group position is updated to the individual optimal weight group position; Step 4: Determining the position of the optimal weight group, for each particle, comparing its current fitness value with the fitness value corresponding to the optimal weight group position. If the current fitness value is higher, then the current particle's weight group position is updated to the optimal weight group position; Step 5: Updating particle velocity, updating the movement speed and beam weight group position of each particle; Step 6: Iteration, if the termination condition is not met, returning to step 2. Usually, the algorithm terminates when it reaches the maximum number of iterations or when the increment of the optimal fitness value is less than a given threshold.

[0042] In this embodiment, the single-cell weight optimization adopts the particle swarm optimization algorithm, and a chaotic operator is introduced into the particle swarm optimization algorithm to improve the optimization speed of the particle swarm optimization algorithm, increase the particle diversity of the particle swarm, and improve the stability, anti-interference and adaptability of the system.

[0043] In the multi-cell antenna weight search process, a fusion algorithm is used for optimization. This fusion algorithm effectively combines the genetic algorithm and the artificial bee colony algorithm. The artificial bee colony algorithm is introduced into the genetic algorithm. First, the artificial bee colony algorithm is used to generate an initial population with a relatively good distribution. Then, the genetic algorithm is used to solve the problem. After completing one round of search, the artificial bee colony algorithm is used to update the population again, and then the genetic algorithm is run again. This process is repeated until the optimal solution is found. During this process, because the initial population has a good distribution, the genetic algorithm is less stressed during crossover and mutation. The artificial bee colony algorithm has strong local search capabilities and can maintain the diversity of the population, which can reduce the possibility of the algorithm getting stuck in local optima and make it easier for the algorithm to find the global optimal solution of the problem. At the same time, it can also accelerate the convergence speed of the algorithm.

[0044] In one possible implementation of this application, a fusion algorithm is used to search for the optimal weight set in a multi-cell beam weight library. This can include: each cell obtaining the optimal weight and candidate weights for a scene through scene identification; determining the population size based on the optimal and candidate weights of multiple cells; sorting the populations within the population size using an artificial bee colony algorithm based on the population size, and determining the crossover probability and mutation probability; and processing the sorted populations using a genetic algorithm based on the crossover probability and mutation probability until a preset condition is met to obtain the optimal weight set.

[0045] In one example, based on population size, the artificial bee colony algorithm is used to rank the population within the population size and determine the crossover and mutation probabilities. This can include: determining the number of leader bees, crossover probability, and mutation probability based on the population size, where the number of leader bees equals the population size, and the population serves as a food source; placing leader bees at the current food source and calculating the fitness of each food source; calculating the probability that a food source will be selected by follower bees based on the fitness, and determining the location of the food source selected by the follower bees; performing a neighborhood search on the leader bees and follower bees to determine new food sources; updating the food source as the new food source if the fitness of the new food source is greater than the fitness of the food source selected by the leader bees and follower bees; converting leader bees into scout bees and generating new food sources if the number of times a food source has not been updated exceeds a coefficient threshold; determining the reward of all food sources and ranking the food sources based on the reward.

[0046] In one example, a genetic algorithm is used to process the sorted population based on crossover and mutation probabilities until a preset condition is met to obtain the optimal weight set. This process may include: using the crossover operator of the genetic algorithm, performing a crossover operation on the population based on the crossover probability, and calculating the payoff of the newly generated population; if the payoff of the newly generated population is greater than the payoff of the population that underwent the crossover operation, then the payoff of the newly generated population is taken as the population's payoff, and the crossover operation is performed on the newly generated population; using the mutation operator of the genetic algorithm, performing a mutation operation on the population based on the mutation probability, and calculating the payoff of the newly generated population; if the payoff of the newly generated population is greater than the payoff of the population that underwent the mutation operation, then the payoff of the newly generated population is taken as the population's payoff, and the mutation operation is performed on the newly generated population; if the crossover and mutation operations meet the preset conditions, the optimal weight set is obtained, and the preset conditions include reaching the maximum number of operations.

[0047] The artificial bee colony algorithm is used to find food sources. After one round of food source search, a genetic algorithm is introduced to perform selection, crossover, and mutation operations to find food sources. Then, the artificial bee colony algorithm is used again in the next round. By combining the artificial bee colony algorithm and the genetic algorithm, the new algorithm not only has the strong global search capability of the genetic algorithm, but also improves the optimization performance of the original algorithm due to the positive feedback mechanism. The detailed steps of the weight optimization algorithm after fusing genetic algorithm and artificial bee colony algorithm are as follows: 1. First, initialize and determine the population size (number of food sources) as... The number of leader bees is equal to the population size. Only one scout bee is selected, and the number of leader bees is equal to the number of observation bees. The fitness function fit(Xi) is set, where Xi is the i-th solution vector. The crossover probability Pc and mutation probability Pr are defined.

[0048] In this process, the corresponding weight optimization is as follows: the food source adopts binary encoding, and the solution to the problem is represented by an n-dimensional vector, where n is the number of cells to be optimized. Its value is the optimization scheme assigned to cell i, including azimuth angle, downtilt angle, horizontal half-power angle and vertical half-power angle, number of horizontal beams, number of vertical beams, etc.

[0049] Each 5G cell, after scene recognition, has an optimal weight and a candidate weight corresponding to that scene. The optimal and candidate weights of multiple cells are randomly combined to generate... food sources ( , ..., ), as the initial population.

[0050] 2. Place the leader bee on the found food source, calculate the fitness of each food source, and use a coverage-first optimization strategy for antenna weight optimization, comprehensively considering indicators such as coverage, capacity, and interference. Simultaneously, it supports comprehensive optimization of multiple cost functions. Among the following three cost functions... Coverage is the primary cost function. yes , To determine the validity of the nectar source (i.e., the feasible solution); =

[0051] =

[0052] =

[0053] 3. Based on each leader bee adapted to the food source, and using a roulette wheel method to recruit observation bees, the selection probability of the leader bee is determined. for: =

[0054] 4. Use a neighborhood search method for the leading bee and calculate the applicability value of the new solution. If the applicability value of the new solution is better than that of the original solution, replace the original solution with the new solution; otherwise, leave it unchanged. = +

[0055] in, For the middle Random numbers between, Let i be the number of feasible numbers. It is the j-th feasible solution.

[0056] According to the formula, a neighborhood search is performed on the following bee, and the fitness value of the new solution is calculated. If the fitness value of the new solution is better than that of the original solution, the new solution is used to replace the original solution; otherwise, it remains unchanged.

[0057] 5. Enter the scout bee phase, and count the update status of the optimal solution of the leader bee. If the number of times the nectar source has not been updated is greater than the Limit value, the leader bee will become a scout bee and generate a new nectar source in a random way to replace the original nectar source.

[0058] 6. Calculate the profitability of all food sources and rank them.

[0059] 7. Introduce the crossover operator of the genetic algorithm, and determine the crossover probability. Perform cross-operations on food sources, calculate the benefit of newly generated individuals, and if the benefit of the parent generation is less than that of the offspring generation, abandon the parent generation's food source and select the offspring generation to enter the next round of operations; otherwise, retain the parent generation to enter the next round of operations.

[0060] 8. Introduce the mutation operator of the genetic algorithm to categorize food sources according to mutation probability. Perform a mutation operation and calculate the benefit of the newly generated individual. If the benefit of the parent generation is less than that of the offspring generation, abandon the parent generation as a food source and select the offspring generation to enter the next round of operation. Otherwise, retain the parent generation to enter the next round of operation.

[0061] 9. Check if the algorithm's termination condition is met. If not, proceed to the next iteration of the loop and continue searching for food sources.

[0062] If the algorithm's termination condition is met, the current optimal solution is output.

[0063] Steps 3, 4, 5, and 6 belong to the artificial bee colony algorithm stage, while steps 7, 8, and 9 belong to the genetic algorithm stage. The organic combination of the two algorithms can achieve the good effect of complementing each other's strengths and weaknesses.

[0064] In one specific embodiment of this application, user feedback was received that 5G data was lagging and abnormally perceived in the outpatient department of a certain hospital. Based on measurement report data, it was found that the worst-case reference signal received power (RSRP) of 5G in the first and second cells of the outpatient department was -108dBm, the signal to interference plus noise ratio (SINR) was 5dB, the download speed was 39Mbps, and the upload speed was 16Mbps. The weak 5G coverage caused the abnormal perception.

[0065] It provides 17 commonly used coverage scenarios, such asFigure 2 As shown.

[0066] The distance between the first residential area (2.6G) and the station building is 280M. The station is 50M high, the building is 90M high, and the building is 120M wide. The second residential area (2.6G) is 180M away from the station building of the hospital. The station is 23M high, the building is 90M high and 120M wide. The weight optimization method, which combines genetic algorithm and artificial bee colony algorithm, is shown below: The horizontal beamwidth, vertical beamwidth, downtilt angle, and azimuth angle of the antenna are selected as variables and named k, x, y, and z, respectively. They are then encoded in binary. For example, if the horizontal beamwidth, vertical beamwidth, downtilt angle, and azimuth angle are 25°, 6°, 18°, and 3°, respectively, their binary encoding words are 11001, 01010, 00110, 10010, and 00011. The number of buildings to be optimized is 10, and the population size is [missing information]. =11, there is 1 scout bee, 5 leader bees and 5 follower bees, the initial value of the crossover probability Pc is 0.9, and the initial value of the mutation probability Pr is 0.1; After scene recognition, the optimal and alternative weights of multiple building areas are randomly combined to generate... food sources ( , ..., The initial population consists of seven actions with horizontal wavewidths of 15°, 25°, 45°, 65°, 90°, 105°, and 110°, three actions with vertical wavewidths of 6°, 12°, and 25°, 95 actions with an azimuth adjustment range of -47° to 47° and a step size of 1°, and 16 actions with a downtilt adjustment range of -2° to 13° and a step size of 1°. The combination of actions of the four variables yields the set of all actions, totaling 31,920 actions, which are the 31,920 food sources. Calculate the fitness of each food source: =f(k, x, y, z)= , in , , These are respectively coverage rate, proportion of weak coverage grids, The coefficient; Then, the selection probability of the leader bee is calculated based on the fitness of each food source; Like the first one It equals 0.7376. It equals 0.6845. It equals 0.4572, obtained by choosing the probability formula. , , The selection probabilities are 39.25%, 36.42%, and 24.33%, respectively. Draw them on a roulette wheel, spin the wheel, and the place where it stops is the fitness of the selected one. The larger the proportion, the greater the probability of being selected. According to the formula, a neighborhood search is performed on the following bee, and the fitness value of the new solution is calculated. If the fitness value of the new solution is better than that of the original solution, the new solution is used to replace the original solution; otherwise, it remains unchanged.

[0067] Once the scout bee phase begins, the optimal solution update status of the leader bee is recorded. If the number of times the nectar source has not been updated exceeds the Limit value, the leader bee becomes a scout bee and generates a new nectar source randomly to replace the original nectar source.

[0068] Calculate the profitability of all food sources and rank them. The first cell (2.6G) is 280 meters away from the hospital's station building. The station is 50 meters high, the building is 90 meters high and 120 meters wide. Calculations show that a horizontal angle of 24 degrees and a vertical angle of 17 degrees are required. Based on the calculation results, the beam pattern was adjusted from the default (horizontal 105 degrees, vertical narrow beam) to 15 (horizontal 25 degrees, vertical wide beam). The second cell (2.6G) is 180 meters away from the hospital's station building. The station is 23 meters high, the building is 90 meters high and 120 meters wide. Calculations show that it requires 36 degrees horizontally and 26 degrees vertically. Based on the calculation results, the beam pattern was adjusted from the default (105 degrees horizontally and narrow beam vertically) to 14 (45 degrees horizontally and wide beam vertically).

[0069] After the adjustment, RSRP increased by 8dBm, SINR increased by 12dB, download speed increased by 113Mbps, and upload speed increased by 8Mbps.

[0070] like Figure 3 The diagram shown is a schematic representation of a 5G network antenna weight optimization device provided in an embodiment of this application. Figure 3 As shown, the 5G network antenna weight optimization device may include: an acquisition module 301, a decoupling module 302, an establishment module 303, a construction module 304, and an optimization module 305.

[0071] The system includes: an acquisition module 301 for acquiring terminal data through measurement report data; a decoupling module 302 for decoupling the network into multiple subnet slices based on the terminal data, each subnet slice including multiple cells; an establishment module 303 for establishing a single-cell beam weight library according to different coverage scenarios; a construction module 304 for constructing a multi-cell beam weight library based on the single-cell beam weight library; and an optimization module 305 for using a fusion algorithm to search for the optimal weight set in the multi-cell beam weight library, wherein the fusion algorithm is a combination of population algorithm and genetic algorithm.

[0072] In this embodiment, the acquisition module 301 first acquires terminal data through measurement report data. Then, the decoupling module 302 decouples the network into multiple sub-network slices based on the terminal data. Each sub-network slice includes multiple cells. Next, the establishment module 303 establishes a single-cell beam weight library according to different coverage scenarios. Finally, the construction module 304 constructs a multi-cell beam weight library based on the single-cell beam weight library. Finally, the optimization module 305 uses a fusion algorithm to search for the optimal weights in the multi-cell beam weight library. The fusion algorithm is a combination of population algorithm and genetic algorithm. This embodiment constructs a multi-cell beam weight library from a single-cell beam weight library and then uses a combination of population algorithm and genetic algorithm to search for the optimal weights. The population algorithm generates a well-distributed initial population, and the genetic algorithm solves the problem, reducing the possibility of the algorithm getting trapped in local optima and making it easier to find the global optimum. This also accelerates algorithm convergence and improves efficiency.

[0073] In one possible implementation of this application, the 5G network antenna weight optimization device may further include: a second optimization module.

[0074] The second optimization module is used to search the single-cell beam weight library using the particle swarm optimization algorithm to obtain the optimal weight group for the single cell.

[0075] In one possible implementation of this application, the second optimization module is configured to: initialize the number of particles in the population, randomize the position and movement speed of the beam weight group; evaluate the fitness value of each particle according to the fitness function; update the movement speed of each particle using a chaotic operator as a random number; update the position of the beam weight group based on the current position of the particle's weight group and the historical best position of the weight group; and terminate the update to obtain the optimal weight group for a single cell if a termination condition is met, wherein the termination condition includes at least one of the following: reaching the maximum number of iterations, or the increment of the fitness value being less than the increment threshold.

[0076] In one possible implementation of this application, the second optimization module is configured to: for each particle, compare the fitness value corresponding to the particle's current weight group position with the fitness value corresponding to the particle's historical best weight group position; if the fitness value corresponding to the current weight group position is greater than the fitness value corresponding to the historical best weight group position, update the current weight group position to the particle's optimal weight group position; for each particle, compare the particle's current fitness value with the fitness value corresponding to the population's optimal weight group position; if the particle's current fitness value is greater than the fitness value corresponding to the population's optimal weight group position, update the particle's current fitness value to the population's optimal weight group position.

[0077] In one possible implementation of this application, the optimization module 305 is configured to: obtain the optimal weight and alternative weight of a scene for each cell through scene recognition; determine the population size based on the optimal weight and alternative weight of multiple cells; sort the populations in the population size using an artificial bee colony algorithm based on the population size, and determine the crossover probability and mutation probability; and process the sorted populations using a genetic algorithm based on the crossover probability and the mutation probability until a preset condition is met to obtain the optimal weight group.

[0078] In one possible implementation of this application, the optimization module 305 is configured to: determine the number of leader bees, crossover probability, and mutation probability based on the population size, wherein the number of leader bees is equal to the population size in the population, and the population serves as a food source; place the leader bees in the current food source and calculate the fitness of each food source; calculate the probability that a food source is selected by follower bees based on the fitness, and determine the position of the food source selected by the follower bees; perform a neighborhood search on the leader bees and follower bees to determine a new food source; update the food source to the new food source if the fitness of the new food source is greater than the fitness of the food source selected by the leader bees and the food source selected by the follower bees; if the number of times a food source has not been updated is greater than a coefficient threshold, convert the leader bees into scout bees and generate a new food source; determine the profitability of all food sources and sort the food sources based on the profitability.

[0079] In one possible implementation of this application, the optimization module 305 is configured to: utilize the crossover operator of a genetic algorithm to perform a crossover operation on the population based on the crossover probability, and calculate the payoff of the newly generated population; if the payoff of the newly generated population is greater than the payoff of the population that underwent the crossover operation, then the payoff of the newly generated population is taken as the payoff of the population, and the crossover operation is performed on the newly generated population; utilize the mutation operator of a genetic algorithm to perform a mutation operation on the population based on the mutation probability, and calculate the payoff of the newly generated population; if the payoff of the newly generated population is greater than the payoff of the population that underwent the mutation operation, then the payoff of the newly generated population is taken as the payoff of the population, and the mutation operation is performed on the newly generated population; and if the crossover operation and the mutation operation satisfy preset conditions, the preset conditions include reaching the maximum number of operations.

[0080] The functionality of the 5G network antenna weight optimization device in this application has already been implemented. Figures 1 to 2 The method embodiments shown are described in detail. Therefore, for any parts not covered in detail in this embodiment, please refer to the relevant descriptions in the foregoing embodiments, which will not be repeated here.

[0081] like Figure 4As shown, this application embodiment also provides an electronic device 400, including a processor 401, a memory 402, and a program or instructions stored in the memory 402 and executable on the processor 401. When the program or instructions are executed by the processor 401, they implement the various processes of the above-described 5G network antenna weight optimization processing method embodiment and can achieve the same technical effect. To avoid repetition, they will not be described again here.

[0082] Optionally, embodiments of this application also provide a computer-readable storage medium storing a computer program. When executed by a processor, this computer program implements the various processes of the above-described 5G network antenna weight optimization method embodiments and achieves the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0083] Optionally, this application embodiment also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, which, when executed by a computer, implement the various processes of the above-described 5G network antenna weight optimization method embodiment and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0084] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0085] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0086] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for optimizing antenna weights in a 5G network, characterized in that, include: Obtain terminal data through measurement report data; Based on the terminal data, the network is decoupled into multiple subnet slices, each subnet slice including multiple cells; Establish a single-cell beam weight library based on different coverage scenarios; Based on the single-cell beam weight library, a multi-cell beam weight library is constructed; The optimal weight set is obtained by searching the multi-cell beam weight library using a fusion algorithm, which is a combination of population algorithm and genetic algorithm.

2. The method according to claim 1, characterized in that, After establishing a single-cell beam weight library according to different coverage scenarios, the method includes: The particle swarm optimization algorithm is used to search the single-cell beam weight library to obtain the optimal weight group for the single cell.

3. The method according to claim 2, characterized in that, The step of using particle swarm optimization to search the single-cell beam weight library to obtain the optimal weight set for a single cell includes: Initialize the population particle count, and randomize the position and speed of the beam weights; The fitness value of each particle is evaluated based on the fitness function; The movement speed of each particle is updated using chaotic operators as random numbers; Update the beam weight group position based on the current weight group position of the particle and the historical best weight group position; If the termination condition is met, the update ends and the optimal weight group for a single cell is obtained. The termination condition includes at least one of the following: reaching the maximum number of iterations, or the increment of the fitness value is less than the increment threshold.

4. The method according to claim 3, characterized in that, The step of updating the beam weight group position based on the current position of the particle weight group and the historical best position of the weight group includes: For each particle, the fitness value corresponding to the particle's current weight set position is compared with the fitness value corresponding to the particle's historical best weight set position. If the fitness value corresponding to the current weight set position is greater than the fitness value corresponding to the historical best weight set position, the current weight set position is updated to the optimal weight set position of the particle. For each particle, the particle's current fitness value is compared with the fitness value corresponding to the position of the population's optimal weight set; If the current fitness value of a particle is greater than the fitness value corresponding to the position of the optimal weight group in the population, the current fitness value of the particle is updated to the position of the optimal weight group in the population.

5. The method according to claim 2, characterized in that, The step of using a fusion algorithm to search for the optimal weight set in the multi-cell beam weight library includes: Each community obtains the optimal and alternative weights for a scene through scene recognition; The population size is determined based on the optimal and alternative weights of multiple cells. Based on the population size, the populations within the population size are sorted using the artificial bee colony algorithm, and the crossover probability and mutation probability are determined. Using a genetic algorithm, the sorted population is processed based on the crossover probability and the mutation probability until a preset condition is met, resulting in the optimal weight group.

6. The method according to claim 5, characterized in that, The step of sorting the populations within the specified population size using an artificial bee colony algorithm and determining the crossover and mutation probabilities, includes: Based on the population size, the number of leader bees, crossover probability, and mutation probability are determined, wherein the number of leader bees is equal to the population size in the population size, and the population serves as a food source. Place the leader bee at the current food source and calculate the fitness of each food source; Based on the fitness, the probability that a food source is selected by a follower bee is calculated, and the location of the food source selected by the follower bee is determined. The leader bee and follower bees conduct neighborhood searches to identify new food sources; If the fitness of the new food source is greater than the fitness of the food source selected by the leader bee and the follower bee, the food source is updated to the new food source; If the number of times the food source has not been updated exceeds a coefficient threshold, the leader bee will be transformed into a scout bee, and a new food source will be generated. Determine the profitability of all food sources and rank them based on the profitability.

7. The method according to claim 5, characterized in that, The process of using a genetic algorithm to process the sorted population based on the crossover probability and the mutation probability until a preset condition is met to obtain the optimal weight set includes: Using the crossover operator of the genetic algorithm, crossover operations are performed on the population based on the crossover probability to calculate the payoff of the newly generated population; If the payoff of the newly generated population is greater than the payoff of the population that underwent the crossover operation, then the payoff of the newly generated population is taken as the payoff of the population, and the crossover operation is performed using the newly generated population. Using the mutation operator of the genetic algorithm, the population is mutated based on the mutation probability, and the payoff of the newly generated population is calculated. If the payoff of the newly generated population is greater than the payoff of the population undergoing mutation, then the payoff of the newly generated population is taken as the payoff of the population, and the mutation operation is performed using the newly generated population. When the crossover operation and the mutation operation meet preset conditions, an optimal weight set is obtained. The preset conditions include the number of operations reaching the maximum number of operations.

8. A 5G network antenna weight optimization device, characterized in that, include: The acquisition module is used to acquire terminal data through measurement report data; The decoupling module is used to decouple the network into multiple subnet slices based on the terminal data, wherein the subnet slices include multiple cells; A module is established to create a single-cell beam weight library based on different coverage scenarios; The construction module is used to construct a multi-cell beam weight library based on the single-cell beam weight library; The optimization module is used to search for the optimal weight set in the multi-cell beam weight library using a fusion algorithm. The fusion algorithm is a combination of population algorithm and genetic algorithm.

9. An electronic device, characterized in that, It includes a processor, a memory, a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions that, when executed by a computer, implement the steps of the method as described in any one of claims 1 to 7.

Images