Multi-frequency operator network parameter intelligent collaboration and adjustment system and method
By dividing network scenarios into different types and optimizing parameters using machine learning and genetic algorithms, the problems of insufficient utilization of spectrum resources and prolonging network switching in the multi-frequency coverage area are solved, and faster and more stable Network switching and higher user experience.
Patent Information
- Application Number
- CN202411481789.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-23
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-10-23
AI Technical Summary
In the multi-frequency coverage areas of 4G and 5G networks, it is difficult for the prior art to effectively coordinate and adjust network parameters, resulting in insufficient utilization of spectrum resources, high delay and failure rates during network switching, affecting user experience.
Divide network scenarios into three types: points, lines, and surfaces, formulate corresponding parameter templates, and predict network switching through machine learning models, pre-allocate resources, build fitness functions to optimize parameters, and use genetic algorithms to optimize parameters to reduce switching Failure rate and network error.
It improves network switching speed and stability, reduces latency, improves user experience and network service satisfaction, and enhances network reliability and adaptability.
Smart Images

Figure CN119012298A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of information technology, and in particular to a multi-frequency operator network parameter intelligent coordination and adjustment system and method. Background Art
[0002] In the field of Internet consumption, 4G technology is widely used. By using the 4G network, users can make mobile payments and instant communications more conveniently. At the same time, the application effect of 4G is also relatively smooth in the transmission of images, videos and other content. In comparison, 5G is mainly used in special tasks such as industrial manufacturing, the Internet, and intelligent driving. The two have a synergistic development and coexistence relationship as a whole. As the 4G and 5G cross-system interoperability solutions gradually mature and even come into being, users will get a better continuous experience.
[0003] In actual use, 4G and 5G have multiple frequencies covering the same area. In order to better utilize spectrum resources and achieve better coverage, this paper proposes a multi-frequency operator network parameter intelligent coordination and adjustment system and method. Summary of the invention
[0004] The present invention proposes a technical solution to solve the problems in the background technology: a multi-frequency operator network parameter intelligent coordination and adjustment method, the method comprising: Different network scenarios are divided into three types: point, line, and surface. Corresponding parameter templates are formulated according to the characteristics of each type of scenario, and a type stamp is added to each parameter template. Obtain network data for each type of scenario to form a historical data set; the network data includes average latency, signal strength, signal-to-noise ratio, total number of handover attempts between operator networks, and number of handover failures; Analyze historical data for each type of scenario through machine learning models to predict whether a handover between operator networks will occur in the scenario; Pre-allocate resources before the predicted handover occurs to reduce the delay and reduce the handover failure rate; Constructing a fitness function 1 to evaluate the impact of reducing the switching failure rate on the network switching performance; Constructing fitness function 2 to evaluate the impact of network error and control input on network switching quality, so as to adjust the parameter templates in three types of scenarios; The parameters of fitness function 1 and fitness function 2 are optimized by genetic algorithm to find the optimal solution, and the corresponding parameters of the parameter template of the corresponding type of scene are modified according to the optimal solution.
[0005] Preferably, the obtaining of network data in each type of scenario to form a historical data set comprises the following steps: Acquire network data in point-type scenarios according to the preset collection cycle and collection frequency to form a point-type historical data set , and according to Constructing feature datasets for point type scenarios ; Acquire network data in face type scenarios according to preset collection cycle and collection frequency to form historical data sets of face type , and according to Constructing feature datasets for line type scenarios ; Acquire network data in face type scenarios according to preset collection cycle and collection frequency to form historical data sets of face type , and according to Feature dataset for building surface type scenarios .
[0006] Preferably, the Constructing feature datasets for point type scenarios , including the following steps: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the point type scenario is ; The basis Constructing feature datasets for point type scenarios , including the following steps: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the line type scenario is ; The basis Feature dataset for building surface type scenarios , including the following steps: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the face type scenario is .
[0007] Preferably, analyzing the historical data in each type of scenario by a machine learning model to predict whether a handover between operator networks will occur in the type of scenario includes the following steps: Prediction Model 1, Prediction Model 2, and Prediction Model 3 are constructed through regression algorithms to predict the switching of 4G and 5G networks in point-type scenarios, line-type scenarios, and surface-type scenarios, respectively. The prediction model 1, prediction model 2 and prediction model 3 are trained by using the historical data set in the point type scenario, the historical data set in the line type scenario and the historical data set in the surface type scenario, specifically including the following steps: After normalizing the data in the historical data set in the point type scenario, the data are divided into a training set and a validation set. The training set is used to train the prediction model 1, and the validation set is used to verify the performance of the prediction model 1. After normalizing the data in the historical data set in the line type scenario, the data are divided into a training set and a validation set. The training set is used to train the prediction model 2, and the validation set is used to verify the performance of the prediction model 2. After normalizing the data in the historical data set under the face type scenario, the data are divided into a training set and a validation set. The training set is used to train the prediction model 3, and the validation set is used to verify the performance of the prediction model 3. The network data information within the preset time span is collected from point, line and surface type networks through the preset sliding window, and the newly collected network data information is preprocessed and normalized and then input into the trained prediction model 1, prediction model 2 and prediction model 3 respectively; Through the trained prediction model 1, prediction model 2 and prediction model 3, it is predicted whether switching between 4G and 5G networks will occur at a certain moment in the future.
[0008] Preferably, the prediction model 1 includes a switching prediction model 1 and a switching failure prediction model 1, and the switching prediction model 1 is ; Switching failure model 1 is ; in, , are the intercept, , are the error terms, Indicates the number of feature vectors before the switch is successful, Indicates the number of feature vectors before the switching failure; The input variable of the prediction model 1 is a vector consisting of signal-to-noise ratio and signal strength. , therefore, switch prediction model 1 to: ;in is a weight matrix 1, which is used to represent the influence of signal-to-noise ratio and signal strength on the input variable. ;in, ; Similarly, the switch failure model is converted to: ; Before the switching time point in the sliding window Seconds of data Input into prediction model 1 to predict whether network switching will occur at a future moment; Setting the switching threshold ,if It is determined that a network switch will occur at a future time; otherwise, a network switch will not occur; If it is predicted that a network switch will occur at a future time, the input variables of the switch prediction model 1 are input into the switch failure model 1 to predict the probability of a successful switch; Setting probability thresholds ,if The switch is judged to fail.
[0009] Preferably, the pre-allocation of resources before the predicted handover occurs to reduce the delay before and after and reduce the handover failure rate includes the following steps: If it is determined that the switch will fail, resources are allocated before the predicted switch occurs; The allocation of resources comprises the following steps: Obtain bandwidth resources and computing resources that can be pre-allocated through the network management system; The real-time network load of 4G and 5G networks is obtained through the network management system; if the 4G network is heavily loaded, priority is given to reserving bandwidth for the 5G network to reduce the pressure on the 4G network.
[0010] Preferably, if it is determined that the switching will fail, the switching threshold can be adjusted dynamically, and the network switching can be prevented by increasing the switching threshold, and the parameter data of the corresponding parameter template can be corrected; The step of modifying the parameter data of the corresponding parameter template comprises the following steps: Modify the power parameters of the device in the parameter template to improve the network signal strength; Modify the number of enabled devices in the parameter template and enable more network base stations to improve network signal coverage and signal strength.
[0011] Preferably, the constructing of a fitness function 1 for evaluating the impact of reducing the switching failure rate on the network switching performance includes: Constructing fitness function 1: ; in, Indicates the switching failure rate; represents the average switching delay, represents the weight coefficient; in, , ; Obtain the total number of network switching attempts, number of switching failures, and each switching delay through the network management system; The fitness function 2 is constructed to evaluate the influence of network error and control input on network switching quality, so as to adjust parameter templates in three types of scenarios, including: Construct fitness function 2: ; in, represents the network error, represents the control input, Represent the weight coefficients of network error and control input respectively; The method of optimizing the parameters of the fitness function 1 and the fitness function 2 by using a genetic algorithm to seek an optimal solution comprises the following steps: By optimizing the parameters of fitness function 1 through genetic algorithm, we seek to make Parameter combination 1 with the minimum value; By optimizing the parameters of fitness function 2 through genetic algorithm, we seek to make The minimum parameter combination 2 is obtained, which is used to indicate that both the network error and the control input are under control; The method of optimizing the parameters of the first fitness function and the second fitness function by a genetic algorithm to seek an optimal solution and modifying corresponding parameters of the corresponding type of scene parameter template according to the optimal solution includes: Get the optimized parameter combination ; Get the optimized parameter combination 2 ; because , modify the switching threshold in the parameter template to adjust the total number of switching attempts; because , by adjusting the cell reselection priority and neighbor relationship configuration in the parameter template, the switching speed can be improved and the switching delay can be reduced.
[0012] The present invention also provides a multi-frequency operator network parameter intelligent coordination and adjustment system, and the system is used to execute the multi-frequency operator network parameter intelligent coordination and adjustment method.
[0013] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the multi-frequency operator network parameter intelligent coordination and adjustment method.
[0014] Beneficial effects of the present invention: 1. The present invention divides different network scenarios into three types of scenarios: point, line, and surface, and formulates corresponding parameter templates according to the characteristics of each type of scenario. A type stamp is added to each parameter template to distinguish different types of usage scenarios, and to facilitate the rapid retrieval of corresponding parameter templates during later parameter modification. By optimizing the switching process and the parameters of the parameter template, network delay can be reduced and user experience can be improved, especially for delay-sensitive applications such as online games and video conferencing. After improving the switching speed, fast and accurate network switching can provide smoother and more stable services, enhancing user satisfaction with network services.
[0015] 2. The present invention sets fitness function 1 and fitness function 2, evaluates the influence of network switching failure rate and average switching delay on the switching network through fitness function 1, and seeks the parameter combination of the optimal solution; evaluates the influence of network error and control input on network switching quality through fitness function 2, and seeks the parameter combination of the optimal solution. In the process of seeking the optimal solution, the genetic algorithm is applied to optimize the parameters, so that the network parameters can be dynamically optimized according to different environmental factors (such as user distribution, time, location, etc.) to adapt to the complex and changeable network environment. After obtaining the optimal parameter combination, the parameters (signal strength, switching delay, base station power, switching control parameters, cell reselection priority and collection relationship configuration, etc.) are adjusted by modifying the parameter template to ensure the quality of network switching.
[0016] 3. The present invention takes into account two key network parameters, signal-to-noise ratio and signal strength, when setting the input variables of prediction model 1, prediction model 2, and prediction model 3, and uses the products of multiplying them with weight matrix 1, weight matrix 2, and weight matrix 3 as the input variables of the three models respectively. The degree of influence of the two variables on network switching is reflected through the weight matrix. This method of integrating multiple parameters makes the prediction more comprehensive and improves the accuracy of the prediction.
[0017] 4. In the present invention, each prediction model includes a switching prediction model and a switching failure prediction model. By predicting the occurrence of switching, resources such as bandwidth are prepared in advance, thereby improving the success rate of switching; by predicting that switching may fail, the switching threshold is adjusted to avoid the switching, thereby avoiding the impact on user use due to switching failure. By reducing the number of switching failures and optimizing network performance, the stability and reliability of the overall network can be improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 The present invention is a flow chart of a method for intelligent coordination and adjustment of network parameters of multiple frequency operators. DETAILED DESCRIPTION
[0019] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles of the present invention defined in the following description can be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not deviate from the spirit and scope of the present invention.
[0020] It is to be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element may be one, while in another embodiment, the number of the element may be multiple, and the term "one" should not be understood as a limitation on the quantity.
[0021] refer to Figure 1 The technical solution provided by the present invention is: a method for intelligent coordination and adjustment of network parameters of multi-frequency operators, comprising the following steps: Step 1: Different network scenarios are divided into three types of scenarios: point, line, and surface. A corresponding parameter template is formulated according to the characteristics of each type of scenario, and a type stamp is added to each parameter template so as to call the parameter template of the corresponding type of scenario.
[0022] The so-called point scenario refers to the network connection of a single device or node, the line scenario refers to the communication link between two devices, and the surface scenario may involve the entire network topology composed of multiple devices.
[0023] Set clear scope and boundaries for each scenario type to ensure that they are operational and measurable in the actual network environment. Point scenarios need to focus on the bandwidth and latency of the device, line scenarios need to focus on the reliability of the link and bandwidth utilization, and surface scenarios need to focus on the stability and scalability of the entire network.
[0024] In this embodiment, appropriate bandwidth limits can be set for each scenario type according to the scenario type and actual needs. For example, for point scenarios, the bandwidth can be set according to the processing capability of the device; for line scenarios, the bandwidth can be set according to the capacity of the link; for surface scenarios, the bandwidth can be set according to the needs of the entire network. For point scenarios, the delay threshold is set according to the processing speed of the device; for line scenarios, the delay threshold is set according to the transmission speed of the link; for surface scenarios, the delay threshold is set according to the transmission speed of the entire network.
[0025] Add a unique type stamp to each parameter template to identify the scenario type to which the template applies. In this way, when you need to call a specific type of scenario parameter, you can quickly find the corresponding template based on the type stamp. Type stamps make it easy to manage and call parameter templates of different scenario types. For example, you can use a database or other data storage system to store templates and provide appropriate interfaces for other systems and services to call. Administrators or automated scripts dynamically update parameter templates based on monitoring results.
[0026] Step 2: Obtain network data for each type of scenario to form a historical data set; the network data includes average latency, signal strength, signal-to-noise ratio, total number of handover attempts between operator networks, and number of handover failures; specifically, the following steps are included: Step 2.1: Obtain network data in point type scenarios according to the preset collection cycle and collection frequency to form a point type historical data set , and according to Constructing feature datasets for point type scenarios , specifically: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the point type scenario is
[0027] Step 2.2: Obtain network data in the face type scenario according to the preset collection cycle and collection frequency to form a historical data set of the face type , and according to Constructing feature datasets for line type scenarios , specifically: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the line type scenario is .
[0028] Step 2.3: Obtain network data in the face type scenario according to the preset collection cycle and collection frequency to form a historical data set of the face type , and according to Feature dataset for building surface type scenarios , specifically: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the face type scenario is .
[0029] Step 3: Analyze historical data for each type of scenario through a machine learning model to predict whether a handover between operator networks will occur in the scenario. Specifically: Construct a prediction model 1 adapted to the point type scenario, including a switching prediction model 1 and a switching failure prediction model 1. The switching prediction model 1 is ; Switching failure model 1 is ;in, , are the intercept, , are the error terms, Indicates the number of feature vectors before the switch is successful, Indicates the number of feature vectors before the switching failure; Since the input variable of prediction model 1 is a vector consisting of signal-to-noise ratio and signal strength, , therefore, switch the prediction model to: ;in is a weight matrix used to represent the influence of signal-to-noise ratio and signal strength on input variables. ;in, ; = = ; ; Similarly, the switch failure model is converted to: ; Before the switching time point in the sliding window Seconds of data Input into prediction model 1 to predict whether network switching will occur at a future moment; Setting the switching threshold ,if It is determined that a network switch will occur at a future moment.
[0030] In this embodiment, the switching threshold is set to 0.5. , then it is predicted that network switching will occur, otherwise, it is predicted that network switching will not occur.
[0031] The so-called future moment is related to the time span of acquiring data. If one hour is used as the collection cycle and data is collected every 5 minutes, the predicted future moment is the moment 5 minutes later.
[0032] If a network switch is predicted to occur at a future moment, the input variables of the switch prediction model 1 are input into the switch failure model 1 to predict the probability of a successful switch, which helps to identify potential network switch problems in advance and take corresponding measures to optimize network performance or user experience, specifically: Setting probability thresholds ,if In this embodiment, the probability threshold is set to 0.8. , then the switch will fail.
[0033] Similarly, we build a prediction model 2 that is adapted to the line type scenario: ; are the intercept, represents the regression coefficient, is the error term, Indicates the number of feature vectors before the switch is successful; Construct prediction model 3 adapted to face type scenarios: ; are the intercept, represents the regression coefficient, is the error term.
[0034] Setting probability thresholds ,if Then it is determined that switching will occur; set the probability threshold ,if It is determined that switching will occur.
[0035] The handover failure prediction in line and surface type scenarios is also achieved by building a handover failure prediction model similar to that in point type scenarios. Since the principles of model construction and judgment are similar, they will not be described here. By predicting the occurrence and success probability of network handover, network administrators can better plan network resources and improve network stability and reliability. It also helps to reduce unnecessary network handover attempts, thereby saving network resources and reducing potential risks.
[0036] Step 4: pre-allocate resources before the predicted handover occurs to reduce the delay and the handover failure rate, specifically: During the network switching process, the increase in latency and the increase in switching failure rate are two major problems. By predicting the time when the switching will occur, resources can be allocated in advance, thus effectively reducing these problems.
[0037] If it is determined that the upcoming handover may fail, resources are pre-allocated (including bandwidth resources and computing resources) before the predicted handover time to ensure a smooth handover process; Obtain the real-time network load of 4G and 5G networks through the network management system; if the 4G network is heavily loaded, give priority to reserving bandwidth for the 5G network to reduce the pressure on the 4G network; In addition, the switching threshold is an important parameter for determining whether to perform network switching. By dynamically adjusting this threshold, the occurrence of network switching can be effectively controlled.
[0038] In this embodiment, if it is determined that the upcoming switch may fail, the switch threshold can be automatically increased to prevent the switch from occurring. At the same time, the parameter data in the corresponding parameter template is corrected to ensure the stability of the network. The switch threshold can be increased by a fixed percentage or value; after updating the switch threshold, check again whether the new threshold meets the requirements. When the switch threshold reaches the set upper limit (increase by more than 10%), stop increasing the threshold to prevent infinite loops or over-adjustment. At this time, other strategies can be selected to handle, such as sending an alarm to notify the administrator to take corresponding treatment measures.
[0039] The step of modifying the parameter data of the corresponding parameter template comprises the following steps: Modify the power parameters of the device in the parameter template to improve the network signal strength; Modify the number of enabled devices in the parameter template and enable more network base stations to improve network signal coverage and signal strength.
[0040] Step 5: Construct a fitness function 1 to evaluate the impact of reducing the switching failure rate on the network switching performance, including: Constructing fitness function 1: ; in, Indicates the switching failure rate; represents the average switching delay, represents the weight coefficient; in, , ; Obtain the total number of network switching attempts, number of switching failures, and each switching delay through the network management system; Step 6: Construct fitness function 2 to evaluate the impact of network error and control input on network switching quality, so as to adjust parameter templates in three types of scenarios; including: Construct fitness function 2: ; in, represents the network error, represents the control input, Represent the weight coefficients of network error and control input respectively; The method of optimizing the parameters of the fitness function 1 and the fitness function 2 by using a genetic algorithm to seek an optimal solution comprises the following steps: By optimizing the parameters of fitness function 1 through genetic algorithm, we seek to make Parameter combination 1 with the minimum value; By optimizing the parameters of fitness function 2 through genetic algorithm, we seek to make The minimum parameter combination 2 is obtained, which is used to indicate that both the network error and the control input are under control; Step 7: Optimize the parameters of fitness function 1 and fitness function 2 through genetic algorithm to find the optimal solution, and modify the corresponding parameters of the corresponding type of scene parameter template according to the optimal solution, including: Get the optimized parameter combination ; Get the optimized parameter combination 2 ; because , modifying the handover threshold in the parameter template can be used to adjust the total number of handover attempts; Combined with the above method of reserving bandwidth and computing resources, the number of failed switching can be reduced to a certain extent, while further improving the switching speed.
[0041] because , by adjusting the cell reselection priority and neighbor relationship configuration in the parameter template, the switching speed can be improved and the switching delay can be reduced. For example, the priority of the cell where the 5G network is deployed can be increased to switch to the 5G network first. Alternatively, the neighbor list can be updated to ensure that the user equipment can promptly identify and switch to the neighbor cell with better signal quality.
[0042] The process of optimizing parameters through genetic algorithms in steps six and seven includes routine steps such as initializing the population, evaluating fitness, selection, crossover, mutation, replacement, and checking termination conditions, which can be specifically implemented through Python.
[0043] In this embodiment, in the optimization of fitness function 1, the main parameters can be set to: population_size = 100, gene_length = 2, max_iterations = 1000, mutation_rate = 0.1, crossover_rate = 0.8.
[0044] In the crossover operation, the single-point crossover method is used to return the offspring after crossover; in the mutation operation, the uniform mutation method is used to return the offspring after mutation.
[0045] The present invention also provides a multi-frequency operator network parameter intelligent coordination and adjustment system, and the system is used to execute the multi-frequency operator network parameter intelligent coordination and adjustment method.
[0046] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the multi-frequency operator network parameter intelligent coordination and adjustment method.
[0047] The embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. The embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part, and / or installed from a removable medium. When the computer program is executed by the central processing unit (CPU), the above-mentioned functions defined in the method of the present application are executed. It should be noted that the above-mentioned computer-readable medium of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of an electrical, magnetic, optical, electromagnetic, infrared segment, or semiconductor, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more wire segments, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, electrical wire, optical cable, RF, etc., or any suitable combination of the foregoing.
[0048] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0049] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments. Without departing from the principles, the implementation methods of the present invention may be deformed or modified in any way.
Claims
1. A method for intelligent coordination and adjustment of network parameters of multi-frequency operators, It is characterized in that The method comprises: Different network scenarios are divided into three types: point, line, and surface. Corresponding parameter templates are formulated according to the characteristics of each type of scenario, and a type stamp is added to each parameter template. Obtain network data for each type of scenario to form a historical data set; the network data includes average latency, signal strength, signal-to-noise ratio, total number of handover attempts between operator networks, and number of handover failures; Analyze historical data for each type of scenario through machine learning models to predict whether a handover between operator networks will occur in the scenario; Pre-allocate resources before the predicted handover occurs to reduce the delay and reduce the handover failure rate; Constructing a fitness function 1 to evaluate the impact of reducing the switching failure rate on the network switching performance; Constructing fitness function 2 to evaluate the impact of network error and control input on network switching quality, so as to adjust the parameter templates in three types of scenarios; The parameters of fitness function 1 and fitness function 2 are optimized by genetic algorithm to find the optimal solution, and the corresponding parameters of the parameter template of the corresponding type of scene are modified according to the optimal solution.
2. A multi-frequency operator network parameter intelligent coordination and adjustment method according to claim 1, It is characterized in that The process of obtaining network data in each type of scenario to form a historical data set includes the following steps: Acquire network data in point-type scenarios according to the preset collection cycle and collection frequency to form a point-type historical data set , and according to Constructing feature datasets for point type scenarios ; Acquire network data in face type scenarios according to preset collection cycle and collection frequency to form historical data sets of face type , and according to Constructing feature datasets for line type scenarios ; Acquire network data in face type scenarios according to preset collection cycle and collection frequency to form historical data sets of face type , and according to Feature dataset for building surface type scenarios .
3. A multi-frequency operator network parameter intelligent coordination and adjustment method according to claim 2, It is characterized in that The basis Constructing feature datasets for point type scenarios , including the following steps: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the point type scenario is ; The basis Constructing feature datasets for point type scenarios , including the following steps: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the line type scenario is ; The basis Feature dataset for building surface type scenarios , including the following steps: from Extract network switching time points within a preset time span , Switch failure time point ; Get the network switching time point forward The signal strength and signal-to-noise ratio data of all point-type scenarios within one second constitute the feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; Switching failure time point forward All signal strength and signal-to-noise ratio data within seconds constitute a feature vector set ,in express The feature vector at time, Respectively The signal-to-noise ratio and signal strength at each moment, ; The historical dataset in the face type scenario is .
4. A multi-frequency operator network parameter intelligent coordination and adjustment method according to claim 1, It is characterized in that The method of analyzing historical data in each type of scenario by a machine learning model to predict whether a handover between operator networks will occur in the type of scenario includes the following steps: Prediction Model 1, Prediction Model 2, and Prediction Model 3 are constructed through regression algorithms to predict the switching of 4G and 5G networks in point-type scenarios, line-type scenarios, and surface-type scenarios, respectively. The prediction model 1, prediction model 2 and prediction model 3 are trained by using the historical data set in the point type scenario, the historical data set in the line type scenario and the historical data set in the surface type scenario, specifically including the following steps: After normalizing the data in the historical data set in the point type scenario, the data are divided into a training set and a validation set. The training set is used to train the prediction model 1, and the validation set is used to verify the performance of the prediction model 1. After normalizing the data in the historical data set in the line type scenario, the data are divided into a training set and a validation set. The training set is used to train the prediction model 2, and the validation set is used to verify the performance of the prediction model 2. After normalizing the data in the historical data set under the face type scenario, the data are divided into a training set and a validation set. The training set is used to train the prediction model 3, and the validation set is used to verify the performance of the prediction model 3. The network data information within the preset time span is collected from point, line and surface type networks through the preset sliding window, and the newly collected network data information is preprocessed and normalized and then input into the trained prediction model 1, prediction model 2 and prediction model 3 respectively; Through the trained prediction model 1, prediction model 2 and prediction model 3, it is predicted whether switching between 4G and 5G networks will occur at a certain moment in the future.
5. A multi-frequency operator network parameter intelligent coordination and adjustment method according to claim 4, It is characterized in that The prediction model 1 includes a switching prediction model 1 and a switching failure prediction model 1. The switching prediction model 1 is ; Switching failure model 1 is ; in, , are the intercept, , are the error terms, Indicates the number of feature vectors before the switch is successful, Indicates the number of feature vectors before the switching failure; The input variable of the prediction model 1 is a vector consisting of signal-to-noise ratio and signal strength. , therefore, switch prediction model 1 to: ;in is a weight matrix 1, which is used to represent the influence of signal-to-noise ratio and signal strength on the input variable. ;in, ; Similarly, the switch failure model is converted to: ; Before the switching time point in the sliding window Seconds of data Input into prediction model 1 to predict whether network switching will occur at a future moment; Setting the switching threshold ,if It is determined that a network switch will occur at a future time; otherwise, a network switch will not occur; If it is predicted that a network switch will occur at a future time, the input variables of the switch prediction model 1 are input into the switch failure model 1 to predict the probability of a successful switch; Setting probability thresholds ,if The switch is judged to fail.
6. A multi-frequency operator network parameter intelligent coordination and adjustment method according to claim 4, It is characterized in that The method of pre-allocating resources before the predicted switching occurs to reduce the delay before and after and reduce the switching failure rate includes the following steps: If it is determined that the switch will fail, resources are allocated before the predicted switch occurs; The allocation of resources comprises the following steps: Obtain bandwidth resources and computing resources that can be pre-allocated through the network management system; The real-time network load of 4G and 5G networks is obtained through the network management system; if the 4G network is heavily loaded, priority is given to reserving bandwidth for the 5G network to reduce the pressure on the 4G network.
7. A multi-frequency operator network parameter intelligent coordination and adjustment method according to claim 1, It is characterized in that If it is determined that the switching will fail, the switching threshold can be adjusted dynamically. By increasing the switching threshold, the network switching is prevented and the parameter data of the corresponding parameter template is corrected; The step of modifying the parameter data of the corresponding parameter template comprises the following steps: Modify the power parameters of the device in the parameter template to improve the network signal strength; Modify the number of enabled devices in the parameter template and enable more network base stations to improve network signal coverage and signal strength.
8. A multi-frequency operator network parameter intelligent coordination and adjustment method according to claim 1, It is characterized in that The first fitness function is constructed to evaluate the impact of reducing the switching failure rate on the network switching performance, including: Constructing fitness function 1: ; in, Indicates the switching failure rate; represents the average switching delay, represents the weight coefficient; in, , ; Obtain the total number of network switching attempts, number of switching failures, and each switching delay through the network management system; The fitness function 2 is constructed to evaluate the influence of network error and control input on network switching quality, so as to adjust parameter templates in three types of scenarios, including: Construct fitness function 2: ; in, represents the network error, represents the control input, Represent the weight coefficients of network error and control input respectively; The method of optimizing the parameters of the fitness function 1 and the fitness function 2 by using a genetic algorithm to seek an optimal solution comprises the following steps: By optimizing the parameters of fitness function 1 through genetic algorithm, we seek to make Parameter combination 1 with the minimum value; By optimizing the parameters of fitness function 2 through genetic algorithm, we seek to make The minimum parameter combination 2 is obtained, which is used to indicate that both the network error and the control input are under control; The method of optimizing the parameters of the first fitness function and the second fitness function by a genetic algorithm to seek an optimal solution and modifying corresponding parameters of the corresponding type of scene parameter template according to the optimal solution includes: Get the optimized parameter combination ; Get the optimized parameter combination 2 ; because , modify the switching threshold in the parameter template to adjust the total number of switching attempts; because , by adjusting the cell reselection priority and neighbor relationship configuration in the parameter template, the switching speed can be improved and the switching delay can be reduced.
9. A multi-frequency operator network parameter intelligent coordination and adjustment system, It is characterized in that The system is used to execute a multi-frequency operator network parameter intelligent coordination and adjustment method as described in any one of claims 1-8 above.
10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement a multi-frequency operator network parameter intelligent coordination and adjustment method as described in any one of claims 1 to 8.
Citation Information
Patent Citations
Cell parameter configuration method and system
CN101998419A
Method and system for optimizing network
CN101998465A
Method and device for data service shunting
CN105828401A
LTE new network access anchor point cell parameter configuration method and device under 5G NSA networking
CN112243249A
Switching parameter configuration method for three-dimensional space mobile unmanned aerial vehicle user
CN113993099A
Cited By
Switching method of ground communication network and satellite communication system
CN119324737A
