Wireless capacity optimization method, device, system and storage medium based on large model

Abstract

The embodiment of the present application provides a wireless capacity optimization method, device, system and storage medium based on a large model, which relates to the field of mobile communication technology. The method includes: obtaining wireless network related data of a first area to which a first cell belongs, the first cell is a cell where an abnormal wireless network indicator occurs, and the first area includes multiple cells; using the wireless network related data of the first area as input data of the first model to obtain a network element snapshot of the first area generated by the first model; using the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization solution for the first area output by the first model. The embodiment of the present application can shorten the time for locating wireless network problems, improve the accuracy of network optimization solutions, avoid the impact of unreasonable optimization solutions on network quality, realize end-to-end closed loop of wireless network optimization, and improve wireless network optimization efficiency and wireless network quality.

Description

Technical Field

[0001] The present application relates to the field of mobile communication technology, and in particular to a large model-based wireless capacity optimization method, device, system and storage medium. Background Art

[0002] Wireless communication network optimization is the key to improving network service quality and an important way to enhance network competitiveness. Current wireless optimization solutions mainly optimize wireless networks manually or through small models. Among them, wireless network optimization through manual methods is usually based on the wireless business experience accumulated by network optimization experts over many years. It uses IT means to identify wireless problems by setting thresholds. Network optimization experts further analyze the causes of the problems manually and give corresponding optimization solutions based on the current network status. Wireless network optimization through small models is usually targeted at specific wireless scenarios, specific regions, and specific businesses. Different small models are used to analyze and identify the effectiveness of the problem, the root cause, and adjust the value of the specified parameter according to the required indicators in the scenario.

[0003] Manual wireless network optimization relies heavily on expert experience and has shortcomings in both optimization efficiency and optimization effect. When wireless network optimization is performed through small models, due to the limited generalization ability of small models, in actual applications, the wireless network environment is complex and changeable, and small models are difficult to accurately predict and adapt to all possible network conditions. In addition, the computing power resources of small models are limited, and the performance of optimization algorithms is limited, resulting in shortcomings in optimization efficiency and optimization effect. Summary of the invention

[0004] At least one embodiment of the present application provides a large-model-based wireless capacity optimization method, device, system and storage medium, which are used to solve the problem of insufficient optimization efficiency and optimization effect of wireless network optimization methods in the prior art.

[0005] In order to solve the above technical problems, this application is implemented as follows:

[0006] In a first aspect, an embodiment of the present application provides a wireless capacity optimization method based on a large model, including:

[0007] Acquire wireless network related data of a first area to which a first cell belongs, the first cell being a cell where an abnormal wireless network indicator occurs, and the first area includes a plurality of cells;

[0008] Using the wireless network related data of the first area as input data of the first model, and obtaining a network element snapshot of the first area generated by the first model;

[0009] Using the network element snapshot of the first area as input data of the first model, and obtaining a wireless network optimization solution for the first area output by the first model;

[0010] The first model is a large model trained using wireless network related data of the mobile network to which the first cell belongs.

[0011] Optionally, the wireless network related data includes at least one of the following:

[0012] Cell performance data;

[0013] Key performance indicator (KPI) data of the community;

[0014] The cell's network resource management (NRM) data;

[0015] Measurement report MR data of the cell;

[0016] Parameter adjustment record of the cell.

[0017] Optionally, the wireless network optimization solution for the first area includes at least one of the following:

[0018] An adjustment plan for cell-level parameters in the first area;

[0019] an adjustment scheme for neighboring cell level parameters in the first cell;

[0020] An adjustment scheme for base station-level parameters in the first cell.

[0021] Optionally, the above method further includes:

[0022] Acquire wireless network related data of the first cell;

[0023] Using the wireless network related data of the first cell as input data of the first model, to obtain a network element snapshot of the first cell generated by the first model;

[0024] The network element snapshot of the first cell is used as input data of the first model to obtain a disturbance analysis result output by the first model as to whether the wireless network anomaly of the first cell is a disturbance problem.

[0025] Optionally, the network element snapshot of the first cell is a feature vector generated by the first model based on wireless network related data of the first cell;

[0026] The network element snapshot of the first area is a feature vector generated by the first model based on the wireless network related data of the first area.

[0027] Optionally, the above method further includes:

[0028] When the disturbance analysis result indicates that the wireless network abnormality of the first cell is not a disturbance problem, the step of obtaining wireless network related data of the first area to which the first cell belongs is performed.

[0029] Optionally, the above method further includes:

[0030] When the disturbance analysis result indicates that the wireless network anomaly of the first cell is not a disturbance problem, the network element snapshot of the first cell is used as input data of the first model to obtain the root cause positioning result of the wireless network anomaly of the first cell output by the first model.

[0031] Optionally, using the network element snapshot of the first area as input data of the first model to obtain the wireless network optimization solution for the first area output by the first model includes:

[0032] The root cause location result of the wireless network anomaly of the first cell and the network element snapshot of the first area are used together as input data of the first model to obtain the wireless network optimization solution for the first area output by the first model.

[0033] Optionally, the root cause positioning result of the wireless network abnormality of the first cell includes multiple levels of root causes, wherein the first level root cause includes one of wireless capacity problems, wireless coverage problems, and wireless interference problems; the second level root cause includes one of high load and low traffic; the third level root cause includes unreasonable power parameter settings, unreasonable load balancing parameter settings, unreasonable mobility parameter settings, unreasonable neighboring cell parameter settings, unreasonable capacity settings, unreasonable energy-saving parameter settings, small number of users, and other reasons that require on-site investigation.

[0034] Optionally, the above method further includes:

[0035] The first area is optimized according to the wireless network optimization solution of the first area.

[0036] Optionally, the above method further includes:

[0037] Using the wireless network optimization solution of the first area and the network element snapshot of the first area as input data of the first model, to obtain the wireless network optimization result of the first area predicted by the first model;

[0038] When the wireless network optimization result of the first area meets the requirements, the first area is optimized according to the wireless network optimization solution of the first area.

[0039] Optionally, the above method further includes:

[0040] Obtain a neighboring cell relationship table of a mobile network to which the first cell belongs, where the neighboring cell relationship table includes a one-hop neighboring cell of each cell;

[0041] According to the neighboring cell relationship table, a first area to which the first cell belongs is determined, wherein the first area includes the first cell and neighboring cells within N hops of the first cell, where N is an integer greater than or equal to 1.

[0042] Optionally, the acquiring a neighboring cell relationship table of a mobile network to which the first cell belongs includes:

[0043] Obtaining handover data of each cell of the mobile network to which the first cell belongs within a preset historical time period, the handover data recording the number of handovers from the current cell to other cells; obtaining measurement report MR data received by each cell, the MR data including location information of a terminal reporting the MR data; determining the overlapping area between any two cells according to the MR data;

[0044] Determine, according to the handover data of the current cell, the number of handovers from the current cell to each other cell, and sort the other cells in descending order of the number of handovers to obtain a queue composed of the other cells; wherein, when at least two other cells have the same number of handovers, sort the at least two other cells according to the size of the overlapping area between the at least two other cells and the current cell;

[0045] The first M other cells are selected from the queue as one-hop neighboring cells of the current cell, and a neighboring cell relationship table of the current cell is obtained.

[0046] In a second aspect, an embodiment of the present application provides a wireless capacity optimization device based on a large model, including:

[0047] A first acquisition module is used to acquire wireless network related data of a first area to which a first cell belongs, wherein the first cell is a cell where an abnormal wireless network indicator occurs, and the first area includes a plurality of cells;

[0048] A first processing module, configured to use the wireless network related data of the first area as input data of a first model to obtain a network element snapshot of the first area generated by the first model;

[0049] A second processing module uses the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization solution for the first area output by the first model;

[0050] The first model is a large model trained using wireless network related data of the mobile network to which the first cell belongs.

[0051] Optionally, the wireless network related data includes at least one of the following:

[0052] Cell performance data;

[0053] Key performance indicator (KPI) data of the community;

[0054] The cell's network resource management (NRM) data;

[0055] Measurement report MR data of the cell;

[0056] Parameter adjustment record of the cell.

[0057] Optionally, the wireless network optimization solution for the first area includes at least one of the following:

[0058] An adjustment plan for cell-level parameters in the first area;

[0059] an adjustment scheme for neighboring cell level parameters in the first cell;

[0060] An adjustment scheme for base station-level parameters in the first cell.

[0061] Optionally, the above device further comprises:

[0062] A second acquisition module, used to acquire wireless network related data of the first cell;

[0063] A third processing module, configured to use the wireless network related data of the first cell as input data of the first model to obtain a network element snapshot of the first cell generated by the first model;

[0064] The fourth processing module is used to use the network element snapshot of the first cell as input data of the first model to obtain a disturbance analysis result of whether the wireless network anomaly of the first cell is a disturbance problem output by the first model.

[0065] Optionally, the network element snapshot of the first cell is a feature vector generated by the first model based on wireless network related data of the first cell;

[0066] The network element snapshot of the first area is a feature vector generated by the first model based on the wireless network related data of the first area.

[0067] Optionally, the above device further comprises:

[0068] The first control module is used to trigger the first acquisition module to acquire wireless network related data of a first area to which the first cell belongs when the disturbance analysis result indicates that the wireless network abnormality of the first cell is not a disturbance problem.

[0069] Optionally, the above device further comprises:

[0070] The second control module is used to use the network element snapshot of the first cell as input data of the first model when the disturbance analysis result indicates that the wireless network anomaly of the first cell is not a disturbance problem, obtain the root cause positioning result of the wireless network anomaly of the first cell output by the first model, and trigger the first acquisition module to obtain the wireless network related data of the first area to which the first cell belongs.

[0071] Optionally, the second processing module is also used to use the root cause positioning result of the wireless network anomaly in the first cell and the network element snapshot of the first area as input data of the first model to obtain the wireless network optimization plan for the first area output by the first model.

[0072] Optionally, the root cause positioning result of the wireless network abnormality of the first cell includes multiple levels of root causes, wherein the first level root cause includes one of wireless capacity problems, wireless coverage problems, and wireless interference problems; the second level root cause includes one of high load and low traffic; the third level root cause includes unreasonable power parameter settings, unreasonable load balancing parameter settings, unreasonable mobility parameter settings, unreasonable neighboring cell parameter settings, unreasonable capacity settings, unreasonable energy-saving parameter settings, small number of users, and other reasons that require on-site investigation.

[0073] Optionally, the above device further comprises:

[0074] The first optimization module is used to optimize the first area according to the wireless network optimization plan of the first area.

[0075] Optionally, the above device further comprises:

[0076] a fifth processing module, configured to use the wireless network optimization solution of the first area and the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization result of the first area predicted by the first model;

[0077] The second optimization module is used to optimize the first area according to the wireless network optimization plan of the first area when the wireless network optimization result of the first area meets the requirements.

[0078] Optionally, the above device further comprises:

[0079] A second acquisition module is used to obtain a neighbor relationship table of the mobile network to which the first cell belongs, the neighbor relationship table including a one-hop neighbor of each cell; and determine a first area to which the first cell belongs according to the neighbor relationship table, wherein the first area includes the first cell and neighboring areas within N hops of the first cell, where N is an integer greater than or equal to 1.

[0080] Optionally, the above device further comprises:

[0081] A third acquisition module is used to obtain switching data of each cell of the mobile network to which the first cell belongs within a preset historical time period, the switching data recording the number of switchings from the current cell to other cells; obtain measurement report MR data received by each cell, the MR data including location information of the terminal reporting the MR data; determine the overlapping area between any two cells according to the MR data; determine the number of switchings from the current cell to each other cell according to the switching data of the current cell, sort the other cells in descending order according to the number of switchings, and obtain a queue composed of the other cells; wherein, when at least two other cells have the same number of switchings, the at least two other cells are sorted according to the size of the overlapping area between the at least two other cells and the current cell; select the first M other cells from the queue as 1-hop neighboring cells of the current cell, and obtain a neighboring cell relationship table of the current cell.

[0082] In the third aspect, an embodiment of the present application provides a large model-based wireless network optimization system, comprising a transceiver, a processor, a memory, and a program or instruction stored in the memory and executable on the processor; when the processor executes the program or instruction, the steps of the method described in any one of the first aspects are implemented.

[0083] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in any one of the first aspects are implemented.

[0084] In a fifth aspect, an embodiment of the present application provides a computer program product, comprising computer instructions, which, when executed by a processor, implement the steps of the method described in any one of the first aspects.

[0085] Compared with the prior art, the wireless capacity optimization method, system, medium and program product based on a large model provided in the embodiments of the present application generate a wireless network optimization plan based on the wireless network related data of the area to which the problem cell belongs through the large model, which can shorten the time for locating wireless network problems, improve the accuracy of the network optimization plan, avoid the impact of unreasonable optimization plans on network quality, realize end-to-end closed loop of wireless network optimization, and improve wireless network optimization efficiency and wireless network quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0087] Figure 1 A schematic diagram of an application scenario of an embodiment of the present application;

[0088] Figure 2 A flow chart of a wireless capacity optimization method or a wireless network optimization method based on a large model according to an embodiment of the present application;

[0089] Figure 3 Another flow chart of a wireless capacity optimization method or a wireless network optimization method based on a large model according to an embodiment of the present application;

[0090] Figure 4 It is another flow chart of the wireless capacity optimization method or wireless network optimization method based on a large model according to an embodiment of the present application;

[0091] Figure 5 A schematic diagram of a technical architecture of a wireless capacity optimization method or a wireless network optimization method based on a large model according to an embodiment of the present application;

[0092] Figure 6 A schematic diagram of the interaction between the network optimization platform and the big model in the wireless capacity optimization method or wireless network optimization method based on the big model in the embodiment of the present application;

[0093] Figure 7 An example diagram of problem work order optimization according to an embodiment of the present application;

[0094] Figure 8 A schematic diagram of the structure of a wireless capacity optimization device or a wireless network optimization device based on a large model according to an embodiment of the present application;

[0095] Fig. 9 Another structural schematic diagram of a large model-based wireless capacity optimization device or a wireless network optimization device according to an embodiment of the present application. DETAILED DESCRIPTION

[0096] The terms "first", "second", etc. of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way are interchangeable where appropriate, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of one type, and the number of objects is not limited, for example, the first object can be one or more. In addition, "or" in the present application represents at least one of the connected objects. For example, "A or B" covers three schemes, namely, Scheme 1: including A but not including B; Scheme 2: including B but not including A; Scheme 3: including both A and B. The character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0097] The term "indication" in this application can be a direct indication (or explicit indication) or an indirect indication (or implicit indication). A direct indication can be understood as the sender explicitly informing the receiver of specific information, operations to be performed, or request results, etc. in the sent indication; an indirect indication can be understood as the receiver determining the corresponding information according to the indication sent by the sender, or making a judgment and determining the operation to be performed or the request result, etc. according to the judgment result.

[0098] It is worth noting that the technology described in the embodiments of the present application is not limited to the Long Term Evolution (LTE) / LTE-Advanced (LTE-A) system, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency Division Multiple Access (SC-FDMA) or other systems. The terms "system" and "network" in the embodiments of the present application are often used interchangeably, and the described technology can be used for the above-mentioned systems and radio technologies as well as other systems and radio technologies. The following description describes a New Radio (NR) system for example purposes, and NR terms are used in most of the following descriptions, but these technologies can also be applied to systems other than NR systems, such as 6th Generation (6G) communication systems.

[0099] Figure 1A block diagram of a wireless communication system applicable to the embodiment of the present application is shown. The wireless communication system includes a terminal 11 and a network side device 12. Among them, the terminal 11 can be a mobile phone, a tablet computer (Tablet Personal Computer), a laptop computer (Laptop Computer), a notebook computer, a personal digital assistant (Personal Digital Assistant, PDA), a handheld computer, a netbook, an ultra-mobile personal computer (Ultra-mobile Personal Computer, UMPC), a mobile Internet device (Mobile Internet Device, MID), an augmented reality (Augmented Reality, AR), a virtual reality (Virtual Reality, VR) device, a robot, a wearable device (Wearable Device), an aircraft (flight vehicle), a vehicle-mounted device (Vehicle User Equipment, VUE), a ship-mounted device, a pedestrian terminal (Pedestrian User Equipment, PUE), a smart home (home appliances with wireless communication functions, such as refrigerators, televisions, washing machines or furniture, etc.), a game console, a personal computer (Personal Computer, PC), a teller machine or a self-service machine and other terminal side devices. Wearable devices include: smart watches, smart bracelets, smart headphones, smart glasses, smart jewelry (smart bracelets, smart bracelets, smart rings, smart necklaces, smart anklets, smart anklets, etc.), smart wristbands, smart clothing, etc. Among them, the vehicle-mounted device can also be called a vehicle-mounted terminal, a vehicle-mounted controller, a vehicle-mounted module, a vehicle-mounted component, a vehicle-mounted chip or a vehicle-mounted unit, etc. It should be noted that the specific type of the terminal 11 is not limited in the embodiment of the present application. The network side device 12 may include an access network device or a core network device, wherein the access network device may also be referred to as a radio access network (Radio Access Network, RAN) device, a radio access network function or a radio access network unit. The access network device may include a base station, a wireless local area network (Wireless Local Area Network, WLAN) access point (Access Point, AP) or a wireless fidelity (Wireless Fidelity, WiFi) node, etc.Among them, the base station can be called Node B (Node B, NB), Evolved Node B (Evolved Node B, eNB), the next generation Node B (the next generation Node B, gNB), New Radio Node B (New Radio Node B, NR Node B), access point, Relay Base Station (RBS), Serving Base Station (SBS), Base Transceiver Station (Base Transceiver Station, BTS), radio base station, radio transceiver, Basic Service Set (Basic Service Set, BSS), Extended Service Set (Extended Service Set, ESS), home Node B (home evolved Node B, HNB), home evolved Node B (home evolved Node B), Transmission Reception Point (Transmission Reception Point, TRP) or other appropriate terms in the field. As long as the same technical effect is achieved, the base station is not limited to specific technical vocabulary. It should be noted that in the embodiments of the present application, only the base station in the NR system is taken as an example for introduction, and the specific type of the base station is not limited.

[0100] In view of the deficiencies of the wireless network optimization solutions in the prior art, the embodiments of the present application provide a wireless capacity optimization method based on a large model, which can optimize the wireless network, improve the optimization efficiency and optimization effect of the wireless network, and thus improve the quality of the wireless network.

[0101] Please refer to Figure 2 , a wireless capacity optimization method based on a large model provided in an embodiment of the present application includes:

[0102] Step 21, obtaining wireless network related data of a first area to which a first cell belongs, wherein the first cell is a cell where wireless network indicators are abnormal, and the first area includes a plurality of cells.

[0103] Here, the first cell may be a cell with a wireless network abnormality. Specifically, the cell with a wireless network abnormality may be determined based on wireless network indicators and / or user complaint information. The wireless network indicators may be indicators such as cell performance data and alarm data. In step 21, the embodiment of the present application determines the first area to which the first cell belongs, and then obtains wireless network-related data of the first area. Typically, the first area may include multiple cells including the first cell. That is, when an abnormality occurs in the first cell and needs to be optimized, the embodiment of the present application performs overall optimization on an area including the first cell, thereby avoiding abnormalities in other cells after the first cell is optimized.

[0104] The wireless network related data of the first area may generally include at least one of the following data:

[0105] (1) Cell performance data, which may include one or more of the following: number of successful intra-system handovers, number of inter-system handovers, number of RRC connection establishment requests, maximum number of RRC connections, carrier shutdown duration, channel shutdown duration, etc.

[0106] (2) Key Performance Indicator (KPI) data of the cell, which may include one or more of the following: uplink average PRB utilization, downlink average PRB utilization, PDCCH channel CCE occupancy rate, wireless connection rate, handover success rate, uplink traffic, downlink traffic, etc.

[0107] (3) Network Resource Management (NRM) data of the cell, which may include one or more of the following: cell DN information, special information, list of associated cells, station type, frequency number, etc.

[0108] (4) Measurement Report (MR) data of the cell, which may include one or more of the following: number of sampling points under cell-level RSRP segment, number of sampling points under cell-level TA segment, number of good sampling points at the cell level, total number of sampling points at the cell level, number of neighboring cell sampling points at the neighboring cell level, average RSRP of neighboring cells, etc.

[0109] (5) Parameter adjustment records of a cell, which may include one or more of the following: the adjusted parameter, adjustment time, parameter values ​​before and after adjustment, etc.

[0110] When determining the first area to which the first cell belongs, the embodiment of the present application can obtain a neighbor relationship table of the mobile network to which the first cell belongs, and the neighbor relationship table includes a 1-hop neighbor of each cell. Then, according to the neighbor relationship table, the first area to which the first cell belongs is determined, wherein the first area includes the first cell and the neighboring areas within N hops of the first cell, and N is an integer greater than or equal to 1. For example, when N is 2, it means that the first cell and the neighboring areas within 2 hops of the first cell are taken as the first area.

[0111] A specific implementation method of obtaining the neighboring area relationship table is as follows:

[0112] Obtaining handover data of each cell of the mobile network to which the first cell belongs within a preset historical time period, wherein the handover data records the number of handovers from the current cell to other cells;

[0113] Acquire measurement report MR data received by each cell, the MR data including location information of a terminal reporting the MR data; determine the overlapping area between any two cells according to the MR data;

[0114] Determine, according to the handover data of the current cell, the number of handovers from the current cell to each other cell, and sort the other cells in descending order of the number of handovers to obtain a queue composed of the other cells; wherein, when at least two other cells have the same number of handovers, sort the at least two other cells according to the size of the overlapping area between the at least two other cells and the current cell;

[0115] The first M other cells are selected from the queue as the one-hop neighboring cells of the current cell, thereby obtaining a neighboring cell relationship table of the current cell.

[0116] The determining, according to the neighbor relationship table, the first area to which the first cell belongs may specifically include: querying the neighbor relationship table of the first cell to obtain the 1-hop neighbor of the first cell; querying the neighbor relationship table of each 1-hop neighbor of the first cell to obtain the 1-hop neighbor of each 1-hop neighbor; obtaining a cell set consisting of the 1-hop neighbor of the first cell and the 1-hop neighbor of each 1-hop neighbor of the first cell, to obtain the first area to which the first cell belongs.

[0117] Step 22: Using the wireless network related data of the first area as input data of the first model, and obtaining a network element snapshot of the first area generated by the first model.

[0118] Here, the first model is a large model trained using the wireless network related data of the mobile network to which the first cell belongs. One function of the large model is to generate a wireless network optimization solution for the area based on the wireless network related data of the area. In step 22, the embodiment of the present application inputs the wireless network related data of the first area into the first model, and the first model generates a network element snapshot of the first area based on the wireless network related data of the first area. Specifically, the network element snapshot of the first area is a feature vector generated by the first model based on the wireless network related data of the first cell.

[0119] Step 23: Using the network element snapshot of the first area as input data of the first model, and obtaining the wireless network optimization solution of the first area output by the first model.

[0120] Here, the first model generates a wireless network optimization solution for the first area based on the network element snapshot of the first area. Specifically, the wireless network optimization solution for the first area includes at least one of the following:

[0121] (1) A plan for adjusting the cell-level parameters in the first area;

[0122] (2) an adjustment plan for neighboring cell-level parameters within the first cell;

[0123] (3) A scheme for adjusting base station-level parameters in the first cell.

[0124] Among them, the cell-level parameters include the main wireless network parameters of the air interface, which include but are not limited to network identification parameters, idle mode parameters, call control parameters, cell selection and reselection parameters, such as same-frequency switching amplitude hysteresis, same-frequency switching offset, same-frequency switching time hysteresis and other parameters. The neighboring cell-level parameters mainly include switching and reselection configuration parameters between cells, which include but are not limited to querying hetero-frequency switching parameter groups, same-frequency switching parameter groups, hetero-frequency adjacent frequency points, same-frequency neighboring cell relationships, and cell offsets (Cell Individual Offset, CIO) in same-frequency neighboring cell relationships. The base station-level parameters mainly involve the configuration and management parameters of the base station, usually including parameters such as base station coverage, transmission power, data transmission priority and security.

[0125] Through the above steps, the embodiment of the present application generates a wireless network optimization plan for the first area to which the first cell belongs through a large model based on the wireless network related data of the first area. The wireless optimization plan can cope with various scenarios of wireless network abnormalities by adjusting the parameters at the cell level, neighboring cell level, and base station level in the first area. The various scenarios include but are not limited to wireless capacity abnormalities, wireless coverage abnormalities, and wireless interference abnormalities, and realize wireless network optimization including wireless capacity optimization, wireless coverage optimization, and wireless interference optimization. Therefore, the wireless capacity optimization method and wireless capacity optimization device provided in the embodiment of the present application can also be referred to as a wireless network optimization method and a wireless capacity optimization device.

[0126] The first cell's wireless network indicators are abnormal, which may be a disturbance problem caused by occasional fluctuations in related indicators. For disturbance problems, usually no attention is needed, and the abnormal indicators can recover on their own. Non-disturbance problems are poor quality problems that require attention in the wireless network. At this time, the root cause of the wireless network abnormality can be further determined by root cause location.

[0127] Another function of the large model of the embodiment of the present application is to determine whether the wireless network anomaly of the cell is a disturbance problem based on the wireless network related data of the cell where the wireless network indicator is abnormal. Figure 3 As shown, the above method of the embodiment of the present application may further include the following steps before step 21:

[0128] Step 31: Acquire wireless network related data of the first cell.

[0129] Step 32: Use the wireless network related data of the first cell as input data of the first model to obtain a network element snapshot of the first cell generated by the first model. Here, the network element snapshot of the first cell is a feature vector generated by the first model based on the wireless network related data of the first cell.

[0130] Step 33: Using the network element snapshot of the first cell as input data of the first model, and obtaining a disturbance analysis result output by the first model as to whether the wireless network anomaly of the first cell is a disturbance problem.

[0131] In the embodiment of the present application, if the disturbance analysis result in step 33 indicates that the wireless network abnormality of the first cell is not a disturbance problem, the following can be performed: Figure 2 Steps 21-23 shown in the figure are performed to generate a wireless network optimization plan for the first area; if the disturbance analysis result in step 33 indicates that the wireless network anomaly of the first cell is a disturbance problem, the process can be terminated at this time and no longer executed. Figure 2 Steps shown.

[0132] The large model of the embodiment of the present application also has the following functions: based on the wireless network related data of the cell where the wireless network indicators are abnormal, generating the root cause positioning result of the wireless network abnormality of the cell, that is, obtaining the root cause or reason of the wireless network abnormality of the cell.

[0133] As another embodiment, in the embodiment of the present application, when the disturbance analysis result indicates that the wireless network anomaly of the first cell is not a disturbance problem, the network element snapshot of the first cell is used as the input data of the first model to obtain the root cause positioning result of the wireless network anomaly of the first cell output by the first model. Furthermore, the large model can also generate the wireless network optimization result of the first area based on the root cause positioning result of the wireless network anomaly of the first cell and the network element snapshot of the first area.

[0134] Specifically, the root cause positioning result of the wireless network abnormality of the first cell includes multiple levels of root causes, wherein the first level root cause includes one of wireless capacity problem, wireless coverage problem, and wireless interference problem, the second level root cause includes one of high load and low traffic, and the third level root cause includes one of unreasonable power parameter setting, unreasonable load balancing parameter setting, unreasonable mobility parameter setting, unreasonable neighboring cell pair parameter setting, insufficient cell capacity, unreasonable energy-saving parameter setting, small number of users, and other reasons that require on-site investigation. Table 1 provides an example of the second and third level root causes when the first level root cause is a wireless capacity problem.

[0135] Table 1

[0136]

[0137] Therefore, another embodiment of the present application, such as Figure 4 As shown, in Figure 2 Before step 21 shown, the following steps may also be included:

[0138] Step 41: Acquire wireless network related data of the first cell.

[0139] Step 42: Use the wireless network related data of the first cell as input data of the first model to obtain a network element snapshot of the first cell generated by the first model. Here, the network element snapshot of the first cell is a feature vector generated by the first model based on the wireless network related data of the first cell.

[0140] Step 43: Using the network element snapshot of the first cell as input data of the first model, and obtaining a disturbance analysis result output by the first model as to whether the wireless network anomaly of the first cell is a disturbance problem.

[0141] Step 44, when the disturbance analysis result indicates that the wireless network abnormality of the first cell is not a disturbance problem, the network element snapshot of the first cell is used as input data of the first model to obtain the root cause positioning result of the wireless network abnormality of the first cell output by the first model.

[0142] After obtaining the root cause location result of the wireless network abnormality of the first cell, execute Figure 2 In addition, when executing the above step 23, the root cause location result of the wireless network anomaly of the first cell and the network element snapshot of the first area can be used as input data of the first model to obtain the wireless network optimization solution of the first area output by the first model.

[0143] Step 45: If the disturbance analysis result indicates that the wireless network anomaly of the first cell is a disturbance problem, the process can be terminated and no further execution is performed. Figure 2 Steps shown.

[0144] After obtaining the wireless network optimization plan for the first area, the embodiment of the present application can optimize the first area according to the wireless network optimization plan for the first area. For example, the adjustment plan of the cell-level parameters, neighboring cell-level parameters, and base station-level parameters in the first area is sent to the corresponding network element to configure the working parameters of the network element, thereby optimizing the wireless network.

[0145] Optionally, the large model of the embodiment of the present application can also predict the wireless network optimization result based on the wireless network optimization scheme of the first area. At this time, before optimizing the first area, the embodiment of the present application can also use the wireless network optimization scheme of the first area and the network element snapshot of the first area as input data of the first model to obtain the wireless network optimization result of the first area predicted by the first model. Then, if the wireless network optimization result of the first area meets the preset requirements, the first area is optimized according to the wireless network optimization scheme of the first area. If the wireless network optimization result of the first area does not meet the preset requirements, the above optimization may not be performed, or the wireless network optimization scheme of the first area may be regenerated by the large model until a wireless network optimization scheme that meets the requirements is generated.

[0146] From the above, it can be seen that the embodiment of the present application adopts a large model to achieve end-to-end optimization of the wireless network. The large model refers to a machine learning model with large-scale parameters and complex computing structures. For example, the embodiment of the present application can learn the general representation method of network status (network element snapshot) based on network structured data, and can build a large model of billion-level network structured data, and deeply learn the intrinsic connection and interaction between parameters and indicators. Based on the large model, the network data such as wireless resources, industrial parameters, parameter change records, alarms, performance, MR, Minimization of Drive-tests (MDT) can be integrated to realize the abnormal disturbance problem identification of wireless network area, the root cause location of wireless network abnormality, the generation of wireless network optimization plan, and the effect estimation of wireless network optimization plan, which shortens the time of wireless network problem location, improves the accuracy of network optimization plan, avoids the impact of unreasonable optimization plan on network quality, realizes the end-to-end closed loop of wireless network optimization, and improves the efficiency of wireless network optimization and wireless network quality.

[0147] The big model of the embodiment of the present application can realize various functions such as disturbance problem identification, root cause location of wireless network anomalies, wireless network optimization solution generation, and wireless network optimization solution effect estimation based on the network element snapshot of the cell or area. When using the big model, the relevant functions of the big model can be called through relevant interfaces and / or instructions to obtain the results generated by the big model.

[0148] Each function of the large model is pre-trained using the wireless network related data of the mobile network to which the first cell belongs. Wherein, in the above steps 33 / 43, when the large model determines whether the wireless network anomaly of the first cell is a disturbance problem based on the network element snapshot of the first cell, and in the above step 44, when the large model determines the root cause location result of the wireless network anomaly of the first cell based on the network element snapshot of the first cell, the large model performs classification tasks. In the above step 23, when the large model obtains the wireless network optimization plan of the first area based on the network element snapshot of the first area, the large model performs the optimization task; when the large model predicts the wireless network optimization result based on the wireless network optimization plan of the first area, it performs the prediction task. The embodiment of the present application can call the large model to perform related tasks through the arrangement design of related interfaces to obtain corresponding results.

[0149] The various functions of the large model described in the embodiment of the present application are all obtained by training based on the wireless network related data of the mobile network to which the first cell belongs. Specifically, supervised training can be used, and the training data includes wireless network related data and its corresponding annotation information. The annotation information can be information obtained by annotation based on expert experience, or information obtained by annotation based on the historical status of the wireless network.

[0150] Figure 5 A technical architecture diagram of a wireless capacity optimization method or a wireless network optimization method based on a large model according to an embodiment of the present application is provided. Figure 5 It demonstrates the complete end-to-end optimization process of wireless networks connected to production, from data dictionary connection to data collection and cleaning, triggering the large model to output intelligent decision-making solutions, and finally applying the intelligent decision-making solutions to the existing network.

[0151] like Figure 5 As shown, the embodiment of the present application is designed with multiple scenarios (including planning, construction, maintenance, optimization), multiple tasks (including supporting algorithms, processes, capabilities), multiple topics (including indicators, parameters, measurements, topologies, events), etc. of wireless network operation and maintenance optimization as the goal, and the data dictionary used by the large model is designed. The data dictionary is used to collect various data required for training the large model, including wireless work order data of the work order system, network optimization data of network optimization platforms at all levels (such as PM data, alarm data, KPI data, etc. with a granularity of 15 minutes, work parameter data, NRM data, soft parameter data, etc. with a granularity of natural days (days), MR data of the MR platform (such as cell-level statistical MR data and cell-pair statistical MR data with a granularity of 15 minutes, etc.), and parameter adjustment record data obtained by the parameter platform (for example, in units of days, the parameter adjustment record of the previous day is obtained every day).

[0152] Figure 5 The data collection and cleaning function in the data dictionary collects and cleans various data in the data dictionary. It can collect data in real time or non-real time, and store the cleaned data in the database for large model training or reasoning. For example, the training data of the large model can be generated based on historical data, and the large model can be trained to obtain a trained large model.

[0153] When using the large model for reasoning, the network optimization platform can call the large model to perform corresponding tasks through the interface between the large model and the large model to generate a wireless network optimization plan (intelligent decision-making plan). The results output by the large model can be transmitted to the network optimization platform through the automation interface, and the parameters that need to be adjusted in the intelligent decision-making plan in the results can be further sent to the devices in the existing network through the wireless operation and maintenance workbench. After the successful issuance, the decision-making plan acts on the existing network equipment to restore the indicators to normal to solve the current problem. The updated indicator data will be input into the large model by the data collection and cleaning function for model iteration optimization and data dictionary optimization.

[0154] Figure 6 Another optimization process for wireless network optimization is provided, in which the network optimization platform schedules the disturbance analysis task of the large model to identify whether the abnormal problem is a disturbance problem. For non-disturbance problems, the network optimization platform will further schedule the root cause location task of the large model to output the detailed root cause. For the root cause, the network optimization platform will schedule the decision control task of the large model to output the final optimization plan. Finally, the network optimization platform schedules the prediction task of the large model to predict the change trend of business indicators after the decision plan is executed in the existing network. Figure 6 The implementation process shown is as follows:

[0155] Step a: The network optimization platform receives a wireless problem ticket

[0156] The network optimization platform extracts necessary information from wireless problem work orders, including the Cell Global Identity (CGI) of the problem cell, the time range when the wireless problem occurred, and other information.

[0157] Step b: Get problem cell data

[0158] The network optimization platform queries and obtains relevant data based on the CGI of the problem cell and the time range when the wireless problem occurred, including performance, KPI, NRM, MR, parameter adjustment records and other data.

[0159] Step c: Generate a network element snapshot

[0160] The cell data obtained by the network optimization platform includes structured data (cell KPI data, cell MR, parameters, some resources, etc.) and unstructured data (some resources, etc.). Unstructured data is usually some text data, including discrete text data (a limited number of base station types: {macro base station, indoor base station, micro base station}) and continuous text data (multiple coverage scenarios, such as residential areas, scenic spots, factory areas, schools, hospitals, etc.). Encoding is performed for this type of data. Different types of data usually use different encoding methods. The data of the cell is divided according to the preset time granularity (such as 15 minutes), so that the network is in a relatively stable state within this time granularity. The data within this time granularity is encoded to form a vector as the state data of the cell at this time granularity. Take the state data of the cell within a period of time (usually including the state data of multiple time granularities), call the network element snapshot capability of the large model, and you can get a high-dimensional vector representation of the cell state data within this period of time, that is, the network element snapshot.

[0161] Step d: Perturbation analysis

[0162] Based on the network element snapshot, the large model classification task is used to perform disturbance analysis and judgment to identify whether the current problem is a disturbance problem or a non-disturbance problem caused by occasional fluctuations in the network element indicators. Disturbance problems do not need to be paid attention to, and abnormal indicators will usually recover on their own. Non-disturbance problems are poor quality problems that need to be solved in wireless networks, and the next step is to locate the root cause.

[0163] Step e: Root cause location

[0164] Based on the network element snapshot, the root cause is located by calling the classification task of the large model. For example, a detailed root cause of the wireless capacity problem that can be located is shown in Table 1. If the third-level root cause is unreasonable setting of related parameters, the next step of decision optimization can be entered. If the third-level root cause is other reasons other than unreasonable setting of related parameters, no decision optimization is required.

[0165] Step f: Decision Optimization

[0166] The decision optimization of the large model is based on the problem cell and the surrounding cells to find the best parameter adjustment plan for the area. The multi-network element joint optimization process is as follows:

[0167] (1) Generate a neighbor table

[0168] The neighboring cells of each cell are determined based on the number of switching times in the performance data table, and a neighboring cell relationship table of each cell is formed. Specifically, the number of switching times from the current cell to each other cell is determined based on the switching data of the current cell, and then the other cells are sorted in descending order according to the number of switching times to obtain a queue composed of the other cells; wherein, when at least two other cells have the same number of switching times, the at least two other cells are sorted according to the size of the overlapping area between the at least two other cells and the current cell. Then, the first M other cells are selected from the queue as the one-hop neighboring cells of the current cell to obtain the neighboring cell relationship table of the current cell.

[0169] (2) Obtain the area to which the problem cell belongs

[0170] Here, the cells within N hops of the problem cell can be used as the area to which the problem cell belongs. For example, when N is equal to 2, the neighboring area relationship table of the problem cell is queried to obtain the 1-hop neighboring area of ​​the problem cell; then, for each 1-hop neighboring area, the neighboring area relationship table of the 1-hop neighboring area is queried again to obtain the 1-hop neighboring area of ​​the 1-hop neighboring area. Then, the union of the 1-hop neighboring area of ​​the problem cell and the 1-hop neighboring area of ​​the 1-hop neighboring area of ​​the problem cell is used as the area to which the problem cell belongs.

[0171] (3) Obtaining area data

[0172] Based on the wireless network related data of each cell in the area where the problem cell belongs, it mainly includes cell performance data, KPI data, NRM data, MR data, parameter adjustment records and other data.

[0173] (4) Generate a snapshot of the network elements in the area

[0174] After acquiring the area data, the data is organized into the input form of the large model, and the high-dimensional representation of the area data is obtained through the network element snapshot acquisition module of the large model.

[0175] (5) Decision optimization

[0176] The decision optimization function of the large model is to optimize within the constraints based on the pre-designed objective function and obtain the parameter adjustment scheme that needs to be adjusted, that is, the wireless network optimization scheme. Usually, different root causes correspond to different objective functions.

[0177] The following are some examples:

[0178] Example 1: Modify neighbor-level parameters of multiple cells based on multiple network elements

[0179] The problem work order of this example is: the 15-minute PRB utilization index of the problem cell exceeded 90% three times in a row on the same day. The large model is based on the inference output of the existing network data: this problem is a non-disturbance problem, the first-level root cause is capacity, the second-level root cause is high load, and the third-level root cause is unreasonable parameter settings of neighboring cells. The decision-making plan adjusts the CellIndividualOffset parameter between the 1-hop neighboring cell and the current cell from 0 to -2, and adjusts the CellIndividualOffset parameter between the current cell and the 1-hop neighboring cell from 0 to 2. It is predicted that after the parameters of this plan are successfully sent to the equipment, the cell PRB utilization will drop to about 49.5%. Figure 7 The optimization goals of the problem work order, parameter adjustment suggestions for the wireless network optimization solution, and comparison of indicators before and after optimization are shown.

[0180] Example 2: Modifying multiple base station-level parameters of a single cell based on multiple network elements.

[0181] The problem work order of this example is: the problem cell has been identified as a high-load cell for three consecutive days. The large model reasoning output based on the existing network data: this problem is a non-disturbance problem, the first-level root cause is capacity, the second-level root cause is high load, and the third-level root cause is unreasonable mobility parameter settings. The decision-making plan is to soft-expand the 40936 frequency cell. It is predicted that after the parameters of this plan are successfully sent to the device, the cell PRB utilization rate will drop to about 57.5%.

[0182] Example 3: Modify cell-level multi-parameters of a single cell based on multiple network elements:

[0183] The problem work order of this example is: the problem cell has been identified as a high-load cell for three consecutive days. The large model infers and outputs based on the existing network data: this problem is a non-disturbance problem, the first-level root cause is capacity, the second-level root cause is high load, and the third-level root cause is unreasonable mobility parameter settings. The decision-making plan is to adjust the cell's A3-based heterofrequency A1 RSRP trigger threshold parameters from -108 to -96, and the A3-based heterofrequency 2 RSRP trigger threshold parameters from -111 to -99. It is predicted that after the parameters of this plan are successfully sent down and acted on the equipment, the cell PRB utilization rate will be reduced to about 65.7%.

[0184] The above describes various methods of the embodiments of the present application. The following further provides devices for implementing the above methods.

[0185] Please refer to Figure 8 The embodiment of the present application further provides a wireless capacity optimization device based on a large model, including:

[0186] A first acquisition module 801 is configured to acquire wireless network related data of a first area to which a first cell belongs, wherein the first cell is a cell where an abnormal wireless network indicator occurs, and the first area includes a plurality of cells;

[0187] A first processing module 802 is configured to use the wireless network related data of the first area as input data of a first model to obtain a network element snapshot of the first area generated by the first model;

[0188] The second processing module 803 uses the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization solution for the first area output by the first model;

[0189] The first model is a large model trained using wireless network related data of the mobile network to which the first cell belongs.

[0190] Through the above modules, the embodiment of the present application realizes end-to-end optimization of the wireless network through a large model. The large model generates a network element snapshot based on the wireless network related data of the area to which the abnormal cell belongs, and then generates a wireless network optimization plan based on the network element snapshot. It can shorten the time for locating wireless network problems, improve the accuracy of network optimization plans, avoid the impact of unreasonable optimization plans on network quality, and improve wireless network optimization efficiency and wireless network quality.

[0191] It should also be noted that the large model-based wireless capacity optimization device of the embodiment of the present application can optimize wireless network problems including wireless capacity optimization, wireless coverage optimization, wireless interference optimization, etc. Therefore, it can also be called a large model-based wireless network optimization device.

[0192] Optionally, the wireless network related data includes at least one of the following:

[0193] Cell performance data;

[0194] Key performance indicator (KPI) data of the community;

[0195] The cell's network resource management (NRM) data;

[0196] Measurement report MR data of the cell;

[0197] Parameter adjustment record of the cell.

[0198] Optionally, the wireless network optimization solution for the first area includes at least one of the following:

[0199] An adjustment plan for cell-level parameters in the first area;

[0200] an adjustment scheme for neighboring cell level parameters in the first cell;

[0201] An adjustment scheme for base station-level parameters in the first cell.

[0202] Optionally, the above device further comprises:

[0203] A second acquisition module, used to acquire wireless network related data of the first cell;

[0204] A third processing module, configured to use the wireless network related data of the first cell as input data of the first model to obtain a network element snapshot of the first cell generated by the first model;

[0205] The fourth processing module is used to use the network element snapshot of the first cell as input data of the first model to obtain a disturbance analysis result of whether the wireless network anomaly of the first cell is a disturbance problem output by the first model.

[0206] Optionally, the network element snapshot of the first cell of the above-mentioned device is a feature vector generated by the first model based on the wireless network related data of the first cell; the network element snapshot of the first area is a feature vector generated by the first model based on the wireless network related data of the first area.

[0207] Optionally, the above device further comprises:

[0208] The first control module is used to trigger the first acquisition module to acquire wireless network related data of a first area to which the first cell belongs when the disturbance analysis result indicates that the wireless network abnormality of the first cell is not a disturbance problem.

[0209] Optionally, the above device further comprises:

[0210] The second control module is used to use the network element snapshot of the first cell as input data of the first model when the disturbance analysis result indicates that the wireless network anomaly of the first cell is not a disturbance problem, obtain the root cause positioning result of the wireless network anomaly of the first cell output by the first model, and trigger the first acquisition module to obtain the wireless network related data of the first area to which the first cell belongs.

[0211] Optionally, the second processing module of the above-mentioned device is also used to use the root cause positioning result of the wireless network anomaly of the first cell and the network element snapshot of the first area as input data of the first model to obtain the wireless network optimization plan for the first area output by the first model.

[0212] Optionally, the root cause positioning result of the wireless network abnormality of the first cell of the above-mentioned device includes multiple levels of root causes, wherein the first level root cause includes one of wireless capacity problem, wireless coverage problem, and wireless interference problem, the second level root cause includes one of high load and low traffic, and the third level root cause includes unreasonable power parameter setting, unreasonable load balancing parameter setting, unreasonable mobility parameter setting, unreasonable neighbor cell pair parameter setting, unreasonable capacity setting, unreasonable energy-saving parameter setting, small number of users, and one of other reasons that require on-site investigation.

[0213] Optionally, the above device further comprises:

[0214] The first optimization module is used to optimize the first area according to the wireless network optimization plan of the first area.

[0215] Optionally, the above device further comprises:

[0216] a fifth processing module, configured to use the wireless network optimization solution of the first area and the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization result of the first area predicted by the first model;

[0217] The second optimization module is used to optimize the first area according to the wireless network optimization plan of the first area when the wireless network optimization result of the first area meets the requirements.

[0218] Optionally, the above device further comprises:

[0219] A second acquisition module is used to obtain a neighbor relationship table of the mobile network to which the first cell belongs, the neighbor relationship table including a one-hop neighbor of each cell; and determine a first area to which the first cell belongs according to the neighbor relationship table, wherein the first area includes the first cell and neighboring areas within N hops of the first cell, where N is an integer greater than or equal to 1.

[0220] Optionally, the above device further comprises:

[0221] A third acquisition module is used to obtain switching data of each cell of the mobile network to which the first cell belongs within a preset historical time period, the switching data recording the number of switchings from the current cell to other cells; obtain measurement report MR data received by each cell, the MR data including location information of the terminal reporting the MR data; determine the overlapping area between any two cells according to the MR data; determine the number of switchings from the current cell to each other cell according to the switching data of the current cell, sort the other cells in descending order according to the number of switchings, and obtain a queue composed of the other cells; wherein, when at least two other cells have the same number of switchings, the at least two other cells are sorted according to the size of the overlapping area between the at least two other cells and the current cell; select the first M other cells from the queue as 1-hop neighboring cells of the current cell, and obtain a neighboring cell relationship table of the current cell.

[0222] It should be noted that the device in this embodiment is a device corresponding to the method applied to the wireless capacity optimization device / wireless network optimization device, and the implementation methods in the above embodiments are applicable to the embodiments of the device, and can also achieve the same technical effect. The above device provided in the embodiment of the present application can implement all the method steps implemented in the above method embodiment, and can achieve the same technical effect. The parts and beneficial effects that are the same as those in the method embodiment in this embodiment will not be specifically described here.

[0223] Another embodiment of the present application provides a wireless capacity optimization system or a wireless network optimization system based on a large model, such as Fig. 9 As shown, it includes a transceiver 910, a processor 900, a memory 920, and a program or instruction stored in the memory 920 and executable on the processor 900; when the processor 900 executes the program or instruction, each process of the above-mentioned large-model-based wireless capacity optimization system or wireless network optimization system embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0224] The transceiver 910 is used to receive and send data under the control of the processor 900 .

[0225] Among them, Fig. 9In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 900 and memory represented by memory 920. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 910 may be a plurality of components, i.e., including a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium. The processor 900 is responsible for managing the bus architecture and general processing, and the memory 920 may store data used by the processor 900 when performing operations.

[0226] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, each process of the above-mentioned wireless capacity optimization method / wireless network optimization method embodiment based on a large model is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0227] The embodiment of the present application also provides a computer program product, including computer instructions. When the computer instructions are executed by a processor, the various processes of the above-mentioned large-model-based wireless capacity optimization method / wireless network optimization method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, they are not repeated here.

[0228] It should be noted that, in this article, the terms "include", "include" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0229] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, and the computer software product is stored in a storage medium (such as ROM / RAM, a disk, or an optical disk), including a number of instructions for a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0230] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims

1. A wireless capacity optimization method based on a large model, characterized in that: include: Acquire wireless network related data of a first area to which a first cell belongs, where the first cell is a cell in which a wireless network indicator is abnormal, the first area includes multiple cells, and the wireless network related data includes at least one of the following: cell performance data; network resource management NRM data of the cell; measurement report MR data of the cell; parameter adjustment record of the cell; The wireless network related data of the first area is used as input data of the first model to obtain a network element snapshot of the first area generated by the first model; wherein the wireless network related data of the first area is divided according to a preset time granularity to obtain data in each time granularity; the data in each time granularity is encoded to obtain state data of each time granularity; the network element snapshot of the first area is a feature vector generated by the first model based on the state data of the first area at multiple time granularities; Using the network element snapshot of the first area as input data of the first model, obtaining a wireless network optimization plan for the first area output by the first model; the wireless network optimization plan for the first area includes at least one of the following: an adjustment plan for cell-level parameters in the first area; an adjustment plan for neighboring cell-level parameters in the first cell; an adjustment plan for base station-level parameters in the first cell; The first model is a large model trained using wireless network related data of the mobile network to which the first cell belongs.

2. The method according to claim 1, characterized in that The wireless network related data also includes: Key performance indicator (KPI) data of the community.

3. The method according to claim 1, characterized in that Also includes: Acquire wireless network related data of the first cell; Using the wireless network related data of the first cell as input data of the first model, to obtain a network element snapshot of the first cell generated by the first model; The network element snapshot of the first cell is used as input data of the first model to obtain a disturbance analysis result output by the first model as to whether the wireless network anomaly of the first cell is a disturbance problem.

4. The method according to claim 3, characterized in that The network element snapshot of the first cell is a feature vector generated by the first model based on the wireless network related data of the first cell.

5. The method according to claim 3, characterized in that: Also includes: When the disturbance analysis result indicates that the wireless network abnormality of the first cell is not a disturbance problem, the step of obtaining wireless network related data of the first area to which the first cell belongs is performed.

6. The method according to claim 3, characterized in that Also includes: When the disturbance analysis result indicates that the wireless network anomaly of the first cell is not a disturbance problem, the network element snapshot of the first cell is used as input data of the first model to obtain the root cause positioning result of the wireless network anomaly of the first cell output by the first model.

7. The method according to claim 6, characterized in that Using the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization solution for the first area output by the first model includes: The root cause location result of the wireless network anomaly of the first cell and the network element snapshot of the first area are used together as input data of the first model to obtain the wireless network optimization solution for the first area output by the first model.

8. The method according to claim 7, characterized in that The root cause positioning result of the wireless network abnormality of the first cell includes multiple levels of root causes, wherein the first level root cause includes one of wireless capacity problem, wireless coverage problem, and wireless interference problem; the second level root cause includes one of high load and low traffic; the third level root cause includes unreasonable power parameter setting, unreasonable load balancing parameter setting, unreasonable mobility parameter setting, unreasonable neighbor cell pair parameter setting, unreasonable capacity setting, unreasonable energy-saving parameter setting, small number of users, and one of other reasons that require on-site investigation.

9. The method according to any one of claims 1 to 8, characterized in that: Also includes: The first area is optimized according to the wireless network optimization solution of the first area.

10. The method according to any one of claims 1 to 8, characterized in that: Also includes: Using the wireless network optimization solution of the first area and the network element snapshot of the first area as input data of the first model, to obtain the wireless network optimization result of the first area predicted by the first model; When the wireless network optimization result of the first area meets the requirements, the first area is optimized according to the wireless network optimization solution of the first area.

11. The method according to claim 1, characterized in that: Also includes: Obtain a neighboring cell relationship table of a mobile network to which the first cell belongs, where the neighboring cell relationship table includes a one-hop neighboring cell of each cell; According to the neighboring cell relationship table, a first area to which the first cell belongs is determined, wherein the first area includes the first cell and neighboring cells within N hops of the first cell, where N is an integer greater than or equal to 1.

12. The method according to claim 11, characterized in that The obtaining a neighboring cell relationship table of a mobile network to which the first cell belongs includes: Obtaining handover data of each cell of the mobile network to which the first cell belongs within a preset historical time period, the handover data recording the number of handovers from the current cell to other cells; obtaining measurement report MR data received by each cell, the MR data including location information of a terminal reporting the MR data; determining the overlapping area between any two cells according to the MR data; Determine, according to the handover data of the current cell, the number of handovers from the current cell to each other cell, and sort the other cells in descending order of the number of handovers to obtain a queue composed of the other cells; wherein, when at least two other cells have the same number of handovers, sort the at least two other cells according to the size of the overlapping area between the at least two other cells and the current cell; The first M other cells are selected from the queue as one-hop neighboring cells of the current cell, and a neighboring cell relationship table of the current cell is obtained.

13. A wireless capacity optimization device based on a large model, characterized in that: include: A first acquisition module is used to acquire wireless network related data of a first area to which a first cell belongs, the first cell is a cell where an abnormal wireless network indicator occurs, the first area includes multiple cells, and the wireless network related data includes at least one of the following: cell performance data; network resource management NRM data of the cell; measurement report MR data of the cell; parameter adjustment record of the cell; A first processing module is configured to use the wireless network related data of the first area as input data of the first model to obtain a network element snapshot of the first area generated by the first model; wherein the wireless network related data of the first area is divided according to a preset time granularity to obtain data within each time granularity; the data within each time granularity is encoded to obtain state data of each time granularity; the network element snapshot of the first area is a feature vector generated by the first model based on the state data of the first area at multiple time granularities; The second processing module uses the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization plan for the first area output by the first model; the wireless network optimization plan for the first area includes at least one of the following: an adjustment plan for cell-level parameters in the first area; an adjustment plan for neighboring cell-level parameters in the first cell; an adjustment plan for base station-level parameters in the first cell; The first model is a large model trained using wireless network related data of the mobile network to which the first cell belongs.

14. The device according to claim 13, characterized in that The wireless network related data also includes: Key performance indicator (KPI) data of the community.

15. The device according to claim 13, characterized in that Also includes: A second acquisition module, used to acquire wireless network related data of the first cell; A third processing module, configured to use the wireless network related data of the first cell as input data of the first model to obtain a network element snapshot of the first cell generated by the first model; The fourth processing module is used to use the network element snapshot of the first cell as input data of the first model to obtain a disturbance analysis result of whether the wireless network anomaly of the first cell is a disturbance problem output by the first model.

16. The device according to claim 15, characterized in that The network element snapshot of the first cell is a feature vector generated by the first model based on the wireless network related data of the first cell.

17. The device according to claim 15, characterized in that Also includes: The first control module is used to trigger the first acquisition module to acquire wireless network related data of a first area to which the first cell belongs when the disturbance analysis result indicates that the wireless network abnormality of the first cell is not a disturbance problem.

18. The device according to claim 15, characterized in that Also includes: The second control module is used to use the network element snapshot of the first cell as input data of the first model when the disturbance analysis result indicates that the wireless network anomaly of the first cell is not a disturbance problem, obtain the root cause positioning result of the wireless network anomaly of the first cell output by the first model, and trigger the first acquisition module to obtain the wireless network related data of the first area to which the first cell belongs.

19. The device according to claim 18, characterized in that The second processing module is also used to use the root cause positioning result of the wireless network anomaly of the first cell and the network element snapshot of the first area as input data of the first model to obtain the wireless network optimization plan for the first area output by the first model.

20. The device according to claim 19, characterized in that The root cause positioning result of the wireless network abnormality of the first cell includes multiple levels of root causes, wherein the first level root cause includes one of wireless capacity problem, wireless coverage problem, and wireless interference problem; the second level root cause includes one of high load and low traffic; the third level root cause includes unreasonable power parameter setting, unreasonable load balancing parameter setting, unreasonable mobility parameter setting, unreasonable neighbor cell pair parameter setting, unreasonable capacity setting, unreasonable energy-saving parameter setting, small number of users, and one of other reasons that require on-site investigation.

21. The device according to any one of claims 13 to 20, characterized in that Also includes: The first optimization module is used to optimize the first area according to the wireless network optimization plan of the first area.

22. The device according to any one of claims 13 to 20, characterized in that Also includes: a fifth processing module, configured to use the wireless network optimization solution of the first area and the network element snapshot of the first area as input data of the first model to obtain a wireless network optimization result of the first area predicted by the first model; The second optimization module is used to optimize the first area according to the wireless network optimization plan of the first area when the wireless network optimization result of the first area meets the requirements.

23. The device according to claim 13, characterized in that Also includes: A second acquisition module is used to acquire a neighboring cell relationship table of a mobile network to which the first cell belongs, wherein the neighboring cell relationship table includes a one-hop neighboring cell of each cell; According to the neighboring cell relationship table, a first area to which the first cell belongs is determined, wherein the first area includes the first cell and neighboring cells within N hops of the first cell, where N is an integer greater than or equal to 1.

24. The device according to claim 23, characterized in that Also includes: A third acquisition module is used to obtain switching data of each cell of the mobile network to which the first cell belongs within a preset historical time period, the switching data recording the number of switchings from the current cell to other cells; obtain measurement report MR data received by each cell, the MR data including location information of the terminal reporting the MR data; determine the overlapping area between any two cells according to the MR data; determine the number of switchings from the current cell to each other cell according to the switching data of the current cell, sort the other cells in descending order according to the number of switchings, and obtain a queue composed of the other cells; wherein, when at least two other cells have the same number of switchings, the at least two other cells are sorted according to the size of the overlapping area between the at least two other cells and the current cell; select the first M other cells from the queue as 1-hop neighboring cells of the current cell, and obtain a neighboring cell relationship table of the current cell.

25. A wireless network optimization system based on a large model, characterized in that: The method comprises a transceiver, a processor, a memory, and a program or instruction stored in the memory and executable on the processor; when the processor executes the program or instruction, the steps of the method according to any one of claims 1 to 12 are implemented.

26. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 12 are implemented.

27. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of the method according to any one of claims 1 to 12.

Images