Network signal optimization method and apparatus, electronic device, and storage medium

Abstract

The application discloses a network signal optimization method and device, electronic equipment and a storage medium, and belongs to the technical field of communication. The method comprises the following steps: in the case that it is detected that the electronic equipment enters a preset scene, collecting scene sensing data of the electronic equipment in the preset scene; determining a corresponding first learning model according to the scene sensing data; in the case that the network signal quality of the electronic equipment meets a preset condition, outputting a network optimization strategy corresponding to the scene sensing data through the first learning model; and controlling the electronic equipment to execute the network optimization strategy.

Description

Technical Field

[0001] This application belongs to the field of communication technology, and specifically relates to a network signal optimization method, apparatus, electronic device and storage medium. Background Technology

[0002] With the widespread adoption of mobile communication technology, the network communication quality of electronic devices in various complex environments has become a key factor affecting user experience. Enclosed mobile environments such as elevators often become weak points in terms of network signal due to the shielding effect of the metal car, frequent cell handovers caused by rapid movement, and the complex propagation environment of base station coverage signals within the shaft. Users frequently encounter problems such as severe signal attenuation, call lag, and data transmission interruptions in these scenarios.

[0003] Currently, network signal optimization solutions for enclosed mobile scenarios such as elevators mostly employ preset, general strategies. For example, when an electronic device detects entering an "elevator" scenario, it triggers a fixed set of parameter adjustment logic, such as increasing signal transmission power or pre-switching to a commonly used frequency band. However, different elevators vary significantly in their physical structure, operating trajectory, surrounding base station deployment, and electromagnetic environment. As a result, existing network signal optimization solutions struggle to adapt to the personalized needs of different scenarios, leading to limited optimization effectiveness. Summary of the Invention

[0004] The purpose of this application is to provide a network signal optimization method, apparatus, electronic device, and storage medium that can take targeted network signal optimization strategies according to the scenario in which the electronic device is located, thereby improving the optimization effect of the network signal.

[0005] In a first aspect, embodiments of this application provide a network signal optimization method, the method comprising: When the electronic device enters a preset scene, scene perception data of the electronic device in the preset scene is collected; Based on the scene perception data, determine the corresponding first learning model; When the network signal quality of the electronic device meets the preset conditions, the first learning model outputs a network optimization strategy corresponding to the scene perception data. Control the electronic device to execute the network optimization strategy.

[0006] Secondly, embodiments of this application provide a network signal optimization device, the device comprising: The data acquisition module is used to acquire scene perception data of the electronic device in the preset scene when the electronic device enters the preset scene; The determining module is used to determine the corresponding first learning model based on the scene perception data; The output module is used to output a network optimization strategy corresponding to the scene perception data through the first learning model when the network signal quality of the electronic device meets the preset conditions. A control module is used to control the electronic device to execute the network optimization strategy.

[0007] Thirdly, embodiments of this application provide an electronic device, which includes a processor and a memory, wherein the memory stores programs or instructions that can run on the processor, and the programs or instructions, when executed by the processor, implement the method as described in the first aspect.

[0008] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the method described in the first aspect.

[0009] Fifthly, embodiments of this application provide a chip, the chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the method as described in the first aspect.

[0010] In a sixth aspect, embodiments of this application provide a computer program product stored in a storage medium, which is executed by at least one processor to implement the method described in the first aspect.

[0011] In this embodiment, when an electronic device is detected to have entered a preset scene, a first learning model corresponding to the preset scene is determined based on the scene perception data of the electronic device in the preset scene. The first learning model corresponding to the preset scene is used to output a dedicated network optimization strategy corresponding to the scene perception data, and then the network signal of the electronic device is optimized according to the network optimization strategy. In this way, the solution of this embodiment can select the first learning model corresponding to the scene in which the electronic device is located, and then take a targeted network signal optimization strategy, rather than using a single general optimization strategy. This can overcome the problem that the traditional "one-size-fits-all" optimization solution cannot adapt to the differences in different mobile scenes, resulting in poor optimization effect. Attached Figure Description

[0012] Figure 1 This is a flowchart illustrating a network signal optimization method provided in some embodiments of this application; Figure 2 This is a flowchart illustrating a network signal optimization method provided in some embodiments of this application; Figure 3 These are schematic diagrams illustrating the structure of a network signal optimization device according to some embodiments of this application; Figure 4 These are schematic diagrams illustrating the structure of an electronic device according to some embodiments of this application; Figure 5 These are schematic diagrams illustrating the hardware structure of an electronic device according to some embodiments of this application. Detailed Implementation

[0013] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0014] The terms "first," "second," etc., used in this application's specification are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, without limiting the number of objects; for example, the first object can be one or N objects. Furthermore, in the specification, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects have an "or" relationship.

[0015] The terminology used in the embodiments of this invention will be explained below.

[0016] Enclosed mobile scenarios refer to specific areas where the physical space is relatively enclosed or semi-enclosed, and objects within the scenario are in motion, resulting in a complex and dynamically changing wireless network signal propagation environment and limited base station coverage. Examples of enclosed mobile scenarios in this application's embodiments include elevators, underground parking garages, enclosed conference rooms, highway tunnels, underground passages, subway cars, high-speed rail cars, and large supermarket shelves.

[0017] Reference Signal Receiving Power (RSRP) is a key parameter representing the strength of a wireless signal in Long Term Evolution (LTE) networks and is one of the physical layer measurement requirements. It is the average signal power received on all resource particles carrying the reference signal within a certain symbol.

[0018] Signal to Interference plus Noise Ratio (SINR): This refers to the ratio of the strength of the received useful signal to the strength of the received interference signal. It can be simply understood as the "signal-to-noise ratio".

[0019] The technical solution of this application embodiment can be applied to scenarios where network signals of electronic devices need to be optimized in enclosed mobile environments. For example, a user's company is located on the 14th floor of office building A. The office building has three elevators: elevator 1, elevator 2, and elevator 3. The user is currently taking elevator 1 from the 1st floor to the 14th floor. During this journey, the user's phone is playing music. Because the user is in an enclosed space and affected by the walls of elevator 1, the network signal of the user's phone may be unstable, and the music playback may experience stuttering or other problems.

[0020] The network signal optimization method provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0021] Figure 1 This is a flowchart illustrating a network signal optimization method provided in an embodiment of this application. The execution subject of the network signal optimization method can be an electronic device, which can be, but is not limited to, a personal computer (PC), a smartphone, a tablet computer, or a personal digital assistant (PDA).

[0022] like Figure 1 As shown, the network signal optimization method provided in this application embodiment may include steps 110-140.

[0023] Step 110: When the electronic device enters a preset scene, collect scene perception data of the electronic device in the preset scene.

[0024] The preset scenario can be an electronic device entering a preset scene. This preset scenario can be a closed mobile scene. For example, in the example above, if the user enters elevator number 1, then elevator number 1 is the preset scenario.

[0025] Scene perception data can be scene data perceived by electronic devices in a preset scene. The scene perception data can include at least: the operating status information corresponding to the preset scene, the spatial location information of the electronic device, and the wireless network environment information corresponding to the preset scene.

[0026] The operating status information corresponding to the aforementioned preset scenario may include at least: the start-stop acceleration of the preset scenario, the time information of the electronic device entering the preset scenario, and the time information of the electronic device leaving the preset scenario. For example, in the example above, when the preset scenario is elevator No. 1, the start-stop acceleration of elevator No. 1, the time of the user's mobile phone entering elevator No. 1, and the time of leaving elevator No. 1.

[0027] Electronic devices can have built-in sensors that can collect movement data of the electronic device and then determine the start-stop acceleration of a preset scenario based on the movement data. For example, the speed and acceleration of the electronic device can be collected by the gyroscope inside the electronic device, and then the start-stop acceleration of the preset scenario can be determined based on the movement data.

[0028] The aforementioned determination of the time information for an electronic device entering and leaving a preset scene can be achieved through the acceleration of the electronic device. For example, the time when the electronic device first detects weightlessness is the time when it enters the preset scene, and the time when it last detects weightlessness is the time when it leaves the preset scene. Specifically, this can be based on sensors collecting motion data from the electronic device, and then using this motion data to determine whether the electronic device is in a weightless state. For instance, the speed and acceleration of the electronic device can be collected using a gyroscope within the device, and then this motion data can be used to determine whether the electronic device is in a weightless state.

[0029] The spatial location information of the aforementioned electronic device can be represented by the latitude and longitude information of the electronic device.

[0030] The wireless network environment information of the aforementioned preset scenario can be network information within the preset scenario, such as base station information, Reference Received Power (RSRP), and Signal-to-Interference-plus-Noise Ratio (SINR). The base station information within the target enclosed scenario can be the base station's identification information, such as its IdentityDocument (ID).

[0031] Step 120: Determine the corresponding first learning model based on the scene perception data.

[0032] The first learning model can be a learning model corresponding to a preset scenario. This first learning model can be a network model trained based on deep learning or reinforcement learning, such as, but not limited to, a neural network model or a support vector machine model. In the example above, the first learning model could be the learning model corresponding to elevator number 1.

[0033] In some embodiments of this application, the electronic device may store at least one learning model, that is, the electronic device may store multiple learning models, and a preset scenario may correspond to one learning model. In other words, the learning model corresponding to each preset scenario is different. For example, in the above example, elevator No. 1 corresponds to one learning model, elevator No. 2 also corresponds to one learning model, and elevator No. 3 also corresponds to one learning model.

[0034] If at least one learning model is stored in the electronic device, step 120 may specifically include: Based on scene perception data, generate identification information corresponding to the preset scene; The learning model corresponding to the identification information in at least one learning model is used as the first learning model.

[0035] Among them, the identification information corresponding to the preset scene can be the identification information of the preset scene, which can be the ID of the preset scene.

[0036] In some embodiments of this application, a preset scenario may correspond to an ID. For example, in the example above, the identification information of elevator No. 1 may be elevator No. 1 in office building A.

[0037] In some embodiments of this application, identification information corresponding to the preset scene can be generated based on scene perception data. The specific method for generating identification information corresponding to the preset scene based on scene perception data will be described in detail in later embodiments.

[0038] Based on the generated identification information corresponding to the preset scenario, the electronic device can be queried to see if there is a learning model corresponding to the identification information. If a learning model corresponding to the identification information is found in the electronic device, the learning model corresponding to the identification information found in at least one learning model can be used as the first learning model.

[0039] Continuing with the example above, by collecting scene perception data for elevator number 1, identification information corresponding to the preset scene can be generated. For example, the identification information for elevator number 1 could be: Elevator 1, Office Building A. The user's phone stores learning model 1 (for elevator number 1), learning model 2 (for elevator number 2), and learning model 3 (for elevator number 3). Then, based on the identification information "Elevator 1, Office Building A," the system checks if a corresponding learning model exists on the user's phone. If learning model 1 is found, it can be used as the first learning model to optimize the user's phone's network signal.

[0040] In the embodiments of this application, by setting a corresponding learning model for each preset scene, when the electronic device enters a preset scene, the corresponding first learning model can be found from the electronic device based on the identification information of the preset scene determined by the scene perception data of that preset scene. Users do not need to manually select or set anything, resulting in a smooth and seamless experience. Furthermore, the scene perception data is used for local matching and calculation, eliminating the need to upload the scene perception data to the cloud for identification or processing, reducing the risk of data leakage and improving the security of the scene perception data. Moreover, the electronic device only needs to load and run this single first learning model, avoiding the huge computational load, memory usage, and power consumption caused by running multiple large or general-purpose models simultaneously. Furthermore, when the electronic device enters a new scene, only an identifier and corresponding model for that scene need to be added locally. The system architecture is clear, easily expandable, and does not affect other deployed models.

[0041] In some embodiments of this application, when the aforementioned scene perception data includes at least: operating status information corresponding to a preset scene, spatial location information of an electronic device, and wireless network environment information corresponding to the preset scene, the step of generating identification information corresponding to the preset scene based on the scene perception data includes: The system integrates operational status information, spatial location information, and wireless network environment information to generate identification information corresponding to a preset scenario.

[0042] In some embodiments of this application, where the aforementioned scene perception data includes at least: operating status information corresponding to a preset scene, spatial location information of electronic devices, and wireless network environment information corresponding to a preset scene, the collected multi-dimensional information, such as operating status information, spatial location information, and wireless network environment information, can be fused to obtain the identification information of the preset scene. Specifically, the operating status information, spatial location information, and wireless network environment information can be encoded separately, for example, using one-hot encoding or other encoding algorithms, which are not limited in this embodiment. Then, the encoding results of the operating status information, spatial location information, and wireless network environment information are fused to obtain the identification information corresponding to the preset scene.

[0043] The above-mentioned method of encoding the operation status information, spatial location information, and wireless network environment information separately, and then fusing the encoding results of the operation status information, spatial location information, and wireless network environment information to obtain the identification information corresponding to the preset scenario, is only one scheme for fusing and processing multi-dimensional information such as collected operation status information, spatial location information, and wireless network environment information provided in this application embodiment. However, this application embodiment is not limited to this only implementation method.

[0044] It should be noted that the above-mentioned operating status information can represent how the preset scene is moving, the above-mentioned spatial location information can represent the approximate location of the electronic device, and the above-mentioned wireless network environment information can represent the signal environment in which the electronic device is located. Thus, a unique identifier for the preset scene can be generated through the above three types of information.

[0045] In the embodiments of this application, since a single information source is prone to repetition, the embodiments of this application integrate multi-dimensional information, which is equivalent to simultaneously verifying "how the electronic device is moving", "where the electronic device is roughly located" and "what signal environment the electronic device is in", forming a complex composite identifier, which greatly reduces the misidentification rate of the preset scenario.

[0046] In some embodiments of this application, the methods described above may further include: If no learning model corresponding to the identification information exists in at least one learning model, an initial learning model corresponding to the preset scene is initialized based on the scene perception data.

[0047] In some embodiments of this application, if no learning model corresponding to the identification information exists in at least one learning model stored in the electronic device, an initial learning model corresponding to a preset scene can be initialized based on scene perception data. Specifically, information strongly correlated with the preset scene in the scene perception data can be used as the state space of the preset scene, and redundant parameters can be removed. For example, the running status information and wireless network environment information can be used as the state space of the preset scene, and spatial location information can be removed. Then, some network optimization strategies corresponding to the state space can be matched to obtain the initial learning model corresponding to the preset scene.

[0048] It should be noted that the network optimization strategies mentioned above that correspond to this state space can be some general optimization strategies, such as improving RSRP, switching frequency bands, or improving SINR, etc.

[0049] It's important to note that using the operational status information and wireless network environment information from the scene-aware data as the state space of this preset scenario, while removing fixed or minimally changing parameters like spatial location information, is to avoid adding invalid dimensions to the stored information of the initial learning model. For example, taking an elevator as the preset scenario, in a single elevator scenario, "latitude and longitude" is a strong identifier rather than a decision-making basis. If the learning model learns a strategy based on this fixed identifier, it might not learn the general "how to optimize when the signal is weak," but rather "what should be done at this latitude and longitude." This would lead to a lack of generalization ability in the learned strategy, making it unable to cope with signal changes caused by different operating stages (such as going up, going down, or stopping) within the same elevator. Removing such fixed parameters forces the learning model to learn to make decisions based on variable operating states and signal quality, making the learned strategy more general and stable. For example, for the same elevator, whether it's on the 1st or 20th floor, its latitude and longitude remain essentially unchanged. If we initialize the learning model with latitude and longitude as a state dimension, the model would need to handle a large number of repetitive or invalid state combinations caused by this invariant dimension during learning and decision-making. This would severely distract the learning model and reduce its learning efficiency. Removing spatial location information is equivalent to allowing the learning model to focus only on dynamically changing key decision variables, such as whether the elevator is accelerating or moving at a constant speed, or the strength of the currently connected base station signal, thus enabling it to converge to an effective strategy more quickly.

[0050] In the embodiments of this application, if there is no learning model corresponding to the identification information in at least one learning model, an initial learning model corresponding to the preset scene is initialized based on the scene perception data. This allows the electronic device to directly obtain the initial learning model corresponding to the preset scene when it re-enters the preset scene, without having to rebuild the learning process. This improves the efficiency of obtaining network optimization strategies and ensures that the electronic device can immediately provide basic services and begin accumulating data in any closed mobile scene. Users will not experience service interruptions or functional loss due to encountering an unrecorded closed mobile scene, ensuring a consistent experience.

[0051] Step 130: When the network signal quality of the electronic device meets the preset conditions, the network optimization strategy corresponding to the scene perception data is output through the first learning model.

[0052] Among them, the preset conditions can be the conditions that the network signal quality of the electronic device must meet when the network signal of the electronic device needs to be optimized. For example, the RSRP of the electronic device may be less than a certain threshold, or the SINR may be less than a certain threshold.

[0053] The network signal quality of the aforementioned electronic devices meeting the preset conditions may include, but is not limited to, at least one of the following: the RSRP of the electronic device is less than a certain threshold, the SINR is lower than a certain threshold, or the electronic device experiences call stuttering, video stuttering, or data transmission interruption.

[0054] The network optimization strategy corresponding to scene-aware data can be a strategy for optimizing the network signals of electronic devices based on scene-aware data.

[0055] In some embodiments of this application, the first learning model may include at least one set of first scene perception data, and each set of first scene perception data corresponds to at least one optimization strategy, with each optimization strategy corresponding to a different reward value. That is, for a learning model, the learning model has different first scene perception data, and the network optimization strategy corresponding to each set of first scene perception data is different. For example, for elevator number 1 in the above example, it may correspond to two sets of first scene perception data. For instance, the first set of first scene perception data might be: running status information: starting acceleration is 2 m / s². 2 The electronic device entered elevator number 1 at 9:00 AM and exited at 9:00:30 AM. The wireless network environment information is: RSRP -120dBm, SINR 1dB. The second set of scene perception data is: Operational status information: Startup acceleration 3m / s². 2 The electronic device enters elevator number 1 at 10:00 AM and exits at 10:00:15 AM. The wireless network environment information is: RSRP -100dBm, SINR 6dB. Therefore, for the first set of scene perception data, there can be at least one optimization strategy, for example, three optimization strategies: the first is to improve startup acceleration, the second is to improve RSRP, and the third is to improve SINR. Similarly, for the second set of scene perception data, there can also be at least one optimization strategy, for example, one strategy, such as improving RSRP.

[0056] The network optimization strategy output by the first learning model corresponding to the scene perception data may specifically include: Among at least one first scene perception data in the first learning model, the optimization strategy with the highest reward value among at least one optimization strategy corresponding to the first scene perception data is taken as the network optimization strategy corresponding to the scene perception data.

[0057] In some embodiments of this application, a learning model corresponding to a preset scene may include multiple sets of first scene perception data. Each set of first scene perception data corresponds to at least one different optimization strategy, and the reward value of each optimization strategy is different. When selecting a network optimization strategy from the first learning model, the first scene perception data corresponding to the scene perception data can be selected from at least one set of first scene perception data in the first learning model. Then, from at least one optimization strategy corresponding to the scene perception data, the optimization strategy with the highest reward value is selected as the network optimization strategy corresponding to the scene perception data. That is, the optimization strategy with the highest reward value among at least one optimization strategy corresponding to the scene perception data in at least one set of first scene perception data in the first learning model is selected as the network optimization strategy corresponding to the scene perception data.

[0058] Continuing with the example above, using learning model 1 as the first learning model to optimize the user's mobile phone network signal, learning model 1 includes two sets of first scene perception data: first scene perception data 1 and first scene perception data 2. For first scene perception data 1, the learning model includes three optimization strategies: startup acceleration, improving RSRP, and improving SINR. The reward value for startup acceleration is 80 points, the reward value for improving RSRP is 85 points, and the reward value for improving SINR is 87 points. For first scene perception data 2, the learning model includes one optimization strategy: improving RSRP, with a reward value of 87 points.

[0059] If, for example, in at least one of the first scene perception data, the first scene perception data corresponding to the scene perception data is first scene perception data 1, then the optimization strategy with the highest reward value is selected from the three optimization strategies corresponding to the first scene perception data 1, that is, improving SINR is selected as the target optimization strategy.

[0060] In the embodiments of this application, by taking the optimization strategy with the highest reward value among the at least one optimization strategy corresponding to the scene perception data in at least one first scene perception data of the first learning model as the network optimization strategy corresponding to the scene perception data, the optimal network optimization strategy corresponding to the scene perception data can be quickly determined based on previous prior experience, thereby improving the optimization efficiency and accuracy of the network signal.

[0061] Step 140: Control the electronic device to execute network optimization strategies.

[0062] In some embodiments of this application, after obtaining the network optimization strategy corresponding to the scene perception data, the electronic device can be controlled to execute the network optimization strategy, that is, the network signal of the electronic device is optimized according to the network optimization strategy.

[0063] In some embodiments of this application, after step 140, the method described above may further include: Obtain network signal quality feedback information from electronic devices; Based on network signal quality feedback information, determine the reward value for the network optimization strategy; Based on the reward value of the network optimization strategy, the reward value of the network optimization strategy corresponding to the scene perception data in the first learning model is updated to obtain the second learning model.

[0064] Among them, network signal quality feedback information can be information fed back by electronic devices to characterize network signal quality, such as the RSRP value, SINR, whether the electronic device experiences call stuttering, video stuttering, data transmission interruption, etc.

[0065] The second learning model can be obtained by updating the reward value of the network optimization strategy corresponding to the scene-aware data in the first learning model.

[0066] In some embodiments of this application, after optimizing the network signal of an electronic device according to a network optimization strategy, the network signal quality of the electronic device can be detected to obtain network signal quality feedback information. Specifically, this can involve checking whether the problems before network optimization still occur, and whether new problems arise. For example, it can detect whether the RSRP value of the electronic device is lower than a certain threshold, whether the SINR value is lower than a certain threshold, and whether the electronic device experiences call stuttering, video stuttering, or data transmission interruption.

[0067] Then, based on the network signal quality feedback information, the reward value of the network optimization strategy can be determined. Subsequently, based on the reward value of the network optimization strategy, the reward value of the network optimization strategy corresponding to the scene perception data in the first learning model can be updated to obtain the second learning model.

[0068] In the embodiments of this application, by detecting the network signal quality of electronic devices and continuously optimizing and iterating the first learning model, the first learning model is guided to generate stable and efficient optimization strategies, which can effectively improve the network signal in the preset scenario, reduce signal interruption, lag and other problems, meet the user's communication needs in the preset scenario, and improve user satisfaction.

[0069] In some embodiments of this application, the aforementioned network signal quality feedback information may include at least one of the following: voice call clarity, video stuttering rate, and data transmission rate of the electronic device. The step of determining the reward value of the network optimization strategy based on the network signal quality feedback information may specifically include: The reward value of the network optimization strategy is determined using an activation function based on voice call clarity, video stuttering rate, and data transmission rate.

[0070] In some embodiments of this application, the reward value of the network optimization strategy can be determined using an incentive function based on voice call clarity, video stuttering rate, and data transmission rate. The incentive function here can be any existing incentive function used for technical incentive values, and is not limited in the embodiments of this application.

[0071] In the embodiments of this application, by using actual user network experience indicators such as voice call clarity, video stuttering rate, and data transmission rate as reward values ​​for network optimization strategies, the optimized network can better meet user needs, satisfy user communication needs in preset scenarios, and improve user satisfaction.

[0072] In some embodiments of this application, if there is no first scene-aware data corresponding to the scene-aware data in the first learning model, then there is no network optimization strategy corresponding to the scene-aware data in the learning model. In this case, a default network optimization strategy can be selected, or available network optimization strategies can be randomly tried. The first learning model can then be updated and expanded based on the default network optimization strategy or the randomly tried available network optimization strategies. Specifically, after adopting the default network optimization strategy or randomly trying available network optimization strategies, the reward value of the optimization strategy can be calculated based on voice call clarity, video stuttering rate, and data transmission rate, thereby updating and expanding the first learning model.

[0073] In some embodiments of this application, the methods described above may further include: Store at least one network optimization strategy for each scene perception data in the first learning model, as well as the reward value corresponding to each network optimization strategy.

[0074] In some embodiments of this application, after initializing the first learning model corresponding to the preset scene, or after updating the reward value of the network optimization strategy corresponding to the scene perception data in the first learning model, the first learning model can be stored on the electronic device so that when the electronic device re-enters the preset scene, the first learning model can be directly called without rebuilding the learning.

[0075] When storing the first learning model on an electronic device, it is possible to store only at least one network optimization strategy for each scene perception data in the preset scenario, as well as the reward value corresponding to each network optimization strategy. No other information needs to be stored, nor is it necessary to cover all theoretically possible scene perception data. Most scene perception data that would not appear in real scenarios can be eliminated.

[0076] In the embodiments of this application, when storing the first learning model on the electronic device, only at least one network optimization strategy for each scenario perception data in the preset scenario, and the reward value corresponding to each network optimization strategy, can be stored. No other redundant information needs to be stored. This keeps the storage volume of the scenario-specific model within a reasonable range, achieving lightweight storage of the learning model. Furthermore, the network optimization model for each preset scenario is stored independently on the local device. Each time the learning model is called, only one preset scenario's network optimization model is invoked. All models are stored independently, without interference, avoiding cross-contamination of model parameters from different preset scenarios and ensuring the independence of each model's optimization strategy.

[0077] To facilitate understanding of the above network signal optimization method, the network optimization method provided in the embodiments of this application will be described below with specific scenarios.

[0078] like Figure 2 As shown, the network signal optimization method provided in this application embodiment may include steps 201-214.

[0079] Step 201: When the electronic device is detected to have entered a preset scene, collect scene perception data of the electronic device in the preset scene.

[0080] Step 201 is the same as step 110 in the above embodiment, and will not be described again here.

[0081] Step 202: Generate identification information corresponding to the preset scene based on the scene perception data.

[0082] Step 202 is consistent with the process of generating identification information corresponding to the preset scene based on scene perception data in the above embodiment, and will not be described again here.

[0083] Step 203: Determine whether there is a learning model corresponding to the identification information in at least one learning model. If yes, proceed to step 204; otherwise, proceed to step 205.

[0084] Step 204: Select at least one learning model that corresponds to the identification information as the first learning model.

[0085] Steps 203-204 are the same as in the above embodiment, where, in the case where there is a learning model in at least one learning model that corresponds to the identification information, the learning model corresponding to the identification information is taken as the first learning model. Therefore, they will not be described again here.

[0086] Step 205: Based on the scene perception data, initialize the initial learning model corresponding to the preset scene.

[0087] Steps 203 and 205 are the same as those in the above embodiments, in which, when there is no learning model corresponding to the identification information in at least one learning model, the initial learning model corresponding to the preset scene is initialized based on the scene perception data, and will not be described again here.

[0088] Step 206: Determine whether the network signal quality of the electronic device meets the preset conditions. If not, proceed to step 212; if yes, proceed to step 207.

[0089] Step 207: Determine whether there is a network optimization strategy in the electronic device that corresponds to the scene perception data. If yes, proceed to step 208; otherwise, proceed to step 209.

[0090] Step 208: Select the network optimization strategy with the highest reward value from at least one network optimization strategy corresponding to the scene perception data as the network optimization strategy corresponding to the scene perception data.

[0091] Steps 206-208 above, that is, in the above embodiment, when it is determined that the network signal quality of the electronic device meets the preset conditions, the optimization strategy with the highest reward value among at least one optimization strategy of the first scene perception data corresponding to the scene perception data in at least one first scene perception data of the first learning model is taken as the network optimization strategy corresponding to the scene perception data.

[0092] Step 209: Select a traditional network optimization strategy.

[0093] In steps 206, 207 and 209, if there is no network optimization strategy corresponding to the scene-aware data in the first learning model, the default general optimization strategy can be adopted or other available optimization strategies can be tried randomly, as in the above embodiments.

[0094] Step 210: Obtain network signal quality feedback information from electronic devices.

[0095] Step 211: Determine the reward value of the network optimization strategy based on the network signal quality feedback information.

[0096] Step 212: Store the scene-aware data and the reward value of the network optimization strategy.

[0097] In step 212, the current scene perception data and the reward value of the corresponding network optimization strategy can be recorded and stored. Specifically, when storing the current scene perception data, only information strongly related to the preset scene can be stored, and redundant parameters can be removed. For example, the running status information and wireless network environment information can be stored, and the spatial location information can be removed.

[0098] Step 213: Determine whether the electronic device has left the preset scene. If yes, return to step 206; otherwise, proceed to step 214.

[0099] Step 214: Update the first learning model to obtain the second learning model.

[0100] In step 214, after determining the reward value of the network optimization strategy based on the network signal quality feedback information, the reward value of the network optimization strategy corresponding to the scene perception data in the first learning model can be updated according to the reward value of the network optimization strategy to obtain the second learning model.

[0101] If the network signal quality of an electronic device does not meet the preset conditions, that is, the network signal quality of the electronic device is very good and does not need to be optimized, the reward value of the network signal quality of the electronic device under the current scene perception data can be calculated. Then, the current scene perception data and the reward value of the network signal quality of the electronic device under the current scene perception data can be recorded. Then, the first learning model can be updated based on the current scene perception data and the reward value of the network signal quality of the electronic device under the current scene perception data.

[0102] In the embodiments of this application, by constructing a dedicated learning model for each mobile closed scenario, and by binding a unique identifier and iteratively optimizing the learning model on the edge, the learning model can accurately adapt to the operating characteristics of different mobile closed scenarios. This avoids problems such as insignificant optimization effects and increased signal fluctuations caused by insufficient scenario adaptation of general models, and ensures that the optimal optimization strategy can be output under various mobile closed scenarios.

[0103] Furthermore, in this embodiment, the learning model initialization, iterative optimization, and storage scheduling are completed using the local computing power of the electronic device, without the need for cloud data transmission and computing power support. On the one hand, this avoids the problem of untimely optimization caused by cloud communication delays, and the learning model wake-up and policy execution delays are controlled within a preset threshold, adapting to dynamic scenarios that run rapidly in closed mobile scenarios. On the other hand, it reduces the cost of cloud resource occupation and avoids the risk of failure when the cloud optimization policy cannot be delivered when the network is interrupted, thus improving the operational stability of the solution.

[0104] In this embodiment, through lightweight storage processing of learning models and independent storage and scheduling mechanisms for learning models, the storage volume of a dedicated learning model for a single closed mobile scenario is controllable. Electronic devices can store multiple learning models corresponding to closed mobile scenarios in parallel without parameter cross-contamination. This will not excessively occupy the storage and computing resources of electronic devices, nor will it affect the normal battery life and other functions of electronic devices, thus achieving a balance between personalized optimization and end-side resource overhead.

[0105] In this embodiment, the reinforcement learning model uses actual network experience metrics as the core reward basis, while incorporating an adaptive penalty term to guide the learning model to generate stable and efficient optimization strategies. This can effectively improve the clarity of voice calls, the smoothness of video, and the data transmission rate in closed mobile scenarios, reduce signal interruptions and stuttering, meet users' communication needs in closed mobile scenarios, and improve user satisfaction.

[0106] The solution in this application embodiment does not require additional hardware deployment. It can be implemented by optimizing the existing hardware resources and software algorithms of electronic devices. It can be adapted to various closed mobile scenarios and is not limited by the age, structure, or base station coverage of the closed mobile scenarios. It does not require modification of the closed mobile scenarios and communication networks, has low implementation costs, and has broad application prospects and promotion value.

[0107] The network signal optimization method provided in this application can be executed by a network signal optimization device. This application uses the example of a network signal optimization device executing an information processing method to illustrate the network signal optimization device provided in this application.

[0108] Figure 3 This is a schematic diagram illustrating the structure of a network signal optimization device according to an exemplary embodiment. For example... Figure 3 As shown, the network signal optimization device 300 may include: The acquisition module 310 is used to acquire scene perception data of the electronic device in the preset scene when the electronic device enters the preset scene; The determining module 320 is used to determine the corresponding first learning model based on the scene perception data; Output module 330 is used to output a network optimization strategy corresponding to the scene perception data through the first learning model when the network signal quality of the electronic device meets the preset conditions. The control module 340 is used to control the electronic device to execute the network optimization strategy.

[0109] In the embodiments of this application, when an electronic device is detected to have entered a preset scene, a first learning model corresponding to the preset scene is determined based on the scene perception data of the electronic device in the preset scene. The first learning model corresponding to the preset scene is used to output a dedicated network optimization strategy corresponding to the scene perception data, and then the network signal of the electronic device is optimized according to the network optimization strategy. In this way, the solution of the embodiments of this application can select the first learning model corresponding to the scene in which the electronic device is located, and then take a targeted network signal optimization strategy, rather than using a single general optimization strategy. This can overcome the problem that the traditional "one-size-fits-all" optimization scheme cannot adapt to the differences in different closed mobile scenes, resulting in poor optimization effect.

[0110] In some embodiments of this application, the electronic device stores at least one learning model; the determining module 320 is specifically used for: Based on the scene perception data, generate identification information corresponding to the preset scene; The learning model corresponding to the identification information in the at least one learning model is used as the first learning model.

[0111] In some embodiments of this application, the scene perception data includes at least: the operating status information corresponding to the preset scene, the spatial location information of the electronic device, and the wireless network environment information corresponding to the preset scene; The determining module 320 is specifically used for: The operating status information, spatial location information, and wireless network environment information are fused together to generate identification information corresponding to the preset scenario.

[0112] In some embodiments of this application, the apparatus further includes: An initialization module is used to initialize an initial learning model corresponding to the preset scene based on the scene perception data when there is no learning model corresponding to the identification information in the at least one learning model.

[0113] In some embodiments of this application, the apparatus further includes: The acquisition module is used to acquire network signal quality feedback information of the electronic device after the electronic device is controlled to execute the network optimization strategy; The determining module 320 is further configured to determine the reward value of the network optimization strategy based on the network signal quality feedback information; The update module is used to update the reward value of the network optimization strategy corresponding to the scene perception data in the first learning model according to the reward value of the network optimization strategy, so as to obtain the second learning model.

[0114] In some embodiments of this application, the network signal quality feedback information includes at least one of the following: the voice call clarity, video stuttering rate, and data transmission rate of the electronic device; The determining module 320 is specifically used for: The reward value of the network optimization strategy is determined using an incentive function based on the voice call clarity, the video stuttering rate, and the data transmission rate.

[0115] The network signal optimization device in this application embodiment can be an electronic device or a component within an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. It can also be a server, network attached storage (NAS), personal computer (PC), television (TV), ATM, or self-service machine, etc. This application embodiment does not specifically limit the device.

[0116] The network signal optimization device in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit the specific operating system used.

[0117] The network signal optimization device provided in this application embodiment can achieve... Figure 1 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0118] Optionally, such as Figure 4 As shown, this application embodiment also provides an electronic device 400, including a processor 401 and a memory 402. The memory 402 stores a program or instructions that can run on the processor 401. When the program or instructions are executed by the processor 401, they implement the various steps of the above-described network signal optimization method embodiment and can achieve the same technical effect. To avoid repetition, they will not be described again here.

[0119] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0120] Figure 5 A schematic diagram of the hardware structure of an electronic device to implement an embodiment of this application.

[0121] The electronic device 500 includes, but is not limited to, components such as: radio frequency unit 501, network module 502, audio output unit 503, input unit 504, sensor 505, display unit 506, user input unit 507, interface unit 508, memory 509, and processor 510.

[0122] Those skilled in the art will understand that the electronic device 500 may also include a power supply (such as a battery) for supplying power to various components. The power supply may be logically connected to the processor 510 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system. Figure 5 The electronic device structure shown does not constitute a limitation on the electronic device. The electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements, which will not be elaborated here.

[0123] The processor 510 is configured to: collect scene perception data of the electronic device in the preset scene when the electronic device enters the preset scene; determine a corresponding first learning model based on the scene perception data; output a network optimization strategy corresponding to the scene perception data through the first learning model when the network signal quality of the electronic device meets preset conditions; and control the electronic device to execute the network optimization strategy.

[0124] Thus, when an electronic device is detected to have entered a preset scene, a first learning model corresponding to the preset scene is determined based on the scene perception data of the electronic device in the preset scene. The first learning model corresponding to the preset scene is used to output a dedicated network optimization strategy corresponding to the scene perception data, and then the network signal of the electronic device is optimized according to the network optimization strategy. In this way, the solution of this application embodiment can select the first learning model corresponding to the scene in which the electronic device is located, and then take a targeted network signal optimization strategy, rather than using a single general optimization strategy. This can overcome the problem that the traditional "one-size-fits-all" optimization scheme cannot adapt to the differences in different closed mobile scenes, resulting in poor optimization effect.

[0125] Optionally, the electronic device stores at least one learning model; the processor 510 is further configured to generate identification information corresponding to the preset scene based on the scene perception data; and to use the learning model corresponding to the identification information among the at least one learning model as the first learning model.

[0126] Optionally, the scene perception data includes at least: the operating status information corresponding to the preset scene, the spatial location information of the electronic device, and the wireless network environment information corresponding to the preset scene; the processor 510 is further configured to fuse the operating status information, the spatial location information, and the wireless network environment information to generate identification information corresponding to the preset scene.

[0127] Optionally, the processor 510 is further configured to initialize an initial learning model corresponding to the preset scene based on the scene perception data if there is no learning model corresponding to the identification information in the at least one learning model.

[0128] Optionally, the processor 510 is further configured to acquire network signal quality feedback information of the electronic device; determine the reward value of the network optimization strategy based on the network signal quality feedback information; and update the reward value of the network optimization strategy corresponding to the scene perception data in the first learning model based on the reward value of the network optimization strategy to obtain a second learning model.

[0129] Optionally, the network signal quality feedback information includes at least one of the following: voice call clarity, video stuttering rate, and data transmission rate of the electronic device; the processor 510 is further configured to determine the reward value of the network optimization strategy using an activation function based on the voice call clarity, the video stuttering rate, and the data transmission rate.

[0130] It should be understood that, in this embodiment, the input unit 504 may include a graphics processing unit (GPU) 5041 and a microphone 5042. The GPU 5041 processes image data of still images or videos obtained by an image capture device (such as a color camera) in video capture mode or image capture mode. The display unit 506 may include a display panel 5061, which may be configured in the form of a liquid crystal display, an organic light-emitting diode, or the like. The user input unit 507 includes at least one of a touch panel 5071 and other input devices 5072. The touch panel 5071 is also called a touch screen. The touch panel 5071 may include a touch detection device and a touch controller. Other input devices 5072 may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, and joysticks, which will not be described in detail here.

[0131] The memory 509 can be used to store software programs and various data. The memory 509 may primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, the memory 509 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DRRAM). The memory 509 in this embodiment includes, but is not limited to, these and any other suitable types of memory.

[0132] Processor 510 may include one or more processing units; optionally, processor 510 integrates an application processor and a modem processor, wherein the application processor mainly handles operations involving the operating system, user interface, and applications, and the modem processor mainly handles wireless communication signals, such as a baseband processor. It is understood that the aforementioned modem processor may also not be integrated into processor 510.

[0133] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described network signal optimization method embodiments and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0134] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0135] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above network signal optimization method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0136] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0137] This application provides a computer program product, which is stored in a storage medium and executed by at least one processor to implement the various processes of the network signal optimization method embodiments described above, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0138] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0139] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0140] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A network signal optimization method, characterized in that, Performed by an electronic device, the method includes: When the electronic device enters a preset scene, scene perception data of the electronic device in the preset scene is collected; Based on the scene perception data, determine the corresponding first learning model; When the network signal quality of the electronic device meets the preset conditions, the first learning model outputs a network optimization strategy corresponding to the scene perception data. Control the electronic device to execute the network optimization strategy.

2. The method according to claim 1, characterized in that, The electronic device stores at least one learning model; determining the corresponding first learning model based on the scene perception data includes: Based on the scene perception data, generate identification information corresponding to the preset scene; The learning model corresponding to the identification information in the at least one learning model is used as the first learning model.

3. The method according to claim 2, characterized in that, The scene perception data includes at least: the operating status information corresponding to the preset scene, the spatial location information of the electronic device, and the wireless network environment information corresponding to the preset scene; The step of generating identification information corresponding to the preset scene based on the scene perception data includes: The operating status information, spatial location information, and wireless network environment information are fused together to generate identification information corresponding to the preset scenario.

4. The method according to claim 2, characterized in that, The method further includes: If no learning model corresponding to the identification information exists in the at least one learning model, the initial learning model corresponding to the preset scene is initialized based on the scene perception data.

5. The method according to claim 1, characterized in that, After controlling the electronic device to execute the network optimization strategy, the method further includes: Obtain network signal quality feedback information from the electronic device; Based on the network signal quality feedback information, determine the reward value of the network optimization strategy; Based on the reward value of the network optimization strategy, the reward value of the network optimization strategy corresponding to the scene perception data in the first learning model is updated to obtain the second learning model.

6. The method according to claim 5, characterized in that, The network signal quality feedback information includes at least one of the following: voice call clarity, video stuttering rate, and data transmission rate of the electronic device; The step of determining the reward value of the network optimization strategy based on the network signal quality feedback information includes: The reward value of the network optimization strategy is determined using an incentive function based on the voice call clarity, the video stuttering rate, and the data transmission rate.

7. A network signal optimization device, characterized in that, Applied to electronic devices, the device includes: The data acquisition module is used to acquire scene perception data of the electronic device in the preset scene when the electronic device enters the preset scene; The determining module is used to determine the corresponding first learning model based on the scene perception data; The output module is used to output a network optimization strategy corresponding to the scene perception data through the first learning model when the network signal quality of the electronic device meets the preset conditions. A control module is used to control the electronic device to execute the network optimization strategy.

8. The apparatus according to claim 7, characterized in that, The electronic device stores at least one learning model; the determining module is specifically used for: Based on the scene perception data, generate identification information corresponding to the preset scene; The learning model corresponding to the identification information in the at least one learning model is used as the first learning model.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing a program or instructions that can run on the processor, the program or instructions being executed by the processor to implement the steps of the network signal optimization method as described in any one of claims 1-6.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the network signal optimization method as described in any one of claims 1-6.

Images