A network communication method and system based on a smart sound box

Abstract

This application provides a network communication method and system based on a smart speaker, applicable to the field of communication technology. The method includes: obtaining current network communication parameter change information and current smart speaker communication requirement information based on smart speaker memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information; generating current network communication requirement adaptation parameter information based on the current network communication parameter change information and the current smart speaker communication requirement information; and generating network communication parameter optimization information based on the current network communication requirement adaptation parameter information, current network communication status information, historical network communication status information, historical network communication requirement adaptation parameter information, a network communication optimization model, and a set of network communication parameters, so as to conduct network communication through the network communication parameter optimization information. This application improves the stability and adaptability of smart speaker network communication.

Description

Technical Field

[0001] This application belongs to the field of communication technology, and in particular relates to network communication methods and systems based on smart speakers. Background Technology

[0002] The smart speaker industry is rapidly evolving towards multimodal interaction, large-scale edge model deployment, and cross-device collaboration. The widespread adoption of communication technologies such as WiFi 6 and Bluetooth 5.3 is driving continuous improvement in its network connectivity capabilities. Meanwhile, the maturity of technologies such as voice recognition and local semantic processing is making smart speakers the core interactive hub of smart homes.

[0003] In existing technologies, network communication of smart speakers mainly revolves around the configuration of fixed network communication parameters, while some adjust the signal transmission power of the smart speaker in conjunction with the real-time network signal strength.

[0004] However, the existing technology lacks dynamic adaptability in adjusting network communication parameters, resulting in low accuracy in optimizing network communication parameters. This can easily lead to excessively high network communication latency, unreasonable bandwidth allocation, and unstable signal transmission, affecting the overall user experience of smart speakers. Summary of the Invention

[0005] In view of this, embodiments of this application provide a network communication method and system based on a smart speaker, aiming to solve the problems in the prior art where the adjustment of network communication parameters cannot keep up with changes in actual user behavior, high network communication latency, unreasonable bandwidth allocation, and unstable signal transmission.

[0006] The first aspect of this application provides a network communication method based on a smart speaker, including:

[0007] Acquire information such as smart speaker memory usage, current network communication status, current user behavior, historical network communication status, historical user behavior, historical network communication demand adaptation parameters, and network communication optimization model.

[0008] Based on the smart speaker's memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information, the current network communication parameter change information and the current smart speaker communication demand information are obtained.

[0009] Based on the current network communication parameter change information and the current smart speaker communication requirement information, generate the current network communication requirement adaptation parameter information;

[0010] Based on the current network communication requirement adaptation parameter information, current network communication status information, historical network communication status information, historical network communication requirement adaptation parameter information, network communication optimization model, and preset network communication parameter set, network communication parameter optimization information is generated to conduct network communication through the network communication parameter optimization information.

[0011] A second aspect of this application provides a network communication system based on a smart speaker, comprising:

[0012] The information acquisition module is used to acquire information such as the smart speaker's memory usage, current network communication status, current user behavior, historical network communication status, historical user behavior, historical network communication demand adaptation parameters, and network communication optimization model.

[0013] The current network communication parameter change and communication demand information generation module is used to obtain current network communication parameter change information and current smart speaker communication demand information based on the smart speaker memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information.

[0014] The current network communication requirement adaptation parameter information generation module is used to generate current network communication requirement adaptation parameter information based on the current network communication parameter change information and the current smart speaker communication requirement information;

[0015] The network communication parameter optimization information generation module is used to generate network communication parameter optimization information based on the current network communication requirement adaptation parameter information, the current network communication status information, the historical network communication status information, the historical network communication requirement adaptation parameter information, the network communication optimization model, and the preset network communication parameter set, so as to conduct network communication through the network communication parameter optimization information.

[0016] A third aspect of this application provides a terminal device, the terminal device including a memory and a processor, the memory storing a computer program executable on the processor, and the processor executing the computer program to implement the steps of the network communication method based on a smart speaker as described in the first aspect above.

[0017] A fourth aspect of this application provides a computer-readable storage medium, comprising: storing a computer program, wherein when executed by a processor, the computer program implements the steps of the network communication method based on a smart speaker as described in the first aspect above.

[0018] The beneficial effects of this application embodiment compared with the prior art are: this application realizes dynamic adaptation and precise optimization of network communication parameters of smart speakers, effectively improves the stability, smoothness and adaptability of network communication, fully meets the diverse needs of users for network communication of smart speakers in different usage scenarios, and at the same time takes into account the balance between hardware resource load of smart speakers and network communication performance, thereby optimizing the overall user experience of smart speakers. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram illustrating the implementation process of the network communication method based on a smart speaker provided in Embodiment 1 of this application;

[0021] Figure 2 This is a schematic diagram illustrating the implementation process of the network communication method based on a smart speaker provided in Embodiment 2 of this application;

[0022] Figure 3 This is a schematic diagram illustrating the implementation process of the network communication method based on a smart speaker provided in Embodiment 3 of this application;

[0023] Figure 4 This is a schematic diagram illustrating the implementation process of the network communication method based on a smart speaker provided in Embodiment 4 of this application;

[0024] Figure 5 This is a schematic diagram illustrating the implementation process of the network communication method based on a smart speaker provided in Embodiment 5 of this application;

[0025] Figure 6 This is a schematic diagram illustrating the implementation process of the network communication method based on a smart speaker provided in Embodiment Six of this application;

[0026] Figure 7 This is a schematic diagram illustrating the implementation process of the network communication method based on a smart speaker provided in Embodiment 7 of this application;

[0027] Figure 8 This is a schematic diagram of the network communication system based on a smart speaker provided in an embodiment of this application;

[0028] Figure 9 This is a schematic diagram of the terminal device provided in the embodiments of this application. Detailed Implementation

[0029] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0030] To illustrate the technical solution described in this application, specific embodiments are provided below.

[0031] Figure 1 A flowchart illustrating the implementation of the network communication method based on a smart speaker provided in Embodiment 1 of this application is shown below in detail:

[0032] Step S101: Obtain the smart speaker's memory usage information, current network communication status information, current user behavior information, historical network communication status information, historical user behavior information, historical network communication demand adaptation parameter information, and network communication optimization model.

[0033] In this embodiment, the smart speaker's memory usage information refers to data such as the currently occupied memory capacity, remaining memory capacity, and memory usage growth rate during the smart speaker's operation. This data reflects the load on the smart speaker's hardware resources, thereby determining whether hardware resources will affect network communication performance. This information can be collected in real time by the smart speaker's built-in memory monitoring module. Current network communication status information refers to real-time communication data during the smart speaker's current network connection process. This data can be used to intuitively reflect the quality and stability of the current network connection. This information can be collected in real time by the smart speaker's network communication module. Current user behavior information refers to the user's current behavior records, including the type of smart speaker function invoked, function usage duration, usage frequency, and operation commands. This information can be captured in real time by the smart speaker's interaction recording module. Historical network communication status information refers to communication data recorded during the smart speaker's past network connections. This data is used to analyze the changing patterns of network communication status and assist in optimizing current network communication parameters. This information can be collected in advance before the system officially starts operating, and then updated daily to generate a continuous historical record. Historical user behavior information refers to records of past user actions with the smart speaker, including the types of smart speaker functions called, usage duration, frequency of use, and operation commands. This information is used to analyze the correlation between user habits and network communication needs. It can be synchronized with current user behavior information and updated daily to generate historical data. Historical network communication demand adaptation parameter information refers to parameter data generated by the smart speaker based on user behavior and network communication status to adapt to the network communication needs at that time. This provides a reference for generating current network communication demand adaptation parameter information. The system can continuously record parameter data from past adaptation processes and dynamically update historical records. The network communication optimization model is an intelligent optimization model built on the Transformer architecture combined with temporal prediction algorithms. It is used to generate network communication parameter optimization information adapted to the current scenario based on multi-dimensional network-related information and user behavior information. This can be achieved by training on network communication data—user behavior datasets—combining a CNN pre-trained model and an LSTM temporal model. A multi-dimensional feature fusion strategy is introduced during training to improve the model's optimization accuracy and adaptation capability for network communication parameters.The current network communication status information may include network communication latency information, network communication bandwidth utilization information, network communication packet loss rate information, and network signal strength information; the current user behavior information may include the user's current behavior of calling the smart speaker's voice interaction function, the user's current behavior of calling the smart speaker's content playback function, and the user's current behavior of calling the smart speaker's cross-device linkage function; the historical network communication status information may include historical network communication latency information, historical network communication bandwidth utilization information, historical network communication packet loss rate information, and historical network signal strength information; the historical user behavior information may include the user's past behavior of calling the smart speaker's voice interaction function, the user's past behavior of calling the smart speaker's content playback function, and the user's past behavior of calling the smart speaker's cross-device linkage function; and the historical network communication demand adaptation parameter information may include historical network communication bandwidth allocation parameter information, historical network communication latency control parameter information, and historical network communication signal transmission power parameter information.Among them, network communication latency information refers to the total time it takes for the smart speaker to send network data to the receiver and receive a response. It is an important indicator for judging the smoothness of network communication. For example, when the network communication latency for the voice interaction function is less than 100 milliseconds, the user will not experience noticeable lag. Network communication bandwidth utilization information refers to the actual percentage of network bandwidth currently being used, reflecting the utilization of network resources. For example, when the bandwidth utilization is greater than 80%, slow content loading is likely to occur. Network communication packet loss rate information refers to the percentage of data packets that are not successfully received in the network data sent by the smart speaker. When the packet loss rate is greater than 1%, it will affect the clarity of voice interaction. Network signal strength information refers to the network signal power received by the smart speaker. When the signal strength is less than -70dBm, the network connection is likely to be unstable. User behavior information for currently calling the smart speaker's voice interaction function refers to the user's current actions of controlling the smart speaker through voice commands, such as voice inquiries and voice control. The corresponding network communication requirements are low latency and high stability. User behavior information for currently calling the smart speaker's content playback function refers to the user's current actions for controlling the smart speaker through voice commands, such as voice inquiries and voice control. The network communication requirements for playing music and radio through smart speakers are high bandwidth and low packet loss. Information about a user's current use of the smart speaker's cross-device linkage function refers to the user's current actions of controlling other smart home devices through the smart speaker, which requires low latency and high reliability. Historical network communication latency, bandwidth utilization, packet loss rate, and signal strength information refer to the smart speaker's past recorded network communication latency, bandwidth utilization, packet loss rate, and signal strength data, reflecting the changing patterns of network communication status. This information can be used to compare the current network status and train personalized network communication optimization models. Records of various function calls in historical user behavior information can be used to analyze the correlation between user habits and network communication requirements, assisting in judging current communication needs and adjusting optimization strategies. Various parameter data in historical network communication requirement adaptation parameter information can be used to refer to past adaptation experience and improve the accuracy of generating current network communication requirement adaptation parameter information. Specifically, the smart speaker's memory usage information can be collected in real time through the built-in memory monitoring module; current and historical network communication status information can be collected in real time and extracted from historical records through the smart speaker's network communication module; current and historical user behavior information can be captured and accumulated through the smart speaker's interaction recording module (including the voice recognition module and operation recording module); historical network communication demand adaptation parameter information can be dynamically updated by automatically recording parameter data from past adaptation processes; and the network communication optimization model can be pre-trained and deployed on the smart speaker's edge or in the cloud, and continuously iterated and optimized based on actual usage data.

[0034] Step S102: Based on the smart speaker memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information, obtain the current network communication parameter change information and the current smart speaker communication requirement information.

[0035] In this embodiment, data preprocessing is first performed on the smart speaker's memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information to remove invalid data and outliers. Then, feature encoding is performed on various types of information to generate multi-dimensional feature vectors, which serve as the basis for subsequent analysis. Next, by combining the current and historical network communication status information, time-series comparative analysis is conducted to capture the changing trends of network communication parameters over time, thereby obtaining the changes in network communication parameters. Simultaneously, by combining the smart speaker's memory usage information, the degree of impact of memory load on network communication parameters is determined. Then, by combining the current and historical user behavior information, the correlation between the user's current usage needs and past usage habits is analyzed to identify the core network communication needs in the user's current usage scenario. Thus, by comprehensively analyzing the results from multiple aspects, the current network communication parameter change information and the current smart speaker communication needs information are obtained. Among them, the current network communication parameter change information can reflect the change range and trend of parameters such as current network communication latency, bandwidth utilization, packet loss rate, and signal strength relative to the historical average level. The current smart speaker communication demand information can clarify the network communication requirements corresponding to the user's current use of various functions of the smart speaker, such as low latency required for voice interaction and high bandwidth required for content playback.

[0036] Step S103: Generate current network communication requirement adaptation parameter information based on the current network communication parameter change information and the current smart speaker communication requirement information.

[0037] In this embodiment, it can be understood that the current network communication parameter changes reflect the actual changes in the current network operating status, and the current smart speaker communication demand information clarifies the user's current network usage needs. The combination of these two can achieve accurate adaptation of network communication parameters. Based on pre-defined feature mapping rules, the current network communication parameter changes can be correlated and matched with the current smart speaker communication demand information. Then, combined with adaptation experience from historical data on similar parameter changes and communication demand combinations, the mapping results are optimized and calibrated. Based on the calibrated results, current network communication demand adaptation parameter information, including network communication bandwidth allocation, latency control, and signal transmission power, is generated, thereby ensuring that this parameter information matches the current network status and user needs.

[0038] Step S104: Based on the current network communication requirement adaptation parameter information, current network communication status information, historical network communication status information, historical network communication requirement adaptation parameter information, network communication optimization model, and preset network communication parameter set, generate network communication parameter optimization information to perform network communication through the network communication parameter optimization information.

[0039] In this embodiment, the preset network communication parameter set can be a dataset covering commonly used parameter ranges and standard parameter values ​​in various network communication scenarios. This dataset can be used to provide a reference benchmark for network communication parameter optimization. The preset network communication parameter set can be manually pre-set. First, the current network communication demand adaptation parameter information, current network communication status information, historical network communication status information, and historical network communication demand adaptation parameter information can be fused to generate a comprehensive feature vector. This comprehensive feature vector is then input into the network communication optimization model, while the preset network communication parameter set is used as the model's reference input. The network communication optimization model then analyzes and processes the comprehensive feature vector and the preset network communication parameter set, combining the network communication optimization rules learned during model training to generate network communication parameter optimization information suitable for the current scenario. This network communication parameter optimization information can specifically adjust parameters such as network communication latency, bandwidth allocation, packet loss rate control, and signal strength. The smart speaker then uses this network communication parameter optimization information to conduct network communication, thereby improving the stability and adaptability of network communication.

[0040] The network communication method based on smart speakers provided in this application embodiment enables dynamic adaptation and precise optimization of network communication parameters of smart speakers, effectively improving the stability, smoothness and adaptability of network communication, fully meeting the diverse needs of users for network communication of smart speakers in different usage scenarios, and balancing the hardware resource load of smart speakers with network communication performance, thereby optimizing the overall user experience of smart speakers.

[0041] Figure 2 The flowchart illustrating the implementation of the network communication method based on a smart speaker provided in Embodiment 2 of this application is shown. The difference between this method and Embodiment 1 is that step S102 specifically includes:

[0042] Step S201: Perform time-series analysis and calculation based on the current network communication status information and historical network communication status information to obtain network communication delay change information, network communication bandwidth utilization change information, network communication packet loss rate change information, and network signal strength attenuation information.

[0043] In this embodiment, network communication latency information, network communication bandwidth utilization information, network communication packet loss rate information, and network signal strength information can be extracted from the current network communication status information first. Then, the corresponding historical data from the historical network communication status information can be extracted. Then, a time-series comparison algorithm is used to compare and analyze the current data with the historical data from the same period one by one, calculate the difference and rate of change of various parameters, and then generate network communication latency change information, network communication bandwidth utilization change information, network communication packet loss rate change information, and network signal strength attenuation information based on the calculation results. Among them, various change information can clearly show the specific changes of various parameters of the current network communication relative to the historical period, such as the increase in latency, the increase in bandwidth utilization, the range of packet loss rate changes, and the degree of signal strength attenuation.

[0044] Step S202: Based on the network communication delay change information, network communication bandwidth utilization change information, network communication packet loss rate change information, and network signal strength attenuation information, obtain the current network communication parameter change information.

[0045] In this embodiment, the network communication latency change information, network communication bandwidth utilization change information, network communication packet loss rate change information, and network signal strength attenuation information can be standardized first to eliminate the dimensional differences between different parameters. Then, based on the pre-set parameter weight allocation rules, corresponding weights are assigned to various types of change information (e.g., the weight of network communication latency change information is higher than that of other parameter change information). Subsequently, a weighted summation algorithm is used to calculate the various standardized change information to obtain a comprehensive network communication parameter change index. Then, based on this index and combined with the pre-set change level classification rules, the current network communication parameter change information is generated, thereby clarifying the overall change trend and degree of change of the current network communication parameters.

[0046] Step S203: Based on the smart speaker memory usage information, current user behavior information, and historical user behavior information, obtain the current smart speaker communication requirement information.

[0047] In this embodiment, the memory usage information of the smart speaker can be analyzed first to determine the current hardware load capacity of the smart speaker, thereby determining the upper limit of network communication performance that the smart speaker can support. Then, the core function call records in the current user behavior information are extracted to clarify the user's current main usage needs. Subsequently, combined with historical user behavior information, the user's past network communication needs preferences under similar memory load conditions are analyzed to explore the correlation between user habits and network communication needs. Then, based on the current hardware load capacity, the user's current core usage needs, and historical needs preferences, the current smart speaker communication needs information is comprehensively generated. This information can clarify the network communication performance requirements that the smart speaker needs to meet in the current scenario.

[0048] The network communication method based on smart speakers provided in this application accurately mines current communication needs, improves the generation accuracy of current network communication parameter change information and current smart speaker communication need information, provides reliable data support for the subsequent generation of current network communication need adaptation parameter information and network communication parameter optimization, makes network communication parameter optimization more targeted, thereby improving the adaptability and stability of smart speaker network communication and optimizing user experience.

[0049] Figure 3 The flowchart illustrating the implementation of the network communication method based on a smart speaker provided in Embodiment 3 of this application is shown. The difference between this method and Embodiment 1 is that step S104 specifically includes:

[0050] Step S301: Based on the preset network communication parameter feature extraction rules, generate network communication parameter feature information according to the preset network communication parameter set.

[0051] In this embodiment, the preset network communication parameter feature extraction rules can refer to a standardized process for processing network communication parameter data through feature standardization, dimensionality reduction, and key parameter screening. These preset rules can be manually defined. The preset network communication parameter set can be a dataset covering commonly used parameter ranges and standard parameter values ​​in various network communication scenarios, providing basic data for network communication parameter feature extraction. Again, the preset network communication parameter set can be manually defined. Multiple typical scenarios' network communication parameter data can be selected from the preset set. Feature standardization is then performed on the network communication parameter data for each scenario to eliminate dimensional differences between parameters. Dimensionality reduction is then performed using principal component analysis to retain the core features that have the greatest impact on network communication performance. Finally, a feature screening algorithm is used to extract key parameter features, generating network communication parameter feature information corresponding to each scenario.

[0052] Step S302: Perform fusion encoding processing based on the historical network communication demand adaptation parameter information and network communication parameter feature information to obtain historical network communication parameter fusion feature information.

[0053] In this embodiment, historical network communication demand adaptation parameter information may include historical network communication bandwidth allocation parameter information, historical network communication latency control parameter information, and historical network communication signal transmission power parameter information, which can reflect the empirical patterns of past network communication adaptation. Network communication parameter feature information can provide standardized core features of network parameters, and the fusion of the two can achieve complementarity between experience and features. The historical network communication demand adaptation parameter information can first be converted into structured values, and then these structured values ​​are aligned with the network communication parameter feature information to ensure that the dimensions are consistent. The aligned two types of information are then input into the encoding module of the CNN pre-trained model. Deep correlation features between the two types of information are extracted through convolution operations. The extracted deep correlation features are then concatenated to generate a feature vector of uniform dimension, thereby obtaining the historical network communication parameter fusion feature information. This information can comprehensively reflect the correlation patterns between historical adaptation parameters and network parameter features.

[0054] Step S303: Perform fusion encoding processing based on the current network communication requirement adaptation parameter information and network communication parameter feature information to obtain the current network communication parameter fusion feature information.

[0055] In this embodiment, the current network communication requirement adaptation parameter information clarifies the adaptation requirements of network communication in the current scenario, while the network communication parameter feature information provides standardized core features of network parameters. The fusion of these two elements can accurately capture the matching relationship between the current adaptation requirements and network parameters. First, the current network communication requirement adaptation parameter information is converted into structured numerical values. Then, these structured numerical values ​​are aligned with the network communication parameter feature information to ensure consistent dimensions. The aligned information is then input into the encoding module of the CNN pre-trained model. Deep correlation features between the two types of information are extracted through convolution operations. The extracted deep correlation features are then concatenated to generate a feature vector of uniform dimension, thus obtaining the current network communication parameter fusion feature information. This information accurately reflects the matching relationship between the current adaptation requirements and network parameter features.

[0056] Step S304: Based on the historical network communication status information, network communication parameter feature information, and historical network communication parameter fusion feature information, the network communication optimization model is trained to generate a trained network communication optimization model.

[0057] In this embodiment, the network communication optimization model can be an intelligent optimization model built based on the Transformer architecture combined with a temporal prediction algorithm. This model generates network communication parameter optimization information adapted to the current scenario based on multi-dimensional network-related information. First, the encoder in the LSTM pre-trained model maps historical network communication state information to a low-dimensional temporal feature vector. Then, this low-dimensional temporal feature vector, network communication parameter features, and historical network communication parameter fusion features are concatenated to generate a comprehensive training feature vector. This comprehensive training feature vector is then input into the network communication optimization model. During training, the optimal adaptation parameters corresponding to the historical network communication state information are used as labels. The error between the predicted adaptation parameters output by the network communication optimization model and the labels is calculated. The learnable parameters of the model are updated through backpropagation. This training process is repeated until the error converges, resulting in a trained network communication optimization model. This model can accurately learn the correlation between historical network communication states, network parameter features, and adaptation parameters.

[0058] Step S305: Based on the current network communication status information, the current network communication parameter fusion feature information, and the trained network communication optimization model, generate network communication parameter optimization information to perform network communication through the network communication parameter optimization information.

[0059] In this embodiment, the current network communication state information can first be input into the encoder of the LSTM pre-trained model to obtain a low-dimensional feature vector of the current network communication state. Then, this low-dimensional feature vector of the current network communication state is concatenated with the feature information of the current network communication parameters to generate a current comprehensive feature vector. Subsequently, the current comprehensive feature vector is input into the trained network communication optimization model. The trained network communication optimization model combines the correlation rules learned during the training process to analyze and process the current comprehensive feature vector to generate network communication parameter optimization information that can adapt to the current network state and the requirements. This network communication parameter optimization information can specifically adjust key parameters such as network communication bandwidth allocation, latency control, packet loss rate control, and signal transmission power. Then, the smart speaker conducts network communication through this network communication parameter optimization information.

[0060] The network communication method based on smart speakers provided in this application enables the trained network communication optimization model to accurately learn the correlation patterns between multi-dimensional information, thereby generating network communication parameter optimization information for the current scenario, effectively improving the accuracy of network communication parameter optimization, and thus optimizing the network communication performance of smart speakers and the user experience.

[0061] Figure 4The flowchart illustrating the implementation of the network communication method based on a smart speaker provided in Embodiment 4 of this application is shown. The difference between this method and Embodiment 3 above is that step S304 specifically includes:

[0062] Step S401: Based on the historical network communication parameter fusion feature information, perform format conversion to generate historical network communication parameter fusion feature vector.

[0063] In this embodiment, the historical network communication parameter fusion feature information is a comprehensive feature information generated through fusion encoding. Its format may have inconsistent dimensions and data types, making it unsuitable for direct training of the network communication optimization model. To address this, the historical network communication parameter fusion feature information can be first cleaned to remove invalid features and outliers. Then, a feature standardization algorithm is used to convert various features into a unified data range. Subsequently, the standardized feature information undergoes vector transformation to convert it into a fixed-dimensional vector form, thereby generating a historical network communication parameter fusion feature vector. This vector eliminates format differences and meets the input requirements of the network communication optimization model.

[0064] Step S402: The historical network communication parameter fusion feature vector and the preset smart speaker communication scenario feature vector are merged to generate the network communication feature vector to be projected.

[0065] In this embodiment, the preset smart speaker communication scenario feature vector refers to the feature vectors corresponding to various typical smart speaker communication scenarios (such as voice interaction scenarios, content playback scenarios, and cross-device linkage scenarios) in the database. It can reflect the differences in the core requirements of network communication under different scenarios. The preset smart speaker communication scenario feature vector can be manually preset. First, the historical network communication parameter fusion feature vector and the preset smart speaker communication scenario feature vector can be dimensionally adapted to ensure that the two dimensions are consistent. Then, the two types of vectors are concatenated row by row, and the concatenated vector is normalized to eliminate the weight differences between different features. Then, the network communication feature vector to be projected is generated. This vector can integrate the historical adaptation parameter features and typical scenario features, enrich the feature dimensions of model training, and improve the adaptability of the network communication optimization model to different scenarios.

[0066] Step S403: Based on the historical network communication status information, network communication parameter feature information, network communication feature vector to be projected, and multiple preset network communication feature projection vectors, the network communication optimization model is trained to generate a trained network communication optimization model.

[0067] In this embodiment, multiple preset network communication feature projection vectors can be learnable parameter vectors used to map high-dimensional feature vectors to a low-dimensional feature space, enabling feature dimensionality reduction and enhancement of key information. These preset network communication feature projection vectors can be manually pre-defined. First, historical network communication state information and network communication parameter feature information are converted into standardized feature vectors. Then, these two standardized feature vectors are fused with the network communication feature vector to be projected to generate a training input feature vector. Next, the training input feature vector is multiplied by multiple preset network communication feature projection vectors to obtain a low-dimensional projection feature vector. This low-dimensional projection feature vector is then input into the network communication optimization model. Using the optimal fitting parameters corresponding to the historical network communication state information as labels, the error between the model's output prediction parameters and the labels is calculated. The parameters of the network communication optimization model and the multiple preset network communication feature projection vectors are updated through backpropagation. This training process is repeated until the error converges, thereby generating the trained network communication optimization model.

[0068] The network communication method based on smart speakers provided in this application makes the training of network communication optimization models more targeted, effectively improves the training accuracy and generalization ability of network communication optimization models, and makes the generated network communication parameter optimization information closely fit the needs of different communication scenarios, thereby optimizing the adaptability and stability of smart speaker network communication.

[0069] Figure 5 The flowchart illustrating the implementation of the network communication method based on a smart speaker provided in Embodiment 5 of this application is shown. Its difference from Embodiment 4 described above lies in:

[0070] Multiple preset network communication feature projection vectors include preset network communication feature query projection vectors, preset network communication feature key projection vectors, and preset network communication feature value projection vectors;

[0071] Step S403 specifically includes:

[0072] Step S501: Calculate the network communication feature correlation information based on the network communication feature vector to be projected, the preset network communication feature query projection vector, and the preset network communication feature key projection vector.

[0073] In this embodiment, the preset network communication feature query projection vector can be a learnable parameter vector used to extract query features from the network communication feature vector to be projected, wherein the preset network communication feature query projection vector can be manually preset; the preset network communication feature key projection vector can be a learnable parameter vector used to extract key features from the network communication feature vector to be projected, wherein the preset network communication feature key projection vector can be manually preset. First, matrix multiplication can be performed between the network communication feature vector to be projected and the preset network communication feature query projection vector to generate a network communication feature query matrix. Then, matrix multiplication can be performed between the network communication feature vector to be projected and the preset network communication feature key projection vector to generate a network communication feature key matrix. Finally, the dot product of the transpose of the network communication feature query matrix and the network communication feature key matrix is ​​calculated to obtain the original relevance matrix. To avoid the gradient vanishing problem caused by excessively high feature dimensions, each element in the original relevance matrix can be divided by the key feature dimension for scaling, thereby obtaining network communication feature relevance information. This information is used to quantify the degree of correlation between different features in the network communication feature vector to be projected.

[0074] Step S502: Normalize the network communication feature correlation information to obtain network communication feature weight information.

[0075] In this embodiment, the normalization process can employ the softmax function, which converts the network communication feature relevance information into weight values ​​conforming to a probability distribution, facilitating subsequent feature weighting and fusion. A softmax operation can be performed on each row of the network communication feature relevance information, converting the relevance value of each row into a weight value between 0 and 1, with the sum of the weight values ​​for each row being 1. This yields the network communication feature weight information, which clearly defines the influence weight of different features in the projected network communication feature vector on the optimization of network communication parameters. Features with higher weight values ​​have a greater impact on parameter optimization, thereby strengthening key features and weakening secondary features.

[0076] Step S503: Calculate the network communication feature value vector based on the network communication feature weight information and the preset network communication feature value projection vector.

[0077] In this embodiment, the preset network communication feature value projection vector can be a learnable parameter vector used to extract the midpoint features of the network communication feature vector to be projected. It can store the core feature information to be fused. The preset network communication feature value projection vector can be manually preset. First, the network communication feature vector to be projected can be multiplied with the preset network communication feature value projection vector to generate a network communication feature value matrix. This matrix stores the core value features in the network communication feature vector to be projected. Then, the network communication feature weight information is weighted and summed with the network communication feature value matrix. Each feature value is multiplied by its corresponding weight value and then summed to generate a fixed-dimensional vector, thus obtaining the network communication feature value vector. This vector can integrate the core features and feature weights, accurately reflecting the feature information that has the greatest influence on the optimization of network communication parameters.

[0078] Step S504: Based on the historical network communication status information, network communication parameter feature information, and network communication feature value vector, the network communication optimization model is trained to generate a trained network communication optimization model.

[0079] In this embodiment, historical network communication state information and network communication parameter feature information can be first converted into standardized feature vectors. Then, these two standardized feature vectors are concatenated with the network communication feature value vector to generate the final training input feature vector. This vector integrates historical network state, network parameter features, and core weighted features, providing comprehensive and accurate input data for model training. The training input feature vector is then input into the network communication optimization model. Using the optimal fitting parameters corresponding to the historical network communication state information as labels, the error between the predicted fitting parameters output by the network communication optimization model and the labels is calculated. The learnable parameters of the network communication optimization model are updated through backpropagation. The training process is repeated until the error converges, thereby generating a trained network communication optimization model. This model can more accurately learn the correlation between core features and network communication parameter optimization.

[0080] The network communication method based on smart speakers provided in this application embodiment effectively improves the training accuracy and parameter optimization of the network communication optimization model, making the generated network communication parameter optimization information fit the actual communication scenario and user needs of smart speakers, greatly improving the stability, smoothness and personalized adaptation capability of smart speaker network communication, and effectively optimizing the user experience.

[0081] Figure 6 The flowchart illustrating the implementation of the network communication method based on a smart speaker provided in Embodiment Six of this application is shown. Its difference from Embodiment One described above lies in:

[0082] The network communication parameter optimization information includes network communication bandwidth allocation parameter information, network communication delay control parameter information, network communication signal transmission power parameter information, and network communication data transmission protocol adaptation parameter information.

[0083] Following step S104, the method further includes:

[0084] Step S601: Based on the network communication bandwidth allocation parameter information, network communication delay control parameter information, network communication signal transmission power parameter information, and network communication data transmission protocol adaptation parameter information, feature extraction and cross-fusion processing are performed to generate network communication bandwidth allocation parameter feature information, network communication delay control parameter feature information, network communication signal transmission power parameter feature information, and network communication data transmission protocol adaptation parameter feature information.

[0085] In this embodiment, the feature extraction and cross-fusion process can employ network parameter feature processing technology. For example, feature standardization algorithms can be used to convert network communication bandwidth allocation parameters, network communication latency control parameters, network communication signal transmission power parameters, and network communication data transmission protocol adaptation parameters into standardized parameters with a unified data range. Key feature filtering algorithms are used to extract the core features that have the greatest impact on the network communication performance and hardware load of the smart speaker from various parameters. Then, cross-correlation analysis is performed on the extracted core features to capture the inherent correlation between different network communication parameter optimization information. Subsequently, the core features of each type of parameter are vector-encoded to generate network communication bandwidth allocation parameter feature information, network communication latency control parameter feature information, network communication signal transmission power parameter feature information, and network communication data transmission protocol adaptation parameter feature information. Among these features, each type of feature information can accurately reflect the core attributes of the corresponding network communication parameter optimization information, providing accurate feature support for the subsequent generation of smart speaker resource scheduling and process management strategies.

[0086] Step S602: Based on the preset mapping relationship between network communication optimization parameter features and smart speaker resource scheduling strategy, generate smart speaker resource scheduling strategy information according to the network communication bandwidth allocation parameter feature information and network communication latency control parameter feature information.

[0087] In this embodiment, the mapping relationship between preset network communication optimization parameter features and smart speaker resource scheduling strategies can be a rule base built based on historical network communication data and smart speaker hardware operation data. This mapping relationship can be manually preset. It can cover two core correspondences: network communication bandwidth allocation parameter features correspond to the smart speaker's CPU computing power allocation ratio, and network communication latency control parameter features correspond to the smart speaker's memory resource allocation priority. For example, if the network communication bandwidth allocation parameter features indicate a need for high bandwidth support, and the network communication latency control parameter features indicate a need for low latency assurance, then a resource scheduling strategy with high CPU computing power allocation and high memory priority is generated. If the network communication bandwidth allocation parameter features indicate a low bandwidth requirement, and the network communication latency control parameter features indicate low latency sensitivity, then a resource scheduling strategy with balanced CPU computing power allocation and normal memory priority is generated. First, the network communication bandwidth allocation parameter feature information and network communication latency control parameter feature information can be matched and queried with the preset mapping relationship. Then, the matching results can be optimized and calibrated by combining the resource scheduling effect corresponding to the same historical features. Subsequently, a smart speaker resource scheduling strategy information containing CPU computing power allocation ratio, memory resource allocation priority, network module power consumption control parameters, etc. can be generated to ensure that the resource scheduling strategy can be accurately adapted to the current network communication parameter optimization requirements and improve the utilization efficiency of smart speaker hardware resources.

[0088] Step S603: Based on the preset mapping relationship between network communication optimization parameter features and smart speaker process management strategy, generate smart speaker process management strategy information according to the network communication signal transmission power parameter feature information and network communication data transmission protocol adaptation parameter feature information.

[0089] In this embodiment, the mapping relationship between the preset network communication optimization parameter features and the smart speaker process management strategy can be a multi-dimensional rule matrix constructed based on historical network communication adaptation experience and the smart speaker process operation rules. This mapping relationship can be manually preset. It can cover two core correspondences: the network communication signal transmission power parameter features correspond to the running priority of network communication-related processes in the smart speaker, and the network communication data transmission protocol adaptation parameter features correspond to the shutdown rules for non-core background processes in the smart speaker. For example, if the network communication signal transmission power parameter features indicate a need for high power to ensure network stability, and the network communication data transmission protocol adaptation parameter features indicate a need to run a specific adaptation protocol process, then a process management strategy is generated that prioritizes network communication-related processes and shuts down all non-core background processes (such as idle push processes). If the network communication signal transmission power parameter features indicate a lower power requirement, and the network communication data transmission protocol adaptation parameter features indicate that the adaptation protocol process consumes fewer resources, then a process management strategy is generated that prioritizes network communication-related processes and selectively shuts down non-core background processes. First, the characteristic information of network communication signal transmission power parameters and network communication data transmission protocol adaptation parameters can be accurately matched with the preset mapping relationship. Then, the matching results can be fine-tuned by referring to the process management effect of similar historical features. Subsequently, process management strategy information for smart speakers, including process running priority sorting, background non-core process shutdown list, process resource usage threshold, etc., can be generated to ensure that the process management strategy can work with network communication parameter optimization to reduce the occupation of hardware resources by invalid processes.

[0090] Step S604: Based on the smart speaker resource scheduling strategy information, perform resource scheduling optimization processing on the smart speaker.

[0091] In this embodiment, the smart speaker resource scheduling strategy information is transmitted to the smart speaker's hardware resource management module. Based on the specific parameters in the strategy information, the hardware resource management module performs targeted scheduling optimization on the smart speaker's CPU computing power, memory resources, and network module power consumption. For example, based on the CPU computing power allocation ratio in the smart speaker resource scheduling strategy information, the CPU usage ratio of the smart speaker's network communication module, interaction processing module, and content decoding module is adjusted to ensure that the network communication module receives sufficient computing power support. Based on the memory resource allocation priority, memory resources are preferentially allocated to core processes related to network communication, releasing memory space occupied by low-priority processes. Based on the network module power consumption control parameters, the operating power of the network module is adjusted to reduce power consumption while ensuring network communication performance. Thus, through resource scheduling optimization, a precise match between the smart speaker's hardware resources and network communication needs is achieved, improving hardware resource utilization efficiency and network communication smoothness.

[0092] Step S605: Based on the smart speaker process management strategy information, perform process management optimization processing on the smart speaker.

[0093] In this embodiment, the process management strategy information of the smart speaker is transmitted to the process management module of the smart speaker. The process management module performs standardized management and optimization of various processes of the smart speaker according to the specific requirements in the strategy information. For example, based on the process execution priority ranking in the smart speaker process management strategy information, the execution priority of core processes related to network communication is increased to ensure that they are not preempted by resources during concurrent multi-process operation; according to the list of background non-core processes to be shut down, idle background non-core processes are shut down one by one to release the CPU, memory, and other hardware resources occupied by these processes; based on the process resource usage threshold, the resource usage of various processes is monitored in real time, and non-core processes exceeding the threshold are restricted or shut down; the resource usage of network communication-related processes is dynamically adjusted. Through process management optimization, the resource consumption of invalid processes is reduced, ensuring the stable operation of network communication-related processes and improving the stability and adaptability of the smart speaker's network communication.

[0094] The network communication method based on smart speakers provided in this application embodiment achieves deep adaptation between smart speaker hardware resources and network communication requirements, effectively improves the utilization efficiency of smart speaker hardware resources, reduces ineffective resource consumption, enhances the effect of network communication parameter optimization, makes smart speaker network communication more stable and smoother, and optimizes the overall operating efficiency and user experience of smart speakers.

[0095] Figure 7 The flowchart illustrating the implementation of the network communication method based on a smart speaker provided in Embodiment Seven of this application is shown. The difference between this method and Embodiment Six is ​​that, after step S605, the method further includes:

[0096] Step S701: Obtain optimized network communication status information.

[0097] In this embodiment, the optimized network communication status information may include optimized network communication latency information, optimized network communication bandwidth utilization information, optimized network communication packet loss rate information, and optimized network signal strength information. This information can intuitively reflect the changes in network communication performance of the smart speaker after resource scheduling optimization and process management optimization. The smart speaker's network communication module can collect various optimized network communication status data in real time while resource scheduling and process management optimization are ongoing. The collected raw data is then preprocessed to remove invalid data and outliers. Data standardization converts various data types into a unified format. The standardized data is then analyzed and calculated to generate optimized network communication status information containing the specific values ​​and trends of each key parameter. This information accurately reflects the combined effects of network communication parameter optimization, resource scheduling optimization, and process management optimization, providing real and effective data support for subsequent iterative optimization of the network communication optimization model.

[0098] Step S702: Generate an optimized network communication model based on the current network communication requirement adaptation parameter information, optimized network communication status information, current network communication status information, network communication parameter optimization information, network communication optimization model, and preset network communication parameter set.

[0099] In this embodiment, the preset network communication parameter set can be a dataset covering commonly used parameter ranges and standard parameter values ​​in various network communication scenarios. This preset network communication parameter set can be manually preset. First, the current network communication demand adaptation parameter information, optimized network communication state information, current network communication state information, and network communication parameter optimization information can be fused to extract core features from each type of information and perform vector encoding to generate a comprehensive iterative feature vector. This vector reflects the correlation between the current adaptation demand, the difference in network state before and after optimization, and the parameter optimization effect. Then, the comprehensive iterative feature vector and the preset network communication parameter set are used as supplementary inputs into the original network communication optimization model. The optimal network communication parameters corresponding to the optimized network communication state information are then used as new training labels. The error between the predicted parameters output by the original network communication optimization model and the new labels is calculated. The learnable parameters of the network communication optimization model are updated using the backpropagation algorithm. The optimization logic of the model is calibrated by combining the preset network communication parameter set, and the iterative training process is repeated until the error converges, thereby generating the optimized network communication optimization model. The optimized network communication model can integrate the actual effects of this optimization, update its own parameter optimization rules, and improve its adaptability to smart speaker network communication scenarios and user needs.

[0100] The network communication method based on smart speakers provided in this application embodiment effectively improves the accuracy and generalization ability of the network communication optimization model, so that the continuously generated network communication parameter optimization information is timely aligned with the actual operating status of the smart speaker and user needs, thereby achieving continuous optimization of the network communication performance of the smart speaker, improving the stability, smoothness and personalized adaptation capabilities of network communication, and greatly optimizing the user experience.

[0101] Corresponding to the method in the above embodiments, Figure 8 The diagram shows a structural block diagram of a network communication system based on a smart speaker provided in an embodiment of this application. For ease of explanation, only the parts related to the embodiments of this application are shown. Figure 8 The example smart speaker-based network communication system can be the execution entity of the smart speaker-based network communication method provided in the aforementioned embodiment 1.

[0102] Reference Figure 8 The smart speaker-based network communication system includes:

[0103] The information acquisition module 810 is used to acquire information such as the smart speaker's memory usage, current network communication status, current user behavior, historical network communication status, historical user behavior, historical network communication demand adaptation parameters, and network communication optimization model.

[0104] The current network communication parameter change and communication demand information generation module 820 is used to obtain the current network communication parameter change information and the current smart speaker communication demand information based on the smart speaker memory usage information, current network communication status information, current user behavior information, historical network communication status information and historical user behavior information.

[0105] The current network communication requirement adaptation parameter information generation module 830 is used to generate current network communication requirement adaptation parameter information based on the current network communication parameter change information and the current smart speaker communication requirement information;

[0106] The network communication parameter optimization information generation module 840 is used to generate network communication parameter optimization information based on the current network communication requirement adaptation parameter information, the current network communication status information, the historical network communication status information, the historical network communication requirement adaptation parameter information, the network communication optimization model, and the preset network communication parameter set, so as to conduct network communication through the network communication parameter optimization information.

[0107] The process by which each module in the network communication system based on a smart speaker, as provided in this application embodiment, implements its respective function, can be found in the foregoing. Figure 1 The description of Embodiment 1 shown will not be repeated here.

[0108] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0109] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0110] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0111] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0112] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0113] The network communication method based on smart speakers provided in this application can be applied to terminal devices such as mobile phones, tablets, wearable devices, in-vehicle devices, augmented reality / virtual reality devices, laptops, super mobile personal computers, netbooks, and personal digital assistants. This application does not impose any restrictions on the specific type of terminal device.

[0114] For example, the terminal device may be a station in a WLAN, a cellular phone, a cordless phone, a session initiation protocol phone, a wireless local loop station, a personal digital processing device, a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a vehicle networking terminal, a computer, a laptop computer, a handheld communication device, a handheld computing device, a satellite wireless device, a wireless modem card, a set-top box, a user premises equipment, and / or other devices for communication over a wireless system, as well as next-generation communication systems, such as mobile terminals in 5G networks or mobile terminals in future evolved public terrestrial mobile networks, etc.

[0115] Figure 9 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. For example... Figure 9 As shown, the terminal device 9 of this embodiment includes: at least one processor 90 ( Figure 9 Only one is shown in the image), and a memory 91 is stored in which a computer program 92 that can run on the processor 90 is stored. When the processor 90 executes the computer program 92, it implements the steps in the above-described embodiments of the network communication method based on a smart speaker, for example... Figure 1 Steps S101 to S104 are shown. Alternatively, when the processor 90 executes the computer program 92, it implements the functions of each module / unit in the above system embodiments, for example... Figure 8 The functions of modules 810 to 840 are shown.

[0116] The terminal device 9 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. The terminal device may include, but is not limited to, a processor 90 and a memory 91. Those skilled in the art will understand that... Figure 9 This is merely an example of terminal device 9 and does not constitute a limitation on terminal device 9. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal device may also include input transmission devices, network access devices, buses, etc.

[0117] The processor 90 may be a central processing unit, or it may be other general-purpose processors, digital signal processors, application-specific integrated circuits, off-the-shelf programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0118] In some embodiments, the memory 91 may be an internal storage unit of the terminal device 9, such as a hard disk or memory of the terminal device 9. The memory 91 may also be an external storage device of the terminal device 9, such as a plug-in hard disk, smart memory card, secure digital card, flash memory card, etc., equipped on the terminal device 9. Furthermore, the memory 91 may include both internal and external storage units of the terminal device 9. The memory 91 is used to store operating systems, applications, bootloaders, data, and other programs, such as the program code of computer programs. The memory 91 can also be used to temporarily store data that has been sent or will be sent.

[0119] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0120] This application also provides a terminal device, which includes at least one memory, at least one processor, and a computer program stored in the at least one memory and executable on the at least one processor. When the processor executes the computer program, it causes the terminal device to implement the steps in any of the above method embodiments.

[0121] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0122] This application provides a computer program product that, when run on a terminal device, enables the terminal device to implement the steps described in the various method embodiments above.

[0123] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0124] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0125] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0126] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A network communication method based on a smart speaker, characterized in that, include: Acquire information such as smart speaker memory usage, current network communication status, current user behavior, historical network communication status, historical user behavior, historical network communication demand adaptation parameters, and network communication optimization model. Based on the smart speaker's memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information, the current network communication parameter change information and the current smart speaker communication demand information are obtained. Based on the current network communication parameter change information and the current smart speaker communication requirement information, generate the current network communication requirement adaptation parameter information; Based on the current network communication requirement adaptation parameter information, current network communication status information, historical network communication status information, historical network communication requirement adaptation parameter information, network communication optimization model, and preset network communication parameter set, network communication parameter optimization information is generated to conduct network communication through the network communication parameter optimization information.

2. The network communication method based on a smart speaker as described in claim 1, characterized in that, The step of obtaining current network communication parameter change information and current smart speaker communication demand information based on the smart speaker memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information specifically includes: Based on the current network communication status information and historical network communication status information, time-series analysis and calculation are performed to obtain network communication latency change information, network communication bandwidth utilization change information, network communication packet loss rate change information, and network signal strength attenuation information; Based on the network communication latency change information, network communication bandwidth utilization change information, network communication packet loss rate change information, and network signal strength attenuation information, the current network communication parameter change information is obtained; Based on the smart speaker's memory usage information, current user behavior information, and historical user behavior information, the current smart speaker's communication requirements are obtained.

3. The network communication method based on a smart speaker as described in claim 1, characterized in that, The step of generating network communication parameter optimization information based on the current network communication demand adaptation parameter information, current network communication status information, historical network communication status information, historical network communication demand adaptation parameter information, network communication optimization model, and preset network communication parameter set, and then performing network communication through the network communication parameter optimization information, specifically includes: Based on preset network communication parameter feature extraction rules, network communication parameter feature information is generated according to a preset set of network communication parameters. Based on the historical network communication demand adaptation parameter information and network communication parameter feature information, a fusion encoding process is performed to obtain historical network communication parameter fusion feature information; Based on the current network communication requirement adaptation parameter information and network communication parameter feature information, a fusion encoding process is performed to obtain the current network communication parameter fusion feature information; Based on the historical network communication status information, network communication parameter feature information, and historical network communication parameter fusion feature information, the network communication optimization model is trained to generate a trained network communication optimization model. Based on the current network communication status information, the current network communication parameter fusion feature information, and the trained network communication optimization model, network communication parameter optimization information is generated to conduct network communication through the network communication parameter optimization information.

4. The network communication method based on a smart speaker as described in claim 3, characterized in that, The step of training the network communication optimization model based on the historical network communication state information, network communication parameter feature information, and historical network communication parameter fusion feature information to generate a trained network communication optimization model specifically includes: Based on the fusion feature information of the historical network communication parameters, the format is converted to generate a fusion feature vector of historical network communication parameters; The historical network communication parameters are fused into a feature vector, and the preset smart speaker communication scenario feature vector is merged to generate a network communication feature vector to be projected. Based on the historical network communication status information, network communication parameter feature information, network communication feature vector to be projected, and multiple preset network communication feature projection vectors, the network communication optimization model is trained to generate a trained network communication optimization model.

5. The network communication method based on a smart speaker as described in claim 4, characterized in that, Multiple preset network communication feature projection vectors include preset network communication feature query projection vectors, preset network communication feature key projection vectors, and preset network communication feature value projection vectors; The step of training the network communication optimization model based on the historical network communication state information, network communication parameter feature information, network communication feature vector to be projected, and multiple preset network communication feature projection vectors to generate a trained network communication optimization model specifically includes: Based on the network communication feature vector to be projected, the preset network communication feature query projection vector, and the preset network communication feature key projection vector, the network communication feature correlation information is calculated. The network communication feature correlation information is normalized to obtain network communication feature weight information; The network communication feature value vector is calculated based on the network communication feature weight information and the preset network communication feature value projection vector. Based on the historical network communication status information, network communication parameter feature information, and network communication feature value vector, the network communication optimization model is trained to generate a trained network communication optimization model.

6. The network communication method based on a smart speaker as described in claim 1, characterized in that, The network communication parameter optimization information includes network communication bandwidth allocation parameter information, network communication delay control parameter information, network communication signal transmission power parameter information, and network communication data transmission protocol adaptation parameter information. After the step of generating network communication parameter optimization information based on the current network communication demand adaptation parameter information, current network communication status information, historical network communication status information, historical network communication demand adaptation parameter information, network communication optimization model, and preset network communication parameter set, and then performing network communication through the network communication parameter optimization information, the method further includes: Based on the network communication bandwidth allocation parameter information, network communication delay control parameter information, network communication signal transmission power parameter information, and network communication data transmission protocol adaptation parameter information, feature extraction and cross-fusion processing are performed to generate network communication bandwidth allocation parameter feature information, network communication delay control parameter feature information, network communication signal transmission power parameter feature information, and network communication data transmission protocol adaptation parameter feature information. Based on the mapping relationship between preset network communication optimization parameter features and smart speaker resource scheduling strategy, smart speaker resource scheduling strategy information is generated according to the network communication bandwidth allocation parameter feature information and network communication latency control parameter feature information. Based on the mapping relationship between preset network communication optimization parameter features and smart speaker process management strategy, smart speaker process management strategy information is generated according to the network communication signal transmission power parameter feature information and the network communication data transmission protocol adaptation parameter feature information. Based on the smart speaker resource scheduling strategy information, resource scheduling optimization processing is performed on the smart speaker; Based on the smart speaker process management strategy information, process management optimization is performed on the smart speaker.

7. The network communication method based on a smart speaker as described in claim 6, characterized in that, After the step of optimizing the process management of the smart speaker based on the smart speaker process management strategy information, the method further includes: Obtain optimized network communication status information; Based on the current network communication requirement adaptation parameter information, optimized network communication status information, current network communication status information, network communication parameter optimization information, network communication optimization model, and preset network communication parameter set, an optimized network communication optimization model is generated.

8. A network communication system based on a smart speaker, characterized in that, include: The information acquisition module is used to acquire information such as the smart speaker's memory usage, current network communication status, current user behavior, historical network communication status, historical user behavior, historical network communication demand adaptation parameters, and network communication optimization model. The current network communication parameter change and communication demand information generation module is used to obtain current network communication parameter change information and current smart speaker communication demand information based on the smart speaker memory usage information, current network communication status information, current user behavior information, historical network communication status information, and historical user behavior information. The current network communication requirement adaptation parameter information generation module is used to generate current network communication requirement adaptation parameter information based on the current network communication parameter change information and the current smart speaker communication requirement information; The network communication parameter optimization information generation module is used to generate network communication parameter optimization information based on the current network communication requirement adaptation parameter information, the current network communication status information, the historical network communication status information, the historical network communication requirement adaptation parameter information, the network communication optimization model, and the preset network communication parameter set, so as to conduct network communication through the network communication parameter optimization information.

9. A terminal device, characterized in that, The terminal device includes a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images