A wireless expansion board for automatically switching networks and a network switching method thereof

Abstract

The present invention provides a wireless expansion board for automatically switching networks and a network switching method thereof, the expansion board comprising: a SIM card module for installing a SIM card; an expansion interface module for externally connecting a general routing board or a CPE device; a wireless interface module electrically connected to the SIM card module through a SIM card interface, and electrically connected to the expansion interface module through a USB protocol; an out-of-band interface module for providing remote control services; a control module electrically connected to the wireless interface module and the out-of-band interface module, for implementing SIM card switching through the wireless interface module, and implementing remote control services through the out-of-band interface module. The present invention uses a plug-in 4G / 5G module wireless expansion board and a standard USB interface as an expansion interface, so that enterprises are no longer restricted by the support capabilities of the general routing operating system for 4G / 5G module products, thereby flexibly realizing independent access and operation and maintenance management of wireless networks.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to a wireless expansion board for automatically switching networks and a network switching method thereof. Background Art

[0002] With the advancement of wireless communication technology, users are increasingly demanding higher network performance from external universal routers or CPE (Customer Premises Equipment). Different usage scenarios and service types require different transmission speeds and power consumption. Existing network access technologies often fail to dynamically adapt to specific user needs, resulting in inadequate device performance and excessive power consumption. Furthermore, switching between networks often requires manual configuration, which is complex and inefficient.

[0003] Currently, wireless communication access network handover technology faces the following challenges: 1. The lack of an automated network handover mechanism prevents automatic adjustment of network settings based on user scenarios or service types. 2. Existing technologies struggle to balance speed and power consumption, resulting in a poor user experience. 3. Device performance and stability cannot be guaranteed during network handovers. Summary of the Invention

[0004] The present invention provides a wireless expansion board and system for automatically switching networks. The present invention can automatically switch networks according to the usage scenario or service type of an external universal routing board or CPE device, optimize transmission rate and power consumption, and improve user experience.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] In a first aspect, the present invention provides a wireless expansion board for automatically switching networks, comprising:

[0007] SIM card module, used to install a SIM card;

[0008] Expansion interface module, used for external universal routing board or CPE equipment;

[0009] a wireless interface module, electrically connected to the SIM card module via a SIM card interface, and electrically connected to the expansion interface module via a USB protocol;

[0010] Out-of-band interface module, used to provide remote control services;

[0011] a control module electrically connected to the wireless interface module and the out-of-band interface module, for implementing SIM card switching through the wireless interface module and implementing remote control services through the out-of-band interface module;

[0012] Among them, the control module implements switching of different access networks in the following manner: network classification is performed on the usage scenarios or business types of the external universal routing board or CPE device, and the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types are obtained; the usage scenario or business type characteristics of the external universal routing board or CPE device are used as input, and the corresponding transmission rate and power consumption data are used as output to construct a training set; a communication network switching model is constructed, and the communication network switching model is trained using the training set, and converged with the maximum transmission rate or minimum power consumption or rate-power consumption balance as the goal to obtain a maximum transmission rate model, a minimum power consumption model and a balance model; for the external device actually connected, its usage scenario or business type is detected, and the transmission rate and power consumption data of the external device are predicted and output using the communication network switching model after the above training is completed; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and the power supply status of the device, its access network is preliminarily selected.

[0013] In some embodiments, the control module performs refined classification of scenarios in which an external universal routing board or CPE device is used, identifies different usage scenarios or business types in which the device is located, and further divides each usage scenario or business type into different network usage levels. Then, for each usage scenario or business type, the device's transmission rate and power consumption data are recorded in detail. In the process of preparing to construct a training set, the collected data is preprocessed, and the usage scenario or business type characteristics of the external universal routing board or CPE device are quantified. According to each usage scenario or business type, they are mapped into descriptive feature vectors, which can represent the device behavior pattern in a specific scenario. Then, for each usage scenario or business type, historical data of the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types is obtained, and the transmission rate and power consumption data corresponding to these feature vectors are extracted as output parameters of the model to facilitate subsequent model training. Data pairing is performed to pair the extracted features with the corresponding transmission rate and power consumption data to ensure that each feature vector has a corresponding set of output data. The paired feature vectors and output data are combined to form a complete training set.

[0014] In some embodiments, the process of constructing a communication network switching model specifically includes: initializing a CNN model structure, which includes multiple convolutional layers, pooling layers, and fully connected layers; training the model using a constructed training set; the training set consists of a feature vector and corresponding label data; inputting the feature vector into the CNN model, calculating the output through forward propagation, and comparing it with the actual label data to calculate the loss function, which sets a loss component and a corresponding trade-off coefficient for each target based on different targets of maximum transmission rate, minimum power consumption, or rate-power balance; by adjusting the value of the trade-off coefficient, optimizing the model to obtain a maximum transmission rate model, a minimum power consumption model, and a balanced model, respectively, and using the Adam optimizer to update the parameters.

[0015] In some embodiments, after completing the training of the communication network switching model, the usage scenario or business type of the external device that is actually connected is detected by analyzing the real-time data stream and behavior pattern of the device; based on the identified usage scenario or business type, a corresponding feature vector is generated, which will be used as the input of the communication network switching model, and the trained communication network switching model is used to predict the transmission rate and power consumption data of the external device; after obtaining the predicted transmission rate and power consumption data of the external device, the optimal wireless communication access network is selected based on these data and the actual rate requirements and power supply status of the device.

[0016] In some embodiments, the SIM card module includes at least two SIM card holders, and the SIM card holders support 4G or 5G SIM cards.

[0017] In some embodiments, the out-of-band interface module includes two TTL serial ports.

[0018] In some embodiments, it further includes: a card adjustment module electrically connected to the SIM card module and the control module; the control module can switch the SIM card through the card adjustment module.

[0019] In a second aspect, the present invention provides a network switching method for a wireless expansion board that automatically switches networks, comprising the following steps:

[0020] Step S1: classify the usage scenarios or service types of the external universal routing board or CPE device into network classes, and obtain the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or service types;

[0021] Step S2: Using the usage scenario or service type characteristics of the external universal routing board or CPE device as input and the corresponding transmission rate and power consumption data as output to construct a training set;

[0022] Step S3: Constructing a communication network switching model, training the communication network switching model using the training set, and converging with the maximum transmission rate or minimum power consumption or rate-power consumption balance as the goal, to obtain a maximum transmission rate model, a minimum power consumption model, and a balance model;

[0023] Step S4: For the external device that is actually connected, its usage scenario or business type is detected, and the communication network switching model after the above training is used to predict and output the transmission rate and power consumption data of the external device; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and device power supply status, its access network is preliminarily selected.

[0024] In some embodiments, it further includes:

[0025] Step S5: Continuously monitor the service type and network transmission rate of the external device actually connected. When it is found that the service type needs to be changed or the network transmission rate is lower than a preset threshold, try to switch to another wireless network.

[0026] In some embodiments, the method also includes: implementing switching between different access networks: classifying the usage scenarios or business types of the external universal routing board or CPE device, and obtaining the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types; taking the usage scenario or business type characteristics of the external universal routing board or CPE device as input, and the corresponding transmission rate and power consumption data as output, to construct a training set; constructing a communication network switching model, using the training set to train the communication network switching model, and converging with the maximum transmission rate or minimum power consumption or rate-power consumption balance as the goal to obtain a maximum transmission rate model, a minimum power consumption model and a balance model; for the external device actually connected, detecting its usage scenario or business type, and using the communication network switching model after the above training is completed, predicting and outputting the transmission rate and power consumption data of the external device; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and the power supply status of the device, preliminarily selecting its access network.

[0027] In some embodiments, the method further includes: performing refined classification of scenarios in which an external universal routing board or CPE device is used, identifying different usage scenarios or business types in which the device is located, and further dividing each usage scenario or business type into different network usage levels, and then recording the transmission rate and power consumption data of the device in detail for each usage scenario or business type; in the process of preparing to construct a training set, preprocessing the collected data, quantifying the usage scenario or business type characteristics of the external universal routing board or CPE device, and mapping them into descriptive feature vectors according to each usage scenario or business type, which can represent the device behavior pattern in a specific scenario; then, for each usage scenario or business type, obtaining historical data on the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types, extracting the transmission rate and power consumption data corresponding to these feature vectors as output parameters of the model to facilitate subsequent model training; performing data pairing, pairing the extracted features with the corresponding transmission rate and power consumption data, ensuring that each feature vector has a corresponding set of output data; combining the paired feature vectors and output data to form a complete training set.

[0028] In some embodiments, the process of constructing a communication network switching model specifically includes: initializing a CNN model structure, which includes multiple convolutional layers, pooling layers, and fully connected layers; training the model using a constructed training set; the training set consists of a feature vector and corresponding label data; inputting the feature vector into the CNN model, calculating the output through forward propagation, and comparing it with the actual label data to calculate the loss function, which sets a loss component and a corresponding trade-off coefficient for each target based on different targets of maximum transmission rate, minimum power consumption, or rate-power balance; by adjusting the value of the trade-off coefficient, optimizing the model to obtain a maximum transmission rate model, a minimum power consumption model, and a balanced model, respectively, and using the Adam optimizer to update the parameters.

[0029] In some embodiments, the method further includes: after completing the training of the communication network switching model, detecting the usage scenario or business type of the external device that is actually connected by analyzing the real-time data stream and behavior pattern of the device; generating a corresponding feature vector based on the identified usage scenario or business type, which will be used as the input of the communication network switching model, and using the trained communication network switching model to predict the transmission rate and power consumption data of the external device; after obtaining the predicted transmission rate and power consumption data of the external device, selecting the optimal wireless communication access network based on these data and the actual rate requirements and power supply status of the device.

[0030] In a third aspect, the present invention further provides a computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the network switching method as described above when executing the computer program.

[0031] In a fourth aspect, the present invention further provides a readable storage medium having a computer program stored thereon, which implements the network switching method as described above when the program is executed by a processor.

[0032] The beneficial effects of this application are:

[0033] The present application provides a wireless expansion board and system for automatically switching networks, which has the following beneficial effects: 1. By using the plug-in 4G / 5G module wireless expansion board product as a unit and the standard USB interface as the docking interface, it is no longer subject to the support capability of the general routing board manufacturer's own operating system for 4G / 5G module products. Even if its general routing board does not support 4G / 5G modules, it can still realize the capability loading of wireless network. At the same time, it can also be installed in 4G or 5G modules in a plug-and-play manner according to customer needs, and the disassembly and installation of 4G and 5G modules for recycled equipment has been realized. 1. Module materials are maximized, and the plug-in 4G / 5G module wireless expansion board product itself also supports the out-of-band management function of the 4G / 5G module and is compatible with its out-of-band management system, which can flexibly realize independent access and operation and maintenance management during the wireless network operation phase; 2. It can switch the communication network of the external device according to the communication network switching model through the network switching method, thereby realizing automatic network switching without manual user intervention; dynamically adjust network settings according to user needs to achieve a balance between rate and power consumption; improve the efficiency and stability of network switching, and ensure device performance and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a schematic diagram of the wireless expansion board of this application;

[0035] Figure 2 This is a flow chart of a network switching method for a wireless expansion board that automatically switches networks. DETAILED DESCRIPTION

[0036] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0037] In order to more clearly understand the above-mentioned objects, features and advantages of the present application, the present application is further described in detail below in conjunction with the accompanying drawings and examples. It will be understood that the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. The specific embodiments described herein are merely used to explain the present application, rather than to limit the present application. Based on the described embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art fall within the scope of protection of this application.

[0038] It should be noted that, in this article, relational terms such as "first" and "second" are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

[0039] Figure 1 This is a schematic diagram of the wireless expansion board of this application.

[0040] The present application provides a wireless expansion board that automatically switches networks. As a plug-in wireless expansion product of a 4G / 5G module, the wireless expansion board can be connected to a universal routing board or CPE device through a standard USB interface as a docking interface, thereby no longer subject to the universal routing board manufacturer's own operating system's support for 4G / 5G module products. Even if its universal routing board does not support 4G / 5G modules, it can still achieve wireless network capability loading. At the same time, it can also load 4G or 5G modules in a plug-and-play manner according to customer needs, and for recycled equipment, the disassembly and assembly of 4G and 5G modules has achieved maximum utilization of module materials. In addition, the plug-in 4G / 5G module wireless expansion board product itself is also compatible with the 4G / 5G module out-of-band management function, which can flexibly achieve independent access and operation and maintenance management during the wireless network operation phase, thereby solving the above-mentioned technical problems.

[0041] Combine Figure 1 , including: a SIM card module for installing a SIM card;

[0042] Expansion interface module, used for external universal routing board or CPE equipment;

[0043] a wireless interface module, electrically connected to the SIM card module via a SIM card interface, and electrically connected to the expansion interface module via a USB protocol;

[0044] Out-of-band interface module, used to provide remote control services;

[0045] a control module electrically connected to the wireless interface module and the out-of-band interface module, for implementing SIM card switching through the wireless interface module and implementing remote control services through the out-of-band interface module;

[0046] Among them, the control module implements switching of different access networks in the following manner: network classification is performed on the usage scenarios or business types of the external universal routing board or CPE device, and the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types are obtained; the usage scenario or business type characteristics of the external universal routing board or CPE device are used as input, and the corresponding transmission rate and power consumption data are used as output to construct a training set; a communication network switching model is constructed, and the communication network switching model is trained using the training set, and converged with the maximum transmission rate or minimum power consumption or rate-power consumption balance as the goal to obtain a maximum transmission rate model, a minimum power consumption model and a balance model; for the external device actually connected, its usage scenario or business type is detected, and the transmission rate and power consumption data of the external device are predicted and output using the communication network switching model after the above training is completed; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and the power supply status of the device, its access network is preliminarily selected.

[0047] Specifically, the embodiments of this application are primarily intended to expand wireless communications for universal routing boards or CPE devices. To this end, this solution primarily includes a SIM card module, an expansion interface module, a wireless interface module, an out-of-band interface module, and a control module. The SIM card module is used to install SIM cards and supports three types of SIM cards, plug-in cards from various carriers, and patch cards from various carriers. The expansion interface module is used to connect to an external universal routing board or CPE device and provide it with wireless network communications through this interface.

[0048] The wireless interface module is a communication module based on the SIM card. It is mainly electrically connected to the SIM card module through the SIM card interface and electrically connected to the expansion interface module through the USB protocol, thereby realizing wireless network communication. The wireless interface module can simultaneously support 4G and 5G wireless network communications. In addition, the wireless interface module in the embodiment of the present application is based on the M.2 interface of the 5G module. The RESERVED pin in the M.2 interface is modified to support the unique function of Mini PCIe compatibility, making the interface compatible with 4G at the same time and realizing communication with external devices through the USB interface.

[0049] Out-of-Band (OOB) is a term used in communications to refer to sending information through an alternative communication channel rather than the primary one. OOB is a technology that combines multiple communication channels to provide more reliable communication services. OOB technology is primarily used as a backup channel. When the primary channel fails, the backup channel can take over to ensure timely information transmission. In remote control, OOB technology provides reliable remote control services, ensuring timely completion of remote control.

[0050] The control module is the core module of the expansion board. The control module is electrically connected to the wireless interface module and the out-of-band interface module, and is used to implement SIM card switching through the wireless interface module, thereby switching different access networks, and to implement remote control services through the out-of-band interface module.

[0051] Among them, the control module implements switching of different access networks in the following manner: network classification is performed on the usage scenarios or business types of the external universal routing board or CPE device, and the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types are obtained; the usage scenario or business type characteristics of the external universal routing board or CPE device are used as input, and the corresponding transmission rate and power consumption data are used as output to construct a training set; a communication network switching model is constructed, and the communication network switching model is trained using the training set, and converged with the maximum transmission rate or minimum power consumption or rate-power consumption balance as the goal to obtain a maximum transmission rate model, a minimum power consumption model and a balance model; for the external device actually connected, its usage scenario or business type is detected, and the transmission rate and power consumption data of the external device are predicted and output using the communication network switching model after the above training is completed; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and the power supply status of the device, its access network is preliminarily selected.

[0052] During the implementation of the control module, the scenarios in which external universal routers or CPE devices are used must be carefully categorized. This step involves identifying the different usage scenarios or service types for the devices, which can include factory automation, remote monitoring, energy management, intelligent transportation systems, and water and wastewater treatment. Each usage scenario or service type is further categorized into different network usage levels to ensure precise control of network switching. Subsequently, detailed data on the device's transmission rate and power consumption is recorded for each usage scenario or service type. More specifically, real-time monitoring and logging can be used to collect operational data on the devices in various usage scenarios. This data includes, but is not limited to, transmission rate and power consumption. To obtain this data, the following steps can be taken: tagging devices to identify their behavior in specific scenarios; using data collection tools to monitor the device's transmission rate and power consumption in real time; and categorizing and storing the collected data by usage scenario.

[0053] Furthermore, in the process of preparing to build the training set, the collected data must first be preprocessed; this includes quantifying the usage scenarios or service type characteristics of the external universal routing board or CPE device so that these characteristics can be used as input parameters of the model. Specifically, each usage scenario or service type is mapped into a descriptive feature vector, which can represent the device behavior pattern in a specific scenario. It can be expressed as , where x represents the feature vector describing the usage scenario or business type, Indicates the expected usage duration of the nth usage scenario or service type.

[0054] Furthermore, for each of the above usage scenarios or service types, historical data on the transmission rate and power consumption of general routing boards or CPE devices in different usage scenarios or service types is obtained. The transmission rate and power consumption data corresponding to these feature vectors are extracted and used as the output parameters of the model to facilitate subsequent model training. Data pairing is performed to pair the extracted features with the corresponding transmission rate and power consumption data to ensure that each feature vector has a corresponding set of output data. The paired feature vectors and output data are combined to form a complete training set, which can be expressed as:

[0055]

[0056]

[0057]

[0058] in and The tag data respectively represent the transmission rate and the power consumption data corresponding to each vector element of the feature vector.

[0059] In order to achieve efficient wireless communication access network switching, a deep learning algorithm is used to build a communication network switching model. Specifically, a convolutional neural network (CNN) is chosen because it performs well in processing time series data and can effectively identify complex patterns. The CNN model structure is initialized, which includes multiple convolutional layers, pooling layers, and fully connected layers, represented as ;in, represents the convolutional layer, represents the pooling layer, Represents the fully connected layer. The model is trained using the training set constructed previously. The training set consists of the feature vector and the corresponding tag data (transmission rate and power consumption ) is composed of; the eigenvector Input into the CNN model, calculate the output through forward propagation, and compare it with the actual label data to calculate the loss function. Regarding the loss function, in order to achieve different goals (maximum transmission rate, minimum power consumption or rate power balance), a multi-task learning framework is adopted to set a loss component for each goal, so that Losses, of which and is the trade-off coefficient; by adjusting and The model is optimized to obtain the maximum transmission rate model, the minimum power consumption model and the balanced model respectively. The Adam optimizer is used for parameter update because it performs well in various tasks and can converge quickly.

[0060] After completing the training of the above models, the actual connected external devices are tested for usage scenarios or business types. This step is achieved by analyzing the real-time data flow and behavior patterns of the devices. The detection process involves real-time monitoring of the data flow of the external devices, including information such as packet size, transmission frequency, application layer protocol, etc.; classifying the behavior patterns of the external devices to identify their usage scenarios or business types; and generating corresponding feature vectors based on the identified usage scenarios or business types. The vectors will serve as the input of the communication network switching model. Using the trained communication network switching model, the feature vectors Input into the model to predict the transmission rate and power consumption data of the external device. This process can be expressed as:

[0061]

[0062] It is a feature vector generated to describe the usage scenario or business type. and are the transmission rate and power consumption predicted by the model, respectively.

[0063] After obtaining the predicted transmission rate and power consumption data of the external device, the optimal wireless communication access network is selected based on these data and the actual rate requirements and power supply status of the device. The specific implementation method is as follows: First, the predicted transmission rate and power consumption The actual speed requirement of the device and power supply status Compare; the purpose of this step is to determine whether the current device performance meets user needs and evaluate whether the device's power supply capacity is sufficient to support the use of the selected network; obtain the actual rate requirements of the device and power supply status ; The predicted transmission rate and power consumption Compare with the actual demand; based on the comparison result, apply the above decision logic to determine the preliminary plan for accessing the network; based on the preliminary plan, adjust the network configuration of the device to access the optimal wireless communication network. For example, suppose the actual rate requirement of the device is 10Mbps and the power supply status is 5W. If the predicted transmission rate is 15Mbps and the power consumption is 4W, then according to the above logic, the system will select a high transmission rate network. If the predicted transmission rate is 8Mbps and the power consumption is 6W, the system will select a balanced network to ensure that the user experience is not affected by insufficient rate, nor the device overheats or the battery is exhausted due to excessive power consumption. In this way, the present invention can intelligently select the most suitable wireless communication access network based on the actual status of the device and user needs.

[0064] In some embodiments, a power interface is also included for connecting to an external power source.

[0065] Specifically, this solution can draw power from an external power source via the power interface, which can be used to draw 5V or 12V power. In addition, the expansion interface module in the above embodiment can also be used for power, allowing direct 5V power draw. The two power supply methods mentioned above can be adaptively adjusted according to actual needs.

[0066] In some embodiments, the SIM card module includes at least two SIM card holders, and the SIM card holders support 4G or 5G SIM cards.

[0067] Specifically, the SIM card module is used to install SIM cards and supports three types of SIM cards, plug-in cards from various operators, and patch cards from various operators; the expansion interface module is used to connect to an external universal routing board or CPE device and provide it with wireless network communication through this interface.

[0068] In some embodiments, the out-of-band interface module includes two TTL serial ports.

[0069] Specifically, the out-of-band interface module of the embodiment of the present application includes two TTL serial ports, which can support a communication rate of (1200~115200bps), support 8 data bits, support 1 stop bit, and support parity check.

[0070] In some embodiments, it further includes: a card adjustment module electrically connected to the SIM card module and the control module; the control module can switch the SIM card through the card adjustment module.

[0071] Specifically, in order to realize the switching of wireless network communications, the embodiment of the present application is further provided with a card adjustment module based on the above-mentioned expansion board. The card adjustment module is electrically connected to the SIM card module and the control module, so that the control module can switch the SIM card through the card adjustment module.

[0072] In some embodiments, an indicator light module is further included, which is electrically connected to the power interface and the control module and is used to indicate whether the power is turned on, as well as the wireless type and signal strength.

[0073] The embodiment of the present application also includes an indicator light module, specifically including a power indicator light, a 4G / 5G module online indicator light, a 4G / 5G module signal strength indicator light, and a card adjustment module heartbeat indicator light, which are respectively used to indicate power supply, network type, signal strength, and network switching.

[0074] In a second aspect, an embodiment of the present application further provides a network switching method for a wireless expansion board that automatically switches networks, which is used for a wireless expansion board that automatically switches networks in any of the above embodiments, comprising the following steps:

[0075] Step S1: classify the usage scenarios or service types of the external universal routing board or CPE device into network classes, and obtain the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or service types;

[0076] Step S2: Using the usage scenario or service type characteristics of the external universal routing board or CPE device as input and the corresponding transmission rate and power consumption data as output to construct a training set;

[0077] Step S3: Constructing a communication network switching model, training the communication network switching model using the training set, and converging with the maximum transmission rate or minimum power consumption or rate-power consumption balance as the goal, to obtain a maximum transmission rate model, a minimum power consumption model, and a balance model;

[0078] Step S4: For the external device that is actually connected, its usage scenario or business type is detected, and the communication network switching model after the above training is used to predict and output the transmission rate and power consumption data of the external device; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and device power supply status, its access network is preliminarily selected.

[0079] First, in step S1, scenarios involving the use of external universal routing boards or CPE devices require detailed classification. This step involves identifying the different usage scenarios or business types of the device, which can include factory automation, remote monitoring, energy management, intelligent transportation systems, water and wastewater treatment, and so on. Each business type is further divided into different network usage levels to ensure precise control of network switching. Subsequently, for each usage scenario or business type, detailed data on the device's transmission rate and power consumption is recorded. More specifically, operational data on the device in various usage scenarios can be collected through real-time monitoring and logging. This data includes, but is not limited to, transmission rate and power consumption data. To obtain this data, the following steps can be taken: tagging the device to identify its behavior in specific scenarios; using data collection tools to monitor the device's transmission rate and power consumption in real time; and categorizing and storing the collected data according to usage scenario.

[0080] Then, in step S2, in the process of preparing to build the training set, the collected data must first be preprocessed; this includes quantifying the usage scenarios or service type characteristics of the external universal routing board or CPE device so that these characteristics can be used as input parameters of the model. Specifically, each usage scenario or service type will be mapped into a descriptive feature vector, which can represent the device behavior pattern in a specific scenario. It can be expressed as , where x represents the feature vector describing the usage scenario or business type, Indicates the expected usage duration of the nth usage scenario or service type.

[0081] Furthermore, for each of the above usage scenarios or service types, historical data on the transmission rate and power consumption of general routing boards or CPE devices in different usage scenarios or service types is obtained. The transmission rate and power consumption data corresponding to these feature vectors are extracted and used as the output parameters of the model to facilitate subsequent model training. Data pairing is performed to pair the extracted features with the corresponding transmission rate and power consumption data to ensure that each feature vector has a corresponding set of output data. The paired feature vectors and output data are combined to form a complete training set, which can be expressed as:

[0082]

[0083]

[0084]

[0085] in and The tag data respectively represent the transmission rate and the power consumption data corresponding to each vector element of the feature vector.

[0086] In step S3, in order to achieve efficient wireless communication access network switching, a deep learning algorithm is used to build a communication network switching model. Specifically, a convolutional neural network (CNN) is selected because it performs well in processing time series data and can effectively identify complex patterns. The CNN model structure is initialized, which includes multiple convolutional layers, pooling layers, and fully connected layers, represented as ;in, represents the convolutional layer, represents the pooling layer, Represents the fully connected layer. The model is trained using the training set constructed previously. The training set consists of the feature vector and the corresponding tag data (transmission rate and power consumption ) is composed of; the eigenvector Input into the CNN model, calculate the output through forward propagation, and compare it with the actual label data to calculate the loss function. Regarding the loss function, in order to achieve different goals (maximum transmission rate, minimum power consumption or rate power balance), a multi-task learning framework is adopted to set a loss component for each goal, so that Losses, of which and is the trade-off coefficient; by adjusting and The model is optimized to obtain the maximum transmission rate model, the minimum power consumption model and the balanced model respectively. The Adam optimizer is used for parameter update because it performs well in various tasks and can converge quickly.

[0087] After completing the training of the above model, in step S4, the usage scenario or business type of the external device that is actually connected is detected. This step is achieved by analyzing the real-time data flow and behavior pattern of the device. The detection process involves real-time monitoring of the data flow of the external device, including information such as data packet size, transmission frequency, application layer protocol, etc.; classifying the behavior pattern of the external device to identify its usage scenario or business type; generating a corresponding feature vector based on the identified usage scenario or business type, which will serve as the input of the communication network switching model. Using the trained communication network switching model, we convert the feature vector Input into the model to predict the transmission rate and power consumption data of the external device. This process can be expressed as:

[0088]

[0089] It is a feature vector generated to describe the usage scenario or business type. and are the transmission rate and power consumption predicted by the model, respectively.

[0090] After obtaining the predicted transmission rate and power consumption data of the external device, the optimal wireless communication access network is selected based on these data and the actual rate requirements and power supply status of the device. The specific implementation method is as follows: First, the predicted transmission rate and power consumption The actual speed requirement of the device and power supply status Compare; the purpose of this step is to determine whether the current device performance meets user needs and evaluate whether the device's power supply capacity is sufficient to support the use of the selected network; obtain the actual rate requirements of the device and power supply status ; The predicted transmission rate and power consumption Compare with the actual demand; based on the comparison result, apply the above decision logic to determine the preliminary plan for accessing the network; based on the preliminary plan, adjust the network configuration of the device to access the optimal wireless communication network. For example, suppose the actual rate requirement of the device is 10Mbps and the power supply status is 5W. If the predicted transmission rate is 15Mbps and the power consumption is 4W, then according to the above logic, the system will select a high transmission rate network. If the predicted transmission rate is 8Mbps and the power consumption is 6W, the system will select a balanced network to ensure that the user experience is not affected by insufficient rate, nor the device overheats or the battery is exhausted due to excessive power consumption. In this way, the present invention can intelligently select the most suitable wireless communication access network based on the actual status of the device and user needs.

[0091] In some embodiments, it further includes:

[0092] Step S5: Continuously monitor the service type and network transmission rate of the external device actually connected. When it is found that the service type needs to be changed or the network transmission rate is lower than a preset threshold, try to switch to another wireless network.

[0093] Specifically, after completing the initial switching of the network type, the embodiment of the present application will continue to monitor the business type and network transmission rate of the external device actually connected. When it is found that the business type needs to be changed or the network transmission rate is lower than the preset threshold, it will attempt to switch to other wireless networks to ensure that the communication network can adaptively switch according to the business type and network transmission rate of the external device.

[0094] In a third aspect, an embodiment of the present application further provides a computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the network switching method as described above when executing the computer program.

[0095] In a fourth aspect, an embodiment of the present application further provides a readable storage medium on which a computer program is stored, and when the program is executed by a processor, the network switching method as described above is implemented.

[0096] Those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is intended to be within the scope of this application and to form different embodiments.

[0097] Those skilled in the art will appreciate that the description of each embodiment has its own focus, and for parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0098] Although the embodiments of the present application are described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present application, and such modifications and variations shall fall within the scope defined by the appended claims. The above are only specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and such modifications or replacements shall be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

[0099] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and such modifications or substitutions are intended to be within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. A wireless expansion board for automatically switching networks, characterized in that: include: SIM card module, used to install a SIM card; Extension interface module, used for external universal routing board or CPE equipment; A wireless interface module, electrically connected to the SIM card module via a SIM card interface, and electrically connected to the expansion interface module via a USB protocol; Out-of-band interface module for providing remote control services; A control module, electrically connected to the wireless interface module and the out-of-band interface module, for implementing SIM card switching through the wireless interface module, thereby switching different access networks, and implementing remote control services through the out-of-band interface module; Among them, the control module implements switching of different access networks in the following manner: network classification of the usage scenarios or business types of the external universal routing board or CPE device, and obtaining the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types; using the usage scenario or business type characteristics of the external universal routing board or CPE device as input, and the corresponding transmission rate and power consumption data as output, to construct a training set; constructing a communication network switching model, using the training set to train the communication network switching model, and converging with the maximum transmission rate or minimum power consumption or rate power consumption balance as the goal to obtain a maximum transmission rate model, a minimum power consumption model and a balance model; for the external device actually connected, its usage scenario or business type is detected, and the transmission rate and power consumption data of the external device are predicted and output using the communication network switching model after the above training is completed; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and the power supply status of the device, its access network is preliminarily selected.

2. A wireless expansion board for automatically switching networks according to claim 1, characterized in that: The control module classifies the scenarios of using an external universal routing board or CPE device in a refined manner, identifies different usage scenarios or business types of the device, and further divides each usage scenario or business type into different network usage levels, and then records the transmission rate and power consumption data of the device in detail for each usage scenario or business type; in the process of preparing to build a training set, the collected data is pre-processed, the usage scenario or business type characteristics of the external universal routing board or CPE device are quantified, and mapped into descriptive feature vectors according to each usage scenario or business type, and these vectors can represent the device behavior pattern in a specific scenario; then, for each usage scenario or business type, the historical data of the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or business types are obtained, and the transmission rate and power consumption data corresponding to these feature vectors are extracted as the output parameters of the model to facilitate subsequent model training; data pairing is performed to pair the extracted features with the corresponding transmission rate and power consumption data to ensure that each feature vector has a corresponding set of output data; the paired feature vectors and output data are combined to form a complete training set.

3. A wireless expansion board for automatically switching networks according to claim 2, characterized in that: The process of constructing a communication network switching model by the control module specifically includes: initializing a CNN model structure, which includes multiple convolutional layers, pooling layers and fully connected layers; training the model using a constructed training set; the training set consists of a feature vector and corresponding label data; inputting the feature vector into the CNN model, calculating the output through forward propagation, and comparing it with the actual label data to calculate the loss function, which sets a loss component and a corresponding trade-off coefficient for each target according to different targets of maximum transmission rate, minimum power consumption or rate-power consumption balance; optimizing the model by adjusting the value of the trade-off coefficient to obtain a maximum transmission rate model, a minimum power consumption model and a balanced model respectively, and using the Adam optimizer to update the parameters.

4. The wireless expansion board for automatically switching networks according to claim 3, characterized in that: After the control module completes the training of the communication network switching model, the control module detects the usage scenario or business type of the external device that is actually connected by analyzing the real-time data flow and behavior pattern of the device; generates a corresponding feature vector based on the identified usage scenario or business type, and the vector will be used as the input of the communication network switching model, and the trained communication network switching model is used to predict the transmission rate and power consumption data of the external device; after obtaining the predicted transmission rate and power consumption data of the external device, the optimal wireless communication access network is selected based on these data and the actual rate requirements and power supply status of the device.

5. The wireless expansion board for automatically switching networks according to claim 4, characterized in that: The SIM card module includes at least two SIM card holders, and the SIM card holders support 4G or 5G SIM cards.

6. The wireless expansion board for automatically switching networks according to claim 5, characterized in that: The out-of-band interface module includes two TTL serial ports.

7. The wireless expansion board for automatically switching networks according to claim 6, characterized in that: Also includes: The card adjustment module is electrically connected to the SIM card module and the control module; the control module can switch the SIM card through the card adjustment module.

8. The wireless expansion board for automatically switching networks according to claim 7, characterized in that: It also includes an indicator light module, which is electrically connected to the power interface and the control module and is used to indicate whether the power is turned on, as well as the wireless type and signal strength.

9. A network switching method for a wireless expansion board for automatically switching networks, used for a wireless expansion board for automatically switching networks according to any one of claims 1 to 8, characterized in that: The following steps are involved: Step S1: network classify the usage scenarios or service types of the external universal routing board or CPE device, and obtain the transmission rate and power consumption data of the universal routing board or CPE device in different usage scenarios or service types; Step S2: Using the usage scenario or service type characteristics of the external universal routing board or CPE device as input, and the corresponding transmission rate and power consumption data as output, to construct a training set; Step S3: construct a communication network switching model, train the communication network switching model using the training set, and converge with the maximum transmission rate or minimum power consumption or rate power consumption balance as the goal to obtain a maximum transmission rate model, a minimum power consumption model and a balance model; Step S4: For the external device actually connected, detect its usage scenario or service type, and use the communication network switching model after the above training to predict and output the transmission rate and power consumption data of the external device; and based on the above predicted transmission rate and power consumption data, combined with the actual rate requirements and the power supply status of the device, preliminarily select its access network.

10. A network switching method for a wireless expansion board that automatically switches networks according to claim 9, characterized in that: Also includes: Step S5: Continuously monitor the service type and network transmission rate of the external device actually connected, and when it is found that the service type needs to be changed or the network transmission rate is lower than a preset threshold, switch to another wireless network.

Images