Multi-mode fusion portable WiFi signal enhancement method

Abstract

The invention discloses a multi-mode fusion portable WiFi signal enhancement method, relates to the technical field of wireless communication, and solves the technical problems that the method cannot adapt to dynamic environment change, cannot dynamically adjust a channel according to real-time signal quality, is difficult to comprehensively analyze multi-source data, and is slow in response speed to load pressure. Signal intensity and operation data of the portable WiFi equipment are collected in real time through the system, and data support is provided for network optimization. Potential network problems can be warned in advance by analyzing the change trend of the signal intensity; through continuous trial and error learning through Q-Learning, a channel selection strategy is adjusted in real time to adapt to a dynamically changing signal environment. Traditional static channel allocation easily causes interference between adjacent grids, and Q-Learning can be dynamically switched to a low-interference channel combination. Through iterative learning in a preset time period, the model can quickly find an optimal channel strategy, and signal attenuation is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communications, and in particular relates to a multi-mode fusion portable WiFi signal enhancement method. Background Art

[0002] Portable Wi-Fi is a device that converts mobile network signals (such as 4G / 5G) into Wi-Fi signals, allowing users to wirelessly access the internet through devices such as mobile phones, tablets, and computers. Its core features include: no fixed broadband required, making it suitable for outdoor use, travel, and temporary work environments. It supports simultaneous connection of multiple devices, meeting the needs of families and teams. Compared to carriers' mobile data plans, some portable Wi-Fi devices offer more flexible data plans. With the widespread adoption of mobile internet, users' demands for network coverage and stability have increased. Early devices supporting only a single frequency band (such as 2.4GHz) have evolved to dual-band (2.4GHz + 5GHz) and multi-mode (4G / 5G) devices, gradually incorporating advanced technologies such as AI optimization and load balancing. Traditional portable Wi-Fi coverage and stability are insufficient in complex environments or areas with severe signal interference. A multi-mode integrated portable Wi-Fi signal enhancement method is needed to improve signal stability.

[0003] Portable Wi-Fi signal management solutions typically use fixed channel allocation, such as default channels, which cannot adapt to dynamic environmental changes and cannot dynamically adjust channels based on real-time signal quality. Furthermore, load analysis relies on manual inspections or user complaints, making it difficult to comprehensively analyze multi-source data and slow to respond to load pressure. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a multi-mode fusion portable WiFi signal enhancement method to address the technical problems of being unable to adapt to dynamic environmental changes, unable to dynamically adjust channels based on real-time signal quality, and having difficulty in comprehensively analyzing multi-source data and slow response to load pressure.

[0005] To solve the above problems, the first aspect of the present invention provides a multi-mode fusion portable WiFi signal enhancement method, the multi-mode fusion portable WiFi signal enhancement method is characterized by comprising the following steps: Obtain the location information of the portable WiFi managed in the area, locate the portable WiFi in the area on the regional map, divide the area into grids, and set mobile relay nodes in the grids; The portable WiFi switches to different signal connection modes according to the environmental mobile data, mobile relay node WiFi signal and the signal strength of the connectable WiFi signal in the environment, and detects the signal strength data and operation data; Receive the signal strength data and operation data of the portable WiFi in the grid through the mobile relay node and upload it to the cloud data analysis module; The cloud data analysis module obtains the portable WiFi background data through mobile relay nodes to analyze the portable WiFi connection device signal demand coefficient, channel interference coefficient and environmental complexity coefficient. Based on the mobile relay node operation data, it generates the environmental interference heat map, 3D signal heat map and device distribution map of the mobile relay nodes in the grid; Based on the location and various coefficients of the portable WiFi connected to the mobile relay node within the grid, and based on the Q-Learning model, the optimal channel combination is selected at preset time intervals. At each preset time interval, the cloud data analysis module analyzes the image data generated by different grids and the various coefficients of the portable WiFi to analyze the network load stability of the mobile relay node and provide grid support.

[0006] Optionally, in an example of the above aspect, the mobile relay node includes: a drone, and a data acquisition module and a signal relay module integrated in the drone; the data acquisition module has a built-in spectrum analysis unit and a performance monitoring interface, the spectrum analysis unit supports 2.4 / 5 GHz frequency band scanning, and detects in real time the signal strength and spectrum occupancy of the WiFi signal emitted by the signal relay module; the performance monitoring interface captures the packet loss rate and delay data of the WiFi signal through the SNMP protocol; The WiFi transmitter of the signal relay module adopts the 802.11ax protocol, sets dual-band concurrent SSID, and supports MU-MIMO antenna array.

[0007] Optionally, in an example of the above aspect, positioning the portable WiFi in the area on the area map and dividing the area into grids includes the following steps: Obtain the MAC addresses of surrounding APs through the portable WiFi device, use the known AP locations as the location of the portable WiFi, and locate it on the area map; Mark commercial and residential areas on the management area map of the portable WiFi, and divide the commercial and residential areas, as well as other unmarked public areas into grids of different densities.

[0008] Optionally, in an example of the above aspect, the portable WiFi switches different signal connection modes according to the signal strength of the environmental mobile data, the mobile relay node WiFi signal, and the environmental connectable WiFi signal, and detects the signal strength data and operation data, including the following steps: Real-time monitoring of mobile data signals, relay node WiFi signal strength, and ambient WiFi signal strength, with a sampling frequency of 200ms / time; When the signal strength of the mobile relay node WiFi signal or the environment can connect to the WiFi signal meets the signal strength requirement of higher than -55dBm and the delay is less than 100ms, the network connection mode of the portable WiFi will be automatically switched from mobile data to WiFi signal connection mode. Otherwise, the mobile data connection will be maintained. When switching to WiFi signal connection mode, select the WiFi signal with the strongest signal among the WiFi signals available in the mobile relay node and the environment; The portable WiFi detects the signal strength data of the ambient signal and the operating data of its own signal. The signal strength data includes: signal strength and spectrum occupancy; the operating data includes: the number of connected devices, the traffic demand data filled in by the user, the packet loss rate and the delay data.

[0009] Optionally, in an example of the above aspect, the cloud data analysis module obtains the portable WiFi background data through the mobile relay node to analyze the portable WiFi connection device signal demand coefficient, channel interference coefficient, and environment complexity coefficient, including the following steps: The cloud data analysis module obtains the portable WiFi background data through the mobile relay node, including: the signal strength and spectrum occupancy of the portable WiFi detection environment signal, the number of connected devices of its own signal, the traffic demand data filled in by the user, the packet loss rate and the delay data; Based on the ratio of the number of connected devices to the theoretical upper limit of the device load (for example, the theoretical upper limit Nmax is 64 devices, and when N=20, the proportion is 31%), and the ratio of the user's total traffic demand to the upper limit of the package, the signal demand coefficient of the connected device of the portable WiFi is analyzed using the following formula: DSDI is the signal demand coefficient of the connected device of the Pocket WiFi, Nmax is the theoretical carrying limit of the device, N is the number of connected devices, Qmax is the upper limit of the Pocket WiFi package, and Q is the total traffic demand filled in by the user; Weight allocation is based on the device concurrent pressure > traffic consumption characteristics; Based on the spectrum occupancy of the 2.4 / 5 GHz bands, if the spectrum occupancy of the 2.4 GHz band channel is greater than 70% or the spectrum occupancy of the 5 GHz band channel is greater than 50%, it is marked as high interference. Combined with the packet loss rate, the channel interference coefficient is analyzed using the following formula: Where CII is the channel interference coefficient of the portable WiFi, U is the weighted average of the spectrum occupancy of the 2.4-band channel and the spectrum occupancy of the 5GHz band channel, and L is the sum of the packet loss rates of the portable WiFi connected devices; According to the mobile data signal, relay node WiFi signal strength and ambient WiFi signal strength of the portable WiFi environment, the signal strength of each signal is normalized and then weighted averaged; Obtain latency data. If the Ping value is greater than 100ms, mark it as a high-latency environment. Otherwise, mark it as a non-high-latency environment. Also, analyze the environment complexity coefficient based on the PCI value using the following formula: Among them, ECI is the environmental complexity coefficient of the portable WiFi, S is the weighted average value after normalization of the signal strength of each signal, D is the Ping value of the delay data, and P is the PCI interference value.

[0010] The lower the PCI value, the less base station interference there is. For example, if PCI<10, it will be 0.9, and if PCI>30, it will be 0.3.

[0011] Optionally, in an example of the above aspect, generating an environmental interference heat map of mobile relay nodes in a grid includes: The cloud data analysis module enables spectrum scanning through the mobile relay node management backend to monitor the channel occupancy rate and interference source intensity of the 2.4GHz / 5GHz frequency band in real time; Use the NetSpot tool to perform a full-scene signal scan and automatically generate an environmental interference heat map that includes channel occupancy and interference intensity for each channel.

[0012] Optionally, in an example of the above aspect, generating a 3D signal heat map of mobile relay nodes in a grid includes: Enable the spectrum scanning function through the mobile relay node management background to monitor the signal strength of each channel in the 2.4GHz / 5GHz frequency band in real time; Import a map of the grid corresponding to the mobile relay node, mark the location of the located portable WiFi and the location of the connected device, set the wall attenuation parameters, and generate a 3D signal heat map of the mobile relay node signal in each channel in the corresponding grid.

[0013] Optionally, in an example of the above aspect, generating a device distribution map of mobile relay nodes in a grid includes: Export the connected portable WiFi list through the mobile relay node background, obtain device coordinates through WiFi positioning technology or Bluetooth beacon, and build a device connection topology map; such as triangulation positioning method; Collect the connection status, signal strength, and data usage of the devices in the portable WiFi list and map them to the corresponding coordinates of the device connection topology map. Mark the signal strength ranges at the corresponding positions of the coordinates on the device connection topology map as a device distribution map.

[0014] Optionally, in an example of the above aspect, selecting an optimal channel combination at a preset time interval based on the location and various coefficients of the portable WiFi connected to the mobile relay node in the grid and based on the Q-Learning model includes the following steps: The signal demand coefficient, channel interference coefficient, and environment complexity coefficient of the portable WiFi connected to the mobile relay node in the grid are weighted and added together to obtain the comprehensive coefficient of the portable WiFi. Defines a set of optional channels for the 2.4 GHz and 5 GHz bands; Set the reward function of the Q-Learning model. The initial value of the reward function is 1, including the incentive mechanism and the penalty mechanism. The incentive mechanism includes: a reward of +5 when the comprehensive coefficient of the portable WiFi decreases after the portable WiFi changes its channel; a penalty mechanism includes: a reward of -5 when the comprehensive coefficient of the portable WiFi increases after the portable WiFi changes its channel; a penalty of -2 when the packet loss rate exceeds 2% due to channel switching; Initialize the Q value table Q(s,a), where s is the value of the reward function and a is the switching channel in the optional channel set; Set the Q-Learning model hyperparameters: learning rate α and exploration rate ε. The exploration rate ε=0.8, decays by 5% per cycle until the steady state ε=0.1‌3. Set α=0.6 / (1+0.01*t) to implement decaying learning rate adjustment, where t represents the model training time. The Q-Learning model is trained using the past 24 hours of historical data on the performance of the portable WiFi channels in the grid; After training is completed, the reward function values ​​s in the Q value of each portable WiFi are sorted from large to small, and the reward function values ​​s in the top 10% and the corresponding a values ​​are selected as the optimal channel combination for switching channels.

[0015] Optionally, in an example of the above aspect, analyzing the network load stability of the mobile relay node and performing grid support includes the following steps: When more than 10% of the portable WiFi devices in a grid report network issues in historical data, the corresponding environmental interference heat map, 3D signal heat map, and device distribution map of the portable WiFi devices in the grid are labeled "Network Unstable" and the remaining images are labeled "Network Stable"; Train the neural network model using labeled images; At preset time intervals, the system obtains the environmental interference heat map, 3D signal heat map, and device distribution map of the portable WiFi within the grid. The trained neural network model uses these maps to identify whether the network status in the grid is stable or unstable. At every preset time interval, the portable WiFi comprehensive coefficient is obtained. If the comprehensive coefficients of more than 10% of the portable WiFi in the grid exceed 0.85, the network status in the grid is judged to be unstable; otherwise, the network status is judged to be stable; If the network status in a grid is judged to be unstable, the backup mobile relay node is called to move to the corresponding grid and provide support as an equal mobile relay node.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention uses Q-Learning, a continuous trial-and-error learning method, to adjust channel selection strategies in real time to adapt to dynamically changing signal environments. Traditional static channel allocation can easily lead to interference between adjacent grids. Q-Learning dynamically switches to low-interference channel combinations. Through iterative learning over a preset time period, the model quickly finds the optimal channel strategy and reduces signal attenuation. By analyzing the network load stability of mobile relay nodes based on various portable WiFi coefficients, it can provide timely support when grid load pressure arises, ensuring the signal strength of the portable WiFi signal within the grid.

[0017] This system collects real-time signal strength and operational data from portable Wi-Fi devices, providing data support for network optimization. By analyzing signal strength trends, it can provide early warning of potential network issues and tailor relay node layout or signal parameters to avoid wasted resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 Schematic diagram of the process of the present invention; Figure 2 This is a schematic diagram of mobile relay node connection in the present invention. DETAILED DESCRIPTION

[0020] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0021] See also Figure 1-Figure 2 The first embodiment of the present invention provides a multi-mode fusion portable WiFi signal enhancement method, comprising the following steps: Obtain the location information of the portable WiFi managed in the area, locate the portable WiFi in the area on the regional map, divide the area into grids, and set mobile relay nodes in the grids; The portable WiFi switches to different signal connection modes according to the environmental mobile data, mobile relay node WiFi signal and the signal strength of the connectable WiFi signal in the environment, and detects the signal strength data and operation data; Receive the signal strength data and operation data of the portable WiFi in the grid through the mobile relay node and upload it to the cloud data analysis module; The cloud data analysis module obtains the portable WiFi background data through mobile relay nodes to analyze the portable WiFi connection device signal demand coefficient, channel interference coefficient and environmental complexity coefficient. Based on the mobile relay node operation data, it generates the environmental interference heat map, 3D signal heat map and device distribution map of the mobile relay nodes in the grid; Based on the location and various coefficients of the portable WiFi connected to the mobile relay node within the grid, and based on the Q-Learning model, the optimal channel combination is selected at preset time intervals. At each preset time interval, the cloud data analysis module analyzes the image data generated by different grids and the various coefficients of the portable WiFi to analyze the network load stability of the mobile relay node and provide grid support.

[0022] Specifically, by locating portable Wi-Fi devices on a regional map and combining it with gridding technology, administrators can monitor device distribution in real time. The deployment of mobile relay nodes further optimizes signal coverage, ensuring device connectivity, especially in complex terrain or signal-dead zones. Portable Wi-Fi devices automatically switch to the optimal connection mode based on ambient mobile data, relay node signal strength, and surrounding Wi-Fi signal quality. This mechanism significantly improves network stability and reliability, particularly in densely populated areas or with significant signal interference. Users no longer need to manually switch networks when using portable Wi-Fi; the system automatically selects the optimal signal source to ensure continuous and stable network connectivity. This feature significantly enhances the user experience, making it particularly suitable for scenarios requiring high network quality, such as mobile office and online education. The system collects real-time signal strength and operational data from portable Wi-Fi devices to provide data support for network optimization. By analyzing signal strength trends, it can provide early warning of potential network issues and make targeted adjustments to relay node layout or signal parameters to avoid wasted resources.

[0023] By analyzing the channel interference coefficient and the environmental complexity coefficient in real time, an environmental interference heat map is generated, which can intuitively present the distribution of signal interference intensity. Interference hotspots, such as microwave ovens and Bluetooth devices, can be clearly displayed, allowing targeted channel adjustments or deployment of relay nodes to reduce signal conflicts. For example, in shopping malls or exhibition scenarios, heat maps can clearly show interference peaks in crowded areas. 3D signal heat maps combine signal strength data with device location to generate a three-dimensional signal coverage model, visually showing how signal strength changes with height and distance. Device distribution maps are updated in real time, visually presenting user density and device distribution patterns, assisting in capacity planning and resource scheduling. Statistical heat maps and device distribution maps provide an intuitive data visualization interface to help managers quickly understand the status of the grid network.

[0024] Based on the location and various coefficients of the portable WiFi connected to the mobile relay node within the grid, and based on the Q-Learning model, the optimal channel combination is selected at preset time intervals. At each preset time interval, the cloud data analysis module analyzes the image data generated by different grids and the various coefficients of the portable WiFi to analyze the network load stability of the mobile relay node and provide grid support.

[0025] Q-Learning uses continuous trial-and-error learning to adjust channel selection strategies in real time, adapting to dynamically changing signal environments. Traditional static channel allocation can easily lead to interference between adjacent grids. Q-Learning dynamically switches to low-interference channel combinations. Through iterative learning over a preset time period, the model quickly finds the optimal channel strategy and minimizes signal attenuation. For example, at large events where user density fluctuates over time, Q-Learning can dynamically switch channels to avoid network lag caused by channel overload.

[0026] By combining various coefficients of the portable WiFi and analyzing the network load stability of the mobile relay node, grid support can be provided in a timely manner when the grid is under load pressure, ensuring the signal strength of the portable WiFi signal within the grid.

[0027] In one embodiment of the present invention, a mobile relay node includes: a drone, and a data acquisition module and a signal relay module integrated in the drone; the data acquisition module has a built-in spectrum analysis unit and a performance monitoring interface. The spectrum analysis unit supports 2.4 / 5GHz frequency band scanning and detects the signal strength and spectrum occupancy of the WiFi signal emitted by the signal relay module in real time; the performance monitoring interface captures the packet loss rate and delay data of the WiFi signal through the SNMP protocol; The WiFi transmitter of the signal relay module adopts the 802.11ax protocol, sets dual-band concurrent SSID, and supports MU-MIMO antenna array.

[0028] In one embodiment of the present invention, locating a portable WiFi device within an area on an area map and dividing the area into grids includes the following steps: Obtain the MAC addresses of surrounding APs through the portable WiFi device, use the known AP locations as the location of the portable WiFi, and locate it on the area map; Mark commercial and residential areas on the management area map of the portable WiFi, and divide the commercial and residential areas, as well as other unmarked public areas into grids of different densities.

[0029] Specifically, in this embodiment, a 10-20m grid is set in commercial areas to capture small signal changes, a 100-200m grid is set in residential areas, and a 500m grid is set in other public areas.

[0030] In one embodiment of the present invention, the portable WiFi switches different signal connection modes according to the environmental mobile data, the mobile relay node WiFi signal, and the signal strength of the connectable WiFi signal, and detects the signal strength data and operation data, including the following steps: Real-time monitoring of mobile data signals, relay node WiFi signal strength, and ambient WiFi signal strength, with a sampling frequency of 200ms / time; When the signal strength of the mobile relay node WiFi signal or the environment can connect to the WiFi signal meets the signal strength requirement of higher than -55dBm and the delay is less than 100ms, the network connection mode of the portable WiFi will be automatically switched from mobile data to WiFi signal connection mode. Otherwise, the mobile data connection will be maintained. When switching to WiFi signal connection mode, select the WiFi signal with the strongest signal among the WiFi signals available in the mobile relay node and the environment; The portable WiFi detects the signal strength data of the ambient signal and the operating data of its own signal. The signal strength data includes: signal strength and spectrum occupancy; the operating data includes: the number of connected devices, the traffic demand data filled in by the user, the packet loss rate and the delay data.

[0031] The signal strength and frequency band occupancy are obtained through the built-in RSSI unit and spectrum scanning function of the WiFi chip.

[0032] For example, use the iwinfo or iw command-line tool of a Qualcomm or MediaTek WiFi chip and periodically call the sitesurvey command to scan the spectrum.

[0033] Use the DHCP client list or MAC address table to count the number of currently connected devices.

[0034] Example: In the OpenWrt system, read the lease file through cat / tmp / dhcp.leases to count the number of active IP addresses.

[0035] Use the chip traffic statistics interface or Netfilter module to monitor packet traffic.

[0036] Example: In OpenWrt, use the vnstat tool to record real-time and historical traffic.

[0037] The built-in performance monitoring interface captures the packet loss rate and latency data of the WiFi signal through the SNMP protocol.

[0038] In one embodiment of the present invention, the cloud data analysis module obtains the portable WiFi background data through the mobile relay node to analyze the portable WiFi connection device signal demand coefficient, channel interference coefficient, and environment complexity coefficient, including the following steps: The cloud data analysis module obtains the portable WiFi background data through the mobile relay node, including: the signal strength and spectrum occupancy of the portable WiFi detection environment signal, the number of connected devices of its own signal, the traffic demand data filled in by the user, the packet loss rate and the delay data; Based on the ratio of the number of connected devices to the theoretical upper limit of the device load (for example, the theoretical upper limit Nmax is 64 devices, and when N=20, the proportion is 31%), and the ratio of the user's total traffic demand to the upper limit of the package, the signal demand coefficient of the connected device of the portable WiFi is analyzed using the following formula: DSDI is the signal demand coefficient of the connected device of the Pocket WiFi, Nmax is the theoretical carrying limit of the device, N is the number of connected devices, Qmax is the upper limit of the Pocket WiFi package, and Q is the total traffic demand filled in by the user; Weight allocation is based on the device concurrent pressure > traffic consumption characteristics; Based on the spectrum occupancy of the 2.4 / 5 GHz bands, if the spectrum occupancy of the 2.4 GHz band channel is greater than 70% or the spectrum occupancy of the 5 GHz band channel is greater than 50%, it is marked as high interference. Combined with the packet loss rate, the channel interference coefficient is analyzed using the following formula: Where CII is the channel interference coefficient of the portable WiFi, U is the weighted average of the spectrum occupancy of the 2.4-band channel and the spectrum occupancy of the 5GHz band channel, and L is the sum of the packet loss rates of the portable WiFi connected devices; According to the mobile data signal, relay node WiFi signal strength and ambient WiFi signal strength of the portable WiFi environment, the signal strength of each signal is normalized and then weighted averaged; Obtain latency data. If the Ping value is greater than 100ms, mark it as a high-latency environment. Otherwise, mark it as a non-high-latency environment. Also, analyze the environment complexity coefficient based on the PCI value using the following formula: Among them, ECI is the environmental complexity coefficient of the portable WiFi, S is the weighted average value after normalization of the signal strength of each signal, D is the Ping value of the delay data, and P is the PCI interference value.

[0039] The lower the PCI value, the less base station interference there is. For example, if PCI<10, it will be 0.9, and if PCI>30, it will be 0.3.

[0040] In one embodiment of the present invention, generating an environmental interference heat map of a mobile relay node in a grid includes: The cloud data analysis module enables spectrum scanning through the mobile relay node management backend to monitor the channel occupancy rate and interference source intensity of the 2.4GHz / 5GHz frequency band in real time; Use the NetSpot tool to perform full-scene signal scanning and automatically generate an environmental interference heat map that includes channel occupancy and interference intensity for each channel. In one embodiment of the present invention, generating a 3D signal heat map of mobile relay nodes in a grid includes: Enable the spectrum scanning function through the mobile relay node management background to monitor the signal strength of each channel in the 2.4GHz / 5GHz frequency band in real time; Import the map of the grid corresponding to the mobile relay node, mark the location of the located portable WiFi and the location of the connected device, set the wall attenuation parameters, and generate a 3D signal heat map of the mobile relay node signal in each channel in the corresponding grid; In one embodiment of the present invention, generating a device distribution map of mobile relay nodes in a grid includes: Export the connected portable WiFi list through the mobile relay node background, obtain device coordinates through WiFi positioning technology or Bluetooth beacon, and build a device connection topology map; such as triangulation positioning method; Collect the connection status, signal strength, and data usage of the devices in the portable WiFi list and map them to the corresponding coordinates of the device connection topology map. Mark the signal strength ranges at the corresponding positions of the coordinates on the device connection topology map as a device distribution map.

[0041] In one embodiment of the present invention, based on the location and various coefficients of the portable WiFi connected to the mobile relay node in the grid and based on the Q-Learning model, the optimal channel combination is selected at a preset time interval, including the following steps: The signal demand coefficient, channel interference coefficient, and environment complexity coefficient of the portable WiFi connected to the mobile relay node in the grid are weighted and added together to obtain the comprehensive coefficient of the portable WiFi. Defines a set of optional channels for the 2.4 GHz and 5 GHz bands; Set the reward function of the Q-Learning model. The initial value of the reward function is 1, including the incentive mechanism and the penalty mechanism. The incentive mechanism includes: a reward of +5 when the comprehensive coefficient of the portable WiFi decreases after the portable WiFi changes its channel; a penalty mechanism includes: a reward of -5 when the comprehensive coefficient of the portable WiFi increases after the portable WiFi changes its channel; a penalty of -2 when the packet loss rate exceeds 2% due to channel switching; Initialize the Q value table Q(s,a), where s is the value of the reward function and a is the switching channel in the optional channel set; Set the Q-Learning model hyperparameters: learning rate α and exploration rate ε. The exploration rate ε=0.8, decays by 5% per cycle until the steady state ε=0.1‌3. Set α=0.6 / (1+0.01*t) to implement decaying learning rate adjustment, where t represents the model training time. The Q-Learning model is trained using the past 24 hours of historical data on the performance of the portable WiFi channels in the grid; After training is completed, the reward function values ​​s in the Q value of each portable WiFi are sorted from large to small, and the reward function values ​​s in the top 10% and the corresponding a values ​​are selected as the optimal channel combination for switching channels.

[0042] In one embodiment of the present invention, analyzing the network load stability of a mobile relay node and performing grid support includes the following steps: When more than 10% of the portable WiFi devices in a grid report network issues in historical data, the corresponding environmental interference heat map, 3D signal heat map, and device distribution map of the portable WiFi devices in the grid are labeled "Network Unstable" and the remaining images are labeled "Network Stable"; Train the neural network model using labeled images; At preset time intervals, the system obtains the environmental interference heat map, 3D signal heat map, and device distribution map of the portable WiFi within the grid. The trained neural network model uses these maps to identify whether the network status in the grid is stable or unstable. At every preset time interval, the portable WiFi comprehensive coefficient is obtained. If the comprehensive coefficients of more than 10% of the portable WiFi in the grid exceed 0.85, the network status in the grid is judged to be unstable; otherwise, the network status is judged to be stable; If the network status in a grid is judged to be unstable, the backup mobile relay node is called to move to the corresponding grid and provide support as an equal mobile relay node.

[0043] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A multi-mode fusion portable WiFi signal enhancement method, characterized in that: The following steps are involved: Obtain the location information of the portable WiFi managed in the area, locate the portable WiFi in the area on the regional map, divide the area into grids, and set mobile relay nodes in the grids; The portable WiFi switches to different signal connection modes according to the environmental mobile data, mobile relay node WiFi signal and the signal strength of the connectable WiFi signal in the environment, and detects the signal strength data and operation data; Receive the signal strength data and operation data of the portable WiFi in the grid through the mobile relay node and upload it to the cloud data analysis module; The cloud data analysis module obtains the portable WiFi background data through mobile relay nodes to analyze the portable WiFi connection device signal demand coefficient, channel interference coefficient and environmental complexity coefficient. Based on the mobile relay node operation data, it generates the environmental interference heat map, 3D signal heat map and device distribution map of the mobile relay nodes in the grid; Based on the location and various coefficients of the portable WiFi connected to the mobile relay node within the grid, and based on the Q-Learning model, the optimal channel combination is selected at preset time intervals. At each preset time interval, the cloud data analysis module analyzes the image data generated by different grids and the various coefficients of the portable WiFi to analyze the network load stability of the mobile relay node and provide grid support.

2. The multi-mode fusion portable WiFi signal enhancement method according to claim 1 is characterized in that: The mobile relay node includes a drone, along with an integrated data acquisition module and signal relay module. The data acquisition module has a built-in spectrum analysis unit and a performance monitoring interface. The spectrum analysis unit supports 2.4 / 5GHz frequency band scanning and detects the signal strength and spectrum occupancy of the WiFi signal emitted by the signal relay module in real time. The performance monitoring interface uses the SNMP protocol to capture WiFi signal packet loss rate and latency data. The WiFi transmitter of the signal relay module adopts the 802.11ax protocol, sets dual-band concurrent SSID, and supports MU-MIMO antenna array.

3. The multi-mode fusion portable WiFi signal enhancement method according to claim 1 is characterized in that: Locating the portable WiFi in the area on the area map and dividing the area into grids includes the following steps: Obtain the MAC addresses of surrounding APs through the portable WiFi device, use the known AP locations as the location of the portable WiFi, and locate it on the area map; Mark commercial and residential areas on the management area map of the portable WiFi, and divide the commercial and residential areas, as well as other unmarked public areas into grids of different densities.

4. The multi-mode fusion portable WiFi signal enhancement method according to claim 1, characterized in that: The portable WiFi switches to different signal connection modes based on the environmental mobile data, the WiFi signal of the mobile relay node, and the signal strength of the connectable WiFi signal in the environment, and detects the signal strength data and operation data, including the following steps: Real-time monitoring of mobile data signals, relay node WiFi signal strength, and ambient WiFi signal strength, with a sampling frequency of 200ms / time; When the signal strength of the mobile relay node WiFi signal or the environment can connect to the WiFi signal meets the signal strength requirement of higher than -55dBm and the delay is less than 100ms, the network connection mode of the portable WiFi will be automatically switched from mobile data to WiFi signal connection mode. Otherwise, the mobile data connection will be maintained. When switching to WiFi signal connection mode, select the WiFi signal with the strongest signal among the WiFi signals available in the mobile relay node and the environment; The portable WiFi detects the signal strength data of the ambient signal and the operating data of its own signal. The signal strength data includes: signal strength and spectrum occupancy; the operating data includes: the number of connected devices, the traffic demand data filled in by the user, the packet loss rate and the delay data.

5. The multi-mode fusion portable WiFi signal enhancement method according to claim 1, characterized in that: The cloud data analysis module obtains the portable WiFi background data through the mobile relay node to analyze the portable WiFi connection device signal demand coefficient, channel interference coefficient and environment complexity coefficient, including the following steps: The cloud data analysis module obtains the portable WiFi background data through the mobile relay node, including: the signal strength and spectrum occupancy of the portable WiFi detection environment signal, the number of connected devices of its own signal, the traffic demand data filled in by the user, the packet loss rate and the delay data; Based on the ratio of the number of connected devices to the theoretical upper limit of the devices, as well as the ratio of the total traffic demand filled in by the user to the upper limit of the package, the signal demand coefficient of the connected devices of the portable WiFi is analyzed using the following formula: DSDI is the signal demand coefficient of the connected device of the Pocket WiFi, Nmax is the theoretical carrying limit of the device, N is the number of connected devices, Qmax is the upper limit of the Pocket WiFi package, and Q is the total traffic demand filled in by the user; Based on the spectrum occupancy of the 2.4 / 5 GHz bands, if the spectrum occupancy of the 2.4 GHz band channel is greater than 70% or the spectrum occupancy of the 5 GHz band channel is greater than 50%, it is marked as high interference. Combined with the packet loss rate, the channel interference coefficient is analyzed using the following formula: Where CII is the channel interference coefficient of the portable WiFi, U is the weighted average of the spectrum occupancy of the 2.4-band channel and the spectrum occupancy of the 5GHz band channel, and L is the sum of the packet loss rates of the portable WiFi connected devices; According to the mobile data signal, relay node WiFi signal strength and ambient WiFi signal strength of the portable WiFi environment, the signal strength of each signal is normalized and then weighted averaged; Obtain latency data. If the Ping value is greater than 100ms, mark it as a high-latency environment. Otherwise, mark it as a non-high-latency environment. Also, analyze the environment complexity coefficient based on the PCI value using the following formula: Among them, ECI is the environmental complexity coefficient of the portable WiFi, S is the weighted average value after normalization of the signal strength of each signal, D is the Ping value of the delay data, and P is the PCI interference value.

6. The multi-mode fusion portable WiFi signal enhancement method according to claim 1, characterized in that: Generates a heat map of the environmental interference for mobile relay nodes in the grid, including: The cloud data analysis module enables spectrum scanning through the mobile relay node management backend to monitor the channel occupancy rate and interference source intensity of the 2.4GHz / 5GHz frequency band in real time; Use the NetSpot tool to perform a full-scene signal scan and automatically generate an environmental interference heat map that includes channel occupancy and interference intensity for each channel.

7. The multi-mode fusion portable WiFi signal enhancement method according to claim 1, characterized in that: Generates a 3D signal heat map of mobile relay nodes in a grid, including: Enable the spectrum scanning function through the mobile relay node management background to monitor the signal strength of each channel in the 2.4GHz / 5GHz frequency band in real time; Import a map of the grid corresponding to the mobile relay node, mark the location of the located portable WiFi and the location of the connected device, set the wall attenuation parameters, and generate a 3D signal heat map of the mobile relay node signal in each channel in the corresponding grid.

8. The multi-mode fusion portable WiFi signal enhancement method according to claim 1, characterized in that: Generate a device distribution map of mobile relay nodes in the grid, including: Export the connected portable WiFi list through the mobile relay node background, obtain device coordinates through WiFi positioning technology or Bluetooth beacon, and build a device connection topology map; Collect the connection status, signal strength, and data usage of the devices in the portable WiFi list and map them to the corresponding coordinates of the device connection topology map. Mark the signal strength ranges at the corresponding positions of the coordinates on the device connection topology map as a device distribution map.

9. The multi-mode fusion portable WiFi signal enhancement method according to claim 5, characterized in that: Based on the location and various coefficients of the portable WiFi connected to the mobile relay node in the grid, and based on the Q-Learning model, the optimal channel combination is selected at preset time intervals, including the following steps: The signal demand coefficient, channel interference coefficient, and environment complexity coefficient of the portable WiFi connected to the mobile relay node in the grid are weighted and added together to obtain the comprehensive coefficient of the portable WiFi. Defines a set of optional channels for the 2.4 GHz and 5 GHz bands; Set the reward function of the Q-Learning model. The initial value of the reward function is 1, including the incentive mechanism and the penalty mechanism. The incentive mechanism includes: a reward of +5 when the comprehensive coefficient of the portable WiFi decreases after the portable WiFi changes its channel; a penalty mechanism includes: a reward of -5 when the comprehensive coefficient of the portable WiFi increases after the portable WiFi changes its channel; a penalty of -2 when the packet loss rate exceeds 2% due to channel switching; Initialize the Q value table Q(s,a), where s is the value of the reward function and a is the switching channel in the optional channel set; Set the Q-Learning model hyperparameters: learning rate α and exploration rate ε. The exploration rate ε=0.8, decays by 5% per cycle until the steady state ε=0.1‌3. Set α=0.6 / (1+0.01*t) to implement decaying learning rate adjustment, where t represents the model training time. The Q-Learning model is trained using the past 24 hours of historical data on the performance of the portable WiFi channels in the grid; After training is completed, the reward function values ​​s in the Q value of each portable WiFi are sorted from large to small, and the reward function values ​​s in the top 10% and the corresponding a values ​​are selected as the optimal channel combination for switching channels.

10. The multi-mode fusion portable WiFi signal enhancement method according to claim 1, characterized in that: Analyzing the network load stability of mobile relay nodes and providing grid support includes the following steps: When more than 10% of the portable WiFi devices in a grid report network issues in historical data, the corresponding environmental interference heat map, 3D signal heat map, and device distribution map of the portable WiFi devices in the grid are labeled "Network Unstable" and the remaining images are labeled "Network Stable"; Train the neural network model using labeled images; At preset time intervals, the system obtains the environmental interference heat map, 3D signal heat map, and device distribution map of the portable WiFi within the grid. The trained neural network model uses these maps to identify whether the network status in the grid is stable or unstable. At every preset time interval, the portable WiFi comprehensive coefficient is obtained. If the comprehensive coefficients of more than 10% of the portable WiFi in the grid exceed 0.85, the network status in the grid is judged to be unstable; otherwise, the network status is judged to be stable; If the network status in a grid is judged to be unstable, the backup mobile relay node is called to move to the corresponding grid and provide support as an equal mobile relay node.

Images