Wireless network optimization system based on artificial intelligence

Abstract

The invention discloses a wireless network optimization system based on artificial intelligence, relates to the technical field of wireless network optimization, and solves the technical problem that the user experience is reduced due to the fact that bandwidth and access equipment are difficult to effectively allocate when excessive equipment is accessed in the existing wireless network application. Comprising a data acquisition module used for acquiring flow data of a plurality of connected devices and monitoring data of a current network; the access optimization module is used for generating a pre-allocated bandwidth according to the equipment characteristic data of new equipment when the new equipment is accessed; generating a current predicted margin bandwidth according to the flow data of each connection device; when the pre-allocated bandwidth is greater than the predicted margin bandwidth, generating a usage priority score of each connection device according to the flow data in the set time period; optimizing the network according to the use priority score; the normal state optimization module is used for generating a network state according to the monitoring data; the network is optimized according to the network state, smooth use of other devices is ensured, and the user experience is improved.

Description

Technical Field

[0001] The present application belongs to the field of wireless network optimization technology, and relates to wireless network optimization technology, and specifically to a wireless network optimization system based on artificial intelligence. Background Art

[0002] With the rapid development and popularization of wireless technology, many smart devices need to access wireless networks, both in public places and at home; wireless networks have become an indispensable part of people's daily life and work.

[0003] When too many devices are connected to an existing wireless network, data transmission congestion and poor data transmission quality will occur. When dealing with this situation, it is difficult for the existing wireless network to reasonably allocate the bandwidth and access devices of the wireless network, resulting in long transmission delays in the wireless network and high data packet loss rates, which in turn leads to a decrease in the user experience of using the wireless network. Therefore, a wireless network optimization system based on artificial intelligence is needed. Summary of the invention

[0004] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes an artificial intelligence-based wireless network optimization system to solve the technical problem that when too many devices are connected to the existing wireless network application, it is difficult to effectively allocate bandwidth and access devices, resulting in a decrease in user experience.

[0005] To achieve the above-mentioned purpose, the first aspect of the present application provides a wireless network optimization system based on artificial intelligence, including: a data acquisition module, an access optimization module, a normal optimization module and a database;

[0006] The data acquisition module is used to obtain the flow data of several connected devices and the monitoring data of the current network;

[0007] The access optimization module: when a new device is connected, the device feature data of the new device is obtained, and the pre-allocated bandwidth is generated according to the device feature data; the flow data of each currently connected device is obtained, and the current estimated surplus bandwidth is generated according to each flow data; when the pre-allocated bandwidth is greater than the estimated surplus bandwidth, the use priority score of each connected device is generated according to the corresponding flow data within the set time period; the network is optimized according to the use priority score;

[0008] The normal optimization module: obtains monitoring data, generates network status according to the monitoring data; and optimizes the network according to the network status.

[0009] The present application obtains device characteristic data of a new device when a new device is connected, and generates a pre-allocated bandwidth according to the device characteristic data; obtains traffic data of each currently connected device, and generates the current estimated surplus bandwidth according to each traffic data; generates a usage priority score for each connected device according to the corresponding traffic data in a set time period when the pre-allocated bandwidth is greater than the estimated surplus bandwidth; optimizes the network according to the usage priority score; and obtains monitoring data in real time during normal operation of the wireless network; generates a network status according to the monitoring data; optimizes the network according to the network status; optimizes the wireless network in different situations through two paths, and promptly disconnects unused devices or devices with low wireless network usage intensity to ensure smooth use of other devices, thereby achieving effective allocation of bandwidth and access devices and further improving user experience.

[0010] Preferably, the data acquisition module communicates and / or is electrically connected to the access optimization module and the normal optimization module; the database communicates and / or is electrically connected to the data acquisition module, the access optimization module and the normal optimization module.

[0011] Preferably, generating the pre-allocated bandwidth according to the device characteristic data includes:

[0012] Extract device type and historical access and usage data from device feature data;

[0013] Input the device type into the bandwidth estimation model to obtain the corresponding estimated bandwidth; the bandwidth estimation model is obtained by training an artificial intelligence model;

[0014] Extract the uplink traffic and the downlink traffic from the historical traffic data corresponding to each recording period in the historical access usage data; generate the pre-allocated bandwidth according to the estimated bandwidth usage, the uplink traffic and the downlink traffic.

[0015] This application first generates an estimated bandwidth usage according to the device type corresponding to the device, and then adjusts the estimated bandwidth usage according to the device's recent traffic usage corresponding to each recording period to obtain the pre-allocated bandwidth of the corresponding device, so that the pre-allocated bandwidth is more matched with the corresponding device, reducing the occurrence of insufficient or over-allocated bandwidth.

[0016] Preferably, the bandwidth estimation model is obtained by training an artificial intelligence model, including:

[0017] Acquire a number of device types and their corresponding estimated bandwidth usage through a database; integrate the number of device types and their corresponding estimated bandwidth usage into a number of sets of training data and test data;

[0018] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data, and finally a bandwidth estimation model is obtained, in which the input is the device type and the output is the corresponding estimated bandwidth usage; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0019] Preferably, generating the pre-allocated bandwidth according to the estimated used bandwidth, a number of uplink flows and downlink flows comprises:

[0020] The upstream traffic and downstream traffic are marked as SLi and XLi respectively; i is the number of the recording period;

[0021] By formula The pre-allocated bandwidth YD is calculated; where SD is the estimated bandwidth, ST is the length of a recording period; β1 is the adjustment coefficient, and 0<β1<1; γ is the conversion adjustment coefficient; α i is the weight coefficient corresponding to the recording period numbered i; i=1, 2, ..., I; I is the total number of set recording periods.

[0022] Preferably, generating the current estimated margin bandwidth according to each flow data includes:

[0023] Extract the upstream and downstream traffic from the traffic data corresponding to each recording period of each connected device, and mark them as YSLjn and YXLjn respectively; where j is the number of the recording period and n is the number of the connected device.

[0024] By formula The estimated margin bandwidth YYD is calculated; where ZD is the wireless network bandwidth, ε j is the weight coefficient corresponding to the recording period numbered j. The specific value is set according to expert experience. The closer the recording period in this embodiment is to the current time, the larger the corresponding weight coefficient is set; j=1, 2, ..., J; J is the total number of set recording periods; n=1, 2, ..., N; N is the total number of connected devices; ST is the time length of a recording period; γ is the conversion adjustment coefficient.

[0025] Preferably, generating a usage priority score for each connected device according to the corresponding traffic data within a set time period includes:

[0026] Extract the upstream flow YSLjn and the downstream flow YXLjn from the flow data corresponding to each recording period within the set period of the connection device;

[0027] By formula The usage priority score SPn corresponding to the connection device numbered n is calculated; where DL is the unit flow, which is used to remove the unit of the upstream flow and the downstream flow; εj is the weight coefficient corresponding to the recording period numbered j.

[0028] Preferably, the optimizing the network according to the usage priority score comprises the following steps:

[0029] S1: Obtaining the usage priority score corresponding to each connected device, and sorting each connected device in ascending order of the usage priority score to obtain a device disconnection sorting table;

[0030] S2: disconnect the connected device with the highest ranking in the device disconnection sorting table;

[0031] S3: Obtain monitoring data, generate network status based on the monitoring data, and determine whether the network status is congested; if yes, remove the disconnected connected device from the device disconnection sorting table; update the device disconnection sorting table and enter S2; if no, generate a network optimization completion signal.

[0032] Preferably, generating the network status according to the monitoring data includes:

[0033] Extract the transmission rate, packet loss rate and bandwidth occupancy rate from the real-time monitoring data;

[0034] Generate data transmission status according to transmission rate and packet loss rate;

[0035] When the data transmission status is transmission congestion, determine whether the bandwidth occupancy rate is greater than the set congestion bandwidth threshold; if yes, set the network status to congestion; if no, set the network status to abnormal; otherwise, set the network status to normal.

[0036] Preferably, generating the data transmission status according to the transmission rate and the packet loss rate includes:

[0037] It is determined whether the transmission rate is lower than the set transmission rate threshold or whether the packet loss rate is higher than the packet loss rate threshold. If yes, the data transmission state is set to a congested state; if no, the data transmission state is set to a smooth state.

[0038] Preferably, the optimizing the network according to the network status comprises the following steps:

[0039] S1: Get the network status and determine whether the network is congested; if yes, go to S2; if no, go to S5;

[0040] S2: Obtaining the usage priority score corresponding to each connected device, and sorting each connected device in ascending order of the usage priority score to obtain a device disconnection sorting table;

[0041] S3: disconnecting the connected device with the highest ranking in the device disconnection ranking table; removing the disconnected connected device from the device disconnection ranking table, and updating the device disconnection ranking table;

[0042] S4: Get the waiting time. After the waiting time is over, get the monitoring data, generate the network status according to the monitoring data, and enter S1;

[0043] S5: Determine whether the network status is abnormal; if yes, generate a network abnormality signal; if no, generate an optimization completion signal.

[0044] Compared with the prior art, the beneficial effects of this application are:

[0045] 1. This application obtains device feature data of new devices when new devices are connected, and generates pre-allocated bandwidth according to the device feature data; obtains traffic data of each currently connected device, and generates the current estimated surplus bandwidth according to each traffic data; when the pre-allocated bandwidth is greater than the estimated surplus bandwidth, generates a usage priority score for each connected device according to the corresponding traffic data in a set time period; optimizes the network according to the usage priority score; and obtains monitoring data in real time during normal operation of the wireless network; generates a network status according to the monitoring data; optimizes the network according to the network status; optimizes the wireless network in different situations through two paths, and promptly disconnects unused or low-intensity wireless network usage devices to ensure smooth use of other devices, thereby achieving effective allocation of bandwidth and access devices and further enhancing user experience.

[0046] 2. This application first generates an estimated bandwidth usage based on the device type corresponding to the device, and then adjusts the estimated bandwidth usage based on the device's recent traffic usage corresponding to each recording period to obtain the pre-allocated bandwidth of the corresponding device, so that the pre-allocated bandwidth is more matched with the corresponding device, reducing the occurrence of insufficient or over-allocated bandwidth. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1 A schematic diagram of the modules of the wireless network optimization system in this application;

[0049] Figure 2 This is a schematic diagram of the network optimization steps when connecting a new device in this application;

[0050] Figure 3 This is a schematic diagram of the network optimization steps under normal conditions in this application. DETAILED DESCRIPTION

[0051] The technical solution of the present application will be described clearly and completely below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0052] See also Figure 1 , a first aspect of the present application provides a wireless network optimization system based on artificial intelligence, including: a data acquisition module, an access optimization module, a normal optimization module and a database;

[0053] Data collection module: obtains the traffic data of several connected devices and the monitoring data of the current network; the traffic data refers to the traffic data generated when the corresponding connected device uses the wireless network to upload or download data, including upload traffic data and download traffic data; the monitoring data refers to the relevant data of network transmission monitored by the wireless network end, including transmission rate, packet loss rate and bandwidth occupancy rate, etc.;

[0054] Access optimization module: when a new device is connected, the device feature data of the new device is obtained. The device feature data is the data related to the bandwidth requirements of the connected device, including the device type and historical traffic data; for example, the bandwidth occupied by a smart TV when in use is generally larger than that occupied by a smart phone; the pre-allocated bandwidth is generated according to the device feature data; the pre-allocated bandwidth is the bandwidth required by the corresponding device generated according to the device feature data; that is, the bandwidth that the access device is expected to occupy when accessing the wireless network; the traffic data of each currently connected device is obtained, and the current estimated surplus bandwidth is generated according to each traffic data; the estimated surplus bandwidth is the bandwidth remaining in addition to the bandwidth used by the currently connected device; when the pre-allocated bandwidth is greater than the estimated surplus bandwidth, it means that after the new device is connected to the wireless network at this time, there is a high probability that network congestion will occur, which may lead to a decrease in the user experience; the use priority score of each connected device is generated according to the corresponding traffic data within the set time period; the use priority score is to score the use status of the connected device according to the use of the wireless network by the connected device. The higher the use priority score, the greater the intensity of the corresponding connected device using the wireless network; the network is optimized according to the use priority score;

[0055] Normal optimization module: obtains monitoring data, generates network status based on the monitoring data, and optimizes the network based on the network status.

[0056] This embodiment obtains device characteristic data of a new device when a new device is connected, and generates a pre-allocated bandwidth according to the device characteristic data; obtains traffic data of each currently connected device, and generates the current estimated surplus bandwidth according to each traffic data; generates a usage priority score for each connected device according to the corresponding traffic data in a set time period when the pre-allocated bandwidth is greater than the estimated surplus bandwidth; optimizes the network according to the usage priority score; and obtains monitoring data in real time during normal operation of the wireless network; generates a network status according to the monitoring data; optimizes the network according to the network status; optimizes the wireless network in different situations through two paths, and promptly disconnects unused devices or devices with low wireless network usage intensity to ensure smooth use of other devices, thereby achieving effective allocation of bandwidth and access devices and further improving user experience.

[0057] The data acquisition module communicates and / or is electrically connected to the access optimization module and the normal optimization module; the database communicates and / or is electrically connected to the data acquisition module, the access optimization module and the normal optimization module.

[0058] Generate pre-allocated bandwidth based on device characteristic data, including:

[0059] Extract the device type and historical access and use data from the device feature data; it is understandable that if the corresponding device accesses the wireless network for the first time, the corresponding historical access and use data is empty;

[0060] Input the device type into the bandwidth estimation model to obtain the corresponding estimated bandwidth; the bandwidth estimation model is obtained by training an artificial intelligence model;

[0061] Extract the uplink traffic and the downlink traffic from the historical traffic data corresponding to each recording period in the historical access usage data; generate the pre-allocated bandwidth according to the estimated bandwidth usage, the uplink traffic and the downlink traffic.

[0062] This embodiment first generates an estimated bandwidth usage according to the device type corresponding to the device, and then adjusts the estimated bandwidth usage according to the device's recent traffic usage corresponding to each recording time period to obtain the pre-allocated bandwidth of the corresponding device, so that the pre-allocated bandwidth is more matched with the corresponding device, reducing the occurrence of insufficient or over-allocated bandwidth; it can be understood that when the corresponding device accesses the wireless network for the first time, its corresponding pre-allocated bandwidth is its corresponding estimated bandwidth usage.

[0063] The bandwidth estimation model is obtained through artificial intelligence model training, including: obtaining several device types and their corresponding estimated bandwidth usage through a database, wherein the estimated bandwidth usage is the bandwidth set by experts according to the normal traffic usage data corresponding to the device type, and allocating corresponding bandwidth to the corresponding device type, which can meet the normal use of most devices of the device type; integrating several device types and their corresponding estimated bandwidth usage into several groups of training data and test data;

[0064] The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data. Specifically, the device type in the test data is input into the trained artificial intelligence model to obtain the existing estimated bandwidth usage output, and the output estimated bandwidth usage is compared with the estimated bandwidth usage recorded in the test data to determine whether the difference is within an acceptable range. If yes, it means that the group of test data has passed the test and the next group of test data is tested; if not, the relevant parameters of the artificial intelligence are adjusted and the group of test data is used again for testing; until a set proportion of test data passes the test, a bandwidth estimation model is finally obtained with the device type as input and the corresponding estimated bandwidth as output; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0065] Generating a pre-allocated bandwidth according to the estimated bandwidth usage, a number of uplink traffic and downlink traffic, including: marking the uplink traffic and the downlink traffic as SLi and XLi respectively; i is the number of the recording period;

[0066] By formula The pre-allocated bandwidth YD is calculated; where SD is the estimated bandwidth, and ST is the length of a recording period; it can be understood that the unit of the time length matches the length of the estimated bandwidth, that is, if the unit of the estimated bandwidth is M / s, then the unit of the time length is s; β1 is the adjustment coefficient, and 0<β1<1; γ is the conversion adjustment coefficient, which is used to adjust the conversion relationship between bandwidth and traffic, that is, the conversion between bits and bytes; here γ=1 / 8; α i is the weight coefficient corresponding to the recording period numbered i, and the specific value is set according to expert experience. The closer the recording period in this embodiment is to the current time, the larger the corresponding weight coefficient is set; i=1, 2, ..., I; I is the total number of set recording periods.

[0067] This embodiment calculates the pre-allocated bandwidth of the corresponding device through the above formula. When the traffic data in the historical usage data of the corresponding device is higher than the normal usage traffic corresponding to the device type, its corresponding bandwidth should also be increased based on the estimated usage bandwidth to ensure smooth use of the device.

[0068] The current estimated surplus bandwidth is generated according to each flow data, including: extracting the uplink flow and downlink flow in the flow data corresponding to each recording period of each connected device, and marking them as YSLjn and YXLjn respectively; wherein j is the number of the recording period, and n is the number of the connected device.

[0069] By formula The estimated margin bandwidth YYD is calculated; where ZD is the wireless network bandwidth, ε j is the weight coefficient corresponding to the recording period numbered j. The specific value is set according to expert experience. The closer the recording period in this embodiment is to the current time, the larger the corresponding weight coefficient is set; j=1, 2, ..., J; J is the total number of set recording periods; n=1, 2, ..., N; N is the total number of connected devices; ST is the time length of a recording period; γ is the conversion adjustment coefficient.

[0070] In this embodiment, the maximum value of the total bandwidth used in the uplink traffic and downlink traffic of each device is selected as the estimated bandwidth usage, and then the estimated bandwidth usage is subtracted from the wireless network bandwidth to obtain the minimum estimated bandwidth surplus, and the bandwidth surplus is recorded as the estimated surplus bandwidth;

[0071] Generate a usage priority score for each connected device according to the corresponding flow data within the set time period, including: extracting the uplink flow YSLjn and the downlink flow YXLjn from the flow data corresponding to each recording time period within the set time period of the connected device;

[0072] By formula The usage priority score SPn corresponding to the connection device numbered n is calculated; where DL is the unit flow, which is used to remove the unit of the upstream flow and the downstream flow; ε j is the weight coefficient corresponding to the recording period numbered j. The specific value is set according to expert experience. The closer the recording period in this embodiment is to the current time, the larger the corresponding weight coefficient is set. It can be understood that the set number of time periods can be the same as the number of recording time periods considered in the estimated margin bandwidth, or it can be different.

[0073] In this embodiment, the usage priority score of each connection device is calculated by the above formula. When the uplink traffic and downlink traffic corresponding to each recording period of the connection device within the set time period are small, it means that the intensity of the connection device using the wireless network is small, that is, the connection device hardly uses the wireless network, so its corresponding usage priority score is set to be small. When the network is congested, the connection device with a small usage priority score can be disconnected first; to ensure that most connection devices and connection devices with higher usage intensity can use the wireless network normally, thereby optimizing the network.

[0074] See also Figure 2 , the network is optimized according to the usage priority score, including the following steps:

[0075] S1: Obtaining the usage priority score corresponding to each connected device, and sorting each connected device in ascending order of the usage priority score to obtain a device disconnection sorting table;

[0076] S2: disconnect the connected device with the highest ranking in the device disconnection sorting table;

[0077] S3: Obtain monitoring data, generate network status based on the monitoring data, and determine whether the network status is congested; if yes, remove the disconnected connected device from the device disconnection sorting table; update the device disconnection sorting table and enter S2; if no, generate a network optimization completion signal.

[0078] When a new device is connected, this embodiment pre-determines whether congestion will occur after the new device is connected. If congestion occurs, the device with less network usage intensity and almost no wireless network use among the connected devices is disconnected first; then the new device is connected to ensure that the new device can be used smoothly after connecting to the network.

[0079] Generate network status based on monitoring data, including: extracting transmission rate, packet loss rate and bandwidth occupancy rate from real-time monitoring data;

[0080] Generate the data transmission status according to the transmission rate and the packet loss rate; specifically, determine whether the transmission rate is lower than the set transmission rate threshold or whether the packet loss rate is higher than the packet loss rate threshold. If yes, set the data transmission status to the congested status; if no, set the data transmission status to the unobstructed status; wherein, the transmission rate threshold and the packet loss rate threshold are set according to experience; if the transmission rate is lower than the transmission rate threshold or the packet loss rate is higher than the packet loss rate threshold, it indicates that the network is congested.

[0081] When the data transmission status is transmission congestion, it is determined whether the bandwidth occupancy rate is greater than the set congestion bandwidth threshold, and the congestion bandwidth threshold is set according to expert experience; if yes, the network status is set to congestion; if no, the network status is set to abnormal; otherwise, the network status is set to normal.

[0082] See also Figure 3 , optimize the network according to the network status, including the following steps:

[0083] S1: Get the network status and determine whether the network is congested; if yes, go to S2; if no, go to S5;

[0084] S2: Obtaining the usage priority score corresponding to each connected device, and sorting each connected device in ascending order of the usage priority score to obtain a device disconnection sorting table;

[0085] S3: disconnecting the connected device with the highest ranking in the device disconnection ranking table; removing the disconnected connected device from the device disconnection ranking table, and updating the device disconnection ranking table;

[0086] S4: Get the waiting time. After the waiting time is over, get the monitoring data, generate the network status according to the monitoring data, and enter S1;

[0087] S5: Determine whether the network status is abnormal; if yes, generate a network abnormality signal; if no, generate an optimization completion signal.

[0088] In this embodiment, when the wireless network is running normally, it is determined whether the network is congested. If congestion occurs, the devices with less network usage intensity and those that hardly use the wireless network are disconnected first, thereby ensuring that most devices connected to the network can be used smoothly.

[0089] It is worth noting that the length of the recording period is set according to the wireless network bandwidth. The larger the wireless network bandwidth, the longer the corresponding recording period. Targeted setting of the length of the recording period makes the analysis results based on the traffic data within the recording period more referenceable.

[0090] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0091] How this application works:

[0092] The present application obtains device characteristic data of a new device when a new device is connected, and generates a pre-allocated bandwidth according to the device characteristic data; obtains traffic data of each currently connected device, and generates the current estimated surplus bandwidth according to each traffic data; generates a usage priority score for each connected device according to the corresponding traffic data in a set time period when the pre-allocated bandwidth is greater than the estimated surplus bandwidth; optimizes the network according to the usage priority score; and obtains monitoring data in real time during normal operation of the wireless network; generates a network status according to the monitoring data; optimizes the network according to the network status; optimizes the wireless network in different situations through two paths, and promptly disconnects unused devices or devices with low wireless network usage intensity to ensure smooth use of other devices, thereby achieving effective allocation of bandwidth and access devices and further improving user experience.

[0093] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, a person of ordinary skill in the art should understand that the technical method of the present application may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. A wireless network optimization system based on artificial intelligence, characterized in that: include: Data collection module, access optimization module, normal optimization module and database; The data acquisition module is used to obtain the flow data of several connected devices and the monitoring data of the current network; The access optimization module: when a new device is connected, the device feature data of the new device is obtained, and the pre-allocated bandwidth is generated according to the device feature data; the flow data of each currently connected device is obtained, and the current estimated surplus bandwidth is generated according to each flow data; when the pre-allocated bandwidth is greater than the estimated surplus bandwidth, the use priority score of each connected device is generated according to the corresponding flow data within the set time period; the network is optimized according to the use priority score; The normal optimization module: obtains monitoring data, generates network status according to the monitoring data; and optimizes the network according to the network status.

2. The wireless network optimization system based on artificial intelligence according to claim 1, characterized in that: The generating of the pre-allocated bandwidth according to the device characteristic data includes: Extract device type and historical access and usage data from device feature data; Input the device type into the bandwidth estimation model to obtain the corresponding estimated bandwidth; the bandwidth estimation model is obtained by training an artificial intelligence model; Extract the uplink traffic and the downlink traffic from the historical traffic data corresponding to each recording period in the historical access usage data; generate the pre-allocated bandwidth according to the estimated bandwidth usage, the uplink traffic and the downlink traffic.

3. The wireless network optimization system based on artificial intelligence according to claim 2, characterized in that: The bandwidth estimation model is obtained by training an artificial intelligence model, including: Acquire a number of device types and their corresponding estimated bandwidth usage through a database; integrate the number of device types and their corresponding estimated bandwidth usage into a number of sets of training data and test data; The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data, and finally a bandwidth estimation model is obtained, in which the input is the device type and the output is the corresponding estimated bandwidth usage; wherein the artificial intelligence model includes a BP neural network model and an RBF neural network model.

4. The wireless network optimization system based on artificial intelligence according to claim 2, characterized in that: The generating of the pre-allocated bandwidth according to the estimated used bandwidth, a number of uplink flows and downlink flows comprises: The upstream traffic and downstream traffic are marked as SLi and XLi respectively; i is the number of the recording period; By formula The pre-allocated bandwidth YD is calculated; where SD is the estimated bandwidth, ST is the length of a recording period; β1 is the adjustment coefficient, and 0<β1<1; γ is the conversion adjustment coefficient; α i is the weight coefficient corresponding to the recording period numbered i; i=1, 2, ..., I; I is the total number of set recording periods.

5. The wireless network optimization system based on artificial intelligence according to claim 1, characterized in that: The generating of the current estimated surplus bandwidth according to each flow data includes: Extract the upstream and downstream traffic from the traffic data corresponding to each recording period of each connected device, and mark them as YSLjn and YXLjn respectively; where j is the number of the recording period and n is the number of the connected device. By formula The estimated margin bandwidth YYD is calculated; where ZD is the wireless network bandwidth, ε j is the weight coefficient corresponding to the recording period numbered j; j=1, 2, ..., J; J is the total number of set recording periods; n=1, 2, ..., N; N is the total number of connected devices; ST is the time length of a recording period; γ is the conversion adjustment coefficient.

6. The wireless network optimization system based on artificial intelligence according to claim 1, characterized in that: The generating of the usage priority score of each connected device according to the corresponding traffic data within the set time period includes: Extract the upstream flow YSLjn and the downstream flow YXLjn from the flow data corresponding to each recording period within the set period of the connection device; By formula Calculate the usage priority score SPn corresponding to the connected device numbered n; where DL is the unit flow; ε j is the weight coefficient corresponding to the recording period numbered j.

7. The wireless network optimization system based on artificial intelligence according to claim 1, characterized in that: The method of optimizing the network according to the use priority score comprises the following steps: S1: Obtaining the usage priority score corresponding to each connected device, and sorting each connected device in ascending order of the usage priority score to obtain a device disconnection sorting table; S2: disconnect the connected device with the highest ranking in the device disconnection sorting table; S3: Obtain monitoring data, generate network status based on the monitoring data, and determine whether the network status is congested; if yes, remove the disconnected connected device from the device disconnection sorting table; update the device disconnection sorting table and enter S2; if no, generate a network optimization completion signal.

8. The artificial intelligence-based wireless network optimization system according to any one of claims 1 or 7, characterized in that: The generating of the network status according to the monitoring data comprises: Extract the transmission rate, packet loss rate and bandwidth occupancy rate from the real-time monitoring data; Generate data transmission status according to transmission rate and packet loss rate; When the data transmission status is transmission congestion, determine whether the bandwidth occupancy rate is greater than the set congestion bandwidth threshold; if yes, set the network status to congestion; if no, set the network status to abnormal; otherwise, set the network status to normal.

9. The wireless network optimization system based on artificial intelligence according to claim 8, characterized in that: Generating the data transmission status according to the transmission rate and the packet loss rate includes: It is determined whether the transmission rate is lower than the set transmission rate threshold or whether the packet loss rate is higher than the packet loss rate threshold. If yes, the data transmission state is set to a congested state; if no, the data transmission state is set to a smooth state.

10. The wireless network optimization system based on artificial intelligence according to claim 1, characterized in that: The step of optimizing the network according to the network status comprises the following steps: S1: Get the network status and determine whether the network is congested; if yes, go to S2; if no, go to S5; S2: Obtaining the usage priority score corresponding to each connected device, and sorting each connected device in ascending order of the usage priority score to obtain a device disconnection sorting table; S3: disconnecting the connected device with the highest ranking in the device disconnection ranking table; removing the disconnected connected device from the device disconnection ranking table, and updating the device disconnection ranking table; S4: Get the waiting time. After the waiting time is over, get the monitoring data, generate the network status according to the monitoring data, and enter S1; S5: Determine whether the network status is abnormal; if yes, generate a network abnormality signal; if no, generate an optimization completion signal.

Images