Channel interference prediction system, method, device, storage medium and program product

Abstract

The invention provides a channel interference prediction system, method and device, a storage medium and a program product, and belongs to the technical field of communication network optimization. The system comprises an acquisition module used for acquiring frequency report data of a working channel supporting a frequency band; the prediction module is used for inputting the frequency report data into the channel interference prediction model to obtain an interference prediction probability value of the working channel output by the channel interference prediction model; the first determination module is used for determining that interference exists in the working channel when the interference prediction probability value of the working channel is greater than a threshold value; wherein the channel interference prediction model is obtained by training a plurality of historical frequency report data of a plurality of channels traversed based on a neural algorithm and a channel interference prediction probability value corresponding to each historical frequency report data. The method is used for solving the problems that the current enhanced distributed channel access cannot predict the future change of the channel and cannot well cope with the channel interference and noise under the complex environment network condition.

Description

Technical Field

[0001] This invention relates to the field of communication network optimization technology, and more specifically to a channel interference prediction system, method, electronic device, machine-readable storage medium, and computer program product. Background Technology

[0002] Currently, Media Access Control (MAC) employs Enhanced Distributed Channel Access (EDCA) to access physical channels and transmit data after gaining access rights. However, key parameters of current EDCA, such as the Contention Window (CW) and Arbitration Inter-Frame Space (AIFS), are usually fixed. These static parameters are mismatched with dynamic networks, making it impossible to predict future channel changes and effectively cope with channel interference and noise under complex network conditions. Summary of the Invention

[0003] The purpose of this invention is to provide a channel interference prediction system, method, device, storage medium, and program product to solve the problem that current enhanced distributed channel access cannot predict future channel changes and cannot effectively cope with channel interference and noise under complex network conditions.

[0004] To achieve the above objectives, embodiments of the present invention provide a channel interference prediction system, comprising: The acquisition module is used to acquire frequency report data of the working channels supporting the frequency band; The prediction module is used to input the frequency report data into the channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model. The first determining module is used to determine that there is interference in the working channel when the interference prediction probability value of the working channel is greater than a threshold. The channel interference prediction model is trained based on a neural algorithm that traverses multiple historical frequency report data for multiple channels and the channel interference prediction probability value corresponding to each historical frequency report data.

[0005] Optionally, there are two or more channels supporting the frequency band, including the working channel and the non-working channel; The system also includes: The second determining module is used to determine the target switching channel according to a preset judgment strategy; The switching module is used to switch the working channel to the target switching channel when the working channel is not the target switching channel.

[0006] Optionally, the preset judgment strategy includes: The channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

[0007] Optionally, there are three or more supported frequency bands, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels of the supported frequency bands; The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. Determine whether multi-link operation is supported; When multi-link operation is not supported, the channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

[0008] Optionally, there are three or more supported frequency bands, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels in the supported frequency bands. The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. When the interference prediction probability value of the non-working channel is not greater than the threshold, the non-working channel does not have interference. Determine whether multi-link operation is supported; When multi-link operation is supported, all non-working channels that are free from interference are identified as target switching channels.

[0009] On the other hand, embodiments of the present invention also provide a channel interference prediction method based on the above-described channel interference prediction system, comprising: Obtain frequency report data for the working channels of the supported frequency bands; The frequency report data is input into the channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model. When the interference prediction probability value of the working channel is greater than the threshold, it is determined that there is interference in the working channel; The channel interference prediction model is trained based on a neural algorithm that traverses multiple historical frequency report data for multiple channels and the channel interference prediction probability value corresponding to each historical frequency report data.

[0010] Optionally, there are two or more channels supporting the frequency band, including the working channel and the non-working channel; After determining that interference exists in the working channel when the interference prediction probability value of the working channel is greater than the threshold, the method further includes: The target switching channel is determined according to a preset judgment strategy; When the working channel is not the target switching channel, the working channel is switched to the target switching channel.

[0011] Optionally, the preset judgment strategy includes: The channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

[0012] Optionally, there are three or more supported frequency bands, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels of the supported frequency bands; The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. Determine whether multi-link operation is supported; When multi-link operation is not supported, the channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

[0013] Optionally, there are three or more supported frequency bands, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels in the supported frequency bands. The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. When the interference prediction probability value of the non-working channel is not greater than the threshold, the non-working channel does not have interference. Determine whether multi-link operation is supported; When multi-link operation is supported, all non-working channels that are free from interference are identified as target switching channels.

[0014] On the other hand, embodiments of the present invention also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the channel interference prediction method described above.

[0015] On the other hand, embodiments of the present invention also provide a machine-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the above-described channel interference prediction method.

[0016] On the other hand, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the above-described channel interference prediction method.

[0017] Through the above technical solution, this invention trains a channel interference prediction model based on multiple historical frequency report data and the channel interference prediction probability value corresponding to each historical frequency report data. Then, it uses the channel interference prediction model to predict channel interference on the frequency report data of the working channel supporting the frequency band. Finally, it determines whether interference exists based on the interference prediction probability value. Therefore, this invention can predict channel state changes, avoid switching to channels that may experience interference or noise in the future, and effectively cope with channel interference and noise under complex network conditions.

[0018] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of the channel interference prediction system provided by the present invention; Figure 2 This is one of the flowcharts illustrating the channel interference prediction process performed by the channel interference prediction system provided by this invention. Figure 3 This is the second schematic diagram of the channel interference prediction system provided by the present invention performing channel interference prediction; Figure 4 This is a flowchart illustrating the channel interference prediction method provided by the present invention; Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0020] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0021] Media Access Control (MAC) is the second layer of protocol communication, responsible for controlling the order of access to logical / physical channels and ensuring Quality of Service (QoS). The current Wi-Fi MAC layer's Access Category (AC) queue scheduling principle (based on the IEEE 802.11e EDCA standard) distinguishes different service types of traffic by priority to ensure QoS. The AC queue scheduling principle divides application data streams into four priority queues as shown in Table 1: Table 1

[0022] Media access control employs an enhanced distributed channel access (EDCA) mechanism to access physical channels. Data is transmitted only after obtaining access rights to the physical channels. The priority of obtaining channel access rights for different AC queues is controlled by the following parameters. Each AC queue independently maintains a set of contention parameters as shown in Table 2: Table 2

[0023] The current enhanced distributed channel access uses the access channel to send data, with each AC queue using default configuration parameters to compete for access to the channel. As shown in Table 2, key parameters of enhanced distributed channel access, such as the contention window (CW) and the arbitration inter-frame space (AIFS), are usually fixed. These static parameters do not match the dynamic network, making it impossible to predict future channel changes and failing to cope well with channel interference and noise under complex network conditions.

[0024] Therefore, the purpose of this invention is to provide a channel interference prediction system, method, device, storage medium, and program product to solve the problem that current enhanced distributed channel access cannot predict future channel changes and cannot effectively cope with channel interference and noise under complex network conditions.

[0025] Please refer to Figure 1 On one hand, embodiments of the present invention provide a channel interference prediction system, including: an acquisition module 101, a prediction module 102, and a first determination module 103.

[0026] The acquisition module 101 is used to acquire frequency report data of the supporting frequency band's operating channel. The acquisition module 101 acquires the frequency report data of the supporting frequency band's operating channel after a spectrum scan. The spectrum scan converts time-domain I / Q data into frequency-domain signal power data using FFT, representing the power spectral density of a specific frequency range within the channel. Generally, chip manufacturers integrate this function into the chip firmware, providing the scan results to the application via a driver in FFT bin data format. In one embodiment, the frequency report data includes metadata and FFT bin data. The FFT bin data is the core output of the supporting frequency band's operating channel after a spectrum scan; each bin represents the signal power of a discrete frequency range within the channel bandwidth. In one embodiment, FFT Bins:[-80,-85,-82,...,-70,...,-90] represents a power array, with each element corresponding to the power (in dBm) of a frequency range. The FFT bin data provided to the kernel is usually accompanied by channel information (channel number, bandwidth), FFT size, noise level, and channel utilization, referred to as metadata, used to deduce which channel the FFT bin data was obtained on and the channel's state. Metadata may include the scanned frequency band (e.g., 2.4G, 5G, 6G), the scanned channel number (e.g., Channel 1, indicating the current band's channel 1), the channel bandwidth (e.g., 20MHz, 40MHZ), the corresponding center frequency (CenterFreq, e.g., 2124), the FFT size (e.g., 64, 128), the noise level (e.g., -95), and the channel utilization percentage (e.g., 30%).

[0027] In one embodiment, the acquisition module 101 can be implemented using a combination of WIFI chip firmware, kernel module, and driver module. The WIFI chip firmware performs a spectrum scanning process and transmits the spectrum scan results (frequency report data) to the kernel module via USB, SDIO, PCIe, or other shared memory methods. The kernel module stores the spectrum scan results (frequency report data) in shared memory (the application can read the / dev / spectral file). The driver module or other kernel threads obtain the spectrum scan results (frequency report data) through the shared memory access interface provided by the WIFI chip firmware.

[0028] The prediction module 102 is used to input the frequency report data into the channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model. The channel interference prediction model is trained based on a neural algorithm that iterates through multiple historical frequency report data points for multiple channels and the corresponding channel interference prediction probability value for each historical frequency report data point. The first determination module 103 is used to determine that interference exists in the working channel when the interference prediction probability value of the working channel is greater than a threshold.

[0029] The prediction module 102 can input the FFT bin data of the channel number of the working channel under the supported frequency band into the channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model. For example, the prediction module 102 inputs the FFT bins [0~127] (FFT bin data) of channel 1 in the 2.4G frequency band into the channel interference prediction model to obtain an interference prediction probability value of 0.8 for the working channel output by the channel interference prediction model. In this embodiment of the invention, a preset threshold of 0.7 can be set. When the interference prediction probability value is greater than 0.7, the first determination module 103 determines that there is channel interference in the working channel. Conversely, when the interference prediction probability value is less than or equal to 0.7, the first determination module 103 determines that there is no channel interference in the working channel. When the interference prediction probability value of the working channel is 0.8, the first determination module 103 determines that there is channel interference in the working channel and channel switching is required. Thus, this embodiment of the invention can predict channel state changes in real time, avoid switching to channels that may have interference or noise in the future, and effectively cope with channel interference and noise under complex network conditions. The channel interference prediction model is trained based on a neural algorithm that iterates through multiple historical frequency report data points for multiple channels and the corresponding channel interference prediction probability value for each historical frequency report data point. The channel interference prediction model can be constructed using a long short-term memory network or a gated recurrent unit.

[0030] In one embodiment, the prediction module 102 and the first determination module 103 may be implemented using a long short-term memory network. The long short-term memory network runs in kernel space as a UMAC (Universal Motion and Automation Controller) functional module, obtains the spectrum scan results (frequency report data) by reading the / dev / spectral file, and processes the results.

[0031] In one embodiment, the network structure of a Long Short-Term Memory (LSTM) network can be defined as follows: Layer 1 LSTM@64 + Layer 2 Dropout@0.3 + Layer 3 LSTM@32 + Layer 4 Dropout@0.3 + Layer 5 Dense@16 + Layer 6 Dense@3. Where Layer 1 LSTM@64 represents 64 hidden states, and Layer 3 LSTM@32 represents 32 hidden states. Training samples can be historical frequency report data from the working channel's historical time steps and the channel interference prediction probability value (interference result label) corresponding to each historical frequency report data. For example, training samples can be from T0 to T... 99 The training objective is to predict whether the current channel will experience interference in the next time step, based on the historical frequency report data of the channel at the current historical time step (Tn, n from 0 to 99) and the hidden states learned from the historical step data stored in the Long Short-Term Memory network (both Layer 1 LSTM@64, 64 hidden states, and Layer 3 LSTM@32, 32 hidden states). If interference is predicted, then this channel will be avoided as the target channel for switching.

[0032] In a typical implementation scenario, please refer to Figure 2 The channel interference prediction system comprises the Wi-Fi chip firmware, a kernel module, and a long short-term memory (LSM) network. The Wi-Fi chip firmware performs the spectrum scanning process and transmits the spectrum scan results (frequency report data) to the kernel module via USB, SDIO, PCIe, or other shared memory methods. The kernel module stores the spectrum scan results (frequency report data) in shared memory (the application can read the / dev / spectral file). The driver module or other kernel threads obtain the spectrum scan results (frequency report data) through the shared memory access interface provided by the Wi-Fi chip firmware. The LSM network runs in kernel space as a UMAC (Universal Motion and Automation Controller) functional module, obtaining the spectrum scan results (frequency report data) by reading the / dev / spectral file and processing the results.

[0033] This invention trains a channel interference prediction model based on multiple historical frequency report data and the corresponding channel interference prediction probability value for each historical frequency report data. The channel interference prediction model is then used to predict channel interference on the frequency report data of the operating channels supporting the frequency band. Finally, the existence of interference is determined based on the interference prediction probability value. Therefore, this invention can predict channel state changes, avoid switching to channels that may experience interference or noise in the future, and effectively cope with channel interference and noise under complex network conditions.

[0034] In other aspects of this invention, the supported frequency band has two or more channels, including the working channel and the non-working channel; the system further includes a second determining module and a switching module. The second determining module is used to determine the target switching channel according to a preset judgment strategy; the switching module is used to switch the working channel to the target switching channel when the working channel is not the target switching channel. The second determining module and the switching module can be implemented using a processor.

[0035] In this embodiment of the invention, all channels supporting a frequency band include the working channel and non-working channels. There are two or more channels supporting a frequency band. All channels supporting a frequency band include multiple channels within the same frequency band. For example, in the 2.4GHz frequency band, there are channel 1 and channel 2. Channel 1 is the working channel. In other embodiments, all channels supporting a frequency band may also include multiple channels within different frequency bands. For example, there are three supporting frequency bands, and each frequency band corresponds to one channel. All channels supporting a frequency band include the working channel and the non-working channel. For example, the 2.4GHz frequency band includes channel 1, the 5GHz frequency band includes channel 36, and the 6GHz frequency band includes channel 109. Channel 1 is the working channel. The preset judgment strategy in this embodiment of the invention includes a target switching channel judgment strategy for multiple channels within the same frequency band and a target switching channel judgment strategy for multiple channels within different frequency bands. Therefore, when there is channel interference on the working channel, the switching module can select the optimal target switching channel from all channels in the supporting frequency band for switching, effectively coping with channel interference and noise under complex network conditions.

[0036] In other aspects of the embodiments of the present invention, the preset judgment strategy includes: selecting the channel with the smallest interference prediction probability value among all channels in the supported frequency band as the target switching channel.

[0037] When there are two or more channels supporting a frequency band, including both active and inactive channels, and when the channels supporting a frequency band include multiple channels within the same frequency band, a preset judgment strategy selects the channel with the lowest interference prediction probability value among all channels in the supported frequency band as the target switching channel. The channel with the lowest interference prediction probability value among all channels in the supported frequency band means that this channel has the lowest probability of channel interference among all channels within the same frequency band. In this embodiment of the invention, the channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target switching channel, and the active channel is switched to the target switching channel. For example, the switching module transmits the channel number and frequency point number corresponding to the target switching channel to the WIFI chip firmware side via a channel switching command, and the WIFI chip firmware side executes the specific channel switching command. This invention avoids switching to channels that may deteriorate in the future, thus effectively dealing with channel interference and noise under complex network conditions.

[0038] It should be noted that the channel with the lowest interference prediction probability value among all channels in the supported frequency band can be a channel whose interference prediction probability value is less than, equal to, or greater than the threshold, as long as it is the channel with the lowest probability of channel interference among all channels within the same frequency band. The channel with the lowest interference prediction probability value among all channels in the supported frequency band can be an operating channel or a non-operating channel. For example, assuming channel 1 in 2.4 GHz is an operating channel and channel 2 in 2.4 GHz is a non-operating channel, if the interference prediction probability value on channel 1 in 2.4 GHz is 0.8 and the interference prediction probability value on channel 2 in 2.4 GHz is 0.5, then channel 2 in 2.4 GHz is determined as the target handover channel. If the interference prediction probability value on channel 1 in 2.4 GHz is 0.8 and the interference prediction probability value on channel 2 in 2.4 GHz is 0.7 or 0.75, then channel 2 in 2.4 GHz is determined as the target handover channel. If the interference prediction probability value on channel 1 of 2.4G is 0.8 and the interference prediction probability value on channel 2 of 2.4G is 0.9, then channel 1 of 2.4G is determined as the target handover channel. This indicates that the channel quality of the non-working channel is inferior to that of the working channel, and channel handover is not necessary.

[0039] In other aspects of the embodiments of the present invention, there are three or more supported frequency bands, and each frequency band corresponds to one channel, and all channels of the supported frequency bands include the working channel and the non-working channel; The preset judgment strategy includes: acquiring frequency report data of the non-working channel in the supported frequency band; inputting the frequency report data of the non-working channel into the channel interference prediction model to obtain the interference prediction probability value of the non-working channel output by the channel interference prediction model; determining whether multi-link operation is supported; when multi-link operation is not supported, selecting the channel with the smallest interference prediction probability value among all channels in the supported frequency band as the target switching channel.

[0040] If there are two or more channels supporting a frequency band, and all channels in the supported frequency band include both active and inactive channels, and there are three or more supported frequency bands, with each frequency band corresponding to one channel, then the frequency report data of the inactive channel is acquired. This inactive channel frequency report data is then input into the channel interference prediction model to obtain the interference prediction probability value of the inactive channel output by the channel interference prediction model. Next, it is determined whether multi-link operation is supported. Multi-link operation refers to the device supporting simultaneous operation in channels of multiple frequency bands (e.g., 2.4 GHz, 5 GHz, and 6 GHz). When multi-link operation is not supported, it means that the device can only operate in one mode of the 2.4 GHz, 5 GHz, or 6 GHz band simultaneously. In this case, the preset judgment strategy selects the channel with the lowest interference prediction probability value among all channels in the supported frequency bands as the target switching channel. For example, with a threshold of 0.7, the interference prediction probability value for 2.4G channel 1 (working channel) is 0.8, for 5G channel 36 (non-working channel) it is 0.5, and for 6G channel 109 (non-working channel) it is 0.3. These interference prediction probability values ​​indicate that interference exists on 2.4G channel 1, but not on 5G channels 36 and 6G channels 109. In this case, 6G channel 109, with the lowest interference prediction probability value among 5G channels 36 and 6G channels 109, is selected as the target handover channel. The fact that 6G channel 109 has the lowest interference prediction probability value indicates that it is the channel with the lowest probability of channel interference among all channels across multiple frequency bands. The handover module transmits the channel number and frequency point number corresponding to 6G channel 109 to the Wi-Fi chip firmware side via a channel handover command, and the Wi-Fi chip firmware side executes the specific channel handover command.

[0041] In one embodiment, please refer to Figure 3The WIFI chip firmware sends frequency report data for all available channels in all frequency bands (2.4G, 5G, 6G) supported by the current site to the kernel module via serial port. The WIFI chip firmware is configured to report data every 10ms, reporting data for 100 time steps within 1 second (T0 to Tn, where Tn is T99), and saving it in memory. This invention reports channel frequency report data every 10ms, i.e., predicts once every 10ms, thereby performing real-time prediction of channel interference. In other embodiments, channel frequency report data within 1 second can also be cached, and prediction can be performed after 1 second, thereby saving power consumption. In this embodiment of the invention, the channel frequency report data is divided into three groups according to frequency band: Spectral_Scan_Data[CHANNEL_2.4G], Spectral_Scan_Data[CHANNEL_5G], and Spectral_Scan_Data[CHANNEL_6G], corresponding to the frequency report data of the 2.4G, 5G, and 6G frequency bands, respectively. The data storage format can be defined as follows in this embodiment of the invention: Spectral_Scan_Data[BAND_INDEX_2DOT4G][MAX_CHANNEL_INDEX_0] { { Band: 2.4GHz Channel: 1Channel Bandwidth: 20 MHz CenterFreq: 2412MHz FFT Size: 128 Noise Floor: -95 dBm Channel Utilization: 30%, FFT Bins[]: [-80, -85, -82, ..., -70, ..., -90] } The code above represents the frequency report data for channel 1 in the 2.4 GHz band. The acquisition module 101 can access the frequency report data according to the structure hierarchy in the programming language. Specifically, it can access the frequency report data for all supported channels of all frequency points by traversing the frequency point index and channel index. That is, the frequency report data for the corresponding channel in the corresponding frequency band can be accessed through Spectral_Scan_Data[BAND_INDEX][CHANNEL_INDEX].FFT Bins.

[0042] Next, frequency report data (FFT bin data and metadata) is retrieved from memory via the Netlink interface message (NL80211_ATTR_SPECTRAL_SCAN). The Long Short-Term Memory (LSTM) network analyzes the frequency report data. In this embodiment, the LSM network groups the frequency report data by frequency band and assigns a value (between 0 and 1) to each channel within the group based on its probability of being selected as the target switching channel. A higher value indicates a lower probability of selection, while a value closer to 0 indicates a higher probability of selection. Referring to Table 3, this embodiment can fuse multi-band data to construct a three-channel (one channel for each of the 2.4G, 5G, and 6G modes) time series matrix, resulting in 3*128 three-channel data. Each channel corresponds to the 2.4G, 5G, and 6G frequency bands respectively. `predict_image = np.dstack((fft_bins_2g, fft_bins_5g, fft_bins_6g))`. Each channel has two dimensions, containing the current channel at the corresponding frequency point and the FFT bin data for that channel. The final input dimension is: [time step * number of supported frequency bands * number of FFT points] = T * 3 * 128, and the output dimension is: [predicted value dimension] = 3.

[0043] Table 3

[0044] In Table 3 above, taking the interference prediction values ​​[0.8, 0.5, 0.4] of the FFT bin data corresponding to time step T2 (2.4g, 5g, 6g) as an example, with a preset threshold of 0.7, [0.8, 0.5, 0.4] means that the interference prediction value of channel 13 in the 2.4g band (0.8) is greater than 0.7, indicating interference exists; the interference prediction value of channel 42 in the 5g band (0.5) is less than 0.7, indicating no interference exists; and the interference prediction value of channel 53 in the 6g band (0.4) is less than 0.7, indicating no interference exists. Since multi-link operation is not supported, the 6G channel 53 with the lowest interference prediction probability is selected as the target switching channel. The fact that the 6G channel 53 has the lowest interference prediction probability indicates that it is the channel with the lowest probability of channel interference among all channels in multiple frequency bands. The switching module transmits the channel number and frequency point number corresponding to the target switching channel 6G channel 53 to the WIFI chip firmware side via a channel switching command, and the WIFI chip firmware side executes the specific channel switching command. This invention employs a Long Short-Term Memory (LSTM) network with predictive capabilities to analyze frequency report data and select the channel to switch to. Based on predictive analysis of frequency report data, this invention can detect channel state changes in real time, assess and select the channel to switch to, and avoid switching to channels that may deteriorate in the future.

[0045] Similarly, the channel with the lowest interference prediction probability value among all channels in the supported frequency band can be a channel whose interference prediction probability value is less than, equal to, or greater than the threshold, as long as it is the channel with the lowest probability of channel interference among all channels in the supported frequency band. The channel with the lowest interference prediction probability value among all channels in the supported frequency band can be an active channel or a non-active channel. For example, suppose 2.4G channel 1 is an active channel, and 5G channel 36 and 6G channel 109 are non-active channels. If the interference prediction probability value on 2.4G channel 1 is 0.8, and the interference prediction probabilities on 5G channel 36 and 6G channel 109 are 0.5 and 0.4 respectively, then 6G channel 109 is determined as the target handover channel. If the interference prediction probability value on 2.4G channel 1 is 0.8, and the interference prediction probabilities on 5G channel 36 and 6G channel 109 are 0.7 or 0.75 respectively, then 5G channel 36 is determined as the target handover channel. With an interference prediction probability of 0.8 on 2.4G channel 1 and 0.9 on 5G channel 36 and 6G channel 109 respectively, 2.4G channel 1 is determined as the target handover channel. This indicates that the channel quality of the non-working channel is inferior to that of the working channel, and channel handover is not necessary.

[0046] When multi-link operation is not supported, this embodiment of the invention selects the channel with the lowest interference prediction probability value among all channels in the supported frequency band as the target switching channel, and switches the working channel to the target switching channel. This embodiment of the invention achieves channel evaluation and switching, timely and effectively avoiding interference and noisy channels, and selecting the optimal channel to send data. This embodiment of the invention can sense the channel congestion in public densely populated areas, avoiding data transmission failures or delays that could compromise the data service quality of the real-time priority queue (e.g., game lag, intermittent voice calls, video call stuttering, blurry images).

[0047] In other aspects of the embodiments of the present invention, there are three or more supported frequency bands, and each frequency band corresponds to one channel, and all channels of the supported frequency bands include the working channel and the non-working channel; The preset judgment strategy includes: acquiring frequency report data of the non-working channels supporting the frequency band, inputting the frequency report data of the non-working channels into the channel interference prediction model, and obtaining the interference prediction probability value of the non-working channels output by the channel interference prediction model; when the interference prediction probability value of the non-working channels is not greater than a threshold, the non-working channels do not have interference; determining whether multi-link operation is supported; when multi-link operation is supported, determining all non-working channels without interference as target switching channels.

[0048] When there are two or more channels supporting a frequency band, and all channels in the supported frequency band include both active and inactive channels, and there are three or more supported frequency bands, with each frequency band corresponding to one channel, the frequency report data of the inactive channels is acquired, and the frequency report data of the inactive channels is input into the channel interference prediction model to obtain the interference prediction probability value of the inactive channels output by the channel interference prediction model. When the interference prediction probability value of the inactive channels is not greater than a threshold, then the inactive channels do not have interference. Next, it is determined whether multi-link operation is supported. When multi-link operation is supported, it means that the device can simultaneously select to work in multiple frequency bands among the 2.4G, 5G, and 6G bands. When multi-link operation is supported, the preset judgment strategy determines all the inactive channels without interference as the target switching channels.

[0049] For example, with a threshold of 0.7, the interference prediction probability value for 2.4G channel 1 (working channel) is 0.8, for 5G channel 36 (non-working channel) it is 0.5, and for 6G channel 109 (non-working channel) it is 0.3. These interference prediction probability values ​​indicate that interference exists on 2.4G channel 1, but not on 5G channels 36 and 6G channels 109. In this case, 5G channels 36 and 6G channels 109 are selected as the target switching channels. The switching module transmits the channel number and frequency point number corresponding to 5G channel 36 and 6G channel 109 to the Wi-Fi chip firmware side via a channel switching command, and the Wi-Fi chip firmware side performs dual-channel switching.

[0050] In Table 3 above, taking the interference prediction values ​​[0.8, 0.5, 0.4] of the FFT bin data corresponding to time step T2 (2.4g, 5g, 6g) as an example, with a preset threshold of 0.7, [0.8, 0.5, 0.4] means that for channel 13 in the 2.4g band, an interference prediction value of 0.8 greater than 0.7 indicates interference; for channel 42 in the 5g band, an interference prediction value of 0.5 less than 0.7 indicates no interference; and for channel 53 in the 6g band, an interference prediction value of 0.4 less than 0.7 indicates no interference. Since multi-link operation is supported at this time, 5G channel 42 and 6G channel 53 are selected as the target switching channels. The switching module transmits the channel number and frequency point number corresponding to 5G channel 42 and 6G channel 53 as the target switching channels to the WIFI chip firmware side through a channel switching command, and the WIFI chip firmware side executes the specific channel switching command.

[0051] In this embodiment of the invention, when multi-link operation is supported, all non-working channels free from interference are identified as target switching channels, and the working channels are switched to the target switching channels. This embodiment of the invention achieves channel evaluation and switching, effectively and promptly avoiding interference and noisy channels, and selecting the optimal channel to transmit data. This embodiment of the invention can detect channel congestion in densely populated public areas, avoiding data transmission failures or delays that could compromise the data service quality of the real-time priority queue (e.g., game lag, intermittent voice calls, and video call stuttering or blurry images).

[0052] In other aspects of this invention, the target switching channel is determined according to a preset judgment strategy. When the working channel is not the target switching channel, the channel utilization of the target switching channel can also be considered. That is, when the target switching channel's utilization is greater than or equal to a set threshold (e.g., 70%), it is considered that the channel utilization is too high and is excluded; in this case, the working channel is not switched to the target switching channel. Therefore, this invention can prevent user equipment in densely populated scenarios from continuously accessing good channels, which is beneficial for improving access capacity and channel utilization.

[0053] The channel interference prediction system includes a processor and a memory. The acquisition module 101, prediction module 102 and first determination module 103 are all stored in the memory as program units. The processor executes the program units stored in the memory to realize the corresponding functions.

[0054] A processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured.

[0055] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0056] Please refer to Figure 4 This invention provides a channel interference prediction method implemented by the above-mentioned channel interference prediction system, comprising: Step 410: Obtain frequency report data for the working channels of the supported frequency bands.

[0057] Step 420: Input the frequency report data into the channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model.

[0058] Step 430: When the interference prediction probability value of the working channel is greater than the threshold, it is determined that there is interference in the working channel.

[0059] This invention trains a channel interference prediction model based on multiple historical frequency report data and the corresponding channel interference prediction probability value for each historical frequency report data. The channel interference prediction model is then used to predict channel interference on the frequency report data of the operating channels supporting the frequency band. Finally, the existence of interference is determined based on the interference prediction probability value. Therefore, this invention can predict channel state changes, avoid switching to channels that may experience interference or noise in the future, and effectively cope with channel interference and noise under complex network conditions.

[0060] Optionally, there are two or more channels supporting the frequency band, including the working channel and the non-working channel; After determining that interference exists in the working channel when the interference prediction probability value of the working channel is greater than the threshold, the method further includes: The target switching channel is determined according to a preset judgment strategy; When the working channel is not the target switching channel, the working channel is switched to the target switching channel.

[0061] Optionally, there are two or more channels supporting the frequency band, including the working channel and the non-working channel; After determining that interference exists in the working channel when the interference prediction probability value of the working channel is greater than the threshold, the method further includes: The target switching channel is determined according to a preset judgment strategy; When the working channel is not the target switching channel, the working channel is switched to the target switching channel.

[0062] Optionally, the preset judgment strategy includes: The channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

[0063] Optionally, there are three or more supported frequency bands, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels of the supported frequency bands; The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. Determine whether multi-link operation is supported; When multi-link operation is not supported, the channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

[0064] Optionally, there are three or more supported frequency bands, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels in the supported frequency bands. The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. When the interference prediction probability value of the non-working channel is not greater than the threshold, the non-working channel does not have interference. Determine whether multi-link operation is supported; When multi-link operation is supported, all non-working channels that are free from interference are identified as target switching channels.

[0065] Figure 5 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 5 As shown, the electronic device may include a processor 510, a communication interface 520, a memory 530, and a communication bus 540, wherein the processor 510, the communication interface 520, and the memory 530 communicate with each other through the communication bus 540. The processor 510 can call logical instructions in the memory 530 to execute a channel interference prediction method, which includes: acquiring frequency report data of the working channel supporting the frequency band; inputting the frequency report data into a channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model; determining that the working channel has interference when the interference prediction probability value of the working channel is greater than a threshold; wherein the channel interference prediction model is trained based on a neural algorithm that traverses multiple historical frequency report data of multiple channels and the channel interference prediction probability value corresponding to each historical frequency report data.

[0066] Furthermore, the logical instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0067] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program that can be stored on a machine-readable storage medium. When the computer program is executed by a processor, the computer is able to execute a channel interference prediction method, the method including: acquiring frequency report data of a working channel supporting a frequency band; inputting the frequency report data into a channel interference prediction model to obtain an interference prediction probability value of the working channel output by the channel interference prediction model; determining that interference exists in the working channel when the interference prediction probability value of the working channel is greater than a threshold; wherein the channel interference prediction model is trained based on a neural algorithm that traverses multiple historical frequency report data of multiple channels and the channel interference prediction probability value corresponding to each historical frequency report data.

[0068] In another aspect, the present invention also provides a machine-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a channel interference prediction method, the method comprising: acquiring frequency report data of a working channel supporting a frequency band; inputting the frequency report data into a channel interference prediction model to obtain an interference prediction probability value of the working channel output by the channel interference prediction model; determining that interference exists in the working channel when the interference prediction probability value of the working channel is greater than a threshold; wherein the channel interference prediction model is trained based on a neural algorithm traversing multiple historical frequency report data of multiple channels and the channel interference prediction probability value corresponding to each historical frequency report data.

[0069] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0070] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A channel interference prediction system, characterized in that, include: The acquisition module is used to acquire frequency report data of the working channels supporting the frequency band; The prediction module is used to input the frequency report data into the channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model. The first determining module is used to determine that there is interference in the working channel when the interference prediction probability value of the working channel is greater than a threshold. The channel interference prediction model is trained based on a neural algorithm that traverses multiple historical frequency report data for multiple channels and the channel interference prediction probability value corresponding to each historical frequency report data.

2. The channel interference prediction system according to claim 1, characterized in that, All channels in the supported frequency bands must have two or more channels, including the working channels and non-working channels; The system also includes: The second determining module is used to determine the target switching channel according to a preset judgment strategy; The switching module is used to switch the working channel to the target switching channel when the working channel is not the target switching channel.

3. The channel interference prediction system according to claim 2, characterized in that, The preset judgment strategy includes: The channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

4. The channel interference prediction system according to claim 2, characterized in that, The number of supported frequency bands is three or more, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels of the supported frequency bands; The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. Determine whether multi-link operation is supported; When multi-link operation is not supported, the channel with the lowest interference prediction probability value among all channels in the supported frequency band is selected as the target handover channel.

5. The channel interference prediction system according to claim 2, characterized in that, The number of supported frequency bands is three or more, and each frequency band corresponds to one channel. All channels of the supported frequency bands include the working channel and the non-working channel. The preset judgment strategy includes: Obtain frequency report data for the non-operating channels in the supported frequency bands. The frequency report data of the non-operating channel is input into the channel interference prediction model to obtain the interference prediction probability value of the non-operating channel output by the channel interference prediction model. When the interference prediction probability value of the non-working channel is not greater than the threshold, the non-working channel does not have interference. Determine whether multi-link operation is supported; When multi-link operation is supported, all non-working channels that are free from interference are identified as target switching channels.

6. A channel interference prediction method based on the channel interference prediction system according to any one of claims 1-5, characterized in that, include: Obtain frequency report data for the working channels of the supported frequency bands; The frequency report data is input into the channel interference prediction model to obtain the interference prediction probability value of the working channel output by the channel interference prediction model. When the interference prediction probability value of the working channel is greater than the threshold, it is determined that there is interference in the working channel; The channel interference prediction model is trained based on a neural algorithm that traverses multiple historical frequency report data for multiple channels and the channel interference prediction probability value corresponding to each historical frequency report data.

7. The channel interference prediction method according to claim 6, characterized in that, All channels in the supported frequency bands must have two or more channels, including the working channels and non-working channels; After determining that interference exists in the working channel when the interference prediction probability value of the working channel is greater than the threshold, the method further includes: The target switching channel is determined according to a preset judgment strategy; When the working channel is not the target switching channel, the working channel is switched to the target switching channel.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the channel interference prediction method according to any one of claims 6 to 7.

9. A machine-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the channel interference prediction method according to any one of claims 6 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the channel interference prediction method according to any one of claims 6 to 7.

Images