Data transmission method and device, electronic equipment and storage medium

Abstract

The invention provides a data transmission method and device, electronic equipment and a storage medium, relates to the technical field of communication networks, and is used for reducing prediction errors in a data transmission process. The method comprises the following steps: acquiring historical channel state data of a wireless channel, the historical channel state data being used for representing channel characteristics of the wireless channel; a lightweight channel prediction model is called to predict the historical channel state data to obtain predicted signal-to-noise ratio data, and the predicted signal-to-noise ratio data is used for representing the channel quality of the wireless channel at the future moment; adjusting the predicted signal-to-noise ratio data based on the signal-to-noise ratio offset to obtain adjusted predicted signal-to-noise ratio data; wherein the predicted signal-to-noise ratio data after adjustment is smaller than the predicted signal-to-noise ratio data before adjustment; determining a first modulation coding strategy based on the adjusted predicted signal-to-noise ratio data; and transmitting data based on the first modulation coding strategy.

Description

Technical Field

[0001] This application relates to the field of communication network technology, and in particular to a data transmission method, apparatus, electronic device and storage medium. Background Technology

[0002] Currently, when deploying deterministic services in wireless networks, the time-varying and random nature of wireless channels leads to problems such as multipath fading and co-channel interference, resulting in drastic fluctuations in channel quality and posing a fundamental challenge.

[0003] In related technologies, predictive adaptive modulation and coding (Pred-AMC) is commonly used to achieve deterministic transmission when channel prediction is employed. However, this method suffers from amplification effects on prediction errors. Specifically, there is a non-linear threshold between channel quality and the optimal modulation and coding scheme, meaning that even small prediction deviations can lead to severe mismatches. Overestimating the channel results in high bit error rates and retransmissions, compromising determinism, while underestimating the channel wastes spectral efficiency, leading to the technical problem of large prediction errors during data transmission. Summary of the Invention

[0004] This application provides a data transmission method, apparatus, electronic device, and storage medium for reducing prediction errors during data transmission.

[0005] In a first aspect, this application provides a data transmission method, the method comprising: acquiring historical channel state data of a wireless channel, wherein the historical channel state data is used to represent the channel characteristics of the wireless channel; calling a lightweight channel prediction model to predict the historical channel state data to obtain predicted signal-to-noise ratio (SNR) data, wherein the predicted SNR data is used to represent the channel quality of the wireless channel at a future time; adjusting the predicted SNR data based on an SNR offset to obtain adjusted predicted SNR data; wherein the adjusted predicted SNR data is less than the original predicted SNR data; determining a first modulation and coding strategy based on the adjusted predicted SNR data; and transmitting data based on the first modulation and coding strategy.

[0006] The technical solution provided in this application offers at least the following benefits: It uses a lightweight channel prediction model to locally predict historical channel state data, and then conservatively adjusts the predicted signal-to-noise ratio (SNR) data based on the SNR offset. Furthermore, based on the adjusted predicted SNR data, a corresponding first modulation and coding strategy is selected for data transmission. In other words, by conservatively adjusting the channel quality of predicted future time slots, this application solves the high bit error rate and retransmission problems caused by overly optimistic channel prediction in existing technologies. Introducing the SNR offset is equivalent to reserving a safety margin, improving link reliability, and reducing the risk of transmission failure. This solves the technical problem of large prediction errors during data transmission and achieves the technical effect of reducing prediction errors during data transmission.

[0007] One possible implementation is that the signal-to-noise ratio offset is the control parameter corresponding to the transmission strategy that meets reliability constraints through multiple simulation scenarios in a virtual environment.

[0008] Another possible implementation involves adjusting the predicted signal-to-noise ratio (SNR) data based on the SNR offset to obtain adjusted predicted SNR data. This includes determining the adjusted predicted SNR data by summing the SNR offset and the predicted SNR data.

[0009] Another possible implementation method further includes: determining the block error rate of data during transmission based on acknowledgment information during data transmission; in response to the block error rate being greater than a block error rate threshold, determining a second modulation and coding strategy based on channel quality indication information in the next transmission time interval of the current transmission time interval; wherein the channel quality indication information is used to indicate the channel quality of the wireless channel at the current moment; and transmitting data based on the second modulation and coding strategy.

[0010] Another possible implementation method further includes: continuing to transmit data based on the first modulation and coding strategy in response to a block error rate less than or equal to a block error rate threshold.

[0011] Another possible implementation method further includes: acquiring historical angle-of-arrival data of the beam received by the base station, wherein the historical angle-of-arrival data is used to represent the incident direction information of the beam; calling a lightweight angle prediction model to predict the historical angle-of-arrival data to obtain predicted angle-of-arrival data, wherein the predicted angle-of-arrival data is used to represent the incident direction information of the beam at a future time; and transmitting wide-beam data based on the predicted angle-of-arrival data.

[0012] Another possible implementation method further includes: determining angular deviation data between predicted angle of arrival data and target angle of arrival data, wherein the target angle of arrival data is obtained by beam scanning of the coverage area of ​​the wide beam; and transmitting data based on the narrow beam in response to the angular deviation data being less than an angle threshold and the gain of the narrow beam nested in the wide beam being greater than a gain threshold.

[0013] Another possible implementation method further includes: continuing to transmit data based on a wide beam in response to angular deviation data being greater than or equal to an angular threshold, or the gain of the narrow beam being less than or equal to a gain threshold.

[0014] Secondly, this application provides a data transmission apparatus, comprising: an acquisition module for acquiring historical channel state data of a wireless channel, wherein the historical channel state data represents the channel characteristics of the wireless channel; a prediction module for calling a lightweight channel prediction model to predict the historical channel state data to obtain predicted signal-to-noise ratio (SNR) data, wherein the predicted SNR data represents the channel quality of the wireless channel at a future time; an adjustment module for adjusting the predicted SNR data based on an SNR offset to obtain adjusted predicted SNR data, wherein the adjusted predicted SNR data is less than the original predicted SNR data; a determination module for determining a first modulation and coding strategy based on the adjusted predicted SNR data; and a transmission module for transmitting data based on the first modulation and coding strategy.

[0015] Thirdly, this application provides an electronic device comprising: a processor and a memory; the memory storing processor-executable instructions; when the processor is configured to execute the instructions, causing the electronic device to implement the method of the first aspect described above.

[0016] Fourthly, this application provides a computer-readable storage medium comprising: computer software instructions; which, when executed in an electronic device, cause the electronic device to implement the method described in the first aspect.

[0017] Fifthly, this application provides a computer program product comprising a computer program; when the computer program is run in an electronic device, it causes the electronic device to implement the method described in the first aspect.

[0018] The beneficial effects of the second to fifth aspects mentioned above are described in the corresponding description of the first aspect and will not be repeated here. Attached Figure Description

[0019] Figure 1 A schematic diagram of the architecture of a data transmission system provided in this application; Figure 2 A flowchart illustrating a data transmission method provided in this application; Figure 3 A schematic diagram illustrating the model training and policy evaluation process in a digital twin network layer provided in this application; Figure 4 A flowchart illustrating another data transmission method provided in this application; Figure 5 A schematic diagram illustrating the interaction between a physical network layer and a digital twin network layer provided in this application; Figure 6 A flowchart illustrating a conservative predictive AMC implementation provided for this application; Figure 7 A schematic diagram of a hierarchical predictive beam management process provided for this application; Figure 8 A schematic diagram of the composition of a data transmission device provided in this application; Figure 9 This is a schematic diagram of the composition of an electronic device provided in this application. Detailed Implementation

[0020] The data transmission method provided in this application will now be described in detail with reference to the accompanying drawings.

[0021] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0022] The terms "first" and "second," etc., used in the specification and drawings of this application are used to distinguish different objects or to distinguish different treatments of the same object, rather than to describe a specific order of objects.

[0023] Furthermore, the terms "comprising" and "having," and any variations thereof, used in the description of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0024] It should be noted that in the embodiments of this application, the words "exemplarily" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplarily" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplarily" or "for example" is intended to present the relevant concepts in a specific manner.

[0025] To facilitate a clear description of the technical solutions of the embodiments of this application, the terms "first" and "second" are used in the embodiments of this application to distinguish the same or similar items with essentially the same function and effect. Those skilled in the art can understand that the terms "first" and "second" are not intended to limit the quantity or execution order.

[0026] In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0027] Currently, emerging applications such as industrial automation, remote control, and autonomous driving place extreme demands on network latency, jitter, and reliability, driving the development of deterministic network technologies. In the wired domain, technologies such as time slot scheduling and frame duplication based on communication protocol standards (e.g., IEEE 802.1TSN) can provide microsecond-level deterministic guarantees. However, deploying deterministic services in wireless networks, such as Ultra-Reliable Low-Latency Communication (5GURLLC) and Wi-Fi 7, faces fundamental challenges.

[0028] Traditional wireless transmission assurance mechanisms, such as adaptive modulation and coding schemes and hybrid automatic repeat requests, are essentially reactive, adjusting only after errors or packet loss are detected. They cannot proactively mitigate risks and struggle to meet stringent requirements for bounded latency and extremely low jitter. Existing deterministic wireless solutions often employ static excess resource reservations, such as reserving maximum redundant bandwidth or consistently using the highest protection level coding or duplicated paths. While these methods can guarantee service delivery under worst-case conditions, they lead to significant spectrum resource waste and network inefficiency.

[0029] In related technologies, research on deterministic transmission achieved through channel prediction mainly includes Predictive Adaptive Modulation and Coding (Pred-AMC) technology, which aims to optimize the physical layer link adaptation process. Traditional Adaptive Modulation and Coding (AMC) relies on current or past channel measurements, such as Channel Quality Indicator (CQI), to select a Modulation and Coding Scheme (MCS). Due to measurement, feedback, and processing delays, the selected MCS may become outdated when the channel changes rapidly, leading to unreliable or inefficient transmission. Pred-AMC, on the other hand, utilizes Artificial Intelligence (AI) models, such as Long Short-Term Memory (LSTM) networks and Gated Recurrent Units (GRUs), to predict the channel quality, such as the signal-to-noise ratio (SNR), for one or more transmission time intervals (TTIs) in the future, and determines the optimal MCS level in advance based on the predicted values. When the data packet actually arrives at the scheduler, the system already knows the channel state and can directly use the best-matching MCS for transmission, thereby improving efficiency. However, as mentioned in the background section, predictive adaptive modulation and coding techniques suffer from a large prediction error during data transmission.

[0030] Based on this, this application uses a lightweight channel prediction model to locally predict historical channel state data, and then conservatively adjusts the predicted signal-to-noise ratio (SNR) data based on the SNR offset. Furthermore, based on the adjusted predicted SNR data, a corresponding first modulation and coding strategy is selected for data transmission. In other words, by conservatively adjusting the channel quality of predicted future time slots, this application can solve the high bit error rate and retransmission problems caused by overly optimistic channel prediction in existing technologies. Introducing the SNR offset is equivalent to reserving a safety margin, improving link reliability, and reducing the risk of transmission failure, thereby solving the technical problem of large prediction errors during data transmission and achieving the technical effect of reducing prediction errors during data transmission.

[0031] The embodiments provided in this application will now be described in detail with reference to the accompanying drawings.

[0032] The data transmission method provided in this application can be applied to, for example... Figure 1 In the system architecture shown. Figure 1 A schematic diagram of the architecture of a data transmission system provided in this application is shown below. Figure 1As shown, the system includes a digital twin network layer 101 and a physical network layer 102. The digital twin network layer 101 and the physical network layer 102 interact by synchronizing real-time status data and distributing lightweight strategy parameters.

[0033] The digital twin network layer 101 includes a high-fidelity channel simulator, a model training and policy evaluation module, and a conservative policy generator. The high-fidelity channel simulator simulates the physical environment, specifically constructing a virtual wireless environment that evolves synchronously with the physical environment based on real-time status data and a geographic information database fed back from the physical network layer 102. The model training and policy evaluation module runs complex models for virtual simulations. Specifically, in the virtual environment, it runs complex prediction models for training and policy exploration, allowing for the safe testing of various aggressive prediction policies without affecting the actual network. The conservative policy generator calculates security parameters based on the virtual evaluation results generated by the model training and policy evaluation module. Specifically, based on the test results of the virtual environment, it calculates dynamic conservative offsets for the physical network layer 102. For example, for AMC (Advanced Multi-Channel Compression), it calculates the SNR safety margin required to guarantee the target block error rate under the current channel conditions; for beamforming, it calculates the optimal wide-beam backup scheme or beam offset angle.

[0034] The physical network layer 102 includes: a real base station, user equipment, and a wireless channel. The physical network layer 102 is responsible for performing the final data transmission and continuously feeding back the actual channel measurement results, acknowledgment (ACK) or negative acknowledgment (NACK) information to the digital twin network layer 101.

[0035] In some embodiments, the workflow within the digital twin network layer includes: starting with synchronizing data from the physical network layer, performing multi-scenario simulations in a virtual environment to evaluate the performance of different strategies, generating specific conservative strategy parameters based on reliability objectives, and finally distributing the generated conservative strategy parameters to each base station in the physical network layer.

[0036] In some embodiments, the interaction process between the digital twin network layer and the physical network layer includes: the physical network layer collecting real-time state data and synchronizing the collected real-time state data to the digital twin network layer; the digital twin network layer updating the virtual environment based on the real-time state data, performing multi-scenario virtual simulations using complex models in the virtual environment, evaluating the performance of different strategies, generating conservative strategy parameters, and distributing the generated conservative strategy parameters to the physical network layer; and the physical network layer updating its local strategy based on the conservative strategy parameters.

[0037] It should be noted that the system architecture described in the embodiments of this application is for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and does not constitute a limitation on the technical solutions provided in the embodiments of this application. Those skilled in the art will understand that as system architectures evolve, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems. See also Figure 2 This is a flowchart illustrating a data transmission method provided in an embodiment of this application. Figure 2 As shown, the data transmission method provided in this application specifically includes the following steps S201~S205: S201. Obtain historical channel status data of the wireless channel.

[0038] Historical channel state data can be used to represent the channel characteristics of a wireless channel, such as channel frequency response, reference signal received power, channel quality indication (CQI), gain, phase, and angle of arrival (AoA). This is merely an example and does not impose specific limitations on the content of historical channel state data. Historical channel state data can also be called historical channel state information (CSI).

[0039] For example, local historical CSI can be collected, and by utilizing the temporal correlation of wireless channels and analyzing the channel status of several past time slots, the channel quality of future time slots can be accurately predicted. Furthermore, based on the prediction results, the base station can select modulation and coding schemes in advance, thereby significantly improving the reliability and efficiency of communication in high-frequency bands (such as millimeter waves) or high-mobility scenarios.

[0040] S202. Call the lightweight channel prediction model to predict the historical channel state data and obtain the predicted signal-to-noise ratio data.

[0041] The predicted signal-to-noise ratio (SNR) data can be used to represent the channel quality of the wireless channel at future times, and can be represented by SNR_pred. The lightweight channel prediction model can be a lightweight prediction model recommended by the digital twin network layer (digital twin layer) received by the base station in the physical network layer (referred to as the physical layer). For example, it can be a small LSTM.

[0042] For example, the lightweight prediction model described above can be used to predict the channel quality SNR_pred for future time slots based on local historical CSI.

[0043] S203. Based on the signal-to-noise ratio offset, adjust the predicted signal-to-noise ratio data to obtain the adjusted predicted signal-to-noise ratio data.

[0044] The adjusted predicted SNR data is less than the unadjusted predicted SNR data. The SNR offset can be a conservative offset received by the physical layer base station from the digital twin layer, represented by Δ, and this conservative offset is a negative dB value. The adjusted predicted SNR data can be represented by SNR_decision.

[0045] It should be noted that the aforementioned conservative offset Δ is dynamically provided by the digital twin layer. For example, when the channel changes rapidly or the prediction confidence is low, the absolute value of Δ will increase.

[0046] For example, the channel quality of the predicted future time slot is conservatively adjusted based on a conservative offset Δ, which aims to reserve performance margin to cope with prediction errors, sudden environmental changes or channel uncertainties, thereby avoiding link failure, increased bit error rate or access interruption due to overly optimistic estimation.

[0047] In some embodiments, adjusting the predicted signal-to-noise ratio (SNR) data based on the SNR offset to obtain adjusted predicted SNR data includes: determining the adjusted predicted SNR data by summing the SNR offset and the predicted SNR data.

[0048] For example, the SNR_decision used for decision-making can be calculated using the following formula: SNR_decision=SNR_pred+Δ In this process, the predicted channel quality of the future time slot is added to the conservative offset Δ. Since the conservative offset Δ is a negative dB value, it can reduce the predicted channel quality.

[0049] S204. Based on the adjusted predicted signal-to-noise ratio data, determine the first modulation and coding strategy.

[0050] For example, based on the determined SNR_decision, a predefined MCS mapping table is searched to select the first modulation and coding scheme.

[0051] S205. Transmit data based on the first modulation and coding strategy.

[0052] For example, data transmission operations are performed based on a first modulation and coding strategy selected from a predefined MCS mapping table.

[0053] In some embodiments, the method further includes: determining the block error rate of data during transmission based on acknowledgment information during data transmission; in response to the block error rate being greater than a block error rate threshold, determining a second modulation and coding strategy based on channel quality indication information in the next transmission time interval of the current transmission time interval; and transmitting data based on the second modulation and coding strategy.

[0054] The acknowledgment information can be used to indicate the transmission result of data transmitted based on the first modulation and coding strategy. The acknowledgment information can include at least ACK and NACK information. The block error rate threshold can be a target threshold set according to the actual data transmission process, and can be expressed as a percentage. For example, the block error rate threshold can be 10%. This is only an example and no specific limit is placed on the value of the block error rate threshold. The transmission time interval (TTI) can be the smallest time unit for data transmission and is also the basic time granularity for link adaptation. The channel quality indicator (CQI) information can be a quantitative indicator fed back from the user equipment to the base station in a wireless communication system. It can be used to indicate the channel quality of the wireless channel at the current moment, so that the base station can select an appropriate modulation and coding scheme for data transmission. The second modulation and coding strategy can be the traditional AMC scheme.

[0055] For example, during data transmission based on the first modulation and coding strategy, actual ACK and NACK feedback can be collected. Based on the collected ACK and NACK feedback, the Block Error Rate (BLER) can be calculated in real time over a short period. If the BLER continuously exceeds a threshold (indicating severely optimistic prediction), a mode switch (fast backoff mechanism) is immediately triggered: the prediction is abandoned in the next transmission time interval (TTI), and data is transmitted directly using the traditional AMC scheme based on the latest real-time measured CQI.

[0056] It should be noted that after the fast rollback mechanism is triggered, the performance anomaly can be reported to the digital twin layer to trigger the digital twin layer to re-evaluate and adjust its strategy.

[0057] This application makes confidence judgments by calculating the BLER in real time over a short period of time. When the BLER exceeds the threshold continuously, a fast backoff mechanism is immediately triggered, which can avoid continuous transmission failures caused by overly optimistic channel prediction, thereby ensuring link reliability, reducing retransmission overhead, and improving the adaptability and robustness of the communication system in highly dynamic scenarios.

[0058] In some embodiments, the method further includes: in response to a block error rate less than or equal to a block error rate threshold, continuing to transmit data based on a first modulation and coding strategy.

[0059] For example, if the BLER does not exceed the threshold (i.e., the prediction is relatively accurate), the predicted channel quality basically matches the actual link conditions, the current modulation and coding scheme is reasonable and reliable, there is no need to trigger the backoff mechanism, and the data transmission operation continues to be performed by the first modulation and coding strategy selected from the predefined MCS mapping table.

[0060] This application continues to use the current first modulation and coding strategy to transmit data when the block error rate is less than or equal to a threshold. This can maintain high spectral efficiency and throughput while ensuring communication reliability, avoid unnecessary down-order scheduling, and thus achieve efficient utilization of system resources and stable, high-performance data transmission.

[0061] In some embodiments, the signal-to-noise ratio offset is a control parameter corresponding to a transmission strategy that satisfies reliability constraints through multiple simulation scenarios in a virtual environment.

[0062] For example, Figure 3 This application provides a flowchart illustrating the model training and policy evaluation process in a digital twin network layer, as shown below. Figure 3 As shown, the model training and policy evaluation process in this digital twin network layer may include the following steps: Step S301: Receive data synchronized from the physical network layer.

[0063] The digital twin layer synchronizes real-time status data from the physical network layer. This real-time status data can include at least user location and coarse channel characteristics.

[0064] Step S302: Load the high-complexity prediction model.

[0065] In the twin environment, a highly complex prediction model is used to extrapolate the future channel under multiple scenarios and strategies. The aforementioned highly complex prediction model can be a large convolutional neural network – long short-term memory network (Large CNN-LSTM).

[0066] Step S303: Select the simulation scene.

[0067] The high-complexity prediction model described above is used to select multiple simulation scenarios for extrapolation. For example, the first scenario can be a normal movement scenario, i.e., a normal channel change simulation; the second scenario can be a sudden blockage scenario, i.e., a sudden blockage simulation; the third scenario can be a high-speed movement scenario, i.e., a high-speed Doppler simulation; and the fourth scenario can be a sudden increase in interference scenario, i.e., a sudden change in interference simulation.

[0068] Step S304: Virtual transmission and performance evaluation.

[0069] In this process, multi-scenario simulations are performed in a virtual environment, and the virtual transmission performance (e.g., throughput, interruption probability) under different strategies (e.g., different MCS, different beamwidths) is evaluated to obtain the evaluation results.

[0070] Step S305: Calculate the performance indicators of each strategy.

[0071] Based on the above evaluation results, the conservative policy generator in the digital twin layer determines a conservative policy parameter that can meet a strict reliability threshold (e.g., 99.999% success rate) in the virtual environment. For example, the target MCS is equal to the target MCS-1 (MCS_target = MCS_optimal-1), or the beamwidth equals the optimal beamwidth. 1.5.

[0072] Step S306: Determine whether the strategy performance indicators have reached the target reliability.

[0073] If the strategy performance index reaches the target reliability, proceed to step S307; otherwise, proceed to step S308.

[0074] Step S307: Generate the first conservative parameter.

[0075] When the performance index of the judgment strategy reaches the target reliability, the first conservative parameter is generated, including: conservative offset Δ=-1dB, beam coefficient=1.2.

[0076] Step S308: Generate the second conservative parameter.

[0077] When the performance index of the judgment strategy fails to reach the target reliability, a second conservative parameter is generated, including: conservative offset Δ = -3dB and beam coefficient = 1.5.

[0078] Step S309: Encapsulate conservative parameters and send them to the physical network layer.

[0079] Specifically, the lightweight conservative strategy parameters (rather than the complex model itself) generated above are sent to the base station controller corresponding to the physical layer.

[0080] For example, as can be seen from the above, the aforementioned conservative offset Δ (i.e., signal-to-noise ratio offset) can be generated by evaluating the performance indicators of different data transmission strategies based on different simulation scenarios in a virtual environment, and by judging whether the evaluation results achieve the target reliability.

[0081] For example, the conservative offset Δ can be generated by the digital twin layer using a highly complex prediction model in the twin environment to extrapolate the future channel under multiple scenarios and strategies, and to evaluate the throughput and outage probability under different MCS. The conservative strategy generator then determines whether the target reliability has been achieved based on the evaluation results. When the evaluation result meets the target reliability, a conservative offset Δ = -1dB is generated; otherwise, a conservative offset Δ = -3dB is generated. In other words, when the transmission strategy performance indicators meet the reliability target, a relatively lenient (e.g., -1dB) conservative strategy parameter is adopted; while when the transmission strategy performance indicators do not meet the reliability target, it reverts to a stricter or more conservative default strategy (e.g., -3dB) to provide a greater safety margin.

[0082] This application uses a digital twin layer to deduce the evolution of channels in multiple scenarios and different MCS strategies using a high-complexity model. Based on the evaluation results of throughput and interruption probability, a conservative offset Δ is dynamically generated. This can maximize spectral efficiency while ensuring the reliability of the target, and realize prediction-driven adaptive robust communication: avoiding performance loss due to excessive conservatism, and preventing link interruption caused by excessive aggressiveness.

[0083] In summary, this application utilizes a conservative offset received from the digital twin layer at the base station, calls a lightweight prediction model for local prediction, and applies the conservative offset to conservatively adjust the predicted channel quality. Furthermore, based on the adjusted channel quality, MCS is selected for data transmission. During transmission, confidence is determined based on the actual BLER: if performance degradation exceeds a threshold (i.e., BLER is greater than the threshold), a fast backoff to traditional AMC mode is triggered, thereby implementing conservative predictive AMC at the physical layer.

[0084] Figure 4 A flowchart of another data transmission method provided in the embodiments of this application is shown below. Figure 4 As shown, the specific steps include S401~S403: S401. Obtain the historical angle of arrival data of the beam received by the base station.

[0085] Historical angle of arrival data can be used to represent the incident direction information of the beam.

[0086] S402. Call the lightweight angle prediction model to predict the historical angle of arrival data and obtain the predicted angle of arrival data.

[0087] The lightweight angle prediction model can be a lightweight prediction model recommended by the digital twin layer, received by the base station in the physical layer. The predicted angle of arrival data can be used to represent the incident direction information of the beam at future times, and can be represented by AoA_pred.

[0088] For example, the lightweight angle prediction model described above can be used to predict a user's future angle of arrival (AoA_pred) based on the collected historical angle of arrival data.

[0089] S403, Wide beam transmission data based on predicted angle of arrival data.

[0090] For example, a wide beam corresponding to the predicted angle of arrival data AoA_pred is always used for initial access and data transmission at the prediction time to ensure basic connection reliability and thus avoid the risk of interruption due to narrow beam misalignment.

[0091] It should be noted that the physical layer base station receives the hierarchical beamcodebook and handover values ​​from the digital twin layer. The hierarchical beamcodebook can contain wide beams (wide coverage, low gain) and narrow beams nested within the wide beams (narrow coverage, high gain). The beamwidth of the wide beam equals the width of the optimal narrow beam. The beam ratio can be 1.2 when the transmission strategy performance meets the reliability target; when the transmission strategy performance does not meet the reliability target, the beam ratio can be 1.5, that is, a wider beam than the theoretical optimal is used to cope with channel estimation errors. The switching value can be the angle threshold for switching from a wide beam to a narrow beam, set according to the actual situation. For example, the switching value can be 10 degrees. This is only an example and no specific limit is imposed on the value of the switching value.

[0092] This application ensures that users can quickly and reliably access the network and maintain basic communication links in high mobility or high uncertainty scenarios by using a wide beam corresponding to the predicted angle of arrival at the predicted time for initial access and data transmission without the need for precise beam alignment. It effectively avoids access failure or connection interruption caused by narrow beam prediction deviation, thereby significantly improving system robustness and service continuity.

[0093] In some embodiments, the method further includes: determining angle deviation data between predicted angle of arrival data and target angle of arrival data; and transmitting data based on the narrow beam in response to the angle deviation data being less than an angle threshold and the gain of the narrow beam nested in the wide beam being greater than a gain threshold.

[0094] The target angle of arrival (AoA) data is obtained by beam scanning the coverage area of ​​the wide beam and can be represented by AoA_measured, also known as the measured optimal angle of arrival data. The angle deviation data represents the angular deviation between the predicted angle of arrival data and the measured optimal angle of arrival data, and can be represented by θ.

[0095] For example, under wide beam coverage, a rapid fine beam scan can be initiated simultaneously, limited to the sector covered by the wide beam, with minimal overhead. During this fine scan, if the angle deviation θ between the measured optimal angle of arrival (OA) and the predicted OA is less than the threshold issued by the digital twin layer, and the channel quality gain of the narrow beam nested within the wide beam exceeds a set threshold, then a smooth switch to the narrow beam can be achieved in the next cycle.

[0096] This application allows for a smooth switch to the narrow beam in the next cycle when the angle deviation between the measured optimal angle of arrival data and the predicted angle of arrival data is less than the angle threshold issued by the digital twin layer, and the channel quality gain of the narrow beam exceeds the threshold. This can efficiently capture the capacity improvement brought by the high-gain narrow beam while ensuring alignment reliability, and achieve an adaptive and low-risk transition from a robust wide beam to a high-performance narrow beam, thereby balancing connection reliability and spectral efficiency.

[0097] In some embodiments, the method further includes: continuing to transmit data based on a wide beam in response to an angle deviation data being greater than or equal to an angle threshold, or a narrow beam gain being less than or equal to a gain threshold.

[0098] For example, if the angle deviation θ between the measured optimal angle of arrival data and the predicted angle of arrival data is too large (i.e. greater than or equal to the threshold), or the gain of the narrow beam is insufficient (i.e. less than or equal to the gain threshold), then the wide beam mode will continue to be maintained.

[0099] It should be noted that if the angle deviation between the measured optimal angle of arrival data and the predicted angle of arrival data is too large, or if the gain of the narrow beam is insufficient, it indicates that the prediction does not match the measurement. In this case, the event of the prediction not matching the measurement is reported to the digital twin layer, which will then determine whether it is an environmental change (e.g., a new obstacle) and update its virtual environment model and prediction strategy.

[0100] When the angle deviation between the measured optimal angle of arrival data and the predicted angle of arrival data is too large, or the gain of the narrow beam is insufficient, this application maintains the wide beam mode, which can avoid link deterioration or interruption caused by forcibly switching to a poor narrow beam. Under the premise of ensuring basic communication reliability, it can robustly cope with prediction inaccuracies or sudden environmental changes, and improve the connection robustness and service continuity in highly dynamic scenarios.

[0101] In summary, this application receives the hierarchical codebook and handover values ​​from the digital twin layer at the base station, calls the lightweight angle prediction model to make predictions, and obtains the predicted angle of arrival (AHA) data. First, a wide beam corresponding to the predicted AHA data is used to ensure connectivity. Simultaneously, a rapid and detailed scan is performed within the coverage area of ​​this wide beam to obtain the measured results. Further, based on the measured results, it is determined whether the narrow beam handover condition is met. If the handover condition is met, the system switches to a high-gain narrow beam; otherwise, the wide beam mode is maintained, and any anomalies are reported to the digital twin layer, thereby achieving hierarchical predictive beam management.

[0102] The technical solutions provided by the above embodiments bring at least the following beneficial effects. The data transmission method provided in this application uses a lightweight channel prediction model to locally predict historical channel state data, and then conservatively adjusts the predicted signal-to-noise ratio (SNR) data based on the SNR offset. Furthermore, based on the adjusted predicted SNR data, a corresponding first modulation and coding strategy is selected for data transmission. In other words, by conservatively adjusting the channel quality of the predicted future time slots, this application can solve the high bit error rate and retransmission problems caused by overly optimistic channel prediction in existing technologies. Introducing the SNR offset is equivalent to reserving a safety margin, improving link reliability, reducing the risk of transmission failure, and thus solving the technical problem of large prediction errors in the data transmission process, achieving the technical effect of reducing prediction errors in the data transmission process.

[0103] The data transmission method of this application embodiment will now be described with reference to a specific example.

[0104] Currently, research on deterministic transmission technologies utilizing channel prediction also includes predictive precoding or beam management in massive MIMO (Multiple-Input Multiple-Output) systems. This technology addresses the challenges of channel acquisition overhead and high-frequency beam tracking in MIMO systems. In MIMO or millimeter-wave systems, acquiring downlink channel state information (CSI) incurs significant uplink feedback or downlink pilot overhead. When users move, CSI becomes rapidly outdated, leading to beam inaccuracies. Prediction can be achieved by leveraging the correlation between the channel in the spatial and temporal domains, treating the multi-antenna channel matrix as a spatiotemporal sequence. Using a CNN-LSTM hybrid model, CNN extracts spatial structure features in the antenna domain, while LSTM captures temporal dynamics. Together, they predict the future complete channel matrix or dominant beam direction. Location-based prediction, combined with user location and velocity information (e.g., from GPS or visual sensors), utilizes Kalman filtering or neural networks to directly predict the user's future angle of arrival, thereby pre-calculating precoding vectors or adjusting beam pointing. The prediction results can be directly used to generate candidate beam sets, narrow the beam scanning range, and significantly reduce the latency of the Synchronization Signal Block (SSB) beam scanning in 5G New Radio (NR).

[0105] However, the drawback of the above methods is their extreme vulnerability in the spatial dimension. In high-frequency narrow-beam systems such as millimeter waves, even a tiny error in the angle of arrival prediction can cause the main beam to completely deviate from the user, triggering link interruption, with consequences far more severe than traditional time-domain prediction errors. Secondly, the above methods introduce unbearable computational complexity. To model the spatiotemporal joint features, deep networks with a huge number of parameters are required to perform real-time prediction and precoding calculations on the high-dimensional channel matrix, leading to a surge in power consumption and latency. Furthermore, the model lacks adaptability in dynamic environments and struggles to cope with non-stationary scenarios such as sudden changes in scatterers. In addition, channel prediction errors in multi-user scenarios are amplified during precoding calculations, causing severe residual interference, and the technology's heavy reliance on auxiliary information such as location introduces privacy risks. Essentially, the above technologies do not eliminate the bottleneck but rather shift it from channel measurement overhead to the accuracy and reliability of prediction calculations.

[0106] In summary, existing predictive technologies suffer from significant shortcomings in terms of reliability, complexity, and environmental adaptability. Therefore, there is an urgent need for a solution that can proactively detect and respond to changes in channel quality, dynamically and precisely adjust resources while ensuring deterministic Service Level Agreements (SLAs), thereby improving wireless spectrum utilization.

[0107] To address the aforementioned technical problems, this application provides an integrated solution that uses a digital twin network as a testing ground and trainer. It introduces conservative decision-making and a fast backoff mechanism into actual physical transmission, thereby significantly improving the determinism and robustness of the system while maintaining prediction gain. This aims to overcome the core shortcomings of wireless determinism in the aforementioned related technologies. Specifically, this application relates to wireless communication, deterministic networking (DetNet), and time-sensitive networking (TSN), particularly a wireless deterministic transmission method that integrates AI channel prediction and dynamically optimizes transmission strategies to achieve bounded delay and high reliability.

[0108] This application constructs a two-layer decision-making architecture consisting of a digital twin and a physical entity. High-complexity, secure model training and policy pre-evaluation are performed at the digital twin layer; lightweight, conservative, and robust transmission is executed at the physical entity layer. Through real-time interaction between the two layers, dynamic compensation for prediction errors and rapid adaptation to environmental changes are achieved. By employing a three-pronged mechanism of offline optimization of digital twins, conservative execution at the physical layer, and rapid feedback and switching, the shortcomings of existing predictive technologies are effectively improved, resulting in at least the following benefits: Combating prediction errors: Through dynamic conservative offsetting and wide-beam backup, the impact of prediction errors is controlled within an acceptable range, prioritizing the determinism and reliability of the connection; Reducing complexity and power consumption: The most complex model training and policy exploration are placed in the offline digital twin layer, while the physical layer only runs lightweight models and executes simple rules, meeting real-time requirements; Enhancing environmental adaptability: Through real-time performance monitoring and rapid mode switching mechanisms at the physical layer, timely responses to environmental changes are possible; Simultaneously, performance anomaly feedback drives continuous learning and updating of the digital twin layer, enabling the system to have long-term evolutionary capabilities; Improving multi-user performance: The digital twin layer can coordinate conservative strategies for multiple users from a global perspective, optimizing resource allocation and suppressing the deterioration of multi-user interference.

[0109] For example, Figure 5 This application provides a schematic diagram of the interaction between a physical network layer and a digital twin network layer, such as... Figure 5 As shown, the interaction process between the physical network layer and the digital twin network layer may include the following steps: S501, the physical network layer collects real-time status data.

[0110] The aforementioned real-time status data may include at least CSI, ACK, NACK, SNR, and location information.

[0111] S502, Synchronize real-time status data to the digital twin network layer.

[0112] S503, the digital twin network layer updates the virtual environment based on real-time status data.

[0113] S504. Use complex models in a virtual environment to perform multi-scenario virtual simulations.

[0114] S505, Evaluate the performance of different strategies.

[0115] The digital twin network layer performs multi-scenario simulations in a virtual environment and evaluates the virtual transmission performance under different strategies. The performance evaluation can include at least throughput, outage probability, and latency.

[0116] S506, Generate conservative strategy parameters.

[0117] The generated conservative strategy parameters may include at least conservative offset, wide beam parameters, and switching threshold.

[0118] S507: Send the conservative strategy parameters to the physical network layer.

[0119] S508, the physical network layer updates the local policy based on conservative policy parameters.

[0120] S509. Determine whether to continue running.

[0121] If it is determined that the system should continue running, proceed to step S501; otherwise, the process ends.

[0122] Figure 6 The flowchart for implementing conservative predictive AMC provided in this application may include the following steps: Step S601: Receive the conservative policy parameters and lightweight prediction model sent by the digital twin network layer.

[0123] In this process, the physical network layer base station receives the current conservative offset (a negative dB value) and the recommended lightweight prediction model, such as a small LSTM, from the digital twin network layer. The conservative policy parameters also include a confidence threshold.

[0124] Step S602: Collect historical CSI data.

[0125] Step S603: Run the lightweight prediction model to output the channel quality of future time slots.

[0126] Among them, a lightweight prediction model is used to predict the channel quality SNR_pred for future time slots based on local historical CSI.

[0127] Step S604: Perform a conservative adjustment on the predicted channel quality to obtain the adjustment result.

[0128] The predicted channel quality is conservatively adjusted to obtain the adjusted channel quality SNR_decision. Specifically, SNR_decision used for decision-making is calculated as SNR_pred + Δ. Here, Δ is dynamically provided by the digital twin network layer, and its absolute value increases when the channel changes rapidly or the prediction confidence is low.

[0129] Step S605: Select the corresponding modulation and coding scheme based on the adjustment results.

[0130] Specifically, the predefined MCS mapping table is searched based on SNR_decision, and the corresponding MCS is selected.

[0131] Step S606: Transmit data using a modulation and coding scheme.

[0132] Step S607: Collect confirmation information feedback during data transmission.

[0133] This involves using a modulation and coding scheme to perform data transmission and collecting actual ACK and NACK feedback.

[0134] Step S608: Calculate the block error rate in the short term.

[0135] Among them, the block error rate (BLER) is calculated in real time over a short period of time.

[0136] Step S609: Determine whether the error block rate is less than or equal to the threshold.

[0137] When the error rate is less than or equal to the threshold, proceed to step S610; otherwise, proceed to step S612.

[0138] Step S610: Maintain the prediction mode.

[0139] Specifically, when the error rate is less than or equal to the threshold, the prediction mode continues.

[0140] Step S611: Report performance data normally.

[0141] Step S612: Trigger the fast rollback mechanism.

[0142] Specifically, when the block error rate (BLER) exceeds a threshold, a fast rollback mechanism is triggered. In other words, if the BLER continuously exceeds the threshold, a mode switch is immediately triggered.

[0143] Step S613: Switch the prediction mode to the traditional AMC scheme.

[0144] In the next TTI, prediction is abandoned, and the traditional AMC scheme based on the latest real-time measurement CQI is directly adopted.

[0145] Step S614: Report the anomaly to the digital twin network layer.

[0146] This performance anomaly was reported to the digital twin layer, triggering a reassessment and adjustment of its strategies.

[0147] In summary, the implementation process of conservative predictive AMC in the physical layer of this application is as follows: the base station receives parameters from the digital twin layer, performs local prediction, applies conservative offset adjustment, selects the MCS, and makes a confidence judgment based on the actual BLER after transmission. If the performance deterioration exceeds the threshold, a fast fallback to the traditional AMC mode is triggered.

[0148] Figure 7 The present application provides a schematic diagram of a hierarchical predictive beam management process, which may include the following steps: Step S701: Receive configuration parameters sent by the digital twin network layer.

[0149] The physical layer base station receives layered beamcodebooks and handover values ​​from the digital twin layer. The codebook contains wide beams (wide coverage, low gain) and nested narrow beams (narrow coverage, high gain).

[0150] Step S702: Collect historical angle of arrival data.

[0151] Step S703: Run the lightweight angle prediction model and output the predicted angle of arrival.

[0152] Among them, a lightweight angle prediction model is used to predict the user's future angle of arrival AoA_pred.

[0153] Step S704: Perform initial transmission using a wide beam.

[0154] Specifically, at the prediction time, a wide beam corresponding to AoA_pred is used for initial access and data transmission to ensure basic connection reliability and thus avoid the risk of interruption due to narrow beam inaccuracy.

[0155] Step S705: Determine the measured optimal angle of arrival.

[0156] Among them, under wide beam coverage, a fast fine beam scan is initiated simultaneously, with the range limited to the sector covered by the wide beam. This process has minimal overhead and is used to measure the actual optimal narrow beam.

[0157] Step S706: Determine the angle deviation between the predicted angle of arrival and the measured optimal angle of arrival.

[0158] Among them, the angular deviation θ between the predicted angle of arrival and the measured optimal angle of arrival AoA_measured can be determined.

[0159] Step S707: Determine whether the switching conditions are met.

[0160] The switching conditions include two conditions: θ is less than a threshold and the narrow beam gain is sufficiently increased. If the switching conditions are met, proceed to step S708; otherwise, proceed to step S710.

[0161] Step S708: Switch from wide beam mode to narrow beam mode.

[0162] If the directional deviation between the measured optimal narrow beam and the predicted optimal narrow beam discovered in the fine scan is less than the threshold issued by the digital twin layer, and the channel quality gain of the narrow beam exceeds the set threshold, then the system smoothly switches to the narrow beam in the next cycle.

[0163] Step S709: Report performance data normally.

[0164] Step S710: Maintain wide beam mode.

[0165] If the directional deviation θ between the measured optimal narrow beam and the predicted optimal narrow beam is too large or the gain is insufficient, the system will remain in wide beam mode.

[0166] Step S711: Report prediction deviation events.

[0167] Among these features, while maintaining the wide beam pattern, events that do not match the predictions will be reported to the digital twin layer.

[0168] Step S712: Update the digital twin network layer model.

[0169] The digital twin layer uses this information to determine whether there is a sudden change in the environment (such as a new obstacle) and updates its virtual environment model and prediction strategy accordingly.

[0170] In summary, the hierarchical predictive beam management process of this application is as follows: The base station receives the hierarchical codebook and values, predicts the angle of arrival, and first uses a wide beam to ensure connectivity while performing a rapid and fine scan. Based on the measured results, it is determined whether the narrow beam switching conditions are met. If the conditions are met, it switches to a high-gain narrow beam; otherwise, it maintains the wide beam mode and reports the anomaly to the digital generation layer.

[0171] This application addresses predictive AMC. The physical layer execution steps specifically include: predicting future channel quality (SNR_pred) using a lightweight model, making conservative adjustments using SNR_decision = SNR_pred + Δ, and then selecting a modulation and coding scheme (MCS); real-time statistics of the short-term block error rate (BER). If the BER exceeds a threshold, a fast backoff mechanism is immediately triggered, switching to a traditional AMC mode based on real-time channel quality indicators. The calculation method for the aforementioned dynamic safety margin Δ is based on the historical prediction error statistical distribution required to achieve the target reliability, obtained through virtual testing in the digital twin layer.

[0172] This application addresses predictive beam management. The physical layer execution steps specifically include: after predicting the user's angle of arrival, prioritizing the use of a wide beam for initial transmission to ensure basic connectivity; performing a rapid and detailed scan within the wide beam coverage sector to determine the optimal narrow beam; switching to the narrow beam only when the deviation between the measured narrow beam and the predicted direction is less than a switching threshold and the gain increase exceeds a threshold; otherwise, maintaining wide beam transmission.

[0173] As can be seen, the above mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, the embodiments of this application provide corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the modules and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0174] In some embodiments, this application also provides a data transmission apparatus. The data transmission apparatus may include one or more functional modules for implementing the data transmission method of the above method embodiments.

[0175] For example, Figure 8 This is a schematic diagram illustrating the composition of a data transmission device provided in an embodiment of this application. Figure 8 As shown, the data transmission device 800 includes: The acquisition module 801 is used to acquire historical channel state data of the wireless channel, wherein the historical channel state data is used to represent the channel characteristics of the wireless channel.

[0176] The prediction module 802 is used to call the lightweight channel prediction model to predict the historical channel state data and obtain the predicted signal-to-noise ratio data, wherein the predicted signal-to-noise ratio data is used to represent the channel quality of the wireless channel at future times.

[0177] The adjustment module 803 is used to adjust the predicted signal-to-noise ratio data based on the signal-to-noise ratio offset to obtain the adjusted predicted signal-to-noise ratio data; wherein the adjusted predicted signal-to-noise ratio data is less than the original predicted signal-to-noise ratio data.

[0178] The determination module 804 is used to determine the first modulation and coding strategy based on the adjusted predicted signal-to-noise ratio data.

[0179] The transmission module 805 is used to transmit data based on the first modulation and coding strategy.

[0180] In some embodiments, the signal-to-noise ratio offset is a control parameter corresponding to a transmission strategy that satisfies reliability constraints through multiple simulation scenarios in a virtual environment.

[0181] In other embodiments, the adjustment module 803 includes a processing unit for determining the adjusted predicted signal-to-noise ratio data by summing the signal-to-noise ratio offset and the predicted signal-to-noise ratio data.

[0182] In some other embodiments, the apparatus is further configured to: determine the block error rate of data during transmission based on acknowledgment information during data transmission; in response to the block error rate being greater than a block error rate threshold, determine a second modulation and coding strategy based on channel quality indication information in the next transmission time interval of the current transmission time interval; wherein the channel quality indication information is used to indicate the channel quality of the wireless channel at the current moment; and transmit data based on the second modulation and coding strategy.

[0183] In some other embodiments, the apparatus is also configured to: continue transmitting data based on a first modulation and coding strategy in response to a block error rate less than or equal to a block error rate threshold.

[0184] In some other embodiments, the device is further configured to: acquire historical angle-of-arrival data of a beam received by a base station, wherein the historical angle-of-arrival data is used to represent the incident direction information of the beam; call a lightweight angle prediction model to predict the historical angle-of-arrival data to obtain predicted angle-of-arrival data, wherein the predicted angle-of-arrival data is used to represent the incident direction information of the beam at a future time; and transmit wide-beam data based on the predicted angle-of-arrival data.

[0185] In some other embodiments, the apparatus is further configured to: determine angular deviation data between predicted angle of arrival data and target angle of arrival data, wherein the target angle of arrival data is obtained by beam scanning of the coverage area of ​​the wide beam; and transmit data based on the narrow beam in response to the angular deviation data being less than an angle threshold and the gain of the narrow beam nested within the wide beam being greater than a gain threshold.

[0186] In some other embodiments, the device is also configured to: continue transmitting data based on a wide beam in response to an angle deviation data being greater than or equal to an angle threshold, or a narrow beam gain being less than or equal to a gain threshold.

[0187] In the case of implementing the functions of the integrated modules described above in hardware, this embodiment of the invention provides a possible structural schematic diagram of the electronic device involved in the above embodiments. For example... Figure 9 As shown, the electronic device 900 includes: a processor 902, a communication interface 903, and a bus 904. Optionally, the electronic device 900 may also include a memory 901.

[0188] Processor 902 may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 902 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 902 may also be a combination that implements computing functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0189] The communication interface 903 is used to connect to other devices via a communication network. This communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0190] The memory 901 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0191] In one possible implementation, the memory 901 can exist independently of the processor 902. The memory 901 can be connected to the processor 902 via a bus 904 and is used to store instructions or program code. When the processor 902 calls and executes the instructions or program code stored in the memory 901, it can implement the data transmission method provided in this embodiment of the invention.

[0192] In another possible implementation, the memory 901 can also be integrated with the processor 902.

[0193] The 904 bus can be an extended industry standard architecture (EISA) bus, etc. The 904 bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0194] Through the above description of the implementation methods, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the service calling device can be divided into different functional modules to complete all or part of the functions described above.

[0195] This application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be executed by computer instructions instructing related hardware. The program can be stored in the aforementioned computer-readable storage medium, and when executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be any of the foregoing embodiments or memory. The aforementioned computer-readable storage medium can also be an external storage device of the aforementioned service invocation device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the aforementioned service invocation device. Further, the aforementioned computer-readable storage medium can include both internal storage units of the aforementioned service invocation device and external storage devices. The aforementioned computer-readable storage medium is used to store the aforementioned computer program and other programs and data required by the aforementioned service invocation device. The aforementioned computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0196] This application also provides a computer program product, which includes a computer program that, when run on a computer, causes the computer to perform any of the data transmission methods provided in the above embodiments.

[0197] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A data transmission method, characterized in that, include: Acquire historical channel state data of a wireless channel, wherein the historical channel state data is used to represent the channel characteristics of the wireless channel; A lightweight channel prediction model is invoked to predict the historical channel state data to obtain predicted signal-to-noise ratio (SNR) data, wherein the predicted SNR data is used to represent the channel quality of the wireless channel at future times; Based on the signal-to-noise ratio offset, the predicted signal-to-noise ratio data is adjusted to obtain the adjusted predicted signal-to-noise ratio data; wherein the adjusted predicted signal-to-noise ratio data is lower than the original predicted signal-to-noise ratio data. Based on the adjusted predicted signal-to-noise ratio data, a first modulation and coding strategy is determined; Data is transmitted based on the first modulation and coding strategy.

2. The method according to claim 1, characterized in that, The signal-to-noise ratio offset is the control parameter corresponding to the transmission strategy that meets reliability constraints through multiple simulation scenarios in a virtual environment.

3. The method according to claim 1, characterized in that, The step of adjusting the predicted signal-to-noise ratio (SNR) data based on the SNR offset to obtain adjusted predicted SNR data includes: The sum of the signal-to-noise ratio offset and the predicted signal-to-noise ratio data is determined as the adjusted predicted signal-to-noise ratio data.

4. The method according to claim 1, characterized in that, The method further includes: Based on the confirmation information during data transmission, the data error rate during transmission is determined; In response to the block error rate being greater than a block error rate threshold, a second modulation and coding strategy is determined in the next transmission time interval of the current transmission time interval based on channel quality indication information; wherein, the channel quality indication information is used to indicate the channel quality of the wireless channel at the current moment; Data is transmitted based on the second modulation and coding strategy.

5. The method according to claim 4, characterized in that, The method further includes: In response to the block error rate being less than or equal to the block error rate threshold, data transmission continues based on the predictive modulation and coding strategy.

6. The method according to claim 1, characterized in that, The method further includes: The historical angle of arrival data of the beam received by the base station is obtained, wherein the historical angle of arrival data is used to represent the incident direction information of the beam; The lightweight angle prediction model is invoked to predict the historical angle of arrival data to obtain predicted angle of arrival data, wherein the predicted angle of arrival data is used to represent the incident direction information of the beam at a future time. Wide-beam transmission data based on the predicted angle of arrival data.

7. The method according to claim 6, characterized in that, The method further includes: Determine the angular deviation data between the predicted angle of arrival data and the target angle of arrival data, wherein the target angle of arrival data is obtained by beam scanning the coverage area of ​​the wide beam; In response to the angle deviation data being less than an angle threshold and the gain of the narrow beam nested within the wide beam being greater than a gain threshold, data is transmitted based on the narrow beam.

8. The method according to claim 7, characterized in that, The method further includes: In response to the angle deviation data being greater than or equal to the angle threshold, and the gain of the narrow beam being less than or equal to the gain threshold, data transmission continues based on the wide beam.

9. A data transmission device, characterized in that, include: An acquisition module is used to acquire historical channel state data of a wireless channel, wherein the historical channel state data is used to represent the channel characteristics of the wireless channel; The prediction module is used to call a lightweight channel prediction model to predict the historical channel state data and obtain predicted signal-to-noise ratio data, wherein the predicted signal-to-noise ratio data is used to represent the channel quality of the wireless channel at a future time. An adjustment module is used to adjust the predicted signal-to-noise ratio data based on the signal-to-noise ratio offset to obtain adjusted predicted signal-to-noise ratio data; wherein the adjusted predicted signal-to-noise ratio data is lower than the original predicted signal-to-noise ratio data. The determining module is used to determine a first modulation and coding strategy based on the adjusted predicted signal-to-noise ratio data; The transmission module is used to transmit data based on the first modulation and coding strategy.

10. An electronic device, characterized in that, It includes a processor and a memory, the processor being coupled to the memory; the memory is used to store computer instructions, which are loaded and executed by the processor to enable the computer device to implement the data transmission method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes computer-executable instructions that, when executed on a computer, cause the computer to perform the data transmission method according to any one of claims 1 to 8.

Images