Low-delay adaptive audio transmission method, system and device and storage medium

Abstract

The invention discloses a low-delay self-adaptive audio transmission method, system and device and a storage medium, and relates to the technical field of audio, and the method comprises the steps: dynamically recognizing an optimal transmission path in wireless dual bands based on an intelligent band detection and self-adaptive channel selection mechanism; decomposing the target audio data through a hierarchical coding algorithm, and executing self-adaptive adjustment of a differential transmission strategy and a quantization parameter; the dual-band performance is monitored in real time based on a load balancing algorithm, the decomposed transmission tasks are dynamically allocated, and path switching is triggered when it is monitored that the transmission quality of the current band is abnormal; the future buffer requirement is predicted based on a predictive buffer algorithm, the size of a buffer area is dynamically adjusted according to the actual playing condition and the network state change so as to ensure the playing continuity and minimize the delay, the technical bottleneck of traditional single-frequency wireless audio transmission is broken through, an effective frequency band coordination mechanism is realized, resource waste or transmission conflicts are avoided, and the transmission efficiency is improved. And the double-frequency advantage is fully exerted.

Description

Technical Field

[0001] The present invention relates to the field of audio technology, and in particular to a low-delay adaptive audio transmission method, system, device and storage medium. Background Art

[0002] Existing wireless audio transmission technologies are primarily based on Bluetooth and WiFi protocols. Traditional Bluetooth audio transmission operates in a single-band mode, using frequency hopping within the 2.4 GHz ISM band to avoid interference. However, due to protocol overhead and codec latency, end-to-end latency is typically over 200ms. While WiFi audio transmission offers higher bandwidth, it consumes significant power and suffers from instability in mobile scenarios. The recent emergence of low-latency audio codecs such as aptXLL and LDAC has improved latency to some extent, but still cannot meet the stringent requirements of professional audio applications. The introduction of dual-band technology offers a new approach to addressing this issue, but existing implementations lack an effective frequency band coordination mechanism, often resulting in wasted resources or transmission conflicts, and failing to fully leverage the advantages of dual-band technology. Summary of the Invention

[0003] The present invention provides a low-latency adaptive audio transmission method, system, device and storage medium, which breaks through the technical bottleneck of traditional single-frequency wireless audio transmission, realizes an effective frequency band coordination mechanism, avoids resource waste or transmission conflicts, and gives full play to the advantages of dual-frequency.

[0004] A first aspect of the present invention provides a low-delay adaptive audio transmission method, comprising: Dynamically identify the optimal transmission path in wireless dual-band based on intelligent frequency band detection and adaptive channel selection mechanism; Decompose the target audio data through a layered coding algorithm, implement differentiated transmission strategies and adaptive adjustment of quantization parameters; Based on the load balancing algorithm, the system monitors the performance of dual frequency bands in real time, dynamically allocates the decomposed transmission tasks, and triggers path switching when abnormal transmission quality of the current frequency band is detected. Based on the predictive buffering algorithm, future buffering needs are predicted, and the buffer size is dynamically adjusted according to the actual playback situation and network status changes to ensure playback continuity while minimizing latency.

[0005] Preferably, the intelligent frequency band detection includes: Perform hardware initialization on the built-in dual-band wireless module; Calculate the available wireless channels in the current environment through dynamic scanning and set operations; The dual-band frequency band is 2.4 GHz and 5 GHz.

[0006] Preferably, the adaptive channel selection mechanism dynamically identifies the optimal transmission path in the wireless dual-band, including: Performing quality assessment on the calculated available wireless channels based on a comprehensive channel quality assessment algorithm; The comprehensive channel quality assessment algorithm calculates the comprehensive score by weighting the original multi-dimensional indicators through a formula, giving priority to wireless channels with high scores. The original multi-dimensional indicators include signal strength, signal-to-noise ratio, packet loss rate and jitter. The weight coefficients in the weighted formula are adaptively adjusted according to the application scenario and historical performance feedback.

[0007] Preferably, the target audio data is decomposed by a layered coding algorithm and a differentiated transmission strategy is implemented, including: Decomposing the original audio data into a base layer and an enhancement layer; When encoding the base layer, the original audio is downsampled and the downsampled signal is subjected to discrete cosine transform; When encoding the enhancement layer, the base layer encoded data is inversely transformed and upsampled to reconstruct the base layer signal. The difference signal obtained by subtracting the reconstructed base layer signal from the original audio signal is the enhancement layer data; Based on the real-time monitored network conditions, the base layer data packets are transmitted first to ensure basic sound quality.

[0008] Preferably, the adaptive adjustment of the quantization parameter includes: Monitor network conditions in real time to obtain current network bandwidth, required bandwidth, actual delay, and target delay; Dynamically calculate the quantization step size based on a weighted formula of bandwidth and delay; The calculated quantization step size is passed to the current audio encoder for dynamic adjustment to maximize audio quality while ensuring latency requirements.

[0009] Preferably, the real-time monitoring of the dual-band performance based on the load balancing algorithm and the dynamic allocation of the decomposed transmission tasks include: Real-time collection of frequency band quality parameters to calculate frequency band quality score Q value and load distribution weight; Compare the allocation weights of the two bands; If the allocation weight of the 2.4 GHz band is greater than the allocation weight of the 5 GHz band, the base layer audio data is allocated to the 2.4 GHz band for transmission, and the enhancement layer audio data is allocated to the 5 GHz band for transmission; If the allocation weight of the 5 GHz band is greater than or equal to the allocation weight of the 2.4 GHz band, the base layer audio data is allocated to the 5 GHz band for transmission, and the enhancement layer audio data is allocated to the 2.4 GHz band for transmission.

[0010] Preferably, triggering path switching when abnormal transmission quality of the current frequency band is monitored includes: Continuously monitor the transmission quality of the current frequency band, including packet loss rate, latency, and signal strength; The trigger is controlled according to the preset switching decision algorithm, and another frequency band is selected as the new transmission path after the trigger; The quality score Q value of the new frequency band is recalculated, and the load balancing algorithm is triggered to optimize the allocation.

[0011] Preferably, the predicting of future buffering requirements based on a predictive buffering algorithm includes: Obtain historical data and changing trends of network latency; Based on the buffer size prediction model, the delay at the next moment is predicted by balancing the current observation value with the historical prediction value through exponential smoothing method; The target buffer size is calculated based on the predicted delay, audio bitrate, and the set safety margin; The buffer size at the receiving end is adjusted according to the target buffer size to achieve smooth playback under network fluctuations.

[0012] Preferably, dynamically adjusting the buffer size according to actual playback conditions and network status changes to ensure playback continuity while minimizing delay includes: Monitor the current buffer level in real time and compare it with the preset threshold; If the buffer level is less than the minimum threshold, the buffer increase strategy is triggered to expand the target buffer; If the buffer level is greater than the maximum threshold, the buffer reduction strategy is triggered to reduce the target buffer; If the buffer level has not changed, the current buffer size is maintained; The buffer size after triggering the adjustment is adjusted to the target size through a smooth transition mechanism.

[0013] Preferably, the method further includes: combining Reed-Solomon error correction coding with a selective retransmission strategy based on a hybrid error recovery mechanism to repair abnormal audio data packets in stages.

[0014] Preferably, the hybrid error recovery mechanism combines Reed-Solomon error correction coding with a selective retransmission strategy to repair abnormal audio data packets in stages, including: Forward error correction is achieved by adding Reed-Solomon error correction code to the audio data packet, automatically recovering from some transmission errors; For lost packets that cannot be recovered by FEC, selective fast retransmission is performed based on a priority algorithm based on packet importance, age, and number of retransmissions, giving priority to recovering critical data.

[0015] Preferably, it also includes: ensuring continuous high-performance transmission by monitoring audio transmission performance in real time and using machine learning to predict optimal parameters.

[0016] A second aspect of the present invention provides a low-delay adaptive audio transmission system, comprising: Frequency band detection module, used to dynamically identify the optimal transmission path in wireless dual-band based on intelligent frequency band detection and adaptive channel selection mechanism; A layered coding module is used to decompose the target audio data through a layered coding algorithm, and to implement differentiated transmission strategies and adaptive adjustment of quantization parameters; The load balancing module monitors the performance of the dual-bands in real time based on a load balancing algorithm, dynamically allocates the decomposed transmission tasks, and triggers path switching when abnormal transmission quality is detected in the current frequency band. The predictive buffering module is used to predict future buffering needs based on the predictive buffering algorithm, and dynamically adjust the buffer size according to the actual playback situation and network status changes to ensure playback continuity while minimizing latency.

[0017] A third aspect of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the above-mentioned low-delay adaptive audio transmission method when executed by the processor.

[0018] A fourth aspect of the present invention provides a computer-readable storage medium, which stores instructions. When the computer-readable storage medium is run on a computer, it enables the computer to execute the above-mentioned low-delay adaptive audio transmission method.

[0019] This invention proposes a low-latency adaptive audio transmission method. Through a dual-band collaborative transmission architecture, it overcomes the technical bottlenecks of traditional single-band wireless audio transmission, achieving multiple key technological breakthroughs and significant performance improvements. First, the system employs intelligent frequency band detection and adaptive channel selection to dynamically identify the optimal transmission path in both the 2.4GHz and 5GHz bands, effectively mitigating electromagnetic interference and improving transmission stability by over 40% compared to traditional solutions. Second, layered coding technology scientifically decomposes audio data into a base layer and an enhancement layer. Through dual-band load-balanced transmission, this method ensures basic sound quality reliability while fully utilizing the high bandwidth of the 5GHz band, improving overall spectrum utilization by 35%. A core predictive buffer management algorithm dynamically adjusts the buffer size based on real-time network latency monitoring and trend prediction, ensuring playback continuity while keeping end-to-end latency below 50 milliseconds, meeting the stringent requirements of professional-grade audio production. The system's integrated Reed-Solomon forward error correction (FEC) coding and intelligent retransmission mechanism provide a robust error recovery system, maintaining smooth playback even in environments with 6% packet loss and reducing audio interruption rates to below 0.1%. Furthermore, the machine learning optimization module, through in-depth analysis of historical transmission data, can predict and automatically adjust transmission parameters, enabling the system to self-learn and continuously optimize. Overall, this technical solution not only addresses the core issues of high latency and poor stability in wireless audio transmission, but also provides a revolutionary solution for extremely latency-sensitive scenarios in the 5G era, such as real-time audio and video applications, professional audio production, online gaming, and virtual reality. It holds broad industrial application prospects and significant technical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Schematic diagram of the process of low-latency adaptive audio transmission method in this embodiment; Figure 2 FIG. 1 is a schematic diagram of an electronic device in this embodiment. DETAILED DESCRIPTION

[0021] The present invention provides a low-latency adaptive audio transmission method, system, device and storage medium, which effectively solves the problem of audio input player reopening capture when switching modes through persistent audio capture session and dynamic audio processing pipeline.

[0022] The terms "first," "second," "third," "fourth," and so on (if any) in the description and claims of the present invention and in the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that shown or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus that includes a series of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, product, or apparatus.

[0023] Example 1 For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the present invention provides a low-delay adaptive audio transmission method, comprising the following steps: S1: Dynamically identifies the optimal transmission path in wireless dual-bands based on intelligent frequency band detection and adaptive channel selection mechanisms; S2: Decompose the target audio data through a layered coding algorithm, perform differentiated transmission strategies and adaptive adjustment of quantization parameters; S3: Based on a load balancing algorithm, it monitors the performance of the dual-bands in real time, dynamically allocates the decomposed transmission tasks, and triggers path switching when it detects abnormal transmission quality in the current band. S4: Based on the predictive buffering algorithm, it predicts future buffering needs and dynamically adjusts the buffer size according to the actual playback situation and network status changes to ensure playback continuity while minimizing latency.

[0024] In this embodiment, through intelligent frequency band detection and adaptive channel selection, the optimal transmission path is dynamically identified, and the frequency band or channel with the best signal and the least interference can be selected for transmission according to the current wireless environment such as interference and signal strength, avoiding problems that may be encountered when using a certain frequency band, such as 2.4G is susceptible to interference and 5G has slightly poor wall penetration, thereby significantly improving the stability and reliability of audio transmission and reducing interruptions or freezes; through layered coding and differentiated transmission, the audio data is processed in layers, and the quantization parameters are adaptively adjusted according to the network conditions, so that the system can transmit high-quality audio when the network is good, and give priority to ensuring the intelligibility of the core audio content when the network is poor, that is, sacrificing some quality to avoid complete interruption. This differentiated strategy not only guarantees the basic experience, but also improves the overall Perceptual quality achieves a "lossy but usable" balance; through load balancing and dynamic task allocation / path switching, dual-band performance is monitored in real time and tasks are dynamically allocated to ensure that both bands are fully utilized and not overloaded. When a band experiences quality anomalies, such as a surge in packet loss or latency, a switch is quickly triggered, shifting transmission tasks to a more reliable band. This greatly enhances the system's ability to cope with sudden network issues and improves overall robustness. A predictive buffering algorithm dynamically adjusts the buffer size based on predicted future buffering needs and actual playback and network status, minimizing unnecessary latency while ensuring playback continuity (avoiding lag). Compared to fixed-size buffers, this approach can better adapt to network fluctuations and provide a more natural real-time experience. Through intelligent selection, dynamic adjustment, and predictive optimization, this method ensures more rational and efficient utilization of wireless bandwidth, processing power, and buffering resources, avoiding waste or insufficiency of resources.

[0025] In one embodiment, the smart frequency band detection in step S1 includes: Perform hardware initialization on the built-in dual-band wireless module; Calculate the available wireless channels in the current environment through dynamic scanning and set operations; The dual-band frequency band is 2.4 GHz and 5 GHz.

[0026] When the system boots up, it first executes a hardware initialization routine. This routine begins by initializing and configuring the dual-band wireless module. This routine communicates with the system's two built-in wireless modules, one operating in the 2.4 GHz band and the other in the 5 GHz band. The initialization process includes setting the module's operating mode (e.g., 802.11b / g / n or 802.11a / n / ac), configuring basic network parameters (such as AP, Station, or Ad-Hoc operation), and enabling the scanning function. The 2.4 GHz module operates under the 802.11b / g / n standard and supports channels 1-14; the 5 GHz module operates under the 802.11a / n / ac standard and supports channels 36-165. After hardware initialization is complete, the system instructs both wireless modules to perform a channel scan. This typically occurs by briefly staying on each channel and listening for other wireless signals (such as other Wi-Fi networks, Bluetooth devices, microwave ovens, etc.). During the scan, the modules collect information about each channel, specifically which channels have strong signals (indicating they are occupied). Based on the collected information, the system calculates the final list of available channels using the formula C_available = C_total - C_occupied - C_restricted. C_total = {2.4GHz: 1-14, 5GHz: 36-165} is known and represents the channel range determined during hardware initialization. C_occupied = currently occupied channels, and C_restricted = channels restricted by regulations. This calculation is essentially a set operation. From the total set of theoretically available channels (C_total), the number of occupied channels (C_occupied) and the number of channels prohibited by regulations or policies (C_restricted) are subtracted. The remaining number of channels truly available for data transmission (C_available) remains. This calculation is not a one-time operation. Although performed once during initialization, C_occupied may change during system operation (as new devices join or leave). Therefore, this detection and calculation process may need to be repeated periodically to maintain an accurate understanding of the current environment. The resulting C_available list serves as the basis for all subsequent channel selection decisions. The system will only select channels from C_available for data transmission, thereby minimizing interference from other wireless devices and achieving adaptability.

[0027] For example, in an office environment, 2.4GHz channels 1, 6, and 11 are occupied by WiFi routers, while 5GHz channels 36 and 40 are occupied by other devices. The system calculates the available channels as {2, 3, 4, 5, 7, 8, 9, 10, 12, 13, 14} for 2.4GHz and {44, 48, 52, 56, 60, 64, 100, 104, 108, 112, 116, 120, 124, 128, 132, 136, 140, 144, 149, 153, 157, 161, 165} for 5GHz. Accurate channel detection avoids interference with existing devices, paving the way for subsequent adaptive transmission. Each channel's available bandwidth is more efficiently utilized because no additional energy is expended to combat interference or retransmit packets lost due to interference.

[0028] After identifying available channels, simply knowing which channels are available is not enough. Different available channels have different current communication qualities, and further quantitative evaluation of each available channel is required to determine which channel is the optimal choice for transmitting audio data in the current environment. The adaptive channel selection mechanism in step S1 dynamically identifies the optimal transmission path in the wireless dual-band, including: Performing quality assessment on the calculated available wireless channels based on a comprehensive channel quality assessment algorithm; The comprehensive channel quality assessment algorithm calculates a comprehensive score using a weighted formula based on multiple raw metrics, including signal strength, signal-to-noise ratio (SNR), packet loss rate (PLR), and jitter, prioritizing wireless channels with high scores. Each available channel is evaluated for quality, and a channel quality database is established. The evaluation metrics include signal strength (RSSI), signal-to-noise ratio (SNR), packet loss rate (PLR), and delay jitter (Jitter). The weight coefficients in the weighted formula are adaptively adjusted based on the application scenario and historical performance feedback.

[0029] The channel quality comprehensive scoring algorithm calculates a composite score, Q_channel, by combining the four raw metrics using a weighted formula: Q_channel = w1 × RSSI_norm + w2 × SNR_norm + w3 × (1-PLR) + w4 × (1-Jitter_norm). Here, w1 + w2 + w3 + w4 = 1 (weight coefficient), RSSI_norm = (RSSI + 100) / 100 (normalized RSSI value), SNR_norm = SNR / 40 (normalized SNR value), and Jitter_norm = Jitter / 100 (normalized jitter value). The normalized values ​​are multiplied by their respective weight coefficients, w1 to w4, and then added together. This algorithm combines multi-dimensional, dimensional metrics into a single comprehensive score, making it easy to compare the pros and cons of different channels and providing a quantitative basis for selecting the optimal channel.

[0030] For example, in the 5 GHz band, channel 44 is evaluated, with an RSSI of -45 dBm, an SNR of 25 dB, a PLR of 0.02, and a jitter of 5 ms. Assuming weights w1 = 0.3, w2 = 0.4, w3 = 0.2, and w4 = 0.1, we obtain: RSSI_norm = (-45 + 100) / 100 = 0.55, SNR_norm = 25 / 40 = 0.625, and Q_channel = 0.3 × 0.55 + 0.4 × 0.625 + 0.2 × (1 - 0.02) + 0.1 × (1 - 0.05) = 0.708. This quantitative channel quality assessment provides a scientific basis for intelligent band selection, improving transmission efficiency.

[0031] In this embodiment, the method for determining the weight coefficient of the comprehensive channel quality assessment algorithm includes the following steps: (1) Basic weight setting: The basic weights of each indicator are determined based on audio transmission requirements: w1 (RSSI weight): 0.2-0.4, to ensure basic connection stability; w2 (SNR weight): 0.3-0.5, to determine the upper limit of transmission quality; w3 (packet loss rate weight): 0.15-0.3, affecting audio continuity; w4 (jitter weight): 0.05-0.2, affecting the real-time experience.

[0032] (2) Scene adaptive adjustment: Real-time conversation scenario: w1=0.25, w2=0.35, w3=0.25, w4=0.15; Music playback scenario: w1=0.3, w2=0.45, w3=0.2, w4=0.05; Professional application scenario: w1=0.35, w2=0.4, w3=0.2, w4=0.05.

[0033] (3) Dynamic optimization mechanism: The weights are adaptively adjusted using performance feedback: w_i(t+1)=w_i(t)+η×∂SPI / ∂w_i, where η is the learning rate, SPI is the system performance indicator, and the constraint condition ∑wi=1.

[0034] During implementation: Initial settings: w1=0.3, w2=0.4, w3=0.2, w4=0.1; Performance monitoring: Collect SPI data for 100 transmission cycles; Sensitivity analysis: Adjust each weight by ±0.05 and calculate the SPI change rate; Weight update: w_i(new)=w_i(old)+0.02×∂SPI / ∂w_i; Normalization: Ensure that the sum of the weights is 1 and wi ≥ 0.05.

[0035] For example, after optimization, the weights of a voice call scenario are adjusted to w1=0.28, w2=0.38, w3=0.24, and w4=0.10, and the system SPI is improved from 0.82 to 0.87. The specific typical scenario configuration table is as follows in Table 1: Table 1 Typical scenario configuration table In order to process the audio data before transmission so that it is more suitable for low-latency and high-reliability transmission in a dynamically changing wireless environment, the target audio data is decomposed by a layered coding algorithm in step S2, and a differentiated transmission strategy is implemented, including: The original audio data is decomposed into a base layer and an enhancement layer, that is, the input high-fidelity audio signal is decomposed into a base layer and an enhancement layer. The base layer contains basic audio information to ensure basic audibility; the enhancement layer contains detailed information to improve sound quality.

[0036] When encoding the base layer, the original audio is downsampled and the downsampled signal is subjected to discrete cosine transform. downsample(audio_signal,factor=2): downsamples the original audio signal and reduces the sampling rate (for example, from 48kHz to 24kHz), directly reducing the amount of data and filtering out some high-frequency information. DCT(audio_downsampled): performs discrete cosine transform (DCT) on the downsampled signal to convert the time domain signal to the frequency domain. The energy is often concentrated in a few coefficients, which facilitates subsequent compression.

[0037] When encoding the enhancement layer, the base layer coded data is inversely transformed and upsampled to reconstruct the base layer signal. The difference signal obtained by subtracting the reconstructed base layer signal from the original audio signal is the enhancement layer data. The layered coding algorithm is based on improved subband coding. For base layer coding, the following code is used: upsample(IDCT(B_layer), factor=2): The base layer coded data (B_layer, i.e., DCT coefficients) is subjected to an inverse DCT (IDCT) to reconstruct the signal, which is then upsampled to the same sampling rate as the original signal. For enhancement layer coding, the following code is used: audio_signal - upsampled_reconstructed_signal: The original audio signal is subtracted from the reconstructed signal from the base layer to obtain the difference between the two. This difference signal contains high-frequency details and information in the original signal that were not fully expressed in the base layer, and is therefore the enhancement layer. DCT = Discrete Cosine Transform, IDCT = Inverse Discrete Cosine Transform, downsample = downsampling function, and upsample = upsampling function. Based on real-time network conditions, base layer data packets are prioritized for transmission to ensure basic sound quality. This method leverages the human ear's auditory characteristics (high-frequency details are less sensitive than low-frequency details) and signal processing techniques (downsampling and transform coding). Downsampling directly reduces data volume, while the DCT method improves compression efficiency. The enhancement layer, acting as a "residual," contains key information for improving sound quality, while the base layer ensures a minimum quality baseline.

[0038] For example, 48kHz / 16-bit stereo audio is layered and encoded. The base layer is downsampled by 2 times to obtain audio with a sampling rate of 24kHz, and 512 coefficients are obtained after DCT transformation. The enhancement layer is the difference between the original signal and the reconstructed signal of the base layer, containing high-frequency detail information. The base layer data volume is about 30% of the original data, and the enhancement layer is 70%. Layered coding allows important basic audio information to be transmitted first, ensuring basic sound quality even under poor network conditions. The layered structure enables the system to flexibly adjust the transmission strategy according to the available bandwidth, which is an effective mechanism to resist packet loss and bandwidth fluctuations.

[0039] Even with layered coding, audio data still needs to be further compressed to accommodate the bandwidth limitations of wireless channels and to control encoding / decoding delays. Quantization is a key step in compression, reducing the amount of data by reducing the number of bits required to represent signal samples. However, this introduces quantization errors, which can affect sound quality. Adaptive quantization means that quantization parameters (such as the quantization step size) are not fixed but are dynamically adjusted based on current network conditions (available bandwidth, actual delay) to maximize sound quality while ensuring the target delay. Therefore, through adaptive adjustment of the quantization parameters, the specific implementation steps include: Real-time monitoring of network conditions to obtain current network bandwidth, required bandwidth, actual delay, and target delay. This allows for dynamic adjustment of quantization parameters and compression ratios based on current network bandwidth and delay requirements. The quantization step size is dynamically calculated based on a weighted formula of bandwidth and delay; the calculated quantization step size is passed to the current audio encoder for dynamic adjustment to maximize audio quality while ensuring delay requirements.

[0040] The calculation formula for the quantization step size is as follows: Q_step=Q_base×(1+α×BW_ratio+β×Delay_ratio), Among them, Q_base = base quantization step, BW_ratio = (BW_available - BW_required) / BW_required, Delay_ratio = (Delay_current - Delay_target) / Delay_target. BW_ratio: measures the surplus degree of the current available bandwidth relative to the required bandwidth. If BW_available > BW_required, then BW_ratio is positive, indicating that there is bandwidth surplus; if BW_available < BW_required, then BW_ratio is negative, indicating that the bandwidth is tight. Delay_ratio: measures the exceeding degree of the current actual delay relative to the target delay. If Delay_current > Delay_target, then Delay_ratio is positive, indicating that the delay exceeds the standard; if Delay_current < Delay_target, then Delay_ratio is negative, indicating that the delay is within the control range. α, β: adjustment coefficients, used to control the influence degree of bandwidth surplus and delay exceeding the standard on the quantization step adjustment. The larger α is, the greater the amplitude of increasing the quantization step (reducing the audio quality to reduce the data volume) when there is bandwidth surplus; the larger β is, the greater the amplitude of reducing the quantization step (improving the audio quality to reduce retransmission, thereby possibly reducing the delay) when the delay exceeds the standard. The principle of the formula is as follows: based on the base quantization step Q_base, dynamic adjustment is performed according to the bandwidth and delay conditions; when there is bandwidth surplus (BW_ratio > 0) and the delay meets the standard (Delay_ratio <= 0), Q_step can be appropriately increased (coarser quantization) to reduce the data volume and improve the spectrum utilization rate; when the bandwidth is tight (BW_ratio < 0) or the delay exceeds the standard (Delay_ratio > 0), Q_step needs to be reduced (finer quantization) to increase data redundancy or improve the coding quality, ensure basic transmission and reduce the delay. The adaptive adjustment of the quantization parameters is realized, taking into account the two key constraint factors of network bandwidth and delay, making the quantization strategy no longer static but able to dynamically respond to network changes. Its core idea is that when the network condition is good, a little sacrifice in audio quality can be made in exchange for higher transmission efficiency; when the network condition is poor or the delay pressure is high, the audio quality and real-time performance are given priority.

[0041] For example, with a target latency of 50ms, a current latency of 65ms, available bandwidth of 1.2Mbps, and a required bandwidth of 1.0Mbps, let Q_base = 0.8, α = 0.3, and β = 0.5: BW_ratio = (1.2-1.0) / 1.0 = 0.2, Delay_ratio = (65-50) / 50 = 0.3, and Q_step = 0.8×(1+0.3×0.2+0.5×0.3) = 0.8×1.21 = 0.968. This value is between the base quantization step size of 0.8 and the maximum possible value, indicating that when the system has sufficient bandwidth but exceeds the latency requirement, it moderately increases quantization accuracy—that is, reduces the quantization step size—to prioritize meeting the latency requirement. In this way, adaptive quantization maximizes audio quality while ensuring latency, achieving a balance between quality and real-time performance.

[0042] In one embodiment, in step S3, real-time monitoring of dual-band performance based on a load balancing algorithm and dynamic allocation of decomposed transmission tasks include: Collect frequency band quality parameters in real time to calculate the frequency band quality score Q value and load distribution weight, dynamically allocate the transmission tasks of the base layer and enhancement layer according to the real-time performance of the two frequency bands, and monitor the network performance indicators connected to the 2.4GHz and 5GHz frequency bands in real time, such as: signal strength (RSSI): the higher the value, the better, signal-to-noise ratio (SNR): the higher the value, the better, Transmission rate: The actual measured throughput, the higher the value, the better; Packet loss rate (PLR): The lower the value, the better (may need to be reverse mapped to the Q value); Latency: The lower the value, the better (may need to be reverse mapped to the Q value).

[0043] Compare the allocation weights of the two frequency bands. Input: Q_2.4G and Q_5G. Use the given weight calculation formulas: W_2.4G = Q_2.4G / (Q_2.4G + Q_5G) and W_5G = Q_5G / (Q_2.4G + Q_5G). Calculate the allocation weights W_2.4G and W_5G for the two frequency bands. If the allocation weight of the 2.4 GHz band is greater than that of the 5 GHz band, the base layer audio data is allocated to the 2.4 GHz band for transmission, and the enhancement layer audio data is allocated to the 5 GHz band for transmission. The 2.4 GHz band generally has good penetration and is more stable, making it suitable for transmitting critical base layer data. The 5 GHz band generally has higher bandwidth and is suitable for transmitting enhancement layer data that has higher bandwidth requirements. If the allocation weight of the 5 GHz band is greater than or equal to the allocation weight of the 2.4 GHz band, the base layer audio data is allocated to the 5 GHz band for transmission, and the enhancement layer audio data is allocated to the 2.4 GHz band for transmission.

[0044] For example, assuming Q_2.4G = 0.75 and Q_5G = 0.68, we calculate: W_2.4G = 0.75 / (0.75 + 0.68) = 0.524, W_5G = 0.68 / (0.75 + 0.68) = 0.476. Since W_2.4G > W_5G, the base layer is assigned to the 2.4 GHz band and the enhancement layer to the 5 GHz band. This uses intelligent load balancing to ensure that critical audio data is transmitted over the most reliable path, improving overall transmission stability. For example, the relative stability of the 2.4 GHz band is used to protect the base layer, while the 5 GHz bandwidth is used to transmit the enhancement layer. This is a static allocation strategy that strikes a balance between stability and quality improvement.

[0045] In one embodiment, in step S3, triggering path switching when abnormal transmission quality of the current frequency band is detected includes: Continuously monitor the transmission quality of the current frequency band, including the packet loss rate (PLR_current), delay (Delay_current), and signal strength (RSSI_current), and obtain the PLR_current, Delay_current, and RSSI_current values ​​of the current path. Triggered by the preset switching decision algorithm. After triggering, select another frequency band as the new transmission path. The preset switching decision algorithm: Input the PLR_current, Delay_current, RSSI_current of the current path, and the preset thresholds PLR_threshold, Delay_threshold, RSSI_threshold; Switch trigger condition: Switch_trigger = (PLR_current > PLR_threshold) OR (Delay_current > Delay_threshold) OR (RSSI_current < RSSI_threshold). Output: If any of the conditions is met, Switch_trigger is true, triggering the switching mechanism. For example, in the example, PLR_threshold = 5% is set. When PLR_current = 6% is detected, Switch_trigger becomes true. Once triggered to switch, another frequency band needs to be selected as the new transmission path. Usually, the currently unused frequency band is selected. For example: If the current 2.4GHz is transmitting the base layer and 5GHz is transmitting the enhanced layer, and the 2.4GHz triggers the switch, then the base layer will try to switch to 5GHz and the enhanced layer will try to switch to 2.4GHz. Or more commonly, all data (base layer + enhanced layer) is switched to another frequency band (5GHz). If all current data is transmitted on one frequency band (e.g., all through 5GHz) and 5GHz triggers the switch, then it switches to 2.4GHz.

[0046] Recalculate the quality score Q value of the new frequency band and trigger the load balancing algorithm for optimal allocation. Monitor the transmission quality of the two frequency bands in real time and automatically switch the transmission path when a quality degradation is detected. After performing the path switch (executed after triggering), stop the data transmission on the currently poorer-performing frequency band, reconfigure the wireless interface, and prepare to transmit through the new target frequency band. This may include re-establishing the connection, obtaining an IP address, etc., and start transmitting data again on the new target frequency band. This process takes time, including the detection time (T_detect), decision calculation time (T_decision), and execution time (T_execution). In the example, the total switching delay Switch_delay = 30ms.

[0047] Switching delay calculation: Switch_delay = T_detect + T_decision + T_execution, where T_detect = quality detection time, T_decision = decision calculation time, T_execution = path switch execution time.

[0048] For example, set the PLR ​​threshold to 5%, the latency threshold to 80ms, and the RSSI threshold to -70dBm. When the packet loss rate in the 2.4GHz band reaches 6%, the switching mechanism is triggered. The detection time is 10ms, the decision time is 5ms, the execution time is 15ms, and the total switching delay is 30ms. The fast path switching mechanism ensures that transmission quality is maintained when the network environment changes and avoids audio interruption. By quickly responding to network deterioration and shifting transmission to a "safe zone," severe audio quality degradation or interruption caused by a single frequency band failure can be avoided. A switching delay of 30ms is generally acceptable for audio transmission and does not cause noticeable lag.

[0049] In one embodiment, predicting future buffering requirements based on a predictive buffering algorithm in step S4 includes: To obtain historical data and changing trends of network delay, the system needs to continuously record network delay, such as the round-trip time (RTT) from the sender to the receiver, or changes in the packet arrival interval, and maintain a historical delay data sequence and the current prediction value of the delay at the next moment, Delay_predict(t). The buffer size prediction model uses exponential smoothing to balance current observations with historical predictions to predict the delay at the next moment. The buffer size prediction model: Delay_predict(t+1)=α×Delay(t)+(1-α)×Delay_predict(t) combines the latest actual delay Delay(t) and the previous prediction value Delay_predict(t). The coefficient α (smoothing coefficient) determines the impact of new data on the prediction result. The larger α is, the more dependent the prediction is on the latest data and the faster the response is. The smaller α is, the smoother the prediction is and the stronger the anti-interference ability is.

[0050] The target buffer size is calculated based on the predicted delay, audio bitrate, and the set safety margin. The specific buffer size is calculated as follows: Buffer_size = Delay_predict × Bitrate + Safety_margin, where: Delay_predict is the predicted delay time, Bitrate is the bit rate of the audio stream, α = smoothing coefficient (0<α<1), Safety_margin is the safety margin used to cope with inaccurate predictions or sudden, unpredicted network fluctuations, and finally the recommended buffer size Buffer_size is calculated.

[0051] Adjust the buffer size at the receiving end according to the target buffer size to achieve smooth playback under network fluctuations. Based on the historical data and changing trends of network latency, predict future buffering requirements. For example, the current latency is 60 ms, the predicted latency is 62 ms, the audio bit rate is 320 kbps, and the safety margin is set at 20%: Buffer_size = 62 ms × 320 kbps × 1.2 = 23.8 kb ≈ 24 kb. Predictive buffering avoids playback interruptions caused by network fluctuations and minimizes additional latency. Predictive buffering reduces the risk of playback interruptions due to sudden network fluctuations by preparing in advance, because there is enough data in the buffer to play. At the same time, by only preparing the data of "predicted requirement + safety margin", the problem of high latency caused by setting too large fixed buffer is avoided.

[0052] In one embodiment, the step S4 of dynamically adjusting the buffer size according to the actual playback situation and network state changes to ensure playback continuity while minimizing latency includes: Real-time monitor the current buffering level and compare it with a preset threshold. The system needs to real-time monitor the filling level of the current buffer, such as the percentage Buffer_level of the received data volume in the total buffer capacity, as well as the actual latency and throughput of the network; Low_threshold (low threshold): for example, 40%. If Buffer_level is lower than this value, it means the buffer is about to be empty and there is a risk of interruption. High_threshold (high threshold): for example, 70%. If Buffer_level is higher than this value, it means there is too much data in the buffer, occupying unnecessary resources and increasing latency.

[0053] If the buffering level is less than the lowest threshold, trigger the buffer increase strategy to expand the target buffer. If Buffer_level < Low_threshold: trigger the "buffer increase" strategy; If the buffering level is greater than the highest threshold, trigger the buffer decrease strategy to shrink the target buffer. Else if Buffer_level > High_threshold: trigger the "buffer decrease" strategy; If the buffering level remains unchanged, keep the current buffer size. Else: keep the current buffer size unchanged; The buffer size after the trigger adjustment is adjusted to the target size through a smooth transition mechanism. Based on the judgment result, a new target buffer size, Buffer_target, is calculated. To avoid playback experience issues caused by sudden changes in buffer size (such as sudden delay jumps), a smooth adjustment method is used to gradually adjust the current buffer size, Buffer_current, to the target size, Buffer_target. The calculation formula is as follows: Buffer_new = Buffer_current + (Buffer_target - Buffer_current) × β, where β is the adjustment rate coefficient and Buffer_new is the new buffer size after the next adjustment. Finally, Buffer_new is used as the new Buffer_current for the next adjustment cycle.

[0054] For example, if the current buffer is 24kb and the player's buffer level is 30% (below the lower threshold of 40%), triggering a buffer increase: Buffer_target = 24kb × 1.5 = 36kb. Assuming β = 0.3, then: Buffer_new = 24kb + (36kb - 24kb) × 0.3 = 27.6kb. The dynamic buffer adjustment mechanism ensures playback continuity while minimizing system latency. When insufficient buffer space is detected, the buffer is increased promptly to avoid lags; when the buffer is excessive, it is gradually reduced to reduce latency. Smooth adjustments ensure that these changes are not too drastic, thereby minimizing system latency while ensuring playback continuity.

[0055] During network transmission, packet loss or corruption is inevitable, especially in wireless environments. To address this, the system needs to be able to detect errors and recover data as much as possible to maximize the integrity and continuity of the audio data. Step S5: Based on a hybrid error recovery mechanism, Reed-Solomon error correction coding is combined with a selective retransmission strategy to repair abnormal audio packets in a hierarchical manner.

[0056] Specifically, in step S5, the Reed-Solomon error correction coding is combined with the selective retransmission strategy based on the hybrid error recovery mechanism to repair the abnormal audio data packet in stages, including: Forward error correction is implemented by adding Reed-Solomon error correction codes to audio data packets, automatically recovering from some transmission errors. Coding parameter selection and initialization: Appropriate RS(n,k) coding parameters are selected based on the audio data packet size, reliability requirements, and acceptable bandwidth overhead. For example, for RS(255,223) in the example, the number of correctable symbols is calculated as t = (nk) / 2, the error correction capability is: Error_correction = t / n × 100%, and the overhead is calculated as: Overhead = (nk) / k × 100%. Data segmentation and encoding: Each audio data packet (containing base layer or enhancement layer data) is segmented into fixed-length symbols. The symbol length must be consistent with the RS encoding library (usually bytes). For each segmented data block (k symbols), the selected RS encoder generates (nk) check symbols. The original k information symbols and the generated (nk) check symbols are combined to form an encoded codeword of length n symbols. Data transmission: The encoded codeword (containing information and redundancy check) is transmitted to the wireless network. Receiving and decoding: The receiver receives the codeword and checks how many of the n received symbols are correct. If the number of erroneous symbols does not exceed t, the RS decoder can automatically correct these errors using the check symbols and recover the original k information symbols. If the number of erroneous symbols exceeds t, the decoder will fail and mark the data packet as "unrecoverable". Error detection: FEC itself also plays a role in error detection. If there are too many errors in the received codeword that cannot be corrected, the receiver knows that there is a problem with this data packet. Through the above steps, even if some data packets are lost or damaged during transmission but the number is within the error correction capability, the receiver can automatically recover the data without the need for the sender to retransmit the request, thereby avoiding the increase in delay and delay fluctuation caused by retransmission. For example, RS(255,223) encoding is selected, which can correct 16 symbols: t = (255-223) / 2 = 16, Error_correction = 16 / 255×100% = 6.27%, Overhead = (255-223) / 223×100% = 14.35%. FEC encoding recovers some transmission errors without retransmission, reducing delay fluctuations.

[0057] Furthermore, for lost packets that FEC cannot recover, selective fast retransmission is performed based on a priority algorithm based on packet importance, age, and retransmission count, prioritizing the recovery of critical data. For lost packets that cannot be recovered through FEC, a selective retransmission strategy is employed. The strategy steps are as follows: Packet loss detection and recording: The receiver detects packet loss by comparing the expected packet sequence number with the actual received packet, or through notification of FEC decoding failure. The lost packet information is recorded, including: whether it is a base layer or enhancement layer (determined by Importance), the time the loss occurred (determined by Age), and whether retransmission has been attempted (determined by Retrans_count). Retransmission priority calculation: For each recorded lost packet, its retransmission priority is calculated according to the formula: Priority = w1 × Importance + w2 × (1-Age) + w3 × (1-Retrans_count). Importance is the packet's importance (base layer > enhancement layer), Age is the normalized packet age, and Retrans_count is the normalized number of retransmissions. Age (how long the packet has been lost) and Retrans_count (how many times it has been retransmitted) are mapped to a range of 0 to 1 to facilitate multiplication with Importance and the weight factor. For example, the newer the Age, the larger the (1-Age) and the higher the priority; the fewer the retransmissions, the larger the (1-Retrans_count) and the higher the priority. Retransmission request generation and sending: A retransmission request queue is maintained, and lost data packets are sorted by the calculated priority. Based on network conditions and policies (for example, sending requests in batches at regular intervals, or sending after reaching a certain threshold), several (or one) lost packets with the highest priority are selected from the head of the queue, and a selective retransmission request (for example, containing the number of the data packet that needs to be retransmitted) is sent to the sender. The sender processes the retransmission request: After receiving the retransmission request, the sender searches for the corresponding data packet (it may need to cache the most recently sent packet) and resends these data packets (FEC encoding is also performed). The receiver processes the retransmitted packets: After receiving the retransmitted data packet, the receiver performs FEC decoding (if applicable) and fills it in the correct position to make up for the previous loss. For example, the importance of the base layer lost packet = 0.9, age = 20ms (normalized to 0.4), number of retransmissions = 1 (normalized to 0.25), weights w1 = 0.5, w2 = 0.3, w3 = 0.2: Priority = 0.5×0.9+0.3×(1-0.4)+0.2×(1-0.25)=0.45+0.18+0.15=0.78. The intelligent retransmission strategy ensures the timely recovery of important data while avoiding congestion caused by excessive retransmissions.By prioritizing the retransmission of packets with high importance (such as the base layer), recent age (recently lost, with a high impact), and low retransmission counts (avoiding infinite retransmissions), the system can more efficiently utilize limited bandwidth for retransmission, prioritizing the recovery of critical data. This ensures excellent audio quality even when FEC coverage is compromised. This also avoids the potential for indiscriminate retransmission of all lost packets, which could lead to further network congestion and latency. Together, these two solutions form a robust error recovery system. FEC provides a foundational layer of automatic error correction, reducing the network's reliance on reliability for round-trip time (RTT). Fast retransmit builds on this foundation by providing intelligent remediation for losses that cannot be automatically recovered, ensuring the timely arrival of critical data. Together, these mechanisms significantly enhance the robustness of wireless audio transmission under harsh network conditions.

[0058] In one embodiment, the low-latency adaptive audio transmission method described in an embodiment of the present invention further includes: S6: ensuring continuous high-performance transmission by real-time monitoring of audio transmission performance and using machine learning to predict optimal parameters. The specific implementation steps are as follows: 1) Data collection: QoS (audio quality score): The decoded audio quality can be evaluated through subjective testing (inviting listeners to score) or objective algorithms (such as PESQ, POLQA). The system needs to calculate this score periodically (such as every second or every few seconds). Efficiency (spectral efficiency): Calculate the ratio of the amount of effective audio data successfully transmitted per unit time to the total bandwidth (or spectrum resources) occupied. This requires statistics on the size and number of transmitted data packets and the channel bandwidth used.

[0059] Stability (transmission stability): It can be measured by monitoring the volatility of key indicators. For example, calculate the variance or standard deviation of latency or packet loss rate. Low variance means more stability. The number of consecutive packet losses or timeouts can also be counted. 2) SPI calculation: The system needs to maintain the current values ​​of these three indicators. When performance needs to be evaluated (for example, regularly or when an optimization request is received), the system reads these three values ​​and calculates them according to the formula SPI=w1*QoS+w2*Efficiency+w3*Stability. The calculated result SPI is stored for subsequent analysis and optimization decisions. 3) Visualization and alarm: The SPI and various sub-indicators can be displayed in real time on the monitoring panel, which makes it easier for operation and maintenance personnel to intuitively understand the system status. Thresholds can be set to trigger alarms when the SPI or a key indicator is lower than / higher than the threshold. For example, the measured QoS = 0.85, Efficiency = 0.78, Stability = 0.92, and the weights are all 1 / 3: SPI = (0.85 + 0.78 + 0.92) / 3 = 0.85. Comprehensive performance monitoring provides quantitative indicators for system tuning to ensure continuous high-quality transmission.

[0060] Furthermore, this embodiment uses historical transmission data to train a neural network model and predict optimal transmission parameters. Data preparation: The system requires a large amount of historical transmission records, including: Input features (X): RSSI_2.4G, RSSI_5G: Real-time detected signal strength (Received Signal Strength Indicator) of the 2.4GHz and 5GHz wireless channels; PLR_2.4G, PLR_5G: Real-time statistics of packet loss rates on the 2.4GHz and 5GHz channels; Delay_2.4G, Delay_5G: Real-time measurements of end-to-end delay on the 2.4GHz and 5GHz channels; Audio_type: The type of audio currently being transmitted (e.g., classical music = 1, pop music = 2, speech = 3, etc.), as different types have different requirements for delay, bitrate, and error resilience. Output labels (Y): The transmission parameters actually used or considered optimal at the time corresponding to these input features: Q_step: The audio coding quantization step used, Buffer_size: The set receiver buffer size, Switch_threshold: The set channel switching threshold, Power_level: The used transmit power level. Data cleaning and labeling: Ensure that the data is accurate and complete. If Y is "considered optimal", it may be necessary to combine the SPI value or other standards at the time to determine it. Model training: Use the collected (X, Y) data pairs to train the neural network, and select a suitable optimization algorithm (such as Adam) and loss function (such as mean square error MSE, which measures the gap between the predicted parameters and the actual optimal parameters). During the training process, adjust the network weights so that the model can accurately predict the output Y from the input X. Cross-validation is required to ensure that the model has strong generalization ability and does not overfit the training data. Model deployment and use: The trained neural network model is deployed in the transmission system's control module. Real-time prediction: During system operation, the control module obtains the current input features X = [RSSI_2.4G, RSSI_5G, PLR_2.4G, PLR_5G, Delay_2.4G, Delay_5G, Audio_type] in real time. X is input into the neural network model to obtain the predicted output Y = [Q_step, Buffer_size, Switch_threshold, Power_level]. Based on the model's output, the system dynamically adjusts parameters such as the audio encoder, buffer management, channel selection strategy, and transmit power. During operation, the system continuously collects new (X,Y) data pairs (Y can be a parameter of the actual application, and its effectiveness is evaluated in conjunction with SPI feedback). The model is regularly fine-tuned or retrained using this new data to adapt to changes in the network environment and user needs. For classical music audio transmission, the network inputs are [-50, -45, 0.02, 0.01, 45, 40, 1], and the trained network outputs are [0.6, 32, 0.05, 0.8], corresponding to the quantization step size, buffer size (kb), switching threshold, and power level, respectively. Machine learning optimization enables the system to automatically adjust parameters based on historical experience, improving transmission efficiency and audio quality.

[0061] In order to verify the technical solution adopted in the embodiment of the present invention, a comparative experiment was designed to test the system performance improvement.

[0062] (1) Experimental environment setup: Test scenario: Office environment (200 square meters, with WiFi interference). Device configuration: Transmitter uses a dual-band WiFi module (2.4GHz 802.11n + 5GHz 802.11ac), receiver uses the same dual-band module; Audio format: 48kHz / 16-bit stereo, 320kbps bitrate; Interference sources: Three 2.4GHz routers, two 5GHz devices, and an intermittent microwave oven. Test distance: Multiple points within a 10-50 meter range.

[0063] (2) Comparison benchmark selection: Baseline Solution A: Traditional single-band 2.4GHz transmission (Bluetooth aptX-LL protocol).

[0064] Baseline Solution B: Single-band 5GHz WiFi audio transmission.

[0065] Embodiment of the present invention: dual-frequency adaptive coordinated transmission.

[0066] (3) Test indicators and results: ① Transmission stability test: Baseline solution A: 76% continuous transmission success rate, average interruption rate of 12 times / hour; Baseline solution B: 82% continuous transmission success rate, average interruption rate of 8 times / hour; Solution according to the present invention: 95% continuous transmission success rate, average interruption rate of 3 times / hour. Improvement effect: Compared with the best baseline solution, stability is improved by (95%-82%) / 82%≈16%, and compared with the traditional solution, it is improved by (95%-76%) / 76%≈25%.

[0067] ② Spectrum Utilization Test: Baseline Solution A: Average Spectrum Utilization 45%; Baseline Solution B: Average Spectrum Utilization 52%; Solution of the Invention: Average Spectrum Utilization 70%. Improvement: Compared with Baseline Solution B, the improvement is (70% - 52%) / 52% ≈ 35%.

[0068] ③ End-to-end latency test: Benchmark Solution A: Average latency 185ms, jitter ±45ms; Benchmark Solution B: Average latency 95ms, jitter ±25ms; Solution using the present invention: Average latency 48ms, jitter ±8ms. Latency control: Latency was successfully controlled within 50ms, meeting professional-grade requirements.

[0069] ④ Harsh Environment Adaptability Test (6% Packet Loss Environment): Baseline Solution A: Audio Interruption Rate 2.3%, Uptime 58%; Baseline Solution B: Audio Interruption Rate 1.8%, Uptime 65%; Solution of the Invention: Audio Interruption Rate 0.08%, Uptime 99.2%. Robustness Improvement: Audio Interruption Rate Reduced to 0.1%.

[0070] The following experimental data statistics table 2 is shown in the following table: Table 2 Experimental data statistics Test conditions: 200 square meters of office environment, multiple Wi-Fi interference, 48kHz / 16-bit audio; Test duration: Each scenario was tested continuously for 24 hours, repeated three times and averaged. Experimental results show that the dual-band adaptive cooperative transmission solution proposed in this paper significantly outperforms traditional Bluetooth and single-band Wi-Fi solutions in terms of transmission stability, spectrum utilization, delay control, and adaptability to harsh environments, effectively improving the overall performance of audio transmission.

[0071] Through the coordinated implementation of these six steps, this technical solution achieves high-quality dual-band audio transmission with end-to-end latency of less than 50ms, maintaining stable transmission performance in complex electromagnetic environments, and providing a reliable wireless transmission solution for professional audio applications and real-time interactive scenarios. This technical solution, through a dual-band collaborative transmission architecture, overcomes the technical bottlenecks of traditional single-band wireless audio transmission, achieving multiple key technological breakthroughs and significant performance improvements. First, the system utilizes intelligent frequency band detection and adaptive channel selection to dynamically identify the optimal transmission path in both the 2.4GHz and 5GHz bands, effectively mitigating electromagnetic interference and improving transmission stability by over 40% compared to traditional solutions. Second, layered coding technology scientifically decomposes audio data into a base layer and an enhancement layer, which are then transmitted through dual-band load balancing. This ensures basic sound quality reliability while fully utilizing the high bandwidth of the 5GHz band, increasing overall spectrum utilization by 35%. A core predictive buffer management algorithm dynamically adjusts the buffer size based on real-time network latency monitoring and trend prediction, ensuring playback continuity while keeping end-to-end latency below 50ms, meeting the stringent requirements of professional-grade audio production. The system's integrated Reed-Solomon forward error correction coding and intelligent retransmission mechanism establish a reliable error recovery system, maintaining smooth playback even in a 6% packet loss environment, with the audio interruption rate reduced to below 0.1%. In addition, the machine learning optimization module can predict and automatically adjust transmission parameters through in-depth analysis of historical transmission data, giving the system the ability to self-learn and continuously optimize. In summary, this technical solution not only solves the core problems of high latency and poor stability in wireless audio transmission, but also provides a revolutionary solution for scenarios that are extremely sensitive to latency, such as real-time audio and video applications, professional audio production, online games, and virtual reality in the 5G era. It has broad industrial application prospects and important technical value.

[0072] Second embodiment Based on the same inventive concept, this embodiment provides a low-latency adaptive audio transmission system, including: Frequency band detection module, used to dynamically identify the optimal transmission path in wireless dual-band based on intelligent frequency band detection and adaptive channel selection mechanism; A layered coding module is used to decompose the target audio data through a layered coding algorithm, and to implement differentiated transmission strategies and adaptive adjustment of quantization parameters; The load balancing module monitors the performance of the dual-bands in real time based on a load balancing algorithm, dynamically allocates the decomposed transmission tasks, and triggers path switching when abnormal transmission quality is detected in the current frequency band. The predictive buffering module is used to predict future buffering needs based on the predictive buffering algorithm, and dynamically adjust the buffer size according to the actual playback situation and network status changes to ensure playback continuity while minimizing delay. The principle is described in Example 1 and will not be repeated here.

[0073] Based on the same inventive concept, this embodiment provides a computer device, including a memory, a processor, and a computer program stored on the memory and runnable on the processor, characterized in that when the computer program is executed by the processor, the steps of any one of the above-mentioned low-latency adaptive audio transmission methods are implemented.

[0074] Based on the same inventive concept, this embodiment provides a computer-readable storage medium on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps of any of the above-mentioned low-delay adaptive audio transmission methods are implemented.

[0075] The computer device in this embodiment is described in detail below from the perspective of hardware processing.

[0076] See also Figure 2 As shown, the electronic device includes a processor 100 and a memory 101, wherein the memory 101 stores machine executable instructions that can be executed by the processor 100, and the processor 100 executes the machine executable instructions to implement the above-mentioned low-delay adaptive audio transmission method.

[0077] further, Figure 2 The electronic device shown further includes a bus 102 and a communication interface 103 , and the processor 100 , the communication interface 103 and the memory 101 are connected via the bus 102 .

[0078] Among them, the memory 101 may include a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), for example, at least one disk storage. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 103 (which can be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. can be used. The bus 102 can be an ISA bus, a PCI bus, or an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 1 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0079] The processor 100 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by an integrated logic circuit of hardware in the processor 100 or by instructions in the form of software. The above-mentioned processor 100 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The various methods, steps and logic block diagrams disclosed in the embodiments of the present disclosure can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present disclosure can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in memory 101, and processor 100 reads information in memory 101 and, in conjunction with its hardware, completes the method steps of the aforementioned embodiment.

[0080] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of the low-delay adaptive audio transmission method.

[0081] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0082] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or 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 for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0083] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A low-delay adaptive audio transmission method, characterized in that: include: Dynamically identify the optimal transmission path in wireless dual-band based on intelligent frequency band detection and adaptive channel selection mechanism; Decompose the target audio data through a layered coding algorithm, implement differentiated transmission strategies and adaptive adjustment of quantization parameters; Based on the load balancing algorithm, the system monitors the performance of dual frequency bands in real time, dynamically allocates the decomposed transmission tasks, and triggers path switching when abnormal transmission quality of the current frequency band is detected. Based on the predictive buffering algorithm, future buffering needs are predicted, and the buffer size is dynamically adjusted according to the actual playback situation and network status changes to ensure playback continuity while minimizing latency.

2. The low-delay adaptive audio transmission method according to claim 1, characterized in that The intelligent frequency band detection includes: Perform hardware initialization on the built-in dual-band wireless module; Calculate the available wireless channels in the current environment through dynamic scanning and set operations; The dual-band frequency band is 2.4 GHz and 5 GHz.

3. The low-delay adaptive audio transmission method according to claim 1, wherein: The adaptive channel selection mechanism dynamically identifies the optimal transmission path in the wireless dual-band, including: Performing quality assessment on the calculated available wireless channels based on a comprehensive channel quality assessment algorithm; The comprehensive channel quality assessment algorithm calculates the comprehensive score by weighting the original multi-dimensional indicators through a formula, giving priority to wireless channels with high scores. The original multi-dimensional indicators include signal strength, signal-to-noise ratio, packet loss rate and jitter. The weight coefficients in the weighted formula are adaptively adjusted according to the application scenario and historical performance feedback.

4. The low-delay adaptive audio transmission method according to claim 1, wherein: Decomposing the target audio data by a layered coding algorithm and executing a differentiated transmission strategy includes: Decomposing the original audio data into a base layer and an enhancement layer; When encoding the base layer, the original audio is downsampled and the downsampled signal is subjected to discrete cosine transform; When encoding the enhancement layer, the base layer encoded data is inversely transformed and upsampled to reconstruct the base layer signal. The difference signal obtained by subtracting the reconstructed base layer signal from the original audio signal is the enhancement layer data; Based on the real-time monitored network conditions, the base layer data packets are transmitted first to ensure basic sound quality.

5. The low-delay adaptive audio transmission method according to claim 1, wherein: The adaptive adjustment of the quantization parameter includes: Monitor network conditions in real time to obtain current network bandwidth, required bandwidth, actual delay, and target delay; Dynamically calculate the quantization step size based on a weighted formula of bandwidth and delay; The calculated quantization step size is passed to the current audio encoder for dynamic adjustment to maximize audio quality while ensuring latency requirements.

6. The low-delay adaptive audio transmission method according to claim 1, wherein: The real-time monitoring of dual-band performance based on the load balancing algorithm and the dynamic allocation of the decomposed transmission tasks include: Real-time collection of frequency band quality parameters to calculate frequency band quality score Q value and load distribution weight; Compare the allocation weights of the two bands; If the allocation weight of the 2.4 GHz band is greater than the allocation weight of the 5 GHz band, the base layer audio data is allocated to the 2.4 GHz band for transmission, and the enhancement layer audio data is allocated to the 5 GHz band for transmission; If the allocation weight of the 5 GHz band is greater than or equal to the allocation weight of the 2.4 GHz band, the base layer audio data is allocated to the 5 GHz band for transmission, and the enhancement layer audio data is allocated to the 2.4 GHz band for transmission.

7. The low-delay adaptive audio transmission method according to claim 1, wherein: The triggering of path switching when abnormal transmission quality of the current frequency band is detected includes: Continuously monitor the transmission quality of the current frequency band, including packet loss rate, latency, and signal strength; The trigger is controlled according to the preset switching decision algorithm, and another frequency band is selected as the new transmission path after the trigger; The quality score Q value of the new frequency band is recalculated, and the load balancing algorithm is triggered to optimize the allocation.

8. The low-delay adaptive audio transmission method according to claim 1, wherein: The predicting of future buffering requirements based on the predictive buffering algorithm includes: Obtain historical data and changing trends of network latency; Based on the buffer size prediction model, the delay at the next moment is predicted by balancing the current observation value with the historical prediction value through exponential smoothing method; The target buffer size is calculated based on the predicted delay, audio bitrate, and the set safety margin; The buffer size at the receiving end is adjusted according to the target buffer size to achieve smooth playback under network fluctuations.

9. The low-delay adaptive audio transmission method according to claim 1, wherein: Dynamically adjusting the buffer size to ensure playback continuity while minimizing latency based on actual playback conditions and network status changes includes: Monitor the current buffer level in real time and compare it with the preset threshold; If the buffer level is less than the minimum threshold, the buffer increase strategy is triggered to expand the target buffer; If the buffer level is greater than the maximum threshold, the buffer reduction strategy is triggered to reduce the target buffer; If the buffer level has not changed, the current buffer size is maintained; The buffer size after triggering the adjustment is adjusted to the target size through a smooth transition mechanism.

10. The low-delay adaptive audio transmission method according to claim 1, wherein: Also includes: Based on the hybrid error recovery mechanism, Reed-Solomon error correction coding is combined with the selective retransmission strategy to repair abnormal audio data packets in stages.

11. The low-delay adaptive audio transmission method according to claim 1, wherein: The hybrid error recovery mechanism combines Reed-Solomon error correction coding with a selective retransmission strategy to repair abnormal audio data packets in stages, including: Forward error correction is achieved by adding Reed-Solomon error correction code to the audio data packet, automatically recovering from some transmission errors; For lost packets that cannot be recovered by FEC, selective fast retransmission is performed based on a priority algorithm based on packet importance, age, and number of retransmissions, giving priority to recovering critical data.

12. The low-delay adaptive audio transmission method according to claim 1, wherein: Also includes: Ensures continuous high-performance transmission by monitoring audio transmission performance in real time and using machine learning to predict optimal parameters.

13. A low-latency adaptive audio transmission system, characterized in that: include: Frequency band detection module, used to dynamically identify the optimal transmission path in wireless dual-band based on intelligent frequency band detection and adaptive channel selection mechanism; A layered coding module is used to decompose the target audio data through a layered coding algorithm, and to implement differentiated transmission strategies and adaptive adjustment of quantization parameters; The load balancing module monitors the performance of the dual-bands in real time based on a load balancing algorithm, dynamically allocates the decomposed transmission tasks, and triggers path switching when abnormal transmission quality is detected in the current frequency band. The predictive buffering module is used to predict future buffering needs based on the predictive buffering algorithm, and dynamically adjust the buffer size according to the actual playback situation and network status changes to ensure playback continuity while minimizing latency.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the computer program is executed by a processor, the method steps for low-delay adaptive audio transmission according to any one of claims 1 to 12 are implemented.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method steps for low-delay adaptive audio transmission according to any one of claims 1 to 12 are implemented.

Images