Low-delay lossless digital audio transmission method and system and storage medium

Abstract

The invention discloses a low-delay lossless digital audio transmission method, a low-delay lossless digital audio transmission system and a storage medium, which relate to the technical field of digital audio transmission and are used for monitoring network environment data in real time and dynamically adjusting audio transmission parameters according to changes of network bandwidth Bw, network delay Dr, packet loss rate Pd and network jitter Wd. Comprising an audio transmission rate Rsv, a data packet size Spd and a redundancy coding ratio Rbf. The adaptive adjustment can ensure that the data transmission quantity is reduced and the packet loss probability is reduced when the network condition is poor, and meanwhile, the transmission efficiency is improved and the audio quality is optimized when the bandwidth condition is good. Compared with a traditional fixed transmission parameter method, lossless audio transmission can be achieved under different network conditions, and meanwhile unnecessary delay is reduced. According to the invention, intelligent compression of the audio data is realized, the data volume is reduced under the condition that the bandwidth is limited, the network burden is reduced, and the stability and integrity of audio transmission in a low-bandwidth environment are ensured.

Description

Technical Field

[0001] This invention relates to the field of digital audio transmission technology, specifically to a low-latency lossless digital audio transmission method, system, and storage medium. Background Technology

[0002] With the rapid development of internet technology, the field of communications has gradually become segmented into several subfields, including network communication, wireless communication, and audio / video communication. Among these, audio communication, as a crucial component of network communication, plays a vital role in various application scenarios, especially in real-time audio transmission (such as voice calls, online meetings, and distance education). In this specific field, low-latency lossless audio transmission technology is particularly important, as it involves minimizing audio transmission latency while ensuring lossless sound quality.

[0003] Despite significant advancements in existing audio transmission technologies and the widespread adoption of numerous audio transmission protocols across various network environments, some shortcomings still exist in specific network conditions. Firstly, bandwidth variations and network instability directly impact audio transmission quality. While existing transmission protocols can generally guarantee audio quality in high-bandwidth environments, in low-bandwidth environments, audio data often requires compression, which can lead to a decline in audio quality.

[0004] The shortcomings of existing audio transmission protocols have led to several problems, the most significant being the instability of voice call quality. When bandwidth is insufficient, excessive compression can cause a significant deterioration in audio quality, especially in the distortion or blurring of speech details, making it difficult for listeners to clearly understand the speech content. Simultaneously, in network environments with fluctuating bandwidth, frequent packet loss and network jitter can cause audio transmission interruptions or stuttering, severely impacting communication continuity and user experience. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a low-latency lossless digital audio transmission method, system, and storage medium, solving the problems mentioned in the background section.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a low-latency lossless digital audio transmission method, comprising the following steps:

[0007] S1. By using network probing techniques and network tools, monitor the current network environment data, fit it into the original dataset YS, and perform preprocessing to obtain the network dataset WS;

[0008] S2. Based on the network dataset WS, dynamically adjust the audio transmission parameters, including audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf, and fit them into a parameter set CW.

[0009] S3. Based on the network bandwidth Bw and network latency Dr in the network dataset WS, adjust the audio compression ratio Cb using an adaptive compression algorithm.

[0010] S4. Based on the parameter set CW and the audio compression ratio Cb, the audio traffic fluctuation F is predicted by using a time series regression prediction algorithm, and a low-latency transmission route is dynamically selected.

[0011] S5. Based on the obtained traffic fluctuation F, the forward error correction technology FEC is combined to add redundant data to the audio data packet through encoding, and the redundancy ratio Rf is obtained.

[0012] S6. Calculate the comprehensive transmission quality index Q based on the obtained network dataset WS, traffic fluctuation F, and redundancy ratio Rf.

[0013] Preferably, S1 includes S11 and S12;

[0014] S11. Using network probing techniques and network tools, monitor current network environment data, including network bandwidth Bw, network latency Dr, packet loss rate Pd, and network jitter Wd, and fit the data to obtain the original dataset YS.

[0015] Among them, network bandwidth Bw was collected using network measurement tools iperf, netflow, and routers;

[0016] Network latency Dr was obtained through latency testing tools ping, traceroute, and router data.

[0017] Packet loss rate Pd is obtained through packet loss detection tools ping and iperf, and data collected from the switch.

[0018] Network jitter (Wd) is collected using latency and packet loss monitoring tools ping and iperf.

[0019] S12. Perform data cleaning and standardization on the obtained raw dataset YS to obtain the network dataset WS;

[0020] Data cleaning includes removing incomplete and outlier values ​​from the original dataset YS, processing missing data packets, and removing noisy data.

[0021] Standardization processing includes standardizing the data in the original dataset YS, converting network data of different dimensions and units into dimensionless data, and obtaining the network dataset WS;

[0022] The network dataset WS is obtained using the following formula:

[0023] ;

[0024] In the formula, WSd represents the d-th data item in the network dataset WS, YSd represents the d-th data item in the original dataset YS, μYSd represents the mean of the d-th data item in the original dataset YS, and σYSd represents the standard deviation of the d-th data item in the original dataset YS.

[0025] Preferably, S2 includes S21 and S22;

[0026] S21. Dynamically adjust the audio transmission rate Rsv based on the network bandwidth Bw and packet loss rate Pd; when the bandwidth is high, a higher audio transmission rate is allowed; while when the bandwidth is low, the transmission rate is reduced to avoid excessive packet loss and delay.

[0027] The audio transmission rate Rsv is obtained using the following formula:

[0028] ;

[0029] In the formula, C1 represents a constant;

[0030] When the packet loss rate Pd remains constant, and the network bandwidth Bw increases, the audio transmission rate Rsv will increase, indicating that more data can be transmitted.

[0031] When the network bandwidth Bw remains constant, as the packet loss rate Pd increases, (1-Pd) will decrease, thereby reducing the transmission rate and avoiding excessive data loss.

[0032] S22. Dynamically adjust the audio data packet size Spd based on network bandwidth Bw and network latency Dr; reduce the data packet size when network bandwidth is low or latency is high to reduce the risk of packet loss during transmission.

[0033] Specifically, the network bandwidth Bw, network latency Dr, and packet loss rate Pd are normalized by using a normalization process to eliminate the dimensions of the data.

[0034] The packet size Spd is obtained using the following formula:

[0035] ;

[0036] In the formula, C2 represents the proportional constant that controls the size of the data packet;

[0037] The proportional constant C2 for controlling the packet size is obtained as follows: First, based on the application type, target bitrate, and historical engineering experience, an initial empirical value C2init can be given; then, a test environment is constructed using the bandwidth range Bw∈[0.5,10], the latency range Dr∈[5,300], and the packet loss rate Pd∈[0.0,0.3]; finally, combined with real-time performance indicators collected from the actual deployed system (such as RTT, Jitter, Packet Loss, voice playback consistency, etc.), a sliding window feedback mechanism is constructed to adjust the proportional constant C2 for controlling the packet size.

[0038] The packet size Spd increases with the increase of network bandwidth Bw. When the network latency Dr increases, the system will actively reduce the packet size to reduce the transmission latency of subsequent data, thereby adapting to the high-latency network environment.

[0039] Adjust the redundancy coding ratio Rbf according to the packet loss rate Pd and network jitter Wd; increase redundancy coding when the packet loss rate is high or the jitter is large to ensure that the audio data can still be transmitted completely even if there is packet loss or network jitter.

[0040] Redundancy coding ratio Rbf is obtained using the following formula:

[0041] ;

[0042] In the formula, C3 represents the redundancy coding ratio coefficient;

[0043] The redundancy coding ratio Rbf is proportional to the packet loss rate Pd and the network jitter Wd, which means that when the packet loss rate Pd and the jitter Wd are high, the redundancy coding ratio Rbf will be increased to ensure that the audio can still be recovered even if packet loss or network fluctuations occur.

[0044] The obtained audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf are fitted to obtain the parameter set CW.

[0045] Preferably, S3 includes S31 and S32;

[0046] S31. Dynamically adjust the audio compression ratio Cb according to the network bandwidth Bw; in a high-bandwidth environment, since more data can be transmitted, we use a lower compression ratio to retain more audio details and reduce compression loss; while in a low-bandwidth environment, we need to increase the compression ratio to reduce the size of the audio data and ensure that the data can be transmitted smoothly in the network.

[0047] The audio compression ratio Cb is obtained using the following formula:

[0048] ;

[0049] In the formula, Bwmax represents the maximum network bandwidth, Cmin represents the minimum compression ratio, and Cmax represents the maximum compression ratio.

[0050] When the network bandwidth Bw approaches the maximum network bandwidth Bwmax, the audio compression ratio Cb will approach the minimum compression ratio Cmin, that is, using a lower compression ratio to preserve audio quality as much as possible.

[0051] When the network bandwidth Bw is low, the audio compression ratio Cb will increase to reach the maximum compression ratio Cmax in order to compress more data and adapt to the low bandwidth environment.

[0052] S32. Adjust the audio compression ratio Cb according to the network latency Dr. When the network latency Dr is greater than the preset latency threshold Tdr, the data transmission delay will increase. At this time, it is necessary to adjust the compression ratio to reduce the impact of latency.

[0053] The formula for adjusting the network latency Dr to the audio compression ratio Cb is as follows:

[0054] ;

[0055] In the formula, α represents the adjustment factor, and Dmax represents the maximum network latency.

[0056] When the network latency Dr is high, the compression ratio Cb will be increased to increase the compression of audio data, reduce the amount of data transmitted, and thus reduce the impact of latency on audio quality.

[0057] When the latency is low, the compression ratio Cb will be slightly reduced to ensure high audio quality as much as possible.

[0058] Preferably, S4 includes S41 and S42;

[0059] S41. Using a time series regression prediction algorithm, based on historical audio traffic data F(t-1), predict the future audio traffic fluctuation F;

[0060] By combining historical audio traffic data F(t-1), network bandwidth Bw, packet loss rate Pd with parameter set CW and audio compression ratio Cb, the traffic fluctuation F(t) at future time t can be predicted.

[0061] ;

[0062] In the formula, Represents the autoregressive coefficient. The coefficient representing the impact of bandwidth variation. This represents the influence coefficient of packet loss rate. Let F(t) represent the influence coefficient of the compression ratio, F(t-1) represent the audio traffic at time t, F(t-1) represent the audio traffic at time t-1, Bw(t) represent the network bandwidth at time t, Pd(t) represent the packet loss rate at time t, and CWi represent the i-th data item in the parameter set CW. This represents the preset weight value of the i-th data item in the parameter set CW.

[0063] Autoregression coefficient The acquisition method is as follows: First, construct an audio traffic time series dataset: collect historical audio traffic data, requiring at least hundreds of time series samples; then, use the least squares (OLS) method for linear fitting, and use the multiple linear regression method to calculate the goodness of fit to obtain the optimal parameters θ1 and θ2; finally, perform the Ljung-Box test on the residual series to obtain the final autoregressive coefficients. ;

[0064] The influence coefficient β2 of bandwidth change is obtained by collecting data pairs corresponding to network bandwidth and audio traffic; keeping other variables (such as packet loss rate, compression ratio, etc.) constant, constructing a single-factor regression model, solving for β2, and finally using the least squares method or gradient descent method to fit and obtain the value of β2.

[0065] The impact coefficient β3 of the packet loss rate was obtained by: collecting historical sample pairs of packet loss rate Pd and audio traffic F; performing partial correlation analysis, controlling for Bw and other variables, and evaluating the net impact of Pd on F; fitting a nonlinear function relationship to obtain the impact coefficient β3 of the packet loss rate.

[0066] The influence coefficient β4 of the compression ratio was obtained by collecting samples of different compression ratios Cb and corresponding F(t) values; constructing a fitting model and testing the effectiveness of β4; and using a t-test to analyze the significance of β4 to ensure that the influence coefficient β4 of the compression ratio is meaningful in the model.

[0067] Preferably, in step S42, a low-latency transmission route is dynamically selected based on the traffic fluctuation F(t) at the obtained future time t;

[0068] Assuming there are N paths, the network bandwidth and network latency of each path are defined, and the comprehensive routing score Rn is calculated based on the predicted traffic fluctuation F(t) at future time t, the latency and bandwidth of each path, and the transmission route is selected.

[0069] Specifically, the network bandwidth nBw and network delay nDr of path n are normalized to unify the dimensions of the data.

[0070] The overall routing score Rn is obtained using the following formula:

[0071] ;

[0072] In the formula, nBw represents the network bandwidth of path n, nDr represents the network latency of path n, and Fmax represents the maximum predicted traffic value, specifically the maximum value among all paths.

[0073] This represents the bandwidth resource per unit of latency, and is one of the classic QoS routing evaluation metrics in communication networks. Higher bandwidth and lower latency indicate that the path has better audio transmission capabilities.

[0074] This represents a dynamic weighting adjustment for the current predicted traffic fluctuations. The greater the predicted fluctuations, the closer this item is to 0, which is used to reduce the score of the path and avoid the impact of high load.

[0075] Calculate the comprehensive routing score Rn for each path n, and select the path with the highest score, which is the path with greater bandwidth and lower latency. Finally, based on the predicted traffic fluctuation F and the comprehensive routing score Rn, we can select the optimal low-latency path in a dynamically changing network environment. This process enables audio data to be transmitted on the most suitable path, reducing latency, minimizing packet loss, and ensuring high-quality audio transmission.

[0076] Preferably, S5, by analyzing the traffic fluctuation F, the amount of redundant data that needs to be added is determined, the redundancy ratio Rf is obtained, and the redundancy ratio Rf is judged by the traffic fluctuation F.

[0077] Traffic fluctuation F reflects the volatility of audio data during transmission. When the traffic fluctuation F is greater than the preset fluctuation threshold Trf, it indicates that the network is unstable and the packet loss rate Pd will increase.

[0078] The redundancy ratio Rf is obtained using the following formula:

[0079] ;

[0080] In the formula, B1 represents the influence coefficient of audio traffic fluctuation on the redundancy ratio;

[0081] The redundancy ratio Rf is determined by matching in the following way:

[0082] When the traffic fluctuation F is greater than the fluctuation threshold Trf, it indicates that an abnormal fluctuation has occurred. As the traffic fluctuation F increases, the redundancy ratio Rf will also increase. Increasing redundant data can help the receiver recover lost data packets.

[0083] When the traffic fluctuation F ≤ the fluctuation threshold Trf, it indicates that there is no abnormal fluctuation. When the traffic fluctuation F decreases, the redundancy ratio Rf will also decrease, thereby reducing the amount of redundant data transmitted and improving transmission efficiency.

[0084] Based on the calculated redundancy ratio Rf, redundant data is added to each audio data packet. The redundant data is embedded into the audio data packet using forward error correction (FEC) technology, ensuring that even if some data packets are lost during transmission, the receiving end can still recover the audio data and obtain the total data volume Dto.

[0085] The total data volume Dto is obtained using the following formula:

[0086] ;

[0087] In the formula, Dau represents the size of the original audio data packet.

[0088] Dto represents the total amount of audio data packets during transmission, including raw audio data and redundant data;

[0089] The redundancy ratio Rf controls the size of redundant data. A higher redundancy ratio Rf will result in a larger amount of redundant data, thereby enhancing data recovery capabilities.

[0090] By embedding redundant data into audio data packets, the receiver can use FEC technology to recover lost data packets; using Reed-Solomon encoding technology, the receiver can recover complete audio from the redundant data even if some data packets are lost.

[0091] Preferably, in step S6, the comprehensive transmission quality index Q is calculated based on the obtained network dataset WS, traffic fluctuation F, and redundancy ratio Rf.

[0092] Specifically, the obtained network bandwidth Bw, network latency Dr, packet loss rate Pd, traffic fluctuation F, and redundancy ratio Rf are normalized to unify the dimensions of the data.

[0093] The overall transmission quality index Q is obtained using the following formula:

[0094] ;

[0095] In the formula, F(t) represents the flow fluctuation at time t, ln represents the logarithmic function, e represents a constant, d1 represents the redundancy gain coefficient, and d2 represents the delay exponential decay factor.

[0096] The steps to obtain the redundancy gain coefficient d1 are as follows:

[0097] Step 1: Define the evaluation time window: Set a fixed time window (e.g., the most recent 30 seconds or 100 data packets); within this window, count the changes in transmission quality before and after enabling FEC.

[0098] Step 2: Collect basic transmission performance indicators: During transmission without redundant coding, collect the following raw parameters: packet loss rate Pd0 (with FEC off); average retransmission count Rt0; end-to-end reconstruction error rate Er0.

[0099] Step 3: Enable FEC mechanism and collect similar indicators: Collect again: packet loss rate Pd1 (FEC enabled); average retransmission count Rt1; reconstruction error rate Er1;

[0100] Step 4: Analyze the quality benefits brought by FEC: Analyze the decrease ratio of Pd0 and Pd1; analyze the decrease trend of Rt0 and Rt1; analyze the actual recovery effect of Er0 and Er1.

[0101] If Pd or Rt decreases significantly while Er improves markedly, it indicates that redundant coding brings actual gains.

[0102] Step 5: Set the redundancy gain coefficient d1 value according to the quality improvement range:

[0103] If Pd and Rt decrease by more than 50%, then the redundancy gain coefficient d1 can be considered to be high (set to 0.8~1.0).

[0104] If the improvement is slight, the redundancy gain coefficient d1 can be set between 0.3 and 0.5;

[0105] If there is no significant improvement (or even a deterioration), then assume the redundancy gain coefficient d1 ≤ 0.2;

[0106] The redundancy gain coefficient d1 can be adjusted by sliding based on the historical multi-window average value.

[0107] The steps to obtain the delay exponential decay factor d2 are as follows:

[0108] Step 1: Set the system response sensitivity level: Determine the current business type (e.g., video stream, IoT heartbeat, file transfer, etc.).

[0109] Set basic response sensitivity levels according to different business needs.

[0110] Step 2: Collect the latency variation dataset: Collect the maximum, minimum, average, and standard deviation of network path latency over the past N windows, and denot them as follows:

[0111] Step 3: Analyze the correlation between latency and jitter / retransmission / packet loss: compare whether the increase in latency is accompanied by a deterioration in QoS indicators (such as increased packet loss, video stuttering, etc.).

[0112] If a high delay period significantly corresponds to a decrease in transmission quality, the delay exponential decay factor d2 can be increased.

[0113] If the system is robust to latency (such as file transfer), the latency exponential decay factor d2 should be relatively low.

[0114] Step 4: Set the delay exponential decay factor d2 value based on the impact of delay volatility and quality: If the standard deviation is >40ms and the impact is significant, the delay exponential decay factor d2 should be ≥1.0;

[0115] If the standard deviation is between 10 and 30 ms and the effect is mild, the delay exponential decay factor d2 can be set to 0.5 to 0.8;

[0116] If the standard deviation is <10ms and the system can tolerate it, the delay exponential decay factor d2 can be set to 0.2~0.4;

[0117] This value can be used for statistical updates via a sliding window, or it can be adaptively adjusted by fitting a long-term historical curve.

[0118] The obtained comprehensive transmission quality index Q is compared with the preset quality threshold TQ to determine whether the transmission quality meets expectations.

[0119] The expected transmission quality is obtained by matching it in the following ways:

[0120] When the overall transmission quality index Q ≥ quality threshold TQ, it indicates that the transmission quality is normal and meets expectations.

[0121] When the overall transmission quality index Q < the quality threshold TQ, it indicates that the transmission quality is abnormal, has not met expectations, has poor transmission quality, and has high latency and packet loss problems. It is necessary to adjust the audio transmission rate Rsv and the redundancy ratio Rf to obtain a new audio transmission rate nRsv and a new redundancy ratio nRf.

[0122] The new audio transmission rate nRsv is obtained using the following formula:

[0123] ;

[0124] The new redundancy ratio nRf is obtained using the following formula:

[0125] ;

[0126] In the formula, X represents the adjustment factor, Rmax represents the maximum value of the redundancy ratio, and min() represents the minimum function. This ensures that the redundancy ratio Rf will not exceed the maximum value Rmax. If the calculated redundancy ratio Rf exceeds the maximum value Rmax, it is then limited to that maximum value.

[0127] A low-latency lossless digital audio system includes a real-time network environment assessment module, a dynamic audio transmission parameter adjustment module, an intelligent audio compression and decompression module, an audio traffic prediction and routing module, a forward error correction and redundancy adjustment module, and a real-time feedback and adjustment module.

[0128] The real-time network environment assessment module monitors the current network environment data through network detection technology and network tools, fits it into the original dataset YS, and performs preprocessing to obtain the network dataset WS;

[0129] The dynamic audio transmission parameter adjustment module dynamically adjusts the audio transmission parameters, including audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf, based on the network dataset WS, and fits them into a parameter set CW.

[0130] The intelligent audio compression and decompression module adjusts the audio compression ratio Cb by using an adaptive compression algorithm based on the network bandwidth Bw and network latency Dr in the network dataset WS.

[0131] The audio traffic prediction and routing module, based on the parameter set CW and the audio compression ratio Cb, uses a time series regression prediction algorithm to predict the traffic fluctuation F of audio traffic and dynamically select low-latency transmission routes.

[0132] The forward error correction and redundancy adjustment module combines the acquired traffic fluctuation F with the forward error correction technology FEC, and adds redundant data to the audio data packets through encoding to obtain the redundancy ratio Rf.

[0133] The real-time feedback and adjustment module calculates and obtains the comprehensive transmission quality index Q based on the acquired network dataset WS, traffic fluctuation F, and redundancy ratio Rf.

[0134] A computer storage medium storing a computer program that, when executed by a processor, enables a low-latency, lossless digital audio transmission method.

[0135] This invention provides a low-latency lossless digital audio transmission method, system, and storage medium, which have the following beneficial effects:

[0136] (1) By monitoring network environment data in real time, and dynamically adjusting audio transmission parameters, including audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf, based on changes in network bandwidth Bw, network latency Dr, packet loss rate Pd, and network jitter Wd, this adaptive adjustment ensures reduced data transmission volume and lower probability of packet loss under poor network conditions, while improving transmission efficiency and optimizing audio quality under better bandwidth conditions. Compared to traditional fixed transmission parameter methods, this invention can achieve lossless audio transmission under different network conditions while reducing unnecessary latency.

[0137] This invention achieves intelligent audio data compression by adjusting the audio compression ratio based on network bandwidth and latency. Specifically, when network bandwidth is high, a lower compression ratio is used to ensure high-fidelity audio quality; while when bandwidth is limited, the compression ratio is increased to reduce data volume, thereby reducing network load and ensuring the stability and integrity of audio transmission in low-bandwidth environments. This adaptive compression technology significantly improves the transmission quality of audio in different network environments.

[0138] (2) By using various network measurement tools to monitor network bandwidth (Bw), network latency (Dr), packet loss rate (Pd), and network jitter (Wd) in real time, and dynamically adjusting various parameters of audio transmission based on these data, this real-time network environment monitoring and dynamic adjustment mechanism ensures that audio transmission can be automatically optimized according to the current network conditions. Compared with traditional fixed parameter settings, this invention can more intelligently respond to changes in network conditions and provide more reliable audio transmission.

[0139] By cleaning and standardizing the original dataset to remove incomplete and noisy data and transforming network parameters into dimensionless data, data consistency and quality during transmission are ensured. This preprocessing allows for more precise subsequent parameter adjustments, preventing audio quality fluctuations caused by inaccurate network data or noisy data influencing decision-making. Furthermore, dynamically adjusting the audio transmission rate and packet size, combined with redundant coding adjustments, further guarantees stable audio transmission even under low bandwidth or unstable network conditions, avoiding packet loss or excessive latency.

[0140] (3) A time-series regression prediction algorithm is introduced to accurately predict future audio traffic fluctuations based on historical audio traffic data, network bandwidth Bw, packet loss rate Pd, and audio compression ratio Cb. Through this prediction, the system can understand the changing trend of audio data traffic in advance and dynamically select a low-latency transmission route according to the prediction results, avoiding audio transmission interruptions or stuttering caused by network bandwidth fluctuations or sudden increases in latency. In this way, the fluctuation of audio traffic and changes in network conditions can be responded to in advance, thereby effectively ensuring the stability and continuity of real-time audio transmission. The impact of latency is reduced by adjusting the audio compression ratio Cb, especially when the network latency is high. The system can automatically adjust the audio compression ratio according to the preset latency threshold. Through this mechanism, the system can reduce the amount of data in a high-latency network environment, thereby reducing the latency of audio transmission and avoiding the impact of high latency on audio smoothness in traditional transmission methods. In addition, dynamically adjusting the compression ratio and transmission parameters also optimizes the balance between audio quality and latency, improving the user's audio experience.

[0141] (4) By analyzing traffic fluctuations in real time, the redundancy ratio Rf is intelligently adjusted, and redundant data is added through forward error correction (FEC) technology to improve the recovery capability of audio data. When traffic fluctuations are large, indicating network instability, the system automatically adds redundant data to ensure that audio data can be recovered even if packet loss occurs during transmission. This redundancy adjustment mechanism based on traffic fluctuations greatly improves the robustness of audio transmission, avoids audio quality loss or transmission interruption caused by network problems, and ensures the continuity and stability of audio calls. By monitoring audio traffic fluctuations and network status in real time, the amount of redundant data is dynamically determined. When the network is stable, the system reduces the redundancy ratio, thereby reducing unnecessary data transmission and improving efficiency; while when network conditions deteriorate, the system automatically adds redundant data to enhance the reliability and recovery capability of audio data. This mechanism effectively avoids bandwidth waste caused by excessive redundant data, and at the same time, it can ensure that audio data is not lost to the greatest extent when the network is unstable. Attached Figure Description

[0142] Figure 1 This is a schematic diagram illustrating the steps of a low-latency lossless digital audio transmission method according to the present invention;

[0143] Figure 2 This is a schematic diagram of a low-latency lossless digital audio transmission system according to the present invention;

[0144] Figure 3 This is a schematic diagram illustrating the execution of the storage medium control flow of the present invention;

[0145] Figure 4 This is a line graph showing the achievement of the comprehensive transmission quality index of the present invention. Detailed Implementation

[0146] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0147] Example 1

[0148] This invention provides a low-latency lossless digital audio transmission method. Please refer to [link / reference]. Figures 1 to 4 This includes the following steps:

[0149] S1. By using network probing techniques and network tools, monitor the current network environment data, fit it into the original dataset YS, and perform preprocessing to obtain the network dataset WS;

[0150] S2. Based on the network dataset WS, dynamically adjust the audio transmission parameters, including audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf, and fit them into a parameter set CW.

[0151] S3. Based on the network bandwidth Bw and network latency Dr in the network dataset WS, adjust the audio compression ratio Cb using an adaptive compression algorithm.

[0152] S4. Based on the parameter set CW and the audio compression ratio Cb, the audio traffic fluctuation F is predicted by using a time series regression prediction algorithm, and a low-latency transmission route is dynamically selected.

[0153] S5. Based on the obtained traffic fluctuation F, the forward error correction technology FEC is combined to add redundant data to the audio data packet through encoding, and the redundancy ratio Rf is obtained.

[0154] S6. Calculate the comprehensive transmission quality index Q based on the obtained network dataset WS, traffic fluctuation F, and redundancy ratio Rf.

[0155] In this embodiment, by monitoring network environment data in real time, audio transmission parameters, including audio transmission rate Rsv, packet size Spd, and redundancy coding ratio Rbf, are dynamically adjusted based on changes in network bandwidth Bw, network latency Dr, packet loss rate Pd, and network jitter Wd. This adaptive adjustment ensures reduced data transmission volume and lower probability of packet loss under poor network conditions, while improving transmission efficiency and optimizing audio quality under good bandwidth conditions. Compared to traditional fixed transmission parameter methods, this invention can achieve lossless audio transmission under different network conditions while reducing unnecessary latency.

[0156] This invention achieves intelligent audio data compression by adjusting the audio compression ratio based on network bandwidth and latency. Specifically, when network bandwidth is high, a lower compression ratio is used to ensure high-fidelity audio quality; while when bandwidth is limited, the compression ratio is increased to reduce data volume, thereby reducing network load and ensuring the stability and integrity of audio transmission in low-bandwidth environments. This adaptive compression technology significantly improves the transmission quality of audio in different network environments.

[0157] This invention employs a time-series regression prediction algorithm to predict audio traffic fluctuations and dynamically selects low-latency transmission routes based on the prediction results. This improvement allows the system to anticipate and adjust transmission paths in advance based on network traffic changes, avoiding network congestion and excessive latency, and ensuring the continuity and real-time performance of audio data during transmission. Compared to traditional methods, this invention can more flexibly respond to network fluctuations, improving the real-time interaction quality in audio calls and online conferences. By calculating the comprehensive transmission quality index Q, this invention enables the system to evaluate the current audio transmission quality in real time. If the system detects excessive latency or a decline in audio quality, it can immediately reduce the transmission rate or add redundant data, thereby adjusting transmission parameters in a timely manner to optimize audio transmission quality. This comprehensive evaluation mechanism ensures that high audio quality is maintained consistently under different network environments and conditions, avoiding quality fluctuations caused by network fluctuations in traditional methods.

[0158] Example 2

[0159] This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically: S1 includes S11 and S12;

[0160] S11. Using network probing techniques and network tools, monitor current network environment data, including network bandwidth Bw, network latency Dr, packet loss rate Pd, and network jitter Wd, and fit the data to obtain the original dataset YS.

[0161] Among them, network bandwidth Bw was collected using network measurement tools iperf, netflow, and routers;

[0162] Network latency Dr was obtained through latency testing tools ping, traceroute, and router data.

[0163] Packet loss rate Pd is obtained through packet loss detection tools ping and iperf, and data collected from the switch.

[0164] Network jitter (Wd) is collected using latency and packet loss monitoring tools ping and iperf.

[0165] S12. Perform data cleaning and standardization on the obtained raw dataset YS to obtain the network dataset WS;

[0166] Data cleaning includes removing incomplete and outlier values ​​from the original dataset YS, processing missing data packets, and removing noisy data.

[0167] Standardization processing includes standardizing the data in the original dataset YS, converting network data of different dimensions and units into dimensionless data, and obtaining the network dataset WS;

[0168] The network dataset WS is obtained using the following formula:

[0169] ;

[0170] In the formula, WSd represents the d-th data item in the network dataset WS, YSd represents the d-th data item in the original dataset YS, μYSd represents the mean of the d-th data item in the original dataset YS, and σYSd represents the standard deviation of the d-th data item in the original dataset YS.

[0171] S2 includes S21 and S22;

[0172] S21. Dynamically adjust the audio transmission rate Rsv based on the network bandwidth Bw and packet loss rate Pd;

[0173] The audio transmission rate Rsv is obtained using the following formula:

[0174] ;

[0175] In the formula, C1 represents a constant;

[0176] S22. Dynamically adjust the audio data packet size Spd based on network bandwidth Bw and network latency Dr;

[0177] The packet size Spd is obtained using the following formula:

[0178] ;

[0179] In the formula, C2 represents the proportional constant that controls the size of the data packet;

[0180] Adjust the redundancy coding ratio Rbf according to the packet loss rate Pd and network jitter Wd;

[0181] Redundancy coding ratio Rbf is obtained using the following formula:

[0182] ;

[0183] In the formula, C3 represents the redundancy coding ratio coefficient;

[0184] The obtained audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf are fitted to obtain the parameter set CW.

[0185] In this embodiment, various network measurement tools are used to monitor network bandwidth (Bw), network latency (Dr), packet loss rate (Pd), and network jitter (Wd) in real time, and the audio transmission parameters are dynamically adjusted based on this data. This real-time network environment monitoring and dynamic adjustment mechanism ensures that audio transmission can be automatically optimized according to the current network conditions. Compared to traditional fixed parameter settings, this invention can more intelligently respond to changes in network conditions and provide more reliable audio transmission.

[0186] By cleaning and standardizing the original dataset to remove incomplete and noisy data and transforming network parameters into dimensionless data, data consistency and quality during transmission are ensured. This preprocessing allows for more precise subsequent parameter adjustments, preventing audio quality fluctuations caused by inaccurate network data or noisy data influencing decision-making. Furthermore, dynamically adjusting the audio transmission rate and packet size, combined with redundant coding adjustments, further guarantees stable audio transmission even under low bandwidth or unstable network conditions, avoiding packet loss or excessive latency.

[0187] This embodiment optimizes the transmission efficiency and robustness of audio data under different network environments by dynamically adjusting the audio transmission rate, data packet size, and redundancy coding ratio. Especially when bandwidth is limited, adjusting the audio transmission rate and data packet size helps reduce network load, while redundancy coding ensures data recovery, thus guaranteeing unaffected audio quality even in the event of packet loss or jitter. This method maximizes network bandwidth utilization and reduces transmission latency while maintaining audio quality. By using a time-series regression prediction algorithm to predict audio traffic fluctuations, network load can be assessed in advance, and low-latency transmission routes can be dynamically selected. This technology avoids the accumulation of network congestion and latency, and allows for timely adjustment of audio transmission paths when network conditions change, ensuring smooth real-time audio data with low latency. It is particularly suitable for scenarios with high requirements for audio quality and latency, such as remote voice calls and online conferences.

[0188] Example 3

[0189] This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically: S3 includes S31 and S32;

[0190] S31. Dynamically adjust the audio compression ratio Cb according to the network bandwidth Bw;

[0191] The audio compression ratio Cb is obtained using the following formula:

[0192] ;

[0193] In the formula, Bwmax represents the maximum network bandwidth, Cmin represents the minimum compression ratio, and Cmax represents the maximum compression ratio.

[0194] S32. Adjust the audio compression ratio Cb according to the network latency Dr. When the network latency Dr is greater than the preset latency threshold Tdr, the data transmission delay will increase. At this time, it is necessary to adjust the compression ratio to reduce the impact of latency.

[0195] The formula for adjusting the network latency Dr to the audio compression ratio Cb is as follows:

[0196] ;

[0197] In the formula, α represents the adjustment factor, and Dmax represents the maximum network latency.

[0198] S4 includes S41 and S42;

[0199] S41. Using a time series regression prediction algorithm, based on historical audio traffic data F(t-1), predict the future audio traffic fluctuation F;

[0200] By combining historical audio traffic data F(t-1), network bandwidth Bw, packet loss rate Pd with parameter set CW and audio compression ratio Cb, the traffic fluctuation F(t) at future time t can be predicted.

[0201] ;

[0202] In the formula, Represents the autoregressive coefficient. The coefficient representing the impact of bandwidth variation. This represents the influence coefficient of packet loss rate. Let F(t) represent the influence coefficient of the compression ratio, F(t-1) represent the audio traffic at time t, F(t-1) represent the audio traffic at time t-1, Bw(t) represent the network bandwidth at time t, Pd(t) represent the packet loss rate at time t, and CWi represent the i-th data item in the parameter set CW. This represents the preset weight value of the i-th data item in the parameter set CW.

[0203] In this embodiment, the audio compression ratio Cb is dynamically adjusted based on network bandwidth Bw and network latency Dr, achieving efficient optimization during audio transmission. When bandwidth is high, a lower audio compression ratio Cb is used to ensure audio quality; conversely, when network latency is high, the compression ratio is appropriately increased to reduce the amount of data transmitted, thereby effectively reducing transmission latency. This method ensures a balance between audio quality and transmission latency under different network conditions, avoiding the problem of audio quality degradation when network bandwidth is insufficient, and preventing excessive latency from affecting the smoothness of voice calls.

[0204] This embodiment introduces a time-series regression prediction algorithm, which accurately predicts future audio traffic fluctuations based on historical audio traffic data, network bandwidth (Bw), packet loss rate (Pd), and audio compression ratio (Cb). Through this prediction, the system can anticipate changes in audio data traffic trends and dynamically select low-latency transmission routes based on the prediction results, avoiding audio transmission interruptions or stuttering caused by network bandwidth fluctuations or sudden increases in latency. In this way, changes in audio traffic fluctuations and network conditions can be addressed in advance, effectively ensuring the stability and continuity of real-time audio transmission. The impact of latency is reduced by adjusting the audio compression ratio (Cb), especially when network latency is high. The system can automatically adjust the audio compression ratio according to a preset latency threshold. Through this mechanism, the system can reduce the amount of data in high-latency network environments, thereby reducing audio transmission latency and avoiding the impact of high latency on audio smoothness in traditional transmission methods. Furthermore, dynamically adjusting the compression ratio and transmission parameters optimizes the balance between audio quality and latency, improving the user's audio experience.

[0205] This embodiment combines an adaptive compression algorithm with a real-time traffic prediction system, enabling the audio transmission system to achieve optimal performance under various network environments. Whether in a high-bandwidth network environment or a low-bandwidth or high-latency environment, the system can dynamically adjust parameters such as audio compression ratio, transmission rate, and data packet size based on real-time network conditions, thereby ensuring lossless and low-latency audio transmission. This high adaptability and robustness significantly improves the system's stability under changing network conditions, reducing common problems such as packet loss and stuttering. By combining factors such as network bandwidth (Bw), packet loss rate (Pd), and audio compression ratio (Cb), and using an autoregressive model to predict future audio traffic fluctuations, the system can adjust audio transmission parameters promptly based on bandwidth changes and packet loss rate fluctuations. This allows the system to accurately adjust audio traffic and transmission paths when facing bandwidth limitations or network fluctuations, ensuring the integrity of audio data and reducing audio quality degradation caused by network condition fluctuations.

[0206] Example 4

[0207] This embodiment is an explanation based on Embodiment 3. Please refer to it. Figure 1 Specifically: S42, based on the traffic fluctuation F(t) at the obtained future time t, dynamically select a low-latency transmission route;

[0208] Assuming there are N paths, the network bandwidth and network latency of each path are defined, and the comprehensive routing score Rn is calculated based on the predicted traffic fluctuation F(t) at future time t, the latency and bandwidth of each path, and the transmission route is selected.

[0209] The overall routing score Rn is obtained using the following formula:

[0210] ;

[0211] In the formula, nBw represents the network bandwidth of path n, nDr represents the network latency of path n, and Fmax represents the maximum predicted traffic value.

[0212] S5. By analyzing the traffic fluctuation F, determine the amount of redundant data that needs to be added, obtain the redundancy ratio Rf, and judge the redundancy ratio Rf based on the traffic fluctuation F.

[0213] Traffic fluctuation F reflects the volatility of audio data during transmission. When the traffic fluctuation F is greater than the preset fluctuation threshold Trf, it indicates that the network is unstable and the packet loss rate Pd will increase.

[0214] The redundancy ratio Rf is obtained using the following formula:

[0215] ;

[0216] In the formula, B1 represents the influence coefficient of audio traffic fluctuation on the redundancy ratio;

[0217] The redundancy ratio Rf is determined by matching in the following way:

[0218] When the flow fluctuation F is greater than the fluctuation threshold Trf, it indicates that an abnormal fluctuation has occurred. As the flow fluctuation F increases, the redundancy ratio Rf will also increase.

[0219] When the flow fluctuation F ≤ the fluctuation threshold Trf, it indicates that there is no abnormal fluctuation. When the flow fluctuation F decreases, the redundancy ratio Rf will also decrease.

[0220] Based on the calculated redundancy ratio Rf, redundant data is added to each audio data packet. The redundant data is then embedded into the audio data packet using forward error correction (FEC) technology to obtain the total data volume Dto.

[0221] The total data volume Dto is obtained using the following formula:

[0222] ;

[0223] In the formula, Dau represents the size of the original audio data packet.

[0224] This embodiment introduces a mechanism for dynamically selecting low-latency transmission routes based on future traffic fluctuation predictions. By evaluating the bandwidth and latency of each available path and combining this with traffic fluctuation predictions, the system can automatically select the path with the lowest latency and most suitable bandwidth for audio data transmission. Compared to traditional fixed routing or simple static path selection methods, this improvement can adapt to changes in network conditions in real time, avoiding audio transmission quality degradation due to network congestion or excessive latency. It ensures that audio transmission latency is kept as low as possible, especially when network conditions change rapidly, effectively mitigating the negative impact of network fluctuations.

[0225] By analyzing traffic fluctuations in real time, the system intelligently adjusts the redundancy ratio Rf and adds redundant data using Forward Error Correction (FEC) technology to improve the recovery capability of audio data. When traffic fluctuations are large, indicating network instability, the system automatically increases redundant data to ensure that audio data can be recovered even if packet loss occurs during transmission. This redundancy adjustment mechanism based on traffic fluctuations greatly improves the robustness of audio transmission, avoiding audio quality loss or transmission interruption caused by network problems, and ensuring the continuity and stability of audio calls. By monitoring audio traffic fluctuations and network status in real time, the system dynamically determines the amount of redundant data. When the network is stable, the system reduces the redundancy ratio, thereby reducing unnecessary data transmission and improving efficiency; while when network conditions deteriorate, the system automatically increases redundant data to enhance the reliability and recovery capability of audio data. This mechanism effectively avoids bandwidth waste caused by excessive redundant data, and at the same time, it can maximize the protection of audio data from loss when the network is unstable.

[0226] By combining traffic fluctuation prediction and redundant data adjustment strategies, this embodiment can intelligently allocate bandwidth resources based on the real-time network status. When network conditions are good, the system automatically reduces redundant data, thereby improving bandwidth utilization efficiency; while in unstable network conditions, it increases redundant data to ensure the integrity and recovery capability of data transmission. This strategy not only improves the stability of audio data transmission but also utilizes network bandwidth more efficiently, avoiding bandwidth waste caused by excessive redundancy in traditional methods. By combining audio traffic prediction and real-time redundant data adjustment, forward error correction (FEC) technology is implemented, effectively solving the packet loss problem. Regardless of network bandwidth stability, the system can ensure the complete transmission of audio data through appropriate redundant data, improving the reliability of audio communication in suboptimal network environments. This technology is particularly suitable for applications with high audio quality requirements, such as remote conferencing and online education, ensuring clear and high-quality audio transmission even under unstable network conditions.

[0227] Example 5

[0228] This embodiment is an explanation based on Embodiment 4. Please refer to it. Figure 1 and Figure 4 Specifically: S6. Calculate and obtain the comprehensive transmission quality index Q based on the obtained network dataset WS, traffic fluctuation F, and redundancy ratio Rf;

[0229] The overall transmission quality index Q is obtained using the following formula, as shown in Table 1:

[0230] ;

[0231] In the formula, F(t) represents the flow fluctuation at time t, ln represents the logarithmic function, e represents a constant, d1 represents the redundancy gain coefficient, and d2 represents the delay exponential decay factor.

[0232] The obtained comprehensive transmission quality index Q is compared with the preset quality threshold TQ to determine whether the transmission quality meets expectations.

[0233] The expected transmission quality is obtained by matching it in the following ways:

[0234] When the overall transmission quality index Q ≥ quality threshold TQ, it indicates that the transmission quality is normal and meets expectations.

[0235] When the overall transmission quality index Q < the quality threshold TQ, it indicates that the transmission quality is abnormal and has not met expectations. It is necessary to adjust the audio transmission rate Rsv and the redundancy ratio Rf to obtain a new audio transmission rate nRsv and a new redundancy ratio nRf.

[0236] The new audio transmission rate nRsv is obtained using the following formula:

[0237] ;

[0238] The new redundancy ratio nRf is obtained using the following formula:

[0239] ;

[0240] In the formula, X represents the adjustment factor, Rmax represents the maximum value of the redundancy ratio, and min(,) represents the minimum function.

[0241] Specific examples:

[0242] Set the quality threshold TQ = 5.0;

[0243] Table 1: Calculation Table of Comprehensive Transmission Quality Index for Audio Transmission Systems;

[0244] Group number Bw Pd Dr F(t) Rf Q Did it meet expectations? 1 0.6 0.12 0.18 0.7 0.35 1.76 Did not meet expectations 2 0.7 0.1 0.15 0.6 0.3 2.97 Did not meet expectations 3 0.8 0.08 0.2 0.55 0.25 4.28 Did not meet expectations 4 0.9 0.05 0.1 0.5 0.2 7.69 Expectations met

[0245] In this embodiment, by calculating the comprehensive transmission quality index Q and comparing it with a preset quality threshold TQ, the quality of audio transmission can be monitored in real time. This mechanism ensures that any anomalies during audio transmission, such as excessive latency, packet loss, or audio quality degradation, can be detected promptly. If the comprehensive transmission quality index Q is less than the quality threshold TQ, the system will automatically adjust and optimize the audio transmission rate Rsv and redundancy ratio Rf to maintain audio quality. Compared to traditional manual adjustment methods, this invention achieves real-time quality assessment and automatic optimization, greatly improving the stability and quality of audio transmission.

[0246] By calculating and adjusting the overall transmission quality index Q, the system can automatically respond to challenges posed by different network conditions. Whether it's network fluctuations, bandwidth changes, packet loss, or latency issues, the system can make corresponding adjustments to ensure high-quality audio transmission under various network environments. Especially when network conditions are unstable, the system can increase redundancy and adjust the transmission rate to avoid audio loss or quality degradation, providing a more reliable audio transmission service. When transmission quality does not meet expectations, the system optimizes transmission quality by automatically adjusting the audio transmission rate Rsv and redundancy ratio Rf. It dynamically selects and acquires new audio transmission rate nRsv and new redundancy ratio nRf to avoid affecting audio quality due to excessively high or low parameter settings. This adaptive adjustment strategy ensures that audio transmission maintains optimal performance under different network conditions.

[0247] The improvements in this embodiment not only enhance the quality and stability of audio transmission but also significantly improve the user experience. Users of real-time audio communication applications will not experience audio interruptions or distortion due to network issues such as network jitter and bandwidth limitations. The system's intelligent optimization ensures that audio quality is virtually unaffected by network fluctuations, providing users with a clearer and smoother voice experience. Furthermore, the system's flexibility allows it to be widely applied to different network environments, such as wireless networks and mobile networks, ensuring the quality and reliability of audio communication.

[0248] Example 6

[0249] A low-latency lossless digital audio transmission system, please refer to Figure 2 Specifically, it includes a real-time network environment assessment module, a dynamic audio transmission parameter adjustment module, an intelligent audio compression and decompression module, an audio traffic prediction and routing module, a forward error correction and redundancy adjustment module, and a real-time feedback and adjustment module.

[0250] The real-time network environment assessment module monitors the current network environment data through network detection technology and network tools, fits it into the original dataset YS, and performs preprocessing to obtain the network dataset WS;

[0251] The dynamic audio transmission parameter adjustment module dynamically adjusts the audio transmission parameters, including audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf, based on the network dataset WS, and fits them into a parameter set CW.

[0252] The intelligent audio compression and decompression module adjusts the audio compression ratio Cb by using an adaptive compression algorithm based on the network bandwidth Bw and network latency Dr in the network dataset WS.

[0253] The audio traffic prediction and routing module, based on the parameter set CW and the audio compression ratio Cb, uses a time series regression prediction algorithm to predict the traffic fluctuation F of audio traffic and dynamically select low-latency transmission routes.

[0254] The forward error correction and redundancy adjustment module combines the acquired traffic fluctuation F with the forward error correction technology FEC, and adds redundant data to the audio data packets through encoding to obtain the redundancy ratio Rf.

[0255] The real-time feedback and adjustment module calculates and obtains the comprehensive transmission quality index Q based on the acquired network dataset WS, traffic fluctuation F, and redundancy ratio Rf.

[0256] In this embodiment, the real-time network environment assessment module continuously monitors the current network bandwidth (Bw), network latency (Dr), packet loss rate (Pd), and network jitter (Wd) using advanced network detection technologies and tools, and fits these parameters into an original dataset. After data cleaning and standardization, a network dataset (WS) is generated, providing reliable real-time data support for subsequent audio transmission optimization. By dynamically adjusting the audio transmission rate (Rsv), packet size (Spd), and redundancy coding ratio (Rbf), the system can optimize transmission parameters in real time according to changes in the network environment. This adaptive adjustment mechanism ensures efficient audio data transmission in high-bandwidth environments, while automatically reducing packet size and transmission rate when network conditions are poor to avoid audio loss or transmission interruption. This optimization method improves the efficiency and stability of audio transmission, performing particularly well in complex network environments.

[0257] Based on an adaptive audio compression algorithm that considers network bandwidth (Bw) and network latency (Dr), the system dynamically adjusts the audio compression ratio (Cb) according to current network conditions. This improvement allows the system to retain more audio details and ensure lossless sound quality when bandwidth is sufficient, while increasing the compression ratio and reducing data volume to minimize transmission latency when bandwidth is limited. The application of intelligent audio compression and decompression technology ensures that audio quality is guaranteed regardless of changes in network conditions, avoiding the risk of sound quality loss due to over-compression in traditional methods.

[0258] This embodiment uses a time-series regression prediction algorithm to predict audio traffic fluctuations and dynamically selects the optimal low-latency transmission route based on the prediction results. This prediction and routing mechanism ensures that the system can automatically switch to a higher-quality transmission path when network congestion or high latency occurs, thereby guaranteeing the stability and low latency of audio data. Through this intelligent routing selection, the system can effectively avoid audio stuttering or packet loss caused by network changes, greatly improving the continuity and real-time performance of audio transmission. During audio data transmission, the system combines forward error correction (FEC) technology with traffic fluctuation prediction to add appropriate redundant data to audio data packets, ensuring that even if packet loss occurs, the receiving end can still recover complete audio data. By adjusting the redundancy ratio according to real-time traffic fluctuations, the system can flexibly respond to network fluctuations, avoid excessive redundant data leading to bandwidth waste, and ensure the integrity and stability of audio transmission. This redundancy adjustment mechanism improves the reliability of the system, ensuring lossless audio transmission even in low-quality network environments.

[0259] Example 7

[0260] A computer storage medium storing a computer program, which, when executed by a processor, implements a method for low-latency lossless digital audio transmission, as described above. Figure 3 .

[0261] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A low-latency lossless digital audio transmission method, characterized in that: Includes the following steps: S1. By using network probing techniques and network tools, monitor the current network environment data, fit it into the original dataset YS, and perform preprocessing to obtain the network dataset WS; S2. Based on the network dataset WS, dynamically adjust the audio transmission parameters, including audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf, and fit them into a parameter set CW. S3. Based on the network bandwidth Bw and network latency Dr in the network dataset WS, adjust the audio compression ratio Cb using an adaptive compression algorithm. S4. Based on the parameter set CW and the audio compression ratio Cb, the audio traffic fluctuation F is predicted by using a time series regression prediction algorithm, and a low-latency transmission route is dynamically selected. S5. Based on the obtained traffic fluctuation F, the forward error correction technology FEC is combined to add redundant data to the audio data packet through encoding, and the redundancy ratio Rf is obtained. S6. Calculate the comprehensive transmission quality index Q based on the obtained network dataset WS, traffic fluctuation F, and redundancy ratio Rf.

2. The low-latency lossless digital audio transmission method according to claim 1, characterized in that: S1 includes S11 and S12; S11. Using network probing techniques and network tools, monitor current network environment data, including network bandwidth Bw, network latency Dr, packet loss rate Pd, and network jitter Wd, and fit the data to obtain the original dataset YS. Among them, network bandwidth Bw was collected using network measurement tools iperf, netflow, and routers; Network latency Dr was obtained through latency testing tools ping, traceroute, and router data. Packet loss rate Pd is obtained through packet loss detection tools ping and iperf, and data collected from the switch. Network jitter (Wd) is collected using latency and packet loss monitoring tools ping and iperf. S12. Perform data cleaning and standardization on the obtained raw dataset YS to obtain the network dataset WS; Data cleaning includes removing incomplete and outlier values ​​from the original dataset YS, processing missing data packets, and removing noisy data. Standardization processing includes standardizing the data in the original dataset YS, converting network data of different dimensions and units into dimensionless data, and obtaining the network dataset WS; The network dataset WS is obtained using the following formula: ； In the formula, WSd represents the d-th data item in the network dataset WS, YSd represents the d-th data item in the original dataset YS, μYSd represents the mean of the d-th data item in the original dataset YS, and σYSd represents the standard deviation of the d-th data item in the original dataset YS.

3. The low-latency lossless digital audio transmission method according to claim 2, characterized in that: S2 includes S21 and S22; S21. Dynamically adjust the audio transmission rate Rsv based on the network bandwidth Bw and packet loss rate Pd; The audio transmission rate Rsv is obtained using the following formula: ； In the formula, C1 represents a constant; S22. Dynamically adjust the audio data packet size Spd based on network bandwidth Bw and network latency Dr; The packet size Spd is obtained using the following formula: ； In the formula, C2 represents the proportional constant that controls the size of the data packet; Adjust the redundancy coding ratio Rbf according to the packet loss rate Pd and network jitter Wd; The redundancy coding ratio Rbf is obtained using the following formula: ； In the formula, C3 represents the redundancy coding ratio coefficient; The obtained audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf are fitted to obtain the parameter set CW.

4. The low-latency lossless digital audio transmission method according to claim 3, characterized in that: S3 includes S31 and S32; S31. Dynamically adjust the audio compression ratio Cb according to the network bandwidth Bw; The audio compression ratio Cb is obtained using the following formula: ； In the formula, Bwmax represents the maximum network bandwidth, Cmin represents the minimum compression ratio, and Cmax represents the maximum compression ratio. S32. Adjust the audio compression ratio Cb according to the network latency Dr. When the network latency Dr is greater than the preset latency threshold Tdr, the data transmission delay will increase. At this time, it is necessary to adjust the compression ratio to reduce the impact of latency. The formula for adjusting the network latency Dr to the audio compression ratio Cb is as follows: ； In the formula, α represents the adjustment factor, and Dmax represents the maximum network latency.

5. The low-latency lossless digital audio transmission method according to claim 4, characterized in that: S4 includes S41 and S42; S41. Using a time series regression prediction algorithm, based on historical audio traffic data F(t-1), predict the future audio traffic fluctuation F; By combining historical audio traffic data F(t-1), network bandwidth Bw, packet loss rate Pd with parameter set CW and audio compression ratio Cb, the traffic fluctuation F(t) at future time t can be predicted. ； In the formula, Represents the autoregressive coefficient. The coefficient representing the impact of bandwidth variation. This represents the influence coefficient of packet loss rate. Let F(t) represent the influence coefficient of the compression ratio, F(t-1) represent the audio traffic at time t, F(t-1) represent the audio traffic at time t-1, Bw(t) represent the network bandwidth at time t, Pd(t) represent the packet loss rate at time t, and CWi represent the i-th data item in the parameter set CW. This represents the preset weight value of the i-th data item in the parameter set CW.

6. The low-latency lossless digital audio transmission method according to claim 5, characterized in that: S42. Dynamically select a low-latency transmission route based on the traffic fluctuation F(t) at the obtained future time t; Assuming there are N paths, the network bandwidth and network latency of each path are defined, and the comprehensive routing score Rn is calculated based on the predicted traffic fluctuation F(t) at future time t, the latency and bandwidth of each path, and the transmission route is selected. The overall routing score Rn is obtained using the following formula: ； In the formula, nBw represents the network bandwidth of path n, nDr represents the network latency of path n, and Fmax represents the maximum predicted traffic value.

7. The low-latency lossless digital audio transmission method according to claim 6, characterized in that: S5. By analyzing the traffic fluctuation F, determine the amount of redundant data that needs to be added, obtain the redundancy ratio Rf, and judge the redundancy ratio Rf based on the traffic fluctuation F. Traffic fluctuation F reflects the volatility of audio data during transmission. When the traffic fluctuation F is greater than the preset fluctuation threshold Trf, it indicates that the network is unstable and the packet loss rate Pd will increase. The redundancy ratio Rf is obtained using the following formula: ； In the formula, B1 represents the influence coefficient of audio traffic fluctuation on the redundancy ratio; The redundancy ratio Rf is determined by matching in the following way: When the flow fluctuation F is greater than the fluctuation threshold Trf, it indicates that an abnormal fluctuation has occurred. As the flow fluctuation F increases, the redundancy ratio Rf will also increase. When the flow fluctuation F ≤ the fluctuation threshold Trf, it indicates that there is no abnormal fluctuation. When the flow fluctuation F decreases, the redundancy ratio Rf will also decrease. Based on the calculated redundancy ratio Rf, redundant data is added to each audio data packet. The redundant data is then embedded into the audio data packet using forward error correction (FEC) technology to obtain the total data volume Dto. The total data volume Dto is obtained using the following formula: ； In the formula, Dau represents the size of the original audio data packet.

8. The low-latency lossless digital audio transmission method according to claim 7, characterized in that: S6. Calculate the comprehensive transmission quality index Q based on the obtained network dataset WS, traffic fluctuation F, and redundancy ratio Rf. The overall transmission quality index Q is obtained using the following formula: ； In the formula, F(t) represents the flow fluctuation at time t, ln represents the logarithmic function, e represents a constant, d1 represents the redundancy gain coefficient, and d2 represents the delay exponential decay factor. The obtained comprehensive transmission quality index Q is compared with the preset quality threshold TQ to determine whether the transmission quality meets expectations. The expected transmission quality is obtained by matching it in the following ways: When the overall transmission quality index Q ≥ quality threshold TQ, it indicates that the transmission quality is normal and meets expectations. When the overall transmission quality index Q < the quality threshold TQ, it indicates that the transmission quality is abnormal and has not met expectations. It is necessary to adjust the audio transmission rate Rsv and the redundancy ratio Rf to obtain a new audio transmission rate nRsv and a new redundancy ratio nRf. The new audio transmission rate nRsv is obtained using the following formula: ； The new redundancy ratio nRf is obtained using the following formula: ； In the formula, X represents the adjustment factor, Rmax represents the maximum value of the redundancy ratio, and min() represents the minimum function.

9. A low-latency lossless digital audio system, characterized in that: It includes a real-time network environment assessment module, a dynamic audio transmission parameter adjustment module, an intelligent audio compression and decompression module, an audio traffic prediction and routing module, a forward error correction and redundancy adjustment module, and a real-time feedback and adjustment module. The real-time network environment assessment module monitors the current network environment data through network detection technology and network tools, fits it into the original dataset YS, and performs preprocessing to obtain the network dataset WS; The dynamic audio transmission parameter adjustment module dynamically adjusts the audio transmission parameters, including audio transmission rate Rsv, data packet size Spd, and redundancy coding ratio Rbf, based on the network dataset WS, and fits them into a parameter set CW. The intelligent audio compression and decompression module adjusts the audio compression ratio Cb by using an adaptive compression algorithm based on the network bandwidth Bw and network latency Dr in the network dataset WS. The audio traffic prediction and routing module, based on the parameter set CW and the audio compression ratio Cb, uses a time series regression prediction algorithm to predict the traffic fluctuation F of audio traffic and dynamically select low-latency transmission routes. The forward error correction and redundancy adjustment module combines the acquired traffic fluctuation F with the forward error correction technology FEC, and adds redundant data to the audio data packets through encoding to obtain the redundancy ratio Rf. The real-time feedback and adjustment module calculates and obtains the comprehensive transmission quality index Q based on the acquired network dataset WS, traffic fluctuation F, and redundancy ratio Rf.

10. A computer storage medium storing a computer program thereon, characterized in that: When the program is executed by the processor, it implements the low-latency lossless digital audio transmission method as described in any one of claims 1-8.

Images