WO2020244327A1

WO2020244327A1 - Data processing method and device

Info

Publication number: WO2020244327A1
Application number: PCT/CN2020/086298
Authority: WO
Inventors: 杨璐; 范志刚
Original assignee: 西安万像电子科技有限公司
Priority date: 2019-06-05
Filing date: 2020-04-23
Publication date: 2020-12-10
Also published as: CN110299963A

Abstract

The present disclosure provides a data processing method and device, relating to the technical field of data error correction code, for solving the problem of current multimedia having relatively simplex forward error correction scheme. The specific technical solution comprises: acquiring data priority of the multimedia and the current network state data during transmission; determining the corresponding redundancy model parameter according to the data priority and the current network state data; and performing error correction coding on the multimedia data according to the redundancy model parameter and sending the multidedia data. According to the disclosure, a technical means of dynamically selecting a redundant model is used for adaptive forward error correction during multimedia transmission.

Description

Data processing method and device

Technical field

The present disclosure relates to the technical field of data error correction codes, and in particular to data processing methods and devices.

Background technique

With the popularization of the Internet and the development of multimedia technology, the application of multimedia technology has become more and more extensive. People usually use voice and video for real-time communication, or upload live videos on the Internet to interact with viewers. At present, multimedia services usually package the code stream data according to fixed rules and send it in the form of data packets.

There are many existing forward error correction methods. The existing previous error correction or redundancy algorithms can basically be divided into three generations of algorithms. The first generation algorithm uses a primary redundancy model, and the error tolerance within the unit data length is low, which is not suitable for large Data transfer. The second-generation algorithm introduces two or more redundant models. However, the algorithm itself is relatively complex, which is not conducive to the calculation and transmission of data such as real-time images when computing resources are limited. The third-generation algorithm requires hardware support for scenarios such as optical fiber data transmission. It is not suitable for the transmission of multimedia data in the existing complicated wired or wireless network under the condition of limited computing resources. The above algorithms are only applicable to specific scenarios, and are not well adapted to the transmission of real-time multimedia data under complex networks.

Summary of the invention

The embodiments of the present disclosure provide a forward error correction method and device, which can solve the problem that the current error correction method cannot be adaptive. The technical solution is as follows:

According to a first aspect of the embodiments of the present disclosure, there is provided a data processing method, the method including:

Acquiring the data priority of the multimedia and current network state data during transmission;

Determine the corresponding redundancy model parameters according to the data priority and current network state data;

Perform error correction coding on the multimedia data according to the redundancy model parameter and send it.

In a possible implementation manner, the acquiring the data priority of the multimedia includes:

The priority of the multimedia data is determined according to the data classification.

In a possible implementation manner, the current network status data includes at least one of the following: packet loss rate, packet error rate, network delay, round-trip time (RTT), life cycle (Time To Live, TTL).

In a possible implementation manner, the determining the corresponding redundancy model parameters according to the data priority and current network state data includes:

Determine the algorithm type adopted by the redundancy model according to the data priority;

Determine the number of effective packets m and the number of redundant packets n of the transmission data according to the network state data.

In a possible implementation manner, the algorithm type used for determining the redundancy model according to the data priority includes:

When the multimedia data is control signaling data, determining that the redundancy model is a primary redundancy model;

When the multimedia data is service signaling data, the amount of data is less than a preset threshold and real-time requirements are high, determining that the redundancy model is a primary redundancy model;

When the multimedia data is service signaling, the amount of data is greater than a preset threshold and real-time requirements are low, and the redundancy model is determined to be a secondary or multiple redundancy model.

In a possible implementation manner, the method further includes:

Check the current hardware and software conditions;

When the current hardware and software conditions meet the preset conditions, select the high-order redundancy model.

According to a second aspect of the embodiments of the present disclosure, there is provided a data processing method, the method including:

Parse the received multimedia data;

Obtaining a corresponding redundancy model according to the multimedia data;

In the case of detecting packet loss or error, the multimedia data is restored according to the redundancy model.

In a possible implementation manner, the method further includes: when the current network state data changes, feeding back the current network data to the sender.

According to a third aspect of the embodiments of the present disclosure, there is provided a data processing device, the device including:

An acquiring unit, configured to acquire the data priority of the multimedia and current network state data during transmission;

A determining unit, configured to determine corresponding redundancy model parameters according to the data priority and current network state data;

The encoding sending unit is configured to perform error correction encoding on the multimedia data according to the redundancy model parameter and send it.

In a possible implementation manner, the acquiring unit is specifically configured to determine the priority of the multimedia data according to data classification.

In a possible implementation manner, the determining unit is specifically configured to:

In a possible implementation manner, the data processing device further includes a detection unit for detecting the current hardware and software conditions; when the current software and hardware conditions meet the preset conditions, the high-order redundancy model is selected.

According to a fourth aspect of the embodiments of the present disclosure, there is provided a data processing device, the device including:

The parsing unit is used to analyze the received multimedia data;

Obtaining a corresponding redundancy model according to the multimedia data;

The detection unit is configured to detect current network state data according to the redundancy model;

The recovery unit, in the case of detecting packet loss, recovers multimedia data according to the redundancy model.

In a possible implementation manner, the data processing device further includes a feedback unit configured to feed back the current network data to the sender when the current network state data changes.

The forward error correction method and device provided in the embodiments of the present disclosure solve the problem that the forward error correction cannot be adaptive by determining the corresponding redundancy model parameters according to the data priority and current network state data. The present invention mainly designs a dynamic forward error correction method for real-time multimedia data transmission, and can recover lost data according to an error correction algorithm without retransmitting data. According to the network status and real-time transmission data volume and data priority, the present invention adopts the method of dynamic redundancy model selection, selects different models, reduces the amount of calculation, and guarantees more effectiveness with less redundant data. Data transmission is correct, suitable for mixed transmission of video, audio, control and other data.

Therefore, the present invention adopts a dynamic redundancy model, introduces multiple redundancy algorithms with lower complexity, introduces network features, data features, computing resource features, and dynamically selects different redundancy methods in real time, not limited to primary and secondary redundancy. For the remaining models, multiple redundancy algorithms can be selected to form three or more redundancy models to adapt to the complex and changeable network environment and hardware environment. It does not cause waste of resources, and at the same time can improve the accuracy of data transmission.

It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure.

Description of the drawings

The drawings herein are incorporated into the specification and constitute a part of the specification, show embodiments in accordance with the disclosure, and together with the specification are used to explain the principle of the disclosure.

FIG. 1 is a first schematic flowchart of a forward error correction method provided by an embodiment of the present disclosure;

FIG. 2 is a second schematic flowchart of a forward error correction method provided by an embodiment of the present disclosure;

FIG. 3 is a first structural diagram of a forward error correction device provided by an embodiment of the present disclosure;

4 is a second structural diagram of a forward error correction device provided by an embodiment of the present disclosure;

Fig. 5 is a schematic diagram of a redundancy model provided by an embodiment of the present invention.

Detailed ways

Here, exemplary embodiments will be described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present disclosure. Rather, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

The embodiment of the present disclosure provides a data processing method. As shown in FIG. 1, the processing method includes the following steps:

101. Acquire the data priority of the multimedia and current network state data during transmission;

In an embodiment, the obtaining the data priority of the multimedia includes: determining the priority of the multimedia data according to a data classification strategy.

For example: According to the data classification strategy, for example, the data can be classified according to the type of data. For example, the data can be classified into control signaling data required to transmit audio and video, video data, audio data, etc. Among them, control signaling data has the highest priority, audio data has the second priority, and video data has the lowest priority. The data volume of control signaling data required to transmit audio and video is the smallest, followed by the data volume of audio data, and the data volume of video data is the largest.

In one embodiment, the current network state data includes at least packet loss rate, packet error rate, network delay, round-trip time (RTT), and time to live (TTL).

102. Determine corresponding redundancy model parameters according to the data priority and current network state data;

In an embodiment, the redundancy model parameters include at least: the number of valid messages m, the number of redundant messages n, and a redundancy check algorithm. Exemplarily, the redundant data format is shown in Figure 5. In essence, the n-th redundancy model is to solve the n-ary linear equations, and the lost packets are unknowns. Under the m+n model, as long as the number of m+n data packets in the current model is less than n, the equation can be solved according to The group mode is restored at the receiving end. Among them, the size of m and n is determined by data priority and network status.

Exemplarily, the redundancy check algorithm includes but is not limited to convolutional code, Hamming code, BCH code, RS code, Turbo code, LDPC code, TPC code, etc. In one embodiment, the selection of the specific verification algorithm depends on the computing capabilities of the software and hardware of the current operating environment.

103. Perform redundant encoding on the multimedia data according to the redundancy model parameter and send it.

The forward error correction method provided by the embodiments of the present disclosure solves the problem of a single forward error correction method in the background art by adopting a technical means of dynamically selecting a redundant model.

A specific embodiment is given below.

Exemplarily, assuming that the current network packet loss rate is 20%, it is simply described that at least 2 redundant packets are added for error correction for every 8 packets, and the effective coding rate is η=4/5. In order to ensure that the receiving end can recover the lost data (of course, the network is fluctuating, 20% is only a measured value, and it may be large or small in practice. The ideal value is used here to illustrate the principle). Introduce data priority and network delay. Important data must be transmitted first and real-time. Therefore, more redundancy is required and the effective coding rate is reduced to η=3/4.

According to the above conditions, after the redundancy is determined, there can be multiple redundancy model options. The following describes the primary and secondary redundancy models, and the calculation of higher-order models is similar.

Primary redundancy model: This redundancy model is suitable for the fastest algorithm in the current operating environment. Here is an example of a punctured convolutional code. When η=2/3, the values of m and n in Figure 5 are respectively 2, 1; each two messages generate a redundancy, namely

c is redundant. There are 3 cases in the transmission process, c is lost, because the original data of a and b are not lost, the result is correct; a or b is lost, according to the convolution algorithm

Both a and b can be restored; when two or more of the three packets of a, b, and c are lost, the data cannot be restored. At this time, it indicates that the packet loss rate is increased, and it needs to be fed back to the sender to update the redundancy model in time.

Quadratic redundancy model: The mathematical model is a binary linear equation, such as:

Among them a/d! =b/e. It can support the data recovery of the loss of two group members. When η=2/3, the values of m and n in Figure 4 are 4 and 2 respectively; therefore, two different redundancy check algorithms are required to generate the equation The irrelevance coefficient of the group. Here we take the punctured convolutional code and BCH code as examples. The finite field of the BCH code is selected as GF(2^8), and the primitive polynomial is taken as p(x)=x8+x4+x3+x2+1(0x11D); A total of 256 elements in the domain are determined. These 256 elements are correspondingly marked as d0, d1, d2...d255, which have an injective relationship with integers from 0 to 255; the results of the four arithmetic operations in this domain are still in this domain, so A two-dimensional table can be generated to store all the results, and the results can be obtained by directly checking the table during calculations, saving a lot of calculation.

The calculation redundancy of the convolutional code is denoted as p, and the calculation redundancy of the BCH code is denoted as q. The 4 valid messages are denoted as x0, x1, x2, x3.

As long as the number of missing (unknown) messages in m+n=6 messages is less than or equal to 2, the binary linear equations or the linear equations in one variable are satisfied, and the data can be recovered.

Actually, when each packet data is being transmitted, there are three situations as follows:

If there is no message loss or only one message is lost, you can directly use the one-time redundancy model. Choose one of the check code algorithms to increase the recoverable data. Generally, choose the one with faster calculation speed. BCH code is fast.

Two messages are lost, and the following three conditions exist

If both p and q are lost, the data area will not be affected;

If p, xn are lost, the p value will not be processed, assuming that the lost is x2, according to the definition of q value,

dn is the inverse element of d2, which can be obtained by looking up the table.

Two x are missing, assuming that the missing is x1, x3, according to the definition of p, q, the calculation process is as follows

dn is

The inverse element of the table is available.

When 3 or more data are lost, the data cannot be recovered. At this time, it indicates that the packet loss rate is increased, and it needs to be fed back to the sender to update the redundant model in time.

The third-order and above redundant models are similar to the above-mentioned first-order and second-order redundant models. In essence, they are solved by a multi-variable linear equation. The coefficients of each equation cannot be linearly correlated, so different check code algorithms are needed to generate this coefficient.

Optionally, when computing power allows, a higher-order redundancy model should be selected as far as possible. Among them, the high-order redundancy model is a redundancy model higher than three times. According to Shannon's theorem, under the same redundancy, the longer the block code length, the stronger the fault tolerance. But at the same time it will increase the average transmission delay, because the longer the code length, the more data must be received to calculate the lost data when recovering data. For example, when the network packet loss is very low, it is necessary to transmit a set of important and high real-time data. At this time, the redundancy model can be selected first to ensure that the data is accurate and not only has strong real-time performance, but also improves the calculation speed and saves Computing resource consumption.

With special hardware support, the present invention can also be extended to implement higher-order redundancy models.

The embodiment of the present disclosure provides a data processing method. As shown in FIG. 2, the processing method includes the following steps:

201. Parse the received multimedia data;

202. Obtain a corresponding redundancy model according to the multimedia data.

203. In the case of packet loss or error, restore the multimedia data according to the redundancy model.

Optionally, the method may further include: when the current network state data changes, feeding back the current network data to the sender.

An embodiment of the present disclosure provides a data processing device 30. As shown in FIG. 3, the processing device includes: an acquiring unit 301, configured to acquire the data priority of the multimedia and current network state data during transmission; and a determining unit 302, It is used to determine the corresponding redundancy model parameter according to the data priority and the current network state data; the encoding sending unit 303 is used to perform error correction coding on the multimedia data and send it according to the redundancy model parameter.

An embodiment of the present disclosure provides a data processing device 40. As shown in FIG. 4, the processing device includes a parsing unit 401 for parsing received multimedia data;

The calculation unit 402 is configured to calculate corresponding redundant model data according to the multimedia data;

The restoring unit 403 restores the multimedia data according to the redundant model data in the case of detecting packet loss or error.

In an embodiment, the forward error correction device may further include a feedback unit 404 for

Based on the forward error correction method described in the above embodiments, the embodiments of the present disclosure also provide a computer-readable storage medium. For example, the non-transitory computer-readable storage medium may be a read-only memory (English: Read Only Memory, ROM), random access memory (English: Random Access Memory, RAM), CD-ROM, magnetic tape, floppy disk and optical data storage device, etc. The storage medium stores computer instructions for executing the forward error correction method described in the above-mentioned embodiment corresponding to FIG. 1, which will not be repeated here.

After considering the specification and practicing the disclosure disclosed herein, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any variations, uses, or adaptive changes of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or conventional technical means in the technical field not disclosed in the present disclosure . The description and embodiments are only regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the claims.

Claims

A data processing method, characterized in that the method includes:

Acquiring the data priority of the multimedia and current network state data during transmission;

Determine the corresponding redundancy model parameters according to the data priority and current network state data;

Perform error correction coding on the multimedia data according to the redundancy model parameter and send it.
The method according to claim 1, wherein said acquiring the data priority of said multimedia comprises:

The priority of the multimedia data is determined according to the data classification.
The method according to claim 1, wherein the current network state data at least includes at least one of the following: packet loss rate, packet error rate, network delay, round trip delay RTT, and life cycle TTL.
The method according to any one of claims 2 or 3, wherein the determining the corresponding redundancy model parameter according to the data priority and current network state data comprises:

Determine the algorithm type adopted by the redundancy model according to the data priority;

Determine the number of effective packets m and the number of redundant packets n of the transmission data according to the network state data.
The method according to claim 4, wherein the determining the algorithm type adopted by the redundancy model according to the data priority comprises:

When the multimedia data is control signaling data of audio and video data, determining that the redundancy model is a primary redundancy model;

When the multimedia data is audio data, determining that the redundancy model is a primary redundancy model;

When the multimedia data is video data, it is determined that the redundancy model is a secondary or multiple redundancy model.
The method of claim 5, wherein the method further comprises:

Check the current hardware and software conditions;

When the current hardware and software conditions meet the preset conditions, select the high-order redundancy model.
A data processing method, characterized in that the method includes:

Parse the received multimedia data;

Obtaining a corresponding redundancy model according to the multimedia data;

In the case of packet loss or error, the multimedia data is restored according to the redundancy model.
The data processing method according to claim 7, wherein the method further comprises: when the current network state data changes, feeding back the current network data to the sender.
A data processing device, characterized in that the device includes:

An acquiring unit, configured to acquire the data priority of the multimedia and current network state data during transmission;

A determining unit, configured to determine corresponding redundancy model parameters according to the data priority and current network state data;

The encoding sending unit is configured to perform error correction encoding on the multimedia data according to the redundancy model parameter and send it.
A data processing device, characterized in that the device includes:

The parsing unit is used to analyze the received multimedia data;

A calculation unit, configured to calculate corresponding redundant model data according to the multimedia data;

The restoring unit restores the multimedia data according to the redundant model data in the case of detecting packet loss or error.