RU2764784C1 - Method for minimising the latency with guaranteed transmission of a packeted smoothed stream of digital compressed images - Google Patents

Abstract

FIELD: communication.

SUBSTANCE: invention relates to the field of optimisation of the size of data packets and can be used to guarantee the quality of service (QoS) requirements when transmitting digital compressed image traffic. Proposed is a method for minimising the delay with guaranteed transmission of a packeted smoothed stream of digital compressed images. The method includes a stage of setting the parameters of the input stream. The parameters of traffic processing apparatuses are then set, the curve of the input stream is formed. Also formed is the curve of the communication network service. The minimum playback delay is calculated, the optimal size of the useful part of the packet is determined.

EFFECT: ensured minimisation of the delay with guaranteed transmission of a packeted smoothed stream of digital compressed images for the purpose of improving the quality of telecommunication traffic service in conditions of optimal smoothing without losses, due to determining the optimal size of the useful part of the packet based on the criterion of minimum playback delay by the method for complete search of possible options, accounting for the supply curve of the input stream of digital compressed images and the end-to-end curve of the communication network service.

1 cl, 3 dwg, 1 tbl

Description

The invention relates to the field of optimizing the size of data packets and can be used to ensure quality of service (QoS) requirements when transmitting digital compressed image traffic.

The ability of this technical solution to work with any data transmission systems, regardless of the compression algorithms used, allows it to be used in any packet transmission systems for digital compressed images, which ensures its wide industrial applicability.

One of the features of modern data transmission systems is the wide distribution of multimedia traffic, which, as you know, is the most "heavy" and demanding in terms of delay and jitter.

Another feature is that packet protocols are widely used in data transmission systems. In this case, an important point is the optimization of the packet size, since this directly affects the quality of service of telecommunication traffic and, in particular, the delay in the transmission of digital compressed images caused by packetization.

The currently used mechanisms for ensuring the quality of service of telecommunications traffic most successfully cope with their functions in the absence of input stream ripples. In view of this, one of the approaches designed to provide maximum guarantees of quality of service is optimal lossless smoothing.

It should also be noted that most of the currently used streaming video transmission protocols do not take into account the delay introduced by packetization and use packet sizes during transmission due to the characteristics of telecommunication equipment, and not the requirement for maximum quality of service. In the context of the proposed method, the maximum quality of service is understood as the minimum delay in the guaranteed transmission of a packetized smoothed stream of digital compressed images.

Moreover, often the packet size for the transmission of all types of traffic in a particular system is fixed.

The methods described in RU 2459373, RU 2370907, RU 2601604, intended for packet data transmission, are focused on ensuring the compatibility of information exchange protocols or eliminating losses and implement various mechanisms for packetizing, routing or smoothing the input stream. At the same time, in these methods, the main emphasis is placed on the current state of the communication network, the delivery route, ensuring the absence of losses and the elimination of jitter.

As a result, analogous methods do not allow minimizing the delay caused by packetization, while eliminating packet loss due to buffer overruns, and preventing buffer underruns.

The closest in technical essence to the claimed method and selected as a prototype is the method described in Jean-Yves Le Boudec and Patrick Thiran. Network Calculus: A Theory of Deterministic Queuing Systems for the Internet. Online Version, December, 2019 (pp. 155-173).

The essence of this method lies in the fact that optimal smoothing of the input stream is performed in a bitwise telecommunication network, while ensuring that there are no losses due to buffer overflow and minimal requirements for transmission delay and allocated buffer memory size. For this:

1. The input stream arrival function is formed.

2. Produce min+convolution of the service curve of the communication network with the shape curve of the smoothing device.

3. The minimum size of the preview delay is determined as the maximum horizontal distance between the input stream arrival function and the curve resulting from the min+convolution of the communication network service curve with the smoother shape curve, which provides optimal smoothing without bitstream loss.

4. The minimum input stream arrival curve is determined by calculating the inverse min+convolution of the input stream function with itself.

5. The minimum playback delay is determined as the maximum horizontal distance between the minimum input stream arrival curve and the curve resulting from min+convolution of the communication network service curve with the curve of the shape of the smoothing device, which provides optimal smoothing without bitstream loss.

6. The minimum size of the playback buffer is determined as the maximum vertical distance between the minimum arrival curve of the input stream and the curve resulting from the min+convolution of the service curve of the communication network with the shape curve of the smoothing device, which provides optimal smoothing without bitstream loss.

7. The communication network is configured to take into account the minimum preview delay, the minimum playback delay, and the minimum playback buffer, which provide optimal bitstream lossless smoothing.

8. Transmit a bit stream.

The prototype method has the following disadvantages.

1. Do not take into account the packetization of the input stream, when obtaining the service curve of the communication network.

2. Do not minimize transmission delay under bundling conditions.

3. Do not take into account the increase in the amount of transmitted data caused by the addition of a header containing service information to each packet.

4. Do not take into account the stepwise nature of the input stream arrival function when transmitting digital compressed images.

5. Do not take into account the degree of compression of the frame-by-frame stream of digital images.

6. Does not take into account the frame rate when transmitting a frame-by-frame video stream.

7. When transmitting a frame-by-frame video stream, the image resolution is not taken into account.

The problem that the proposed method solves is to improve the quality of service of telecommunication traffic by minimizing the delay in the guaranteed transmission of a packetized smoothed stream of digital compressed images by determining the optimal packet size.

The technical result consists in minimizing the delay in the guaranteed transmission of a packetized smoothed stream of digital compressed images in order to improve the quality of service of telecommunication traffic under conditions of optimal smoothing without loss, by determining the optimal size of the useful part of the packet according to the criterion of minimum playback delay by the method of a complete enumeration of possible options, taking into account the arrival curve of the input stream of digital compressed images and the end-to-end service curve of the communication network.

Graphic materials used to illustrate the proposed solution:

1 is a flow diagram of a method for minimizing delay in the transmission of a packetized smoothed digital compressed image stream;

fig. 2 - dependence of the minimum playback delay, providing lossless smoothing of the stream of digital compressed images, on the size of the useful part of the packet;

fig. 3 - dependence of the minimum size of the playback buffer, which provides lossless smoothing of the stream of digital compressed images, on the size of the useful part of the packet.

The figure 1 shows a functional diagram of a method for minimizing delay in the transmission of a packetized smoothed stream of digital compressed images, which includes the following elements:

1. Parameters of the input stream.

2. The minimum size of the useful part of the package.

3. The maximum size of the useful part of the package.

4. Performance parameters of the communication network.

5. Block for forming the curve of receipt of the stream of digital compressed images.

6. Block for enumeration of possible packet sizes.

7. The size of the service part of the package.

8. Block for forming the curve of receipt of the input stream of digital compressed images after bundling with the addition of a service part.

9. Block for forming the service curve of the communication network.

10. A block for calculating the minimum playback delay that provides optimal lossless smoothing.

11. Block for optimizing the size of the useful part of the package.

12. Block for determining the minimum buffer size.

13. Digital compressed image stream.

14. Smoothing device.

15. Transmitter

16. Receiver.

17. Decoder.

18. Playback device.

To solve the problem of improving the quality of service of telecommunications traffic, a method is proposed to minimize the delay in the guaranteed transmission of a packetized smoothed stream of digital compressed images, which consists in the following:

1. The parameters of the input stream 1 of images are set in the block for forming the curve for the arrival of the stream of digital compressed images 5.

Moreover, the values of the following quantities are set as parameters:

_Image L - is the maximum size of the compressed image;

fps is the frame rate;

τ is the possible jitter of the image stream.

2. The minimum size of the useful part of the package 2 is set in the enumeration of the possible sizes of the useful part of the package 6.

3. The maximum size of the useful part of the package 3 is set in the enumeration of the possible sizes of the useful part of the package 6.

4. The performance parameters (processing speed or bandwidth) of the traffic processing devices 4 are set along the route of the digital compressed image stream in the communication network service curve generation unit 9.

5. The input stream arrival curve with parameters 1 is formed in the digital compressed images stream arrival curve forming unit 5 and transferred to the digital compressed images input stream arrival curve forming unit after bundling with the addition of a service part 8.

In this case, an affine curve of the form is used as the input flow arrival curve:

wherein: L _fig - is the maximum size of the compressed image, fps - this frame rate, τ - a possible jitter picture stream.

This affine curve is the minimum arrival curve for a digital compressed image stream.

In the generated form, the input flow arrival curve is a two-dimensional array, into which the values of the flow level are recorded with a time step in accordance with expression (1).

6. The current value of the size of the useful part of the packet is set in the block for enumerating the possible sizes of the useful part of the packet 6 and its block for forming the arrival curve of the input stream of digital compressed images after packetization with the addition of the service part 8 is transmitted and to the block for forming the service curve of the communication network 9.

7. The size of the service part of the packet 7 is set in the block for generating the arrival curve of the input stream of digital compressed images after packetization with the addition of the service part 8 and in the block for forming the service curve of the communication network 9.

8. In the block for forming the arrival curve of the input stream of digital compressed images after bundling with the addition of the service part 8, the curve for the arrival of the input stream of digital compressed images after bundling with the addition of the service part is formed and transferred to the block for calculating the minimum playback delay 10.

In the generated form, the arrival curve of the input stream of digital compressed images after packetization with the addition of a service part is a two-dimensional array in which the values of the stream level are written with a time step in accordance with the expression of the form:

where: l _pl - the size of the useful part of the package, 1 _sl - the size of the service part of the package.

Expression (2) is the result of the inverse min+convolution of the affine arrival curve (1) and the burst-delay function

9. A communication network service curve is formed in the communication network service curve generation unit 9 and transferred to the minimum playback delay calculation unit 10, which provides optimal lossless smoothing.

To form the service curve of the communication network, the service device concatenation rule is used, which is mathematically expressed in terms of min+convolution of the service curves of all service devices along the path of the digital compressed image stream. In the generated form, the service curve of the communication network is a two-dimensional array, in which the values of the level of the serviced flow are written with a time step in accordance with the expression of the form:

where: β ⁽¹⁾ transmitter service curve (15), β ⁽²⁾ - receiver service curve (16), β ⁽³⁾ - decoder service curve (17), l _n - packet size including service and useful parts, R ₁ - transmitter performance (15), R ₂ - receiver performance (16), R ₃ - decoder performance (17).

In this case, the service curves β ⁽¹⁾ β ⁽²⁾ and β ⁽³⁾ characterize the data packetizer and are described in accordance with expressions of the form:

10. Calculate in the minimum playback delay calculation block 10 the minimum playback delay that provides lossless smoothing.

At the same time, the maximum horizontal deviation is calculated between the arrival curve of the input stream of digital compressed images after bundling with the addition of a service part and the service curve of the communication network, in accordance with the expression of the form:

11. The obtained minimum value of the playback delay and the size of the useful part of the packet corresponding to it are transmitted from the block for calculating the minimum playback delay 10 to the block for optimizing the size of the useful part of the packet 11.

12. The block for calculating the minimum playback delay 10 is looped to the block for enumerating the possible sizes of the useful part of the packet 6.

In this case, enumeration means the use of all possible packet sizes in the range from the minimum packet size 2 to the maximum packet size 3.

13. In the enumeration block of possible sizes of the useful part of the package 6, all possible sizes of the useful part of the package are enumerated in the range between the minimum size of the useful part of package 2 and the maximum size of the useful part of package 3.

At the same time, steps (5-12) are repeated at each iteration, and thus the dependency shown in figure 2 is formed in the optimization block for the size of the useful part of the package 11.

The analysis of figure 2 allows us to conclude that the optimal solution for the minimum playback delay criterion exists and can be found by exhaustive search in the optimization block for the size of the useful part of the package 11.

In the block for optimizing the size of the useful part of the packet 11, the data of figure 2 are presented in the form of a two-dimensional array reflecting the dependence of the minimum playback delay on the size of the useful part of the packet.

14. Determine the size of the useful part of the package in the optimization block of the size of the useful part of the package 11, which ensures a minimum delay in playback.

To find the minimum, all elements of the array are sorted out (the delay minimum) and compared in pairs with each other.

15. Configure the transmitter 13, the receiver 14 and the decoder 15 in accordance with the optimal size of the useful part of the package.

16. From the block for determining the optimal size of the useful part of the packet 11 to the block for determining the minimum size of the buffer 12, the optimal size of the useful part of the packet is transmitted.

17. Calculate the minimum size of the playback buffer in the block for determining the minimum size of the buffer 12.

To calculate the minimum buffer size, use the expression

type:

In this case, dependence (8) has the form in accordance with figure 3. Analysis of figure 3 allows us to conclude that the optimal values for delay and buffer size are achieved with the same packet size. Therefore, the buffer size in the block for determining the minimum buffer size 12 is calculated by expression (8) with the optimal size of the useful part of the packet from block 11.

18. Configure the image playback device 16 in accordance with the minimum buffer size.

19. Transmit a frame-by-frame stream of digital compressed images 17 to the anti-aliasing device 18.

20. The stream of digital compressed images 17 is smoothed in the smoothing device 18 and transmitted to the transmitter 13.

21. The smoothed stream of digital compressed images is divided into packets and the packetized smoothed stream of digital compressed images is transmitted over the communication channel to the receiver 14.

22. A packetized smoothed stream of digital compressed images is received over a communication channel in the receiver 14 and transmitted to the decoder 15.

23. The packetized smoothed stream of digital compressed images is decompressed and transferred to the playback device 16.

24. Play back a frame-by-frame stream of digital images on the screen of the playback device 16.

The "industrial applicability" of the method is due to the possibility of implementing it using both programmable integrated circuits and program-controlled telecommunications equipment.

Comparison of the claimed method for minimizing delay with guaranteed transmission of a packetized smoothed stream of digital compressed images with the prototype shows that the claimed method differs significantly from the prototype.

General features of the proposed method and prototype:

1. The input stream arrival function is formed.

2. Produce min+convolution of the service curve of the communication network with the shape curve of the smoothing device.

3. Determine the minimum preview delay size.

4. Determine the minimum inlet flow curve.

5. Determine the minimum playback delay.

6. Determine the minimum size of the playback buffer.

7. Configure the communication network taking into account the minimum playback delay and the minimum playback buffer, which provide lossless smoothing.

8. Smooth out the flow.

9. Pass the stream.

10. Accept the stream.

11. Produce flow decompression.

12. Play the stream.

Distinctive features of the proposed solution.

1. Take into account the packetization of the input stream, when obtaining the service curve of the communication network, by using the concatenation of the service curves (4, 5, 6) for the packetizers in accordance with the expression (3).

2. Minimize the transmission delay under packetization conditions for a stream of digital compressed images by finding the size of the useful part of the packet that provides the minimum minimum playback delay with lossless smoothing.

3. Take into account the increase in the amount of transmitted data caused by the addition of a header containing service information to each packet by using the "burst-delay" function, which converts the input stream arrival curve (1) to the form (2).

4. The stepwise nature of the input stream arrival function is taken into account when transmitting digital compressed images, by applying the affine curve of the input stream arrival described by expression (1).

5. The degree of compression of the frame-by-frame stream of digital images is taken into account by taking as the maximum burst in expression (1) the maximum size of the compressed frame.

6. When transmitting a frame-by-frame video stream, the frame rate is taken into account by using the product of the maximum image size and their repetition rate in expressions (1, 2, 7, 8).

7. When transmitting a frame-by-frame video stream, the image resolution is taken into account by using the maximum size of the compressed frame when forming the input stream arrival curve, which directly depends on the resolution of the original image.

To experimentally test the performance of the proposed method, a telecommunications network was taken, through which digital compressed images were transmitted in accordance with the initial data presented in Table 1.

For the initial data presented in Table 1, the optimal packet size is

When transmitting digital compressed images traffic with a packet size of 744 bytes over the telecommunications network selected for the experiment, guaranteed data delivery with a minimum delay was ensured, which confirms the efficiency of the proposed method for minimizing delay with guaranteed transmission of a packetized smoothed stream of digital compressed images.

Claims (1)

A method for minimizing delay in the guaranteed transmission of a packetized smoothed stream of digital compressed images, which consists in the fact that: specifying the parameters of the input stream; set parameters of traffic processing devices; form the curve of receipt of the input stream; forming a communication network service curve; calculating a minimum playback delay; determine the optimal size of the useful part of the package; configuring a communication network; calculate the minimum buffer size; configuring a playback device; transmitting a frame stream; smooth the flow breaking the stream of compressed images into packets and transmitting over a communication channel; receiving a stream over a communication channel and transmitting it to a decoder; decompressing the stream and transmitting it to the playback device; reproducing a stream, characterized in that: setting the parameters of the input image stream consisting of: the maximum size of the compressed image, the frame rate, the possible jitter of the image stream; set the parameters of the traffic processing devices in the composition of: the minimum size of the useful part of the packet, the maximum size of the useful part of the packet, the size of the service part of the packet and performance; forming a curve of receipt of the input stream of digital compressed images, while the curve of receipt is an affine curve, which is a two-dimensional array obtained taking into account the parameters of the input stream of digital compressed images after packetization with the addition of a service part; form an end-to-end service curve of the communication network, using the rule of concatenation of service devices, expressed in terms of min+convolution of the service curves of all servers along the path of the stream of digital compressed images, while in the generated form the service curve of the communication network is a two-dimensional array; determine the optimal size of the useful part of the package, providing a minimum minimum playback delay when smoothing without loss, by enumeration of all possible sizes of the useful part of the package in the range between the minimum size of the useful part of the package and the maximum size of the useful part of the package; configuring the transmitter, receiver and decoder, taking into account the optimal size of the useful part of the package.

Images