KR20210053157A - Method and server for stream scheduling - Google Patents

Abstract

According to an aspect of the present application, a stream scheduling method performed by a server comprises the following steps of: assigning data to a plurality of streams according to preset conditions if a data transmission request is received from a client; measuring a size and request time of data assigned to a stream; and arranging the stream in a stream queue and selecting a stream to be transmitted based on the measured data size and request time to transmit the same to the client. A stream is a continuous flow of data that an application layer wants to transmit when data is transmitted. Streams arranged in the stream queue are sorted from the earliest request time. The measured data sizes are sorted in a descending order and data are transmitted to the client in the sorted order.

Description

Stream scheduling method and server {METHOD AND SERVER FOR STREAM SCHEDULING}

The present invention relates to a stream scheduling method and a server, and more particularly, when a transport layer protocol in an OSI (Open System Interconnection) 7 layer model supports stream multiplexing, when and how much data is transmitted. It relates to a stream scheduling method that determines whether to do so.

The transport layer of the OSI 7 layer model is responsible for reliable data transmission between both ends, and in the case of TCP, the representative protocol of this layer, provides reliable communication by retransmitting packets that have failed to be transmitted. However, HOL blocking (HoLB: Head-of-Line Blocking) delays the transmission of other data due to the nature of TCP that provides only one byte stream when trying to send multiple independent data using one TCP connection. ) Problems may occur. To solve this problem, a transport layer protocol has been proposed that supports a plurality of streams as a transmission unit and supports a stream multiplexing function that simultaneously transmits these streams within a single connection. Representatively, QUIC and SCTP are two protocols.

The stream of the transport layer protocol is a unit for transmitting independent data. Data transmitted within one stream must be received in the order of transmission, but the transmission order and time of data from one stream and other streams are not determined. Accordingly, in a situation where a plurality of streams exist in one connection, a stream scheduling problem occurs in which a stream is determined when and how much to send.

When multiple subjects share and use shared resources, such as a packet switch network, fair use of resources among them becomes a general consideration. Therefore, as a result of allocating resources, it is possible to check whether the resource allocation method is fair by measuring whether an individual subject has received a fair service in terms of bandwidth and delay time. There is also a need for a method to prevent a problem in which a specific subject is maliciously using a lot of resources and other subjects are allocated fewer resources. The Fair Queuing method and its variants are commonly used because they are known to have an advantage in terms of bandwidth and latency fairness in situations where multiple subjects try to share network resources.

However, the stream scheduling problem in a transport layer protocol supporting stream multiplexing is different from the resource allocation problem in a general network. The general network resource allocation problem assumes that an individual subject wants to allocate more of the shared resources. However, individual streams of the transport layer protocol are originally used by one application, and there is no contention between streams for more resource allocation. Performance indicators are also measured at the application level, which is the upper layer that uses these streams, not for individual streams.

Republic of Korea Patent Publication No. 10-2006-0096077 (title of the invention: method and apparatus for data communication through multiple channels)

The present invention is to solve the above-described problem, in order to reduce the delay time of individual data transmission when an application transmits a plurality of independent data using a transport layer protocol supporting stream multiplexing, streams for each data transmission are transmitted. It is a technical problem to provide a stream scheduling method that determines which stream to send in a waiting situation.

Based on the fact that all streams are transmission units used by higher-level applications that are one subject rather than different subjects, it is based on the principle of transmitting the stream from which the transmission is expected to be completed first, rather than prioritizing fairness between streams. Provides a stream scheduling method.

However, the technical problem to be achieved by the present embodiment is not limited to the technical problem as described above, and other technical problems may exist.

As a technical means for solving the above-described technical problem, the stream scheduling method performed by the server according to the first aspect of the present disclosure, when a data transmission request is received from a client, transfers data to a plurality of streams according to a preset condition. Allocating; Measuring the size of data allocated to the stream and a request time; And arranging the stream in the stream queue, selecting a stream to be transmitted based on the measured data size and the request time, and transmitting the selected stream to the client. Includes. Stream is a continuous flow of data that the application layer wants to transmit when transmitting data, and streams placed in the stream queue are sorted in order from the earliest request time, and sorted in order from the smallest measured data size. It is sent to the client in order.

As a technical means for solving the above technical problem, a stream scheduling server performed by a server according to a second aspect of the present disclosure includes: a memory storing a stream scheduling program; And a processor that executes a program stored in the memory. When a data transmission request is received from a client by execution of the program, the processor allocates data to a plurality of streams according to a preset condition, measures the size and request time of the data allocated to the stream, and performs a stream queue. The streams are arranged in the stream queue, and the streams arranged in the stream queue are arranged in order from the earliest to the request time, and the measured data size is arranged in order from the smallest, and sorted based on the measured data size and request time. Streams to be transmitted are selected in the order specified and transmitted to the client. The stream is a continuous flow of data to be transmitted by the application layer during data transmission.

According to any one of the above-described problem solving means of the present application, as the transmission completion time of individual data decreases, an application receiving the data may quickly utilize or process the data that arrives first. Although there is no difference in the time at which the transmission of the last data ends regardless of which stream scheduling method is used, in the case of data other than the last, the transmission completion time is reduced through the present invention, and thus the efficiency of using the entire data can be improved.

1 is a block diagram showing the configuration of a stream scheduling system according to an embodiment of the present invention.
2 is a block diagram showing the configuration of a stream scheduling server according to an embodiment of the present invention.
3 is a diagram illustrating a stream scheduling method according to an embodiment of the present invention.
4 is a flowchart illustrating a stream scheduling method according to an embodiment of the present invention.
5 is a diagram illustrating a delay in transmitting a stream in a stream scheduling method according to an embodiment of the present invention.
6 is a diagram illustrating application of a stream scheduling method according to an embodiment of the present invention.
7 is a diagram illustrating an effect of a stream scheduling method according to an embodiment of the present invention.
FIG. 8 is a diagram illustrating the effect of a First-Come, First-Served (FCFS) technique to explain the effect of a stream scheduling method according to an embodiment of the present invention.
9 is a view for explaining the effect of the round robin technique to explain the effect of the stream scheduling method according to an embodiment of the present invention.

Hereinafter, embodiments of the present application will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art may easily implement the present application. However, the present application may be implemented in various different forms and is not limited to the embodiments described herein. In addition, in the drawings, parts irrelevant to the description are omitted in order to clearly describe the present application, and similar reference numerals are attached to similar parts throughout the specification.

Throughout this specification, when a part is said to be "connected" with another part, this includes not only the case that it is "directly connected", but also the case that it is "electrically connected" with another element interposed therebetween. do.

Throughout this specification, when a member is positioned “on” another member, this includes not only the case where a member is in contact with the other member, but also the case where another member exists between the two members.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

The OSI 7 layer divides the process of communication in the network into 7 stages, and is a model for identifying and managing the flow at a glance by grasping the process of communication in stages. The OSI 7 layer consists of a physical layer, a data link layer, a network layer, a transport layer, a session layer, a presentation layer, and an application layer.

The application layer directly relates to the application process and performs general application services. Typical application services provide switching between related application processes.

The transport layer allows end-to-end users to send and receive reliable data, so that upper layers do not think about the effectiveness or efficiency of data transfer. It uses a sequence number-based error control method. The transport layer controls the validity of a particular connection, and some protocols are stateful and connection oriented. This means that the transport layer checks whether the transmission of packets is valid and retransmits the packets that failed to be transmitted. In addition, the transport layer is the lowest layer that deals with end-to-end communication and transmits reliable and efficient data from end to end, and its functions perform error detection and recovery, flow control, and redundancy check.

An example of the most well-known transport layer is TCP, and the transport layer protocol supporting stream multiplexing is QUIC or SCTP. Stream multiplexing is a function of simultaneously transmitting multiple streams by supporting multiple streams as a transmission unit within one connection. In this case, the stream is an abstraction of the continuous flow of bytes that the application layer wants to transmit.

Stream Control Transmission Protocol (SCTP) is one of the transport layer protocols in computer networking and plays a similar role to well-known protocols such as Transmission Control Protocol (TCP) and User Datagram Protocol (UDP). It provides some of the same service functions of TCP and UDP. Like TCP, it is a connection-oriented protocol and guarantees reliable sequential message transmission through congestion control.

QUIC (Quick UDP Internet Connections) protocol is an Internet protocol with low latency. It provides the same reliability as TCP by adding a layer, and simultaneously transmits data in multiple parallel streams with one connection, and performs transmission processing in accordance with the order. There are features.

1 is a block diagram showing the configuration of a stream scheduling system according to an embodiment of the present invention.

The stream scheduling method may be implemented in the server 100 or the client 200. The server 100 transmits data to the client 200 according to the request of the client 200. The client 200 requests multiple data in a single request sent to the server 100, and these data are independent data and may be transmitted through different streams. Since the server 100 has data, the size of each data may be known in advance. In addition, there is no limit to the number of times the client 200 requests data from the server 100.

2 is a block diagram showing the configuration of a stream scheduling server according to an embodiment of the present invention.

As shown, the stream scheduling server 100 may include a communication module 110, a memory 120, a processor 130, a database 140, and an input module 150.

The communication module 110 may transmit and receive data with the connected client 200. The communication module 110 may be a device including hardware and software necessary for transmitting and receiving a signal such as a control signal or a data signal through a wired or wireless connection with another network device.

A stream scheduling program is stored in the memory 120. When a data transmission request is received from a client, the stream scheduling program allocates data to a stream according to a preset condition, measures the size and request time of the data allocated to the stream, places the stream in a stream queue, and measures The stream to be transmitted is selected and transmitted to the client based on the data size and the requested time.

Various types of data generated in the process of executing an operating system for driving the stream scheduling server 100 or a stream scheduling program are stored in the memory 120.

In this case, the memory 120 collectively refers to a nonvolatile storage device that continuously maintains stored information even when power is not supplied, and a volatile storage device that requires power to maintain the stored information.

In addition, the memory 120 may perform a function of temporarily or permanently storing data processed by the processor 130. Here, the memory 120 may include a magnetic storage media or a flash storage media in addition to a volatile storage device requiring power to maintain stored information, but the scope of the present invention is limited thereto. It does not become.

The processor 130 executes a program stored in the memory 120, but controls the entire process according to the execution of the stream scheduling program. Each operation performed by the processor 130 will be described in more detail later.

The processor 130 may include all types of devices capable of processing data. For example, it may refer to a data processing device embedded in hardware having a circuit that is physically structured to perform a function represented by a code or command included in a program. As an example of a data processing device built into the hardware as described above, a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, and an application-specific application-specific device (ASIC) Integrated circuit) and processing devices such as a field programmable gate array (FPGA) may be covered, but the scope of the present invention is not limited thereto.

The database 140 stores or provides data necessary for the stream scheduling system under the control of the processor 130. The database 140 may be included as a separate component from the memory 120 or may be built in a partial area of the memory 120.

3 is a diagram illustrating a stream scheduling method according to an embodiment of the present invention.

When data is requested by a client, the stream allocating unit may allocate data to a specific stream according to whether or not the data is independent. Data that are determined to be independent of each other and are not related are allocated to different streams, and data that have dependence and that the transmission order of data must be observed are allocated to the same stream. Data included in one message may be determined as relevant data. In this case, the message is a logical unit of data, not a control or modulated signal for transmission processing, but a signal that is divided into a certain semantic unit and contains actual information therein.

번째 요청에 대해, 요청이 이루어진 시간을

, 요청된 데이터의 수를

라고 하면, 데이터 측정부는

번째 요청에 포함된

개의 데이터의 각 크기인

을 계산할 수 있다. 한편, 각 요청이 이루어진 시간

는 자연히 하기 수학식1을 따르게 된다.The data measurement unit measures the size and request time of the requested data and allocates it to each data. Client's

For the first request, the time the request was made

, The number of requested data

If you say, the data measurement unit

Included in the second request

The size of each of the data

Can be calculated. Meanwhile, the time each request was made

Naturally follows Equation 1 below.

[Equation 1]

를 부여한다. 스트림 할당부와 데이터 측정부에 의해 스트림의 할당과 데이터의 크기가 측정된 데이터는 스트림 대기열에 배치된다. 그 이후 스케줄러는 스트림 대기열의 데이터에 대해

의 값이 작은 것부터 오름차순으로, 그 다음은

의 값이 작은 것부터 오름차순으로 전송할 데이터를 선택한다. 여러 데이터에 부여된

의 값이 같은 경우 그 중 하나를 임의로 선택할 수 있다.Based on this, the data measurement unit

Is given. Data for which the stream allocation and data size are measured by the stream allocating unit and the data measuring unit are placed in the stream queue. After that, the scheduler reads the data in the stream queue.

From smallest to ascending, then

Select the data to be transmitted in ascending order from the smallest value of. Given to multiple data

If the values of are the same, one of them can be arbitrarily selected.

When retransmission occurs because packet loss is detected, the retransmission packet must be transmitted first, so retransmission is not considered. In addition, the target of this stream scheduling is for the stream that delivers application data, and the data transmission for connection establishment and management is not the target. After the data to be transmitted is determined, it is delivered to the packet data queue.

4 is a flowchart illustrating a stream scheduling method according to an embodiment of the present invention.

When a data transmission request is received from the client 200 (S110), the server 100 allocates data to the stream according to a preset condition (S120). The client 200 may request transmission of a plurality of data, and the server 100 allocates data that has dependence between data or the transmission order of data must be observed for the requested plurality of data to the same stream, and is independent of each other. Irrelevant data can be assigned to different streams.

The server 100 measures the size of the data allocated to the stream and the request time (S130).

The server 100 places a stream in the stream queue (S140). The stream queue is a place to wait before the streams to which data is allocated are selected by the stream scheduler, and are not transmitted in the order stored in the stream queue.

The stream to be transmitted is selected and transmitted to the client based on the measured data size and request time (S150).

The server 100 may first select a stream having a small measured request time and transmit it to the client, or select a stream having a small measured data size first and transmit it to the client.

In this case, the stream scheduling method may be performed through a transport layer protocol of an OSI 7 layer model supporting stream multiplexing. For example, the transport layer protocol may be either QUIC or SCTP. In addition, the stream scheduling method performed in the present invention may target data and streams requested by a single application of a client.

5 is a diagram illustrating a delay in transmitting a stream in a stream scheduling method according to an embodiment of the present invention.

The present invention provides a stream scheduling method to provide a low-latency application service. When an individual stream sends a message with a fixed size, the time from the time the message is requested to the transport layer to the completion of transmission to the communication counterpart is called latency, and reducing the sum of the delays of all messages is a method of the present invention. It's the goal. As shown in FIG. 5, since the size of data allocated for each stream is different, the total sum of delays for each stream may vary according to the order in which the streams are transmitted.

6 is a diagram illustrating application of a stream scheduling method according to an embodiment of the present invention.

In order to confirm the effect of the present invention, data is transmitted using each of the stream scheduling, first-in-first-out scheduling, and round-robin scheduling of the present invention. Each transmission is performed in a round unit, and the data size transmitted in each round can be set to 10, 20, 40, 40, 40.

The stream size is A: 16seg, B: 9seg, C: 14seg, D: 4seg, E: 31seg, F: 11 seg, G: 15 seg, H: 18 seg, I: 3seg, J: 7 seg, K: It is 11seg.

Streams A and B are requested before round 1, streams C and D are requested before round 2, streams E, F, and G are requested before round 3, streams H, I and J are requested before round 4, and stream K Is requested before round 5.

The change in the size of transmitted data per round is similar to the congestion control technique (AIMD-based) actually used such as TCP and QUIC. The time taken for each transmission can be assumed to be 100 ms. Only data requested to be transmitted before the round starts may be included in the scheduling target of the round. For example, data of streams C and D of FIG. 6 cannot be transmitted in round 1 and will be scheduled in round 2.

Low Latency (LL) scheduling according to the present invention transmits data in order of small size. The results of the present invention are described in FIG. 7.

First-in-first-out scheduling is transmitted in order of request time, and if the request time is the same, it is transmitted in random order or in the order of stream identification number. The results of the first-in-first-out processing scheduling are described in FIG. 8.

Round-robin scheduling divides all streams into a certain unit and transmits them in a fair manner. The results according to the round robin scheduling will be described in FIG. 9.

Regardless of which scheduling technique is used, the time point at which the last data transmission ends is not changed. However, it can be seen that an advantage of reducing the delay time of individual data can be obtained by applying the present invention.

7 is a diagram illustrating an effect of a stream scheduling method according to an embodiment of the present invention.

In round 1, 1 of stream A and 9 of stream B are transmitted, and in round 2, A:15, C:1, and D:4 are transmitted. Round 3 sends C:13, E:1, F:11, G:15 Round 4 sends E:30, I:3, J:7. In round 5, H:18,K:11 are transmitted and all streams are transmitted.

At this time, the delay time for each stream is A: 200 ms, B: 100 ms, C: 200 ms, D: 100 ms, E: 200 ms, F: 100 ms, G: 100 ms, H: 100 ms, I: 100 ms, J: 100 ms, K: 100 ms. The total delay time combined with the delay times for each stream is 1400 ms.

8 is a diagram illustrating an effect of a first-in-first-out processing technique to explain the effect of a stream scheduling method according to an embodiment of the present invention.

In round 1, 10 of stream A is transmitted, and in round 2, A:6, B:9, and C:5 are transmitted. Round 3 sends C:9, D:4, E:27 Round 4 sends E:4, F:11, G:15, H:10. In round 5, H:8, I:3, J:7, K11 are transmitted, and all streams are transmitted.

At this time, the delay time for each stream is A: 200 ms, B: 100 ms, C: 200 ms, D: 100 ms, E: 200 ms, F: 100 ms, G: 100 ms, H: 200 ms, I: 100 ms, J: 100 ms, K: 100 ms. The total delay time combined with the delay times for each stream is 1500 ms.

9 is a diagram for explaining the effect of the round robin technique in order to explain the effect of the stream scheduling method according to an embodiment of the present invention.

In round 1, streams A:5 and B:5 are transmitted, and in round 2, A:6, B:4, C:6, and D:4 are transmitted. Round 3 sends A:5, C:8, E:9, F:9, G:9 Round 4 sends E:11, F:2, G:6, H:11, I:3, J:7 do. In round 5, E:11, H:7, and K:11 are transmitted, and all streams are transmitted.

At this time, the delay time for each stream is A: 300 ms, B: 200 ms, C: 200 ms, D: 100 ms, E: 300 ms, F: 200 ms, G: 200 ms, H: 200 ms, I: 100 ms, J: 100 ms, K: 100 ms. The total delay time combined with the delay times for each stream is 2000 ms.

An embodiment of the present invention may also be implemented in the form of a recording medium including instructions executable by a computer, such as a program module executed by a computer. Computer-readable media can be any available media that can be accessed by a computer, and includes both volatile and nonvolatile media, removable and non-removable media. Further, the computer-readable medium may include a computer storage medium. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

Although the methods and systems of the present invention have been described in connection with specific embodiments, some or all of their components or operations may be implemented using a computer system having a general-purpose hardware architecture.

The foregoing description of the present application is for illustrative purposes only, and those of ordinary skill in the art to which the present application pertains will be able to understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present application. Therefore, it should be understood that the embodiments described above are illustrative in all respects and are not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as being distributed may also be implemented in a combined form.

The scope of the present application is indicated by the claims to be described later rather than the detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present application.

100: server
110: communication module
120: memory
130: processor
140: database
150: input module
200: client

Claims (13)

In the stream scheduling method performed by the server,
(a) when a data transmission request is received from a client, allocating data to a plurality of streams according to a preset condition;
(b) measuring the size of data allocated to the stream and a request time; And
(c) placing the stream in a stream queue, selecting a stream to be transmitted based on the measured data size and request time, and transmitting the selected stream to the client; Including,
The stream is a continuous flow of data to be transmitted by the application layer during data transmission,
The streams arranged in the stream queue in step (c) are sorted in order from the earliest to the request time, and are sorted in order from the smallest measured data size, and are transmitted to the client in the sorted order,
Stream scheduling method. The method of claim 1,
Steps (a) to (c) are performed through a transport layer protocol of an OSI 7 layer model supporting stream multiplexing,
The stream multiplexing is a technique in which different streams are transmitted through one connection,
Stream scheduling method. The method of claim 2,
The transport layer protocol is either QUIC or SCTP,
Stream scheduling method. The method of claim 1,
In step (a), a plurality of data transmission requests are received from a client,
To measure the size and request time of each of the plurality of data received in the transmission request in step (b),
Stream scheduling method. The method of claim 1,
In the step (a), the data transmission request is a request by a single application of the client,
In the step (c), the stream to be transmitted is transmitted to a single application of the client,
Stream scheduling method. The method of claim 4,
When a plurality of data transmission requests are received in step (a), data that depend on data or that the transmission order of data is to be observed are allocated to the same stream, and data that are independent of each other and are not related to each other are allocated to different streams. That,
Stream scheduling method. In the stream scheduling server,
A memory in which a stream scheduling program is stored;
And a processor that executes a program stored in the memory,
When a data transmission request is received from a client by execution of the program, the processor allocates data to a plurality of streams according to a preset condition, measures the size and request time of the data allocated to the stream, and performs a stream queue. The streams are arranged in the stream queue, and the streams arranged in the stream queue are arranged in order from the earliest to the request time, and the measured data size is arranged in order from the smallest, and sorted based on the measured data size and request time. Select the stream to be transmitted in the order specified and send it to the client,
The stream is a continuous flow of data to be transmitted by the application layer during data transmission,
Stream scheduling server. The method of claim 7,
The processor performs stream scheduling through a transport layer protocol of an OSI (Open System Interconnection) 7 layer model supporting stream multiplexing,
The stream multiplexing is a technique in which different streams are transmitted through one connection,
Stream scheduling server. The method of claim 8,
The transport layer protocol is either QUIC or SCTP,
Stream scheduling server. The method of claim 7,
The processor receives a plurality of data transmission requests from a client, and measures the size and request time of each of the plurality of data from which the transmission request is received,
Stream scheduling server. The method of claim 7,
The data transmission request is a request by a single application of the client,
The stream to be transmitted is transmitted to a single application of the client,
Stream scheduling server. The method of claim 10,
When a plurality of data transmission requests are received from a client, the processor allocates data that depends on data or that the transmission order of data is to be observed in the same stream, and allocates independent and unrelated data to different streams. sign,
Stream scheduling server. A non-transitory computer-readable recording medium on which a computer program for performing the stream scheduling method according to claim 1 is recorded.

Images