JPH02285740A - Data transmission and reception control system - Google Patents

Abstract

PURPOSE:To improve the transfer speed by using a work part of lower 4-layer of a transport, network, data link and physical layer on a memory map in common and generating data to be transferred to host 3 layers of application, presentation and session at a batch. CONSTITUTION:The layer is divided into host 3-layer and a lower 4-layer, and the low-order 4-layer is formed as an intelligent board by front end processors FEP 51, 58, they are located entirely on the same memory space to avoid the independency of each higherarchy. That is, the data transferred from low-order 4-layer 11-13 to the high order layer are always placed to a common work section and the data transfer required time between low-order layers is omitted. At transmission, the n-th layer of the common work section 14 at transmission gives the head address (n) of the data and the data length to the (n-1)th layer, (n-1)th header is added and forms the data to given to the host layer up to the (n-2)th layer and the data length is calculated while the header is removed at each layer at the reception and the result is sent sequentially to the n-th layer. Since the memory transfer of the low-order 4-layer is omitted, the transfer efficiency is remarkably improved.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a data transmission/reception control method between networks using a one-layered communication protocol, and in particular to a data transmission/reception control method that can improve data transfer speed. Regarding the method.

[Conventional technology]

As the demand for communication between different types of networks increases with the advancement of information technology in communication networks, the development of a standard communication protocol based on a common logical structure for networks is being considered.1 The International Organization for Standardization (ISO) and the International Telegraph and Telephone Advisory Committee The standardization of Open System Interconnection (O8I) is progressing with the cooperation of CCITT (CCITT).

The basic elements of this O8I are an open system, a physical medium for O8I, and an application entity, and its basic structure has a hierarchical structure as shown in FIG.

In Figure 2, the physical layer activates, maintains, and
For example, LAN (C8MA/CD (Ca
rrier 5ense Multiple Acce
ss with C.

Transmission of the bit string is realized using a VAN (circuit switching method, packet switching method, etc.) or VAN (circuit switching method, packet switching method, etc.). In addition, the data link layer uses the transmission function of the physical layer to transmit, for example) ([) I, C (High
-1 level Data Link Control)
This enables transparent and reliable data transfer between devices. Furthermore, the network layer uses the services of the lower two layers to realize transparent and highly reliable data transfer. In other words, through open systems and communication networks, relaying, route selection, and data transfer between open systems at the end,
The transport layer realizes high-quality and efficient data transfer between processes, making the user unaware of the communication network.
In addition, the session layer synchronizes and controls transmission so that conversations between processes occur in an orderly manner.
In addition, the presentation layer controls the syntax (code format, presence or absence of encryption and compression) of information handled by the application layer, and controls the syntax used to transfer the information.
The application layer controls the meaning and content of data exchanged between application processes in order to enable business-oriented communications between users, such as file transfer and terminal screen control, and management of resources within the network. .

Furthermore, in O8I, independence of each layer is maintained. This is to clarify the definition of each communication function, facilitate implementation, clarify the division of memory and management, and provide extensibility.

For example, the lower layers from the physical layer to the transport layer are shown on the memory map as shown in FIG.

In this case, the configuration is the program section 3 of the transport layer.
1a and a work unit 31b, a network layer program unit 32a and a work unit 32b.

Program section 33a and work section 3 of data link layer
4b, and a physical layer 34.

Therefore, messages between layers must be received and delivered by memory transfer, and data is passed from the upper layer to the lower layer.

In other words, as shown in Figure 4, in the case of nNM, memory transfer is performed n-1 times, data in the n layer is passed to the n-1 layer, and an n-1 header is attached in the n-1 layer. , passed to the n-2 layer. Furthermore, in the n-2 layer, the n-
2 headers can be attached. In this way, headers are sequentially added to the data as it goes to the lower layers and it is sent out to the line. Also, at the time of reception, contrary to transmission, in the n-2 layer, n-2
The header is removed, and in the n-1 layer, the n-1 header is removed and passed to the upper layer.

On the other hand, when configuring a LAN, the distance between computers is short, so there are fewer transmission errors and high-speed communication is possible, but it is expected that large amounts of image data will be communicated in the future, and faster communication There is a need for a communication protocol that can support this.

Regarding the O5I protocol, see 1. For example, "Standardization of Information Technology, supervised by the Agency of Industrial Science and Technology, Ohmsha (1986)"
It is discussed in

[Problem to be solved by the invention]

In the above-mentioned conventional technology, when each layer is created with independence maintained, the memory space is different for each layer, so it is necessary to send and receive messages between layers by memory transfer. Furthermore, when communicating a large amount of image data over a LAN, it is desirable to shorten the data transmission time.

An object of the present invention is to improve such problems and to place each layer in the same memory space, thereby eliminating the need for memory transfer and shortening data transmission and reception time.

An object of the present invention is to provide a data transmission/reception control method that can improve processing efficiency.

[Means to solve the problem]

In order to achieve the above object, the data transmission/reception control method of the present invention is layered into three upper layers: application layer, presentation layer, and session layer, and lower four layers: transport layer, network layer, data link layer, and physical layer. In a data transmission/reception control method that controls communication between networks using a communication protocol, the lower four layers have a common work section on the memory map that constitutes each layer, and the data to be transferred to the upper three layers is transferred to the common work section. It is characterized by being created in bulk.

[Effect]

In the present invention, a data link layer, a network layer,
By making the work unit of the transport layer common, there is no need to transfer data between each layer, which was conventionally performed, reducing data transmission time and improving processing efficiency.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a memory map diagram of a transport layer, network layer, and data link layer in one embodiment of the present invention, and FIG. 5 is a configuration diagram of a network system in one embodiment of the present invention.

In FIG. 5, reference numerals 51 and 58 refer to front end processors (FEPs) that share the responsibility with the CPUs 53 and 60 and process transmission control procedures and terminal control procedures for communication between LANs;
2.59 is main memory (MEM), 53.60 is CP
U, 54, 61 are external storage devices, 62-64.55-5
7 is a terminal device. Also, outfit! 51 to 57 and devices 58 to 64 each constitute a LAN and are connected by communication lines.

In this example, the 08I7 layer is the upper three layers (application, presentation, session layer) and the lower four layers (transport, network, data link, physical layer).
In particular, the lower four layers are configured as an intelligent board based on FEP51.58, and are all placed in the same memory space. In this way, the independence of each layer is eliminated, making it possible to communicate a large amount of image data at high speed.

That is, when configuring layers lower than the transport layer on the memories of the FEPs 51 and 58, as shown in FIG. 14. As a result, the data to be transferred to the upper layer is always stored in the common work section 14, and there is no need to transfer data between each layer in the lower layer, thereby eliminating the time required to transfer data between each layer in the lower layer. I can do it.

Therefore, at the time of transmission, the common work unit 14 in the lower layer
, the n layer passes the data start address n and data length to the n-1 layer, and the n-1 layer attaches an n-1 header, and transfers the start address n-1 and data length + n-1 header length to the n-2 layer. Create data to be passed to upper layers. Also, when receiving, the common work unit 14
, the n-2 layer removes the n-2 header and passes the start address n-1 and length to the n-1 layer, and the n-1 layer removes the n-1 header and transfers the n-1 header from the length to the n-1 layer. Subtract the head length from ny/! Give it to j.

In addition, in the conventional method, when receiving, for example, 2M bytes of data from the upper layer, the memory transfer is performed by the Intel 801.
When using DMA in 86 (intelligent board manufactured by Ricoh Co., Ltd.), 1 second is required each between the transport layer and the network layer, and between the network layer and the data link layer, resulting in a memory transfer rate of 8 Mbit/sec. . Furthermore, in the case of an Ethernet (C8MA/CD) LAN, the physical layer is 10 Mbit/sec, so even just memory transfer is slower than the physical speed, resulting in wasted resources.

According to this embodiment, data transfer efficiency can be greatly improved by eliminating memory transfer in the lower four layers.

〔Effect of the invention〕

According to the present invention, the memory transfer speed between each layer can be reduced, and the efficiency of data transfer between terminals can be improved.

[Brief explanation of drawings]

FIG. 1 shows a transport layer in an embodiment of the present invention,
A memory map diagram of the network layer and data link layer. Figure 2 is an explanatory diagram of the hierarchical structure of O8I. Figure 3 is a memory map diagram of the conventional transport layer, network layer and data link layer. Figure 4 is an illustration of the connections between each layer of O8I. FIG. 5 is a configuration diagram of a network system in an embodiment of the present invention. 11.31a Nidoran spot layer program section, 12
, 32a: Network layer program section, 13, 33
a: Program section of data link layer, 14: Common work section, 15.34: Physical layer. 31b: Nidor Spot layer work unit; 32b: Network layer work unit; 33b: Data link layer work unit; 51.58: Front end processor (FEP); 52°; 59: Main memory (MEM). 53°60: CPU. 54.61: External storage device. 62-64° 55-57: Terminal device. Representative Patent Attorney Masatoshi Isomura Diagram

Claims (1)

[Claims] (1) Data that controls communication between networks using communication protocols layered into the upper three layers: application layer, presentation layer, and session layer, and the lower four layers: transport layer, network layer, data link layer, and physical layer. The data transmission/reception control method is characterized in that, on the memory map constituting each layer, the lower four layers have a common work section, and the data to be transferred to the upper three layers is collectively created in the common work section. control method.