JP2002077291A - Header-processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem that conventionally in header processing of communication data through network, high rate processing incident to increase of communication traffic cannot be follow up, because a software processing has been employed, and to realize turning into higher rate header-processing, while suppressing cost increase. SOLUTION: In the header-processing method, a next protocol field register 2 and an address translator 3 are configured by hardware and the type field of Ethernet (R) header and the protocol field (next protocol field 1) of IP header in communication data are taken out and stored in the next protocol field register 2. The address translator 3 regards registered values as addresses, and distributes appropriate processing systems selectively.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a transmission control technique used in a data communication system in a network, and more particularly to a header processing method for permitting communication data to be processed in the system.

[0002]

2. Description of the Related Art In general, as a data communication system using a network, a typical Internet is known as a wide area communication, and a LAN (LAN) constructed in a narrow area such as in the same premises or the same building. local area network)
The Ethernet (registered trademark) is the most widespread LAN system.

[0003] In network communication, a head portion of communication data exchanged between terminals includes an IP address consisting of information such as an address number of a transmission destination and its own address number.
A P header and a TCP header are provided. The header information is provided at the head of each packet, and a predetermined process is performed each time communication data is fetched.

FIG. 2A shows an example of a hierarchical structure of a LAN, and FIG. 2B shows an example of header information in Ethernet. The hierarchical structure of the LAN is a physical layer,
Data link layer, network layer, transport layer,
It is hierarchically constructed by a session layer, a presentation layer, an application layer, and the like. At each level,
In order to identify what type of packet is on the upper layer, a field is provided in the header to identify what the upper layer packet is.

[0005] Specific examples of the field for identification include the destination MAC address, source MAC address and packet length / type fields included in the Ethernet header shown in FIG. A protocol field included in the header is typical.

[0006]

In a conventional communication system, when it is necessary to process a field as described above, the communication system is usually constructed on the assumption that a complicated process is performed. Had higher software support. That is, using a processing device such as a CPU,
Such information is identified by software and predetermined processing is performed.

[0007] However, with the recent increase in the amount of communication performed on the Internet, traffic has exploded, and it is difficult for servers and clients to continue to handle all processing with software in the future. is assumed. Various methods have been considered to improve these, but it is desired to realize higher-speed processing while suppressing an increase in cost.

Accordingly, the present invention provides a header processing method in which processing of an upper-layer protocol in communication data of a packet is processed by hardware, thereby suppressing an increase in communication cost and realizing higher-speed processing. With the goal.

[0009]

According to the present invention, in order to achieve the above object, a field for identifying the contents of an upper layer is extracted from header information of communication data received from a network and registered in a register. A header processing method in which a field is regarded as an address, and a processing system in which a corresponding address is defined is selected for a plurality of processing systems for performing header processing to which an address is previously assigned, and information of the field is distributed. I will provide a. The field includes identification information included in the data link layer, the network layer, and the transport layer.

[0010] The header processing method having the above structure is as follows.
The hardware extracts the identification information (next protocol field) included in the data link layer, network layer, and transport layer for identifying the upper layer protocol from the communication data, regards the next protocol field as an address, and For the processing system classified by the address according to the processing content, the corresponding processing system is selected and the field information is sorted.

[0011]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 shows an example of a configuration applied to a header processing apparatus according to an embodiment for realizing the header processing method of the present invention. In the present embodiment, as an example of the communication data, a header in the Ethernet communication will be described and described. However, the communication data is not limited to the Ethernet. Normally, identification information indicating the state of a packet in an upper layer from communication data in packet units received via a network is included in a data link layer, a network layer, a transport layer, and the like. Typical examples are a type field of an Ethernet header, a protocol field of an IP header, which is referred to as a next protocol field, and a protocol field of a TCP header. These fields will be described as an example.

In this configuration, a next protocol field register 2 provided with a plurality of registers for taking out and temporarily registering the next protocol field 1 and a value registered in the next protocol field register 2 as an address Considering this, an address translator 3 that selects a corresponding processing system in a processing unit 4 to be described later and transfers the information of the next protocol field field to each other, and a processing system that performs processing according to each protocol is a predetermined processing system It is composed of a plurality of processing units 4 which are distinguished by addresses. The next protocol field register 2 has a register so that the next protocol field 1 can be registered for each item. Of course, this register may be provided in a plurality of stages. In addition, the configuration may further include a control unit 5 including a CPU or the like so that the user can change the association between the address in the translator and the processing system.

In the header processing method having such a configuration, a translator for performing a distribution to a processing system based on a register for registering a next protocol field and an address is implemented by HBOX (for example, Japanese Patent Application No. 2000-21351 proposed by the present applicant). (Disclosed in Japanese Patent Application Laid-Open No. H10-26095), and a part of the processing of the upper-layer protocol is left to hardware. Other hardware functions may be used for encrypting and decrypting data and the like, and for marking, exchanging and rewriting part of header information, and the like.

Next, a header processing method will be described. First, the processing systems 4a, 4b, 4c,.
Is composed of a circuit for processing input data (type field, protocol field, etc.) for each field, and a circuit for decoding an enable signal for selecting the processing. In the address translator 3, the address and each processing system are associated with each other by hardware, and the processing system is selected based on the input address.

Then, communication data arrives through the network. In this communication data, the type field in the Ethernet header and the protocol field (next protocol field 1) in the IP header as shown in FIG. Is done. Next, the value registered in the register 2 is output to the address translator 3. The address translator 3 regards the registered value as an address, selects a corresponding processing system according to the address extracted from the next protocol field register 2, and
Address data to the processing system. At this time, the type field and the protocol field in the IP header are not selected and processed by one and the same processing system. In principle, a processing system is selected for each field and the processing is performed on each field. In this manner, the type field in the next protocol field 1 from the received next protocol field register 2 and the protocol field of the IP header are processed respectively.

As described above, according to the present embodiment, since the processing system related to the header processing is selected using the configuration constructed by the hardware, necessary processing is automatically performed from the received communication data. Since the next protocol field can be read and the processing system can be designated, high-speed processing can be realized as compared with the case where all processing is performed by software. In addition, since the necessary next protocol field is automatically read and the processing system is selected, the work load on the user is reduced. Furthermore, by consolidating the processing for selecting the processing system into the address translator, the degree of freedom can be increased in the selection by changing the settings. As a result, multi-protocol, multi-layer support is facilitated.

Further, a modified example of the embodiment will be described. In the embodiment described above, the next protocol field
Although the field value registered in the register 2 is used as an address, similar processing can be realized by using an ID instead of this address. That is, when the address translator and the processing system are associated in advance by ID, and each item of the next protocol field (type field in the Ethernet header or protocol field in the IP header) registered in the next protocol field register is registered, An ID may be assigned to each of them, and the processing system may be selected by the address translator using the assigned ID.

When the processing system requires complicated processing, it may be better to perform the processing in the processor. However, the same header processing is performed in the packet in each layer by the processing system. In the case of a process in which a monotonous process is performed, the header processing method according to the present embodiment using a hardware configuration is useful.

[0019]

As described above in detail, according to the present invention, the processing of the upper layer protocol in the communication data of the packet is processed by the hardware, so that the increase in the communication cost can be suppressed and the high-speed processing can be performed. It is possible to provide a header processing method to be realized.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration example applied to a header processing device according to an embodiment for realizing a header processing method of the present invention.

FIG. 2 is a diagram showing a hierarchical structure of a general LAN and header information in Ethernet.

[Explanation of symbols]

 1. Next protocol field 2. Next protocol field register 3. Address translator 4. Processing unit 5. CPU (control unit)

Claims (2)

[Claims] 1. A header process for extracting a field for identifying the contents of an upper layer from header information of communication data received from a network, registering the field in a register, treating this field as an address, and preliminarily assigning an address. A header processing method characterized by selecting a processing system in which a corresponding address is defined for a plurality of processing systems for performing the above, and distributing the information of the field. 2. The method according to claim 1, wherein in the header processing method, the field for identifying the contents of an upper layer comprises identification information included in a data link layer, a network layer, and a transport layer. The header processing method described.