JPH07273794A - Packet communication method - Google Patents

Abstract

PURPOSE:To adjust the balance between the storage capacity of the reception buffer and the transmission processing time in the packet communication. CONSTITUTION:The logic channel number, multiple number, and release time to be used in the transmission of the storage address and reception data to be allocated for the storage of the reception data are stored in a reception buffer pointer table 6. Every time when the reception request is accepted, a reception pointer table 6 is referred to and the discrimination of the presence or absence of the storage area in a reception buffer 7 is performed. When there is no storage area, the allocation of the storage area of the reception buffer is performed. When there is a storage area, the release time is prolonged. When the release time comes, the storage area of the reception buffer 7 is released as an idle area.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a packet communication method for transmitting data by packet switching using a receiving buffer (also called a buffer memory) of a packet switch or a packet switching node, and more particularly, on a logical channel. The present invention relates to a packet communication method for multiplexing data.

[0002]

2. Description of the Related Art In packet communication, data transmission is performed in packet units using channels and a plurality of communication channels called logical channels included therein. A packet communication method for multiplexing data on a logical channel within this channel is known. In such a packet communication method, the lower layer does not care whether the upper layer continuously transfers a large amount of file data or discontinuously transfers short data such as command execution. There is a concept. Based on this concept, a packet communication method is known in which a storage area in a receiving buffer is allocated to transmission of bucket data in a lower layer for every multiplex number on a logical channel from the beginning of system operation. Alternatively, there is also known a packet communication method in which a storage area is allocated in the reception buffer every time there is received packet data from the upper layer, the packet data is received, and the reception buffer is released as a blank area after packet exchange.

[0003]

In the packet communication method in which a storage area is allocated in the receiving buffer for every multiple number on the logical channel from the beginning, the storage capacity of the memory used as the receiving buffer increases. Occurs.

On the other hand, in the packet communication method in which a storage area is allocated in the reception buffer each time there is a reception request from the upper layer, the memory usage of the reception buffer can be small, but the reception buffer in the reception buffer is received every time one packet data is received. It searches for a free area, allocates a storage area, and then releases it. As a result, the number of times the free area is detected increases, which causes a problem that the data transmission process takes time.

In view of the above-mentioned problems, it is an object of the present invention to provide a packet communication method that balances the memory usage of the reception buffer and the data transmission time.

[0006]

In order to achieve such an object, the invention of claim 1 is the same in a packet communication method for transmitting packet data identified by a logical channel number and a multiplex number through a buffer. It monitors whether packet data having the logical channel number and multiplex number of is continuously transmitted in a time series, and keeps a storage area secured in the buffer for the packet data continuously transmitted. When the temporary interruption or the end of the continuous transmission of the packet data is detected, the storage area of the buffer is released as a free area.

According to a second aspect of the present invention, it is determined whether or not the time interval for transmitting the bucket data exceeds a predetermined time, and when a positive determination is obtained, the continuous transmission of the packet data is interrupted or ended. And a negative determination is obtained, it is determined that continuous transmission is in progress.

[0008]

According to the first aspect of the invention, for packet data that is continuously transmitted, the same storage area in the reception buffer is used during the transmission, so a free storage area is searched for each time. No need. Further, since only the transmitted packet data is stored in the buffer, it is not necessary to prepare a storage area for all logical channel numbers and multiplex numbers.

According to the second aspect of the present invention, it is possible to determine the continuity of packet data by a simple method of detecting the transmission interval of packet data.

[0010]

Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 shows the processing contents of the embodiment of the present invention in the form of functional blocks.

In FIG. 1, reference numeral 1 denotes a distribution module, which receives a reception request received from the upper layer 2 in the job queue 3
Accumulate in. The upper layer 2 includes, for example, processes 2a, 2b,
2c, these processes send out the receive request with the same logical channel number (LCN) using separate multiplex numbers. These reception requests are accumulated in the job queue 3 according to the logical channel number and the multiplex number, as indicated by reference numerals 3a, 3b and 3c in FIG. At this time, the address (pointer) of the receiving buffer 7 corresponding to the logical channel number and the multiplex number of the process requesting the reception is stored in the pointer area to the receiving buffer in the job queue 3. Also, a reception request is issued from the job queue 3 to the lower layer 4 in units of logical channel numbers.

Reference numeral 6 is a receive buffer pointer table relating to the present invention, and as shown in the storage contents of FIG. 3, it is associated with a logical channel number and a multiple number to a receive buffer 7 for receiving the receive data corresponding to these numbers. Is stored, that is, a pointer address and a time (hereinafter, referred to as a timeout time) for holding the storage content of the address are stored. A receiving buffer 7 temporarily stores the received packet data 5. In this embodiment, the storage destination in the reception buffer 7 is detected using the reception buffer pointer table 6, and the storage destination is released after a predetermined time (timeout time in the reception buffer pointer table 6) has elapsed. It is characterized by this.

The distribution processing and release processing of the reception buffer 7 executed in such a functional configuration will be described below.

FIG. 2 shows the storage contents of the constituent circuits in the distribution module 1 when the storage area of the reception buffer 7 is allocated. Inside the distribution module 1 of FIG. 2, when a reception request is received, the reception request is accumulated in the job queue 3. Further, the reception buffer pointer table 6 is searched using the logical channel number and the multiplex number when the reception request is received as a keyword, and the pointer address and the timeout time corresponding to the keyword are extracted. When the pointer address and the timeout time are NULL, that is, code information indicating that nothing is stored, the first reception request (logical channel number and multiplex number sent when connecting a connection)
Or, it was received after the timeout time passed. Therefore, a storage area is secured in the reception buffer for the reception data sent later.

The processing for this is performed as follows. The head data of the packet data is sequentially read from the head address of the reception buffer 7, and it is determined whether or not the head data is NULL. NU when the top data is NULL
The area in which LL is stored is determined to be blank. Therefore, the leading address 7a of the blank area is stored in the pointer address corresponding to the keyword (logical channel number and multiplex number), that is, the area indicated by reference numeral 6a in the example of FIG. Further, the blank head address is also stored in the job queue 3 as a pointer to the reception buffer. Then, the distribution module 1 issues a reception request 100 to the lower layer 4.

When packet data is received from the lower layer 4 for each logical channel number in response to this reception request, the distribution module 1 searches the reception buffer pointer table 6 based on the logical channel number and the multiplex number. Since it is indicated that the pointer address and the timeout time obtained as a search result of the reception buffer pointer table 6 are not NULL, it is determined that the storage area of the reception buffer 7 has already been secured. After that, the distribution module 1 stores the received packet data in the storage area of the reception buffer 7 which is already secured, and subsequently, the corresponding packet data 2 in the reception buffer 7.
01 (see FIG. 3) is passed to the upper layer 2. Further, as shown in FIG. 3, the holding time of the reception buffer pointer table 6 is updated to time 202. More specifically, a time obtained by adding a time-out time to the current time is set as a new holding time.

Next, the processing for detecting the end and temporary interruption of continuous packet data will be described with reference to FIG.

The distribution module 1 manages the time-out time of the reception buffer pointer table 6, but refers to the corresponding time-out time of the reception buffer pointer table 6 when a reception request is received from the upper layer 2. When the time-out time is other than NULL, the time-out time is set to NULL prior to the above-mentioned setting of the hold time for initialization. Further, the timeout time (time) of the reception buffer pointer table 6 other than the logical channel number and the multiplex number for which the reception request is made, for example, 3
The data No. 03 is compared with the current time, and if the current time exceeds the time-out time, the time-out time 303 is initialized to NULL. The data related to this, that is, the pointer address is also initialized to NULL. Further, a code indicating a blank area is stored at the head address of the storage area of the reception buffer 7 indicated by this pointer address, and this storage area is released.

When such processing is executed, the storage area of the receiving buffer 7 is secured when the first bucket data reception request is issued, and thereafter, within the fixed time, the same logical channel number and multiplexing are performed. While the packet data using the numbers are continuously transmitted, the reserved storage area of the reception buffer 7 is used for packet exchange. When the transmission of the packet data is completed or temporarily stopped, a new reception request is not generated even after the lapse of the certain time, and thus the secured storage area of the reception buffer 7 is released as a blank area. Receive buffer 7
The fixed time used to determine the release of the storage area can be determined by the following equation.

[0021]

[Equation 1] Hold time = reception response packet size (byte) / line speed (byte / second) XNN is a numerical value between 1 <N <3. The processing explained above is the high-speed arithmetic processor or logic circuit in the distribution module 1. Is realized by accessing the job queue 3, the reception buffer pointer table 6, and the reception buffer 7. The processing procedure for this is shown in FIGS. 5 and 6. FIG. 5 shows a processing procedure for referring to the reception buffer pointer table when a reception request or packet data is acquired, and securing and maintaining the storage area of the reception buffer 7 in accordance with the reference result. FIG. 6 shows a processing procedure for releasing the storage area of the reception buffer 7 which is no longer needed. Note that this processing procedure is activated in an interrupted manner at regular intervals.

In addition to the embodiments described above, the following examples can be carried out.

1) When the logical channel number and the multiplex number are fixed, it is not necessary to describe the logical channel number and the multiplex number in the reception buffer pointer table 6. In this case, the pointer address and the timeout time are stored in the address determined by the logical channel number and the multiplex number.

2) Plural kinds of timeout times can be provided. In this case, it is preferable to selectively set from among a plurality of types corresponding to the communication speed of the logical channel.

3) In this embodiment, the upper layer and the lower layer indicate the processing defined by the protocol, and the lower layer is assumed to be the network layer of the third layer of the OSI model, but the upper layer is the transformer of the fourth layer. It need not be limited to the port layer.

[0026]

As described above, according to the first aspect of the present invention, since the same storage area in the receiving buffer is used for the packet data transmitted continuously during the transmission, There is no need to search for a free storage area each time. Further, since only the transmitted packet data is stored in the buffer, it is not necessary to prepare a storage area for all channel numbers and multiplex numbers. Therefore, the storage capacity of the buffer and the transmission processing time can be balanced.

According to the second aspect of the present invention, since it is possible to determine whether or not the packet data is continuous by detecting the packet data transmission interval, it is possible to reduce the manufacturing cost of the device.

[Brief description of drawings]

FIG. 1 is a block diagram showing a functional configuration of an embodiment of the present invention.

FIG. 2 is a block diagram showing stored contents of the module 1 when receiving a reception request according to the embodiment of the present invention.

FIG. 3 is a block diagram showing stored contents in the middle of the module 1 according to the embodiment of the present invention.

FIG. 4 is a block diagram showing stored contents when the module 1 according to the embodiment of the present invention is released.

FIG. 5 is a flowchart showing a processing procedure of an embodiment of the present invention.

FIG. 6 is a flowchart showing a processing procedure of an embodiment of the present invention.

[Explanation of symbols]

 1 Module 2 Upper Layer 3 Job Queue 4 Lower Layer 5 Received Data 6 Receive Buffer Pointer Table 7 Receive Buffer

Claims (2)

[Claims] 1. A packet communication method for transmitting packet data identified by a logical channel number and a multiple number via a buffer, wherein packet data having the same logical channel number and multiple number are continuously transmitted in time series. Whether the packet data is continuously transmitted, the storage area secured in the buffer is maintained, and the storage area of the buffer is detected when temporary interruption or end of the continuous transmission of the packet data is detected. Is released as a free area. 2. A determination is made as to whether or not a time interval for transmitting the bucket data exceeds a predetermined time, and when an affirmative determination is obtained, it is determined that continuous transmission of the packet data is interrupted or ended, and a negative determination is made. The packet communication method according to claim 1, wherein when the determination is obtained, it is determined that continuous transmission is in progress.