JPH01276939A - Packet transmission and reception system in local area network system - Google Patents

Abstract

PURPOSE:To relieve the load of a high order layer and to improve the response by adopting the constitution such that a deficiency request (NAK response) to missing of a packet and data reception confirmation are entirely implemented in a LAN interface. CONSTITUTION:A LAN interface 13 is provided with a LAN controller 21, a microprocessor 22, a program memory 23 and a buffer memory 24. Then a series of packets served for the transmission of the unit data of an optional size handled in the high order layer of a device connected to the LAN are sent sequentially from the head packet while they are provided with a packet number consecutive from the final packet sequentially. Moreover, a packet addressed to its own equipment from the packet of other device is received and when the received packet is not the head packet of a series of packets, whether or not the received packet is the succeeding packet from the same sender is discriminated based on the packet number of the received packet. When missing of packet is detected, an error response is returned to the sender and when the received packet while missing of packet is not detected is the final packet, the reception end response is returned to the sender.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) It relates to a system in which several devices are interconnected by a local area network, and in particular relates to a transmission/reception method for dividing unit data of arbitrary size handled at the upper layer (application level) into several packets and transmitting and receiving them between devices. .

(Prior Art) Conventionally, when data is transferred via a local area network (hereinafter referred to as LAN), on the sending side, as shown in FIG. )A, B are completely unrelated to small packet units, e.g. al4 ~ al
CAmaL4+-...+al), b8 ~bl
(B-b3+・+bl), and further divide one group G1 into a certain number of packets, for example, 7 packets.
It was common to create On the other hand, on the receiving side, ACK (reception complete) or N
An AK (error received) response was returned to the sending side, and this unit was passed to the upper layer. Group G passed in the upper device
The contents of all data are checked to see if all 1s are present, and if there is a shortage of group members, that is, if there is a shortage in the number of received packets, retransmission is requested at this point.

(Problems to be Solved by the Invention) As mentioned above, in the conventional method, the upper layer needs to confirm data reception and furthermore request packet shortage.
There were problems in that the load on the upper layer was large and the responsiveness to the sending side was poor. In addition, since ACK or NAK (response) is returned for each group, that is, the response transmission interval is uniquely determined, so tuning according to the transmission path is not possible, and the condition of the transmission path is very poor. However, the number of responses could not be reduced.

Therefore, this invention solves the problem of being able to confirm data reception and also request packet shortages at the LAN interface (local area network interface), thereby reducing the load on the upper layer of the network and improving responsiveness to the sending side. This should be the first issue to be addressed. Furthermore, the second problem to be solved by the present invention is to reduce the number of responses to the packet transmitting side.

[Structure of the Invention] (Means for Solving the Problems) The present invention sequentially processes a series of packets used for transmitting unit data of arbitrary size handled in an upper layer of equipment connected to a LAN, starting from the last packet. Packet transmitting means for sequentially transmitting packets starting from the first packet with consecutive packet numbers added; packet receiving means for receiving packets addressed to the device from packet transmitting means of other devices connected to the LAN; and received packets. is not the first packet in a series of packets, it is determined based on the packet number of the received packet whether this received packet is the next packet of the previous received packet from the same source as the same packet. Packet dropout detection means for detecting packet dropouts is provided in the LAN interface of each device, and when a packet dropout is detected by the packet dropout detection means, the LAN
An error response is returned from the interface to the source, and if the received packet when no packet loss is detected is the final packet, that is, if all the series of packets have been received normally, the reception from the LAN interface to the source is completed. The feature is that it returns a response.

(Operation) According to the above configuration, data reception confirmation and insufficient request (error response) are all performed on the LAN interface, so the load on the upper layer is reduced and responsiveness is improved. Also, regardless of the number of packets, a reception completion response (ACK response) is returned only when data reception is confirmed (that is, only when a series of packets corresponding to units of arbitrary size handled by the upper layer are successfully received). The number of responses is reduced compared to the conventional method in which responses are returned in groups.

(Embodiment) Fig. 1 is a flowchart showing a packet reception control processing procedure according to an embodiment of the present invention, and Fig. 2 shows an embodiment of a LAN (local area network) system to which the procedure of Fig. 1 is applied. The block configuration diagram, FIG. 3, is a diagram showing a packet format applied in the LAN system of FIG. 2.

In FIG. 2, 11 is a LAN cable such as a coaxial cable, which is a LAN transmission line. This LAN cable 11 has a transceiver (TR) 12-1=12-
A host device 14 such as a workstation or computer via the n5LAN interfaces 11-1 to 13-n.
-1-14-n are connected.

The LAN interface 13-2 is connected to the transceiver 12-
A LAN that is connected to 2 and has a serial/parallel conversion function.
Controller 21, LAN side (LAN cable 11 side)
A microprocessor 22 performs interface processing between the host device 14-2 and the host device 14-2, a program memory 23 stores a control program for the microprocessor 22, and a buffer memory 2 temporarily stores transmitted and received data.
4 as the main configuration. The main configuration of this LAN interface 13-2 is that other LAN interfaces
The same applies to -1, 11-3 to II-n.

The host device 14-2 mainly includes a host CPU 31 that executes application programs and the like, and a main memory 32 that stores the application programs and the like. The main configuration of this host device 14-2 is almost the same for the other host devices 14-1, 14-3 to 14-n.

The packets applied in the LAN system in Figure 2 are
As shown in the format shown in the figure, it has a well-known field configuration including a field in which a destination address DA indicating the destination is set, a field in which a source address SA indicating the source is set, a type field, and a data field.

However, for example, a predetermined area at the beginning of the data field is allocated to the packet number field in which the packet number (packet number 4Hn) of the corresponding packet is set.
Note that the remaining data fields are used as conventional data fields. Assuming that the number of packets in a series corresponding to unit data of arbitrary size handled at the application level (upper layer) is N, the packet number field of each packet will contain 1, 2, etc. in order from the last packet. Packet number is set. That is, in each packet field of N packets,
From the first packet (N, N as the kerato number n)
-1, N-2, . . . 1 are set respectively.

Next, the operation of an embodiment of the present invention will be described. As shown in FIG. 4, the units (unit data, here files) A and B handled at the application level are 14 packets consisting of a14 to al. It is divided into three packets consisting of ba to bi, and each packet is sent to the LAN interface 13.
-1. The case where data is transmitted from If-3 to the LAN interface 13-2 will be explained with reference to the flowchart in FIG. 1 and the state diagram of the management table in the buffer memory 24 in FIG. 5 as appropriate.

First, the LAN controller 21 of the LAN interface 13-2 receives a packet on the LAN cable 11 that is taken in via the transceiver 12-2, and after receiving the packet, checks whether the number of frames in the received packet is as specified. If it is as specified, it is checked whether the packet is addressed to itself (by checking whether the destination address DA in the received packet matches its own address).

Microprocessor 2 of LAN interface 13-2
2 detects that the packet addressed to itself has been normally received by the LAN controller 21 (step SL).
-33), the source address S in the received packet
A is the source address field (S) of the management table 51 as shown in FIG.
Determine whether it is registered in A field) (
Step S4), each entry in the management table 51 includes, in addition to the SA field described above, a field (total packet number field) indicating the total number N of a series of packets from the source indicated by the same field; A field (received packet number field) indicating the number l of packets actually received is prepared.

If the determination in step S4 is NO, the microprocessor 22 selects the first packet (bucket in the example of FIG. 4) of a series of packets from the source whose received packet is indicated by the source address SA in the packet a14. or bl)
, and registers this source address SA in the SA address field of an empty entry in the management table 51 shown in FIG. 5, and clears the value (N) of the number of packets field of that entry to "0#". (Step S5
).

On the other hand, if the determination in step S4 is YES, the microprocessor 22 determines that the received packet is the second or subsequent packet in a series of packets from the source indicated by the source address SA in the same packet, The values Nl l of the total number of packets field and the number of received packets field of the corresponding entry in the management table 51 are loaded (step S6).

After completing step S5 or step S6, the microprocessor 22 determines the packet number n in the received packet.
(see Figure 3) (step S7), and reads the above N (N cleared to 101 in step S5 or step S6).
Compare the size with N) loaded in (step S8)
. If N<n, the microprocessor 22 assumes that the first packet has been received, and enters N (total number of packets) in the total number of packets field of the entry in the management table 51 in which the source address SA in the received packet is registered. Packet number n in the received packet (first packet)
At the same time, the value i of the received packet number field of the same entry is cleared to "0" (step S9), and the data area (received data area) corresponding to the total number of packets N is buffered corresponding to the same entry. Allocate it on the memory 24 (step 810).

When the microprocessor 22 completes step 89°310, it proceeds to step Sit. Also, in step S8
If it is determined that ≧n, the microprocessor 22
It is determined that the packet is received after the th packet, and the process skips steps S9 and 810 and proceeds to step Sll. The microprocessor 22 performs N-L in step Sll.
and n, and if N-ten, that is, if the value obtained by subtracting the number of received packets l from the total number of packets N does not match n even though the packet number is the reception of a packet with n, It is determined that there is an omission in the series of packets sent from the sender, and a NAK response is sent to the sender via the LAN controller 21.

On the other hand, if it is N-1-n, the microprocessor 22 determines that there is no missing packet, and in step S1
Proceed to step 2.

The microprocessor 22 increments the number i of received packets by 1 in step S12. This indicates the number of packets received from the first packet (with the same source address SA) up to the present time. Next, the microprocessor 22
determines whether the packet number n in the received packet is 1 (step 513), and if it is not n-1, that is, the received packet is not the final packet, the data in the received packet is mapped to the packet number n. The data is registered in the data area in the buffer memory 24 (step 514).

The microprocessor 22 then receives +1 in step 812.
The number l of received packets is saved in the number of received packets field of the corresponding entry in the management table 51 (step 515), and the process returns to step Sl. On the other hand, if n-1 is determined in step 813, that is, if the received packet is the final packet, the microprocessor 22 sends a series of messages from the source indicated by the source address SA in the received packet. It is determined that all the packets (packets af4 to al of file A or packets b3 to bl of file B in the example in FIG. 4) have been received normally without any omissions, and the data in the received packets are assigned packet number n (
-1) in the data area in the buffer memory 24 (step 816), the LAN controller 21
An ACK response is sent to the sender via
7). The microprocessor 22 then notifies the host CPU 31 of the host device 14-2 of the completion of file reception (step 81g).

As a result, the host CPU 31 transfers the received file data from the buffer memory 24 to, for example, the main memory 32.
Read into.

In addition, in FIG. 4, packet a1 corresponding to file A
After the transmission of 4 to at is completed (by LAN interface 13-1), packet b3 corresponding to file B is sent.
Although a case is shown in which transmission of ~bl (by LAN interface 13-1) is performed, for example, a14 to a1
It is also possible to transmit b3 to bl in parallel using the gap between transmissions.

[Effects of the Invention] As described in detail above, according to the present invention, data reception confirmation and missing request (NAK response) in response to packet loss can be performed.
Since all of this is done on the LAN interface, the load on the upper layer is reduced and responsiveness is improved. In addition, regardless of the number of packets, a reception completion response (A
CK response) is returned, the number of responses is reduced compared to the conventional method of returning responses in groups of a predetermined number of packets, making it possible to increase the number of transmitted packets per unit time. Therefore, network tuning can be easily performed.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing a packet reception control processing procedure according to an embodiment of the present invention, and FIG. 2 is a block diagram showing an embodiment of a LAN (local area network) system to which the procedure of FIG. 1 is applied. Fig. 3 is a diagram of the packet format applied in the LAN system of Fig. 2, Fig. 4 is a diagram for explaining the packet transmission/reception processing in the system of Fig. 2, and Fig. 5 is a diagram used to manage packet reception. A diagram showing an example of the structure of a management table compared to FIG. 4, and FIG. 6 is a diagram for explaining conventional packet transmission/reception processing. 11-LAN cable, 13-1-13-n -L
A N interface, 14-1-14-n... host device, 21... LAN controller, 22... microprocessor, 24... buffer memory, 31...
...Host CPU, 51...Management table. Applicant's agent Patent attorney Takehiko Suzue Figure 1 Flute 2F! ! i Figure 4

Claims (2)

[Claims] (1) In a system in which multiple devices are interconnected by a local area network, a series of packets used to transmit unit data of arbitrary size to be handled by the upper layer is sent to the local area network interface of each of the devices. packet transmitting means for transmitting a series of packets in which sequential packet numbers are set in the predetermined area starting from the last packet, starting from the first packet; packet receiving means that receives the same packet when the packet is a packet of a packet dropout detection means for detecting a packet dropout by determining whether or not the packet is the next packet of the previous received packet containing the same source address as the same packet; a first response means for returning an error response to the sender when the packet dropout is detected by the packet dropout detection means; and a reception completion response is returned to the sender when the received packet is the final packet when no packet dropout is detected by the packet dropout detection means; A packet transmission and reception method in a local area network, characterized in that a second response means is provided, and when all of the series of packets are successfully received, the second response means returns a reception completion response to the source. . (2) The above local area network of each of the above devices.
The interface has a management table for registering the source address indicating the source of the above series of packets for each source, and the above depending on whether the source address included in the received packet is registered in this management table. a first packet determining means for determining whether a received packet is a first packet;
1. The apparatus according to claim 1, further comprising: table registration means for registering a source address included in the packet in the management table when the first packet is determined by the first packet discriminating means. Packet transmission and reception method in local area networks.