JP3340162B2 - Network coupling device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an inter-network coupling device such as a bridge or a router required for network connection, for example, the passage of packets between at least two networks connected by a thin line having a limited communication capacity. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an inter-network coupling device having a function of controlling a packet passing amount, which is suitable for efficiently distributing the amount and giving priority to communication with high importance and urgency.

[0002]

2. Description of the Related Art An inter-network coupling device is used to connect networks. For example, as shown in FIG. 7A, when connecting the networks N ₁ and N ₂ , the interface unit 41 of the network N ₁ and the network N _{1
The two} interface units 42 are connected by an inter-network coupling device (hereinafter referred to as a coupling device) 51 such as a bridge or a router, and the transfer of packets between N ₁ and N ₂ is controlled.

[0003] As the coupling device, as shown in FIG. 7 (A), and that of the local-type network N _1, N ₂ is at a short distance, as shown in FIG. 7 (B), the network N _1,
N ₂ is at a long distance, and the transmission units 53, 5
There is a remote type in which a line 4 is provided and the line 4 is connected by a line 55. Further, as illustrated in FIG. 7C, there is a type in which a single coupling device 56 controls packet transfer between a plurality of networks N _{1 to} N ₄ . In FIG. 7C, reference numerals 43 and 44 denote interface units of the networks N ₃ and N ₄ , respectively.

Here, the bridge is an L such as Ethernet.
When combining an AN, MAC (Media Acce) as the source address and destination address of the packet
ss Control) controls the packet in which each address is described, and the router further adds an IP (Internet Protocol) to the data portion of the packet having the MAC address.
tocol) Controls the transfer of a packet in which a source address and an IP destination address are written.

By the way, in such a coupling device,
In order to efficiently use a small communication capacity between networks connected by a line having a limited communication capacity, it is necessary to (a) compress data and (b) permit or prohibit the passage of a specific packet. Two methods were used.

[0006]

However, in the method (a), although the communication capacity of the line is apparently increased, the communication with low urgency and the communication with high urgency cannot be distinguished at the packet level. In a situation where the usage is increasing due to the presence of many low-degree communications and the lines are congested, communication with high urgency cannot be performed smoothly.

Further, in the method (b), since only the setting of permitting or not permitting the passage of the packet can be performed, only the emergency communication can be permitted in a situation where the lines are congested. Conversely, when the line is sufficiently free, communication with low urgency cannot be permitted.

[0008] Accordingly, an object of the present invention is to solve the above-mentioned problems by allowing communication of high urgency to be performed without interruption when lines connecting the networks are congested. The present invention is to provide a coupling device having a function of controlling the amount of packets passed so as to make communication with a low degree of urgency to be performed when a line for connecting the line is free and to reduce waste of the line.

[0009]

To achieve the above object, according to an aspect of the present invention, as shown in FIG. 1, the network N _1,
The coupling device 10 arranged between the N _2, packet throughput measurement unit 1-1 and 1-2, the packet passes through amount setting unit 2, the packet throughput comparator 3, packet throughput controller 4-1,4-
2, the buffer units 5-1 and 5-2, the packet transfer unit 6, and the like are provided.

In the packet passing amount setting unit 2, a pattern of a packet passing through the coupling device 10, a passing amount for the pattern, and the like are set in advance. The pattern can be set, for example, by classifying the source address and the destination address of the packet, or by classifying the packet by the size of the packet. The setting of the passing amount is a method of always giving priority to the pattern A and setting the remaining pattern to another pattern.

The pattern is set, for example, by M
A packet coming from the AC address 1 and going to the MAC address 2 is set as pattern A, and other packets are set as other patterns, or a packet having a packet size of 64 octets or less is set as pattern A, and other packets are set as other patterns.

The pattern is not limited to the two categories as described above, but may be of course classified into three or more appropriate types. The setting of the passing amount may be classified into three or more when the pattern is classified into three or more.

[0013]

The coupling device 10 previously sets the packet passing amount data indicating the packet pattern and the passing amount for the pattern in the packet passing amount setting unit 2 as described above. When the coupling device 10 starts operation, the network N is transmitted by the packet passing amount measuring units 1-1 and 1-2, respectively.
_1, starts to measure the passing amount with respect to the pattern and the pattern of packets communicated between the N _2. The packet sent from the interface 7 of the network N ₁ to the coupling device 10 is accumulated in all temporarily buffer 5-2 by the packet transfer unit 6. Similarly packets sent from the interface 8 of the network N ₂ are stored temporarily in the buffer unit 5-1 by the packet transfer unit 6.

The packet passing amount comparing unit 3 compares the actual packet passing amount data transmitted from the packet passing amount measuring units 1-1 and 1-2 with a set value set in the packet passing amount setting unit 2 in advance. By comparing the packets temporarily stored in the buffer units 5-1 and 5-2, the one with the highest priority is first transmitted so that the actual packet transfer control becomes the same as the set value. The packet passing amount control units 4-1 and 4-2 are controlled so as to transfer the packets.

As described above, according to the present invention, when the lines connecting the networks are congested, the communication with high urgency is prioritized, and when the line is free, the communication with low urgency is given priority. It is possible to provide a coupling device that can perform the operation without waste of the line.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described with reference to FIGS. FIG. 2 is a configuration diagram of one embodiment of the present invention, FIG. 3 is an explanatory diagram of a pattern setting example in the present invention, and FIG. 4 is an explanatory diagram of a passing amount setting example in the present invention.

In FIG. 2, the same symbols as those in the other figures indicate the same parts, 6-1 is a classification unit, and 9 is a passage amount setting unit. Packet throughput measurement unit 1-1 packet throughput and buffer unit 5-1 which is sent to the coupling device 10 from the network N ₁
And measures the packet throughput sent to the network N ₁ from. Here, the packet passing amount is to detect not only the number of packets and their octet length but also the data of the source and destination of the packet such as the source and the destination. Packet throughput measuring unit 1-2 is also intended to measure this Similarly packet throughput sent to the coupling device 10 from the network N ₂ and the packet throughput sent to the network N ₂ from the buffer section 5-2 is there.

The packet passing amount setting section 2 stores control data for controlling packets to be passed through the coupling device 10, and includes pattern setting data, which will be described in detail later with reference to FIGS. Setting data is stored.

The packet passing amount comparison unit 3 is connected to the network N
_A comparison is made as to whether or not the packet passing amount transmitted or transmitted from ₁ or the packet passing amount transmitted or transmitted from the network N ₂ is equal to the packet passing amount stored in the packet passing amount setting unit 2. based controls the packet throughput control unit 4-1 and 4-2 controls the packet fed into the network N _1, N _2.

The packet throughput control unit 4-1 is for controlling the packet fed into the network N _1, packet throughput control unit 4-2 controls a packet fed into the network N ₂ It is.

The buffer unit 5-1 is to hold temporarily the packet to be sent to the network N _1, the buffer unit 5-2 is intended for temporarily storing packets to be sent to the network N _2.

The packet transfer section 6 sends the transmitted packet to a corresponding buffer section on the network side based on the destination address.
For example, with a packet to be transmitted from the network N ₁ to N ₂ and sends to the buffer unit 5-2 and sends the packet to be transmitted from the network N ₂ to N ₁ in the buffer unit 5-1, separating the type of the packet Sorting section 6-1
Having. If the coupling device 10 is a bridge, the classification unit 6-1 sorts the MAC address of the source, the MAC address of the destination, and the like.
Is an IP router, the source IP address, destination IP address, TCP (Transmission
Control Protocol) / UDP (User Datagram Protocol)
ol) / ICMP (Internet control Message Protocol)
), The type of packet is classified based on TCP or UDP source port numbers, destination port numbers, and the like.

The interface 7 transmits and receives a packet to be transmitted from the network N ₁ or a packet to be transmitted to the network N ₁ . Interface 8 is for transmitting and receiving a packet to be sent into packets or network N ₂ to be transmitted from the network N _2.

The passing amount setting section 9 is for writing pattern setting data and passing amount setting data stored in the packet passing amount setting section 2. Next, based on FIG. 3 and FIG. Pattern setting and B. The setting of the passing amount will be described.

A. As shown in FIG. 3, for example, the pattern is set by (1) a method of classifying by a source address / destination address of a packet, (2) a method of classifying by an interface, and (3) a classification by a source port / destination port of a packet. And (4) a method of classifying based on packet size.

(1) A method of classifying packets by source address / destination address is such that a packet between specific hosts on different networks or a packet of a specific host is designated, and priority is given to the packet. When the coupling device is a bridge, a packet such as a packet coming from the MAC address 1 and going to the MAC address 2 or a packet from the MAC address 1 is written, and when the coupling device is a router, for example, Comes from IP address 1 and IP address 2
Or a packet from the IP address 1 or the like.

(2) The method of classifying by interface is such that, as shown in FIG. 7C, when there are three or more interfaces (in this example, there are four interfaces), control is performed so that priority is given based on the interfaces. ,
For example, items such as a packet coming from the interface 1 and going to the interface 2 and a packet passing through the interface 1 are written.

(3) A method of classifying packets by source port / destination port is a case where the coupling device is a router, and the packet is further transmitted by TCP (Transmission Control).
ol Protocol) or UDP (User Datagram Protocol)
In the case of the packet of l), the packet can be classified by the port number. Thus, as described above, the packet whose source port comes from the port number 1024 or more and whose destination port goes to the port number 23 is prioritized. Can be controlled. Here, the port having a port number of 1024 or more is a port used when calling the other party, and the port number 23 indicates a port number of the other party at the time of login. Therefore, this case is an example of calling the other party for login.

(4) The method of classifying by packet size is as follows:
This classifies whether priority should be given to the packet length. For example, it is conceivable that a short one is determined to be for controlling a protocol or the like and given priority, or a long one is determined to be a file transfer and controlled so that it may be delayed, such as a packet having a size of 64 octets or less, Size 10
The control state can be determined by classifying the packet into packets of 00 octets or more.

The pattern setting is performed in the above (1) to
(4) can be used not only alone but also in combination. For example, it can be defined as a packet going from IP address 1 to IP address 2 and a source port coming from port number 1024 or more and a destination port going to port number 23.

B. The passing amount is, for example, as shown in FIG.
(1) A method of setting a passing amount relative to a communication capacity of a line, (2) A method of setting a passing amount by an absolute value, (3) A method of setting a maximum value, and (4) A setting of a minimum value. A method is conceivable.

As a pattern setting, for example, as a pattern A, a packet coming from IP address 1 and going to IP address 2 and having a source port of port number 10
24 or more, and the destination port is a packet with the port number 23, and the pattern B has the IP address 1
When the packet is transmitted to the IP address 2 and the port number is defined as a pattern other than the pattern A and the pattern C is defined as another pattern, the passing amount can be set as follows.

(1) As a method of setting the passing amount relative to the communication capacity of the line, for example, a pattern A
Is 50% of the circuit capacity, Pattern B is 20% of the circuit capacity,
Other patterns are set to 30% of the line capacity.

(2) As a method of setting the passing amount by the absolute value, for example, when the line capacity is 64 kbs, the pattern A is 30 kbs, the pattern B is 20 kbs, and the other patterns are the remaining 14 kbs. is there.

(3) As a method of setting the maximum value, for example, as shown in FIG.
Up to 0%. As a result, half of the line capacity can always be kept free for other patterns.

(4) As a method of setting the minimum value, for example, as shown in FIG.
%. This guarantees that pattern B can communicate at 20%, no matter how busy the line.

By the way, when calculating the amount of passing packets,
It is practically meaningless to find the average amount of packet passing for a long time from the start of the operation of the coupling device 10 to the present.
It is sufficient to check the amount of packets that have passed within the time obtained from the following equation according to the communication speed (communication capacity) of the interface for which the coupling device 10 is trying to control the amount of packets passing.

[0038] Here, N is the number of packets, which may be about 10 to 1000, but preferably about 100. Although the above expression is an example in which the number of octets, that is, the size of a packet, is expressed, the amount of passing packets can be measured by either the number of packets or the size of a packet.

The operation of FIG. 2 showing the configuration of one embodiment of the present invention will be described. In this case, if the coupling device 10 is an IP router, the pattern setting unit 9 sets, for example, the pattern of FIG. FIG.
(2) is set. Here, the packet passing amount measuring units 1-1 and 1-2, the packet passing amount control units 4-1 and 4-
2. Buffer units 5-1 and 5-2 are provided for each interface.

Since the coupling device 10 cannot control the incoming packet, the packet passing amount control unit 4-1,
The control is performed only for the packet going out according to 4-2 based on the control signal from the packet passing amount comparison unit 3.
However, the packet passing amount measuring units 1-1 and 1-2 measure the passing amount of both incoming packets and outgoing packets.

When the coupling device 10 starts operating, the packet passing amount measuring units 1-1 and 1-2 measure the packet passing amount, and notifies the packet passing amount comparing unit 3 of this. All packets received by the coupling device 10 are temporarily stored in the buffer unit 5.
-1, 5-2. The packet passing amount comparing unit 3 compares the actual packet passing amount (including the pattern) with the passing amount (including the pattern) set in the packet passing amount setting unit 2 in advance, and the result is the same as the set value. Thus, control is performed such that a packet having a higher priority among packets stored in the buffer units 5-1 and 5-2 is transferred first.

At this time, the passing amount of the packet is set in advance for each type of packet, and it is checked which type of each passed packet belongs.
The passing amount is measured for each type of packet.

As described above, all the packets received by the coupling device 10 are temporarily stored in the buffer units 5-1 and 5-2. Then, the actual packet passing amount for each type of packet is compared with a preset set value, and the buffer unit 5-
By transferring the packet with the highest priority among the packets stored in the packets 1 and 5-2 first, the amount of passage of the packet is controlled for each type of the packet.

When storing the packets in the buffer units 5-1 and 5-2, the classification unit 6-1 determines in advance which of the packet pattern settings set by the packet passing amount setting unit 2 belongs to the classification unit 6-1. By storing the packets together with the pattern type information in the buffer units 5-1 and 5-2, it is not necessary to determine the type of the packet at the time of actual packet transfer, and the processing speed is improved.

A second embodiment of the present invention will be described with reference to FIGS. In the second embodiment, a clock 11 is provided, and a packet passing amount comparing unit 3 is provided with a time comparing unit 3-1 and a time passing amount setting unit 3-2. In addition, a grace time determination unit 3-3 and a timer 12 are provided in addition to these.

When no grace time is set in the packet passing amount setting unit 2, for example, FIGS.
As shown in (1), pattern setting, passing amount setting, time setting and the like are performed.

As shown in FIG. 6 (A), the pattern setting includes a packet having a size of 1000 octets or more as a pattern P, a packet having a size of 64 octets or less as a pattern Q, a pattern P and a pattern R as a pattern R.
Define packets other than Q.

As shown in FIG. 6B, the passing amount is set to 50% for the pattern P, 20% for the pattern Q, and 30% for the pattern R. Then, as shown in FIG. 6C, the time is set as shown in (B) from 9:00 to 17:00, and the pattern P is set to 70% from 17:00 to 9:00. When the pattern Q is 10% and the pattern R is 20%, for example, the file transfer size is 1
The allocation of packets of 000 octets or more is increased, and the setting is made in consideration of the convenience of file processing.

In the second embodiment of the present invention shown in FIG. 5, the clock 11 for knowing the current time and the time comparing section 3-1 are used by the time comparing section 3-1 based on the time zone information obtained from the clock. Then, the packet passing amount setting data shown in FIG. 6C or as shown in FIG. 6C is extracted, and based on this, the time passing amount setting unit 3-2 controls the packets output from the buffer units 5-1 and 5-2.

As shown in FIG. 5, a timer 12 is provided, and if the packet passing amount setting data illustrated in FIG. And the line capacity is 5
When the state exceeds 0%, the grace time determination unit 3-3
Operates immediately and performs timer monitoring. If the packet transfer of the pattern P is completed within 60 seconds of the grace time, the operation can be continued as it is. In this way, for example, in the case of file transfer, even if the setting is temporarily violated on the way, the process can be smoothly performed by permitting this. In this case, if the setting violation state continues even after the grace period has passed, the transfer process is interrupted, but the packet can be retransmitted because the packet remains in the buffer unit.

In the above description, an example of a coupling device for coupling two networks has been described. However, the present invention is not limited to this, but is also applicable to a coupling device for coupling three or more networks. Of course, it can be applied.

[0052]

According to the present invention, when setting the amount of passing packets, if it is known that communication between a specific host and a specific host is low in importance but high in communication, By setting an upper limit value for the amount of passing packets of a specific communication, it is possible to prevent the line from becoming 100% congested.

On the other hand, if the communication line is usually congested no matter how congested, but when a particular communication occurs, the communication is so important that it must be performed even if other communication is excluded. By setting a lower limit value for the amount of packet passing of the specific communication, the specific communication can always be performed at a certain rate or more regardless of the degree of line congestion.

In addition, it is necessary to give priority to a specific communication from 8:00 am to 5:00 pm. This can be realized by setting a limit value of the packet passing amount together with the information.

Further, for a sudden increase in the amount of passing packets, the passing of the packets is allowed for a short time.
This can be realized even when it is desired to limit the amount of packet passing when communication is performed over a certain period of time.

[Brief description of the drawings]

FIG. 1 is a principle configuration diagram of the present invention.

FIG. 2 is a configuration diagram of an embodiment of the present invention.

FIG. 3 is an explanatory diagram of pattern setting in the present invention.

FIG. 4 is an explanatory diagram of a passing amount setting in the present invention.

FIG. 5 is a configuration diagram of a second embodiment of the present invention.

FIG. 6 is an explanatory diagram of setting a packet passing amount in the present invention.

FIG. 7 is an explanatory diagram of a network connection state.

[Explanation of symbols]

 1-1, 1-2 Packet passing amount measuring unit 2 Packet passing amount setting unit 3 Packet passing amount comparing unit 4-1, 4-2 Packet passing amount control unit 5-1, 5-2 Buffer unit 6 Packet transfer unit

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04L 12/46 H04L 12/56 H04L 12/66

Claims (3)

(57) [Claims] 1. An inter-network connecting apparatus for connecting a plurality of networks, a clock unit for knowing a current time, and a packet for setting a packet passing amount together with time zone information.
A packet passing amount setting unit, a packet passing amount measuring unit for measuring a packet passing amount, a packet pattern, a packet passing amount setting unit for setting a passing amount, and a packet passing amount setting unit and a packet passing amount. A packet passing amount comparison unit that compares the setting value entered in the packet passing amount setting unit; a packet passing amount control unit that controls the packet passing amount so that the comparison result is within the setting value ; An inter-network coupling device comprising a time comparison unit for comparing with a current time . 2. The time from the start of a specific communication is counted.
Set the packet passing amount along with the timer to be transferred and the grace time information of the timer
Characterized by having a packet passing amount setting unit
The network coupling device according to claim 1. 3. A network for connecting a plurality of networks.
In the interworking coupling device, a classification unit for classifying a packet type and a packet passage amount meter for measuring a passage amount for each packet type
Measurement unit and a packet passing amount setting that sets the passing amount for each packet type.
And the actual packet passing amount for each packet type,
Compare with the setting value set in the packet passing amount setting unit
A packet passing amount comparing unit and a packet passing amount comparing unit based on the comparison result.
To control the amount of passing packets by packet type.
And a packet passing amount control unit.
An interworking coupling device.