JPH07115421A - Atm exchange - Google Patents

Abstract

PURPOSE:To easily extend the scale by providing a first mounting unit where a first partial ATM switch is stored and a second mounting unit where a second partial ATM switch which realizes the switching function extended to the first mounting unit is stored. CONSTITUTION:A 16X16-scale exchange switch consists of two 8X8 switch cards 2 stored in two interface racks 1 and one inter-rack switch card 5 for 16X16 stored in a switch rack 4. Intra-switch cells inputted to or outputted from the extension terminal of the card 2 is the output to the rack other than its own rack or the input from the rack other than its own rack. Since only two interface racks exist, it is sufficient if extension output terminals of racks 0 and 1 are connected to extension input terminals of racks 1 and 0 respectively in 1:1. Thus, the switch function is not required in the card 5, and only a wiring exists there.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a switching switch used in an ATM communication system, and particularly to the expansion switch A.
Regarding the TM switch.

[0002]

2. Description of the Related Art Broadband ISDN (Integrated)
(Services Digital Network)
ATM (Asynchronous ran)
In the fer mode) communication system, all data to be communicated is converted into a fixed length packet of 53 bytes (this is called a cell) and exchanged / transmitted. In order to exchange this cell between lines, between terminals or between lines and terminals, the AT
An M switch is used. FIG. 10 shows a block diagram of an ATM exchange. As can be seen from the figure, the ATM exchange functions at least as an interface with the public line / dedicated line and terminals, and exchanges cells between the interface card that executes connection with the ATM switch and the interface cards. And an interface card and a control unit for controlling the ATM switch unit.

Now, consider the case where a cell output from a terminal is output to a public line or a leased line. The terminal is a 150M-TEI (Terminal Interface) of an ATM exchange.
ACE) interface card, and the public line or leased line is 150M- of an ATM exchange.
UNI (User-Network Interface)
e) is connected to the interface card.

In response to a call connection request from the terminal, 150M-
150M from the TEI to the control unit (CPU) of the exchange
-Notify to connect with UNI. CPU is 150M
-The connection path (routing) of the ATM switch is calculated from the accommodation information in the exchange of TEI and 150M-UNI, and the connection identifier VPI / V to 150M-TEI is calculated.
The value of the routing tag indicating the routing information is notified together with the CI and the like.

Here, the routing tag will be described. In the ATM communication system, while the conventional circuit switching network distinguishes the unit of data transmission and switching by the time position represented by the time slot, it is distinguished by the label represented by the destination or the connection identifier. However, since the label of the cell handled in the network is different each time a call is made, the label that is different for each call is not changed as it is.
In the M switch, it is necessary to translate into routing information, which is difficult to handle inside the exchange. Therefore, the processing inside the exchange is simplified by adding a routing tag indicating the routing information.
At this time, the cells input at each interface point are UN
A cell to which an I cell and a routing tag are added is called an in-switch cell. FIG. 11 shows the format of the in-switch cell to which this routing tag is added.

[0006] 150M-TEI rewrites the header of the cell input from the terminal based on the routing information, adds a routing tag, and inputs the in-switch cell to the switch. In the switch, the switch refers to the above-mentioned routing tag to move the cell in the switch to a predetermined output position,
Here, the channel is connected to 150M-UNI. In this way, a method in which the switch itself is autonomously connected by the switch to be exchanged by the routing tag shown in the header of the cell in the switch itself is called the self-routing method, without the direct operation of the control unit. .

FIG. 12 shows an example of an IC unit switch. This includes 8 input channels for cells in the switch.
There are eight output channels, and there is an input terminal of a stage number for designating a reference bit position of a stage number of a routing tag, which will be described later, when guiding an input in-switch cell to a predetermined output.

The stage number means, for example, when a unit switch is connected to a Banyan, 1, 2 from the input side,
3 ... The switch position information is given a number for each stage. In the unit switch, the bit position to be referred to for each stage is determined in the routing tag part of the cell in the switch, and the bit position corresponding to the stage number given to itself is referred to,
The input switch cell is guided to the output channel.

For example, in the case where the in-switch cell input to the switch whose stage number is 1 is output to the channel 5, the routing tag (R-ta) shown in FIG. 11 is used.
(10) at the reference bit position (ST1) of stage 1 of (g).
You can enter 1) 2. This routing tag (R-
As shown in the figure, the tag) is added to the head of the UNI cell 53 bytes at the interface section of the terminal or line.

Now, assuming that the above-mentioned unit switch has an exchange capacity of 8 inputs and 8 outputs (8 × 8), for example,
A plurality of this unit switch can be used to configure a 16 × 16 exchange or a 32 × 32 exchange.

When an exchange is actually constructed, an interface card for mounting an interface circuit, a switch card for mounting a switch circuit, a power supply, etc. are called a rack with eight interface cards as a unit in order to be efficiently housed in a housing. It is housed in one implementation group. Since the number of interface cards accommodated in one rack is eight, two racks are used for a 16 × 16 scale, and four racks are used for a 32 × 32 scale.
13 and 14 are block diagrams of the switch portion of this exchange, and FIGS. 15 and 16 show the external appearance of the housing having this configuration.

In the configuration diagrams of FIGS. 13 and 14, four unit switches are connected in two stages in FIG. 13 to form a 16 × 16 exchange. In FIG. 14, 16 unit switches are connected in four stages to form a 32 × 32 exchange. The method of connecting the unit switches shown in these figures is called shuffle connection. In this connection method, when the function of the switch is expanded to the above-described configuration, the connection between the unit switches is extremely simple, as can be seen from the cable connection shown by the thick lines in FIGS. 13 and 14, The number of cable connection points between switch cards can be reduced, which is superior in terms of equipment maintenance.

Since the exchange is composed of a plurality of racks as shown in FIGS. 15 and 16, the portion surrounded by the one-dot chain line in FIGS.
X16 switch, one card for each rack, 3
In a 2 × 32 exchange, function allocation is realized by mounting two of these cards for each rack. This switch card is the one used in the 16x16 switch,
Among those used in the 32 × 32 exchange, the one sharing the input side stage is called SW2, and the one sharing the output side stage is called SW3.

Next, a method of adding a routing tag will be described.

First, a 16 × 16 structure will be described with reference to FIG. In FIG. 17A, as a first example,
A case of routing a cell in a switch from 7 channels of rack 1 to a channel of 3 in rack 1 and a second example, a case of routing a cell in a switch from channel 7 of rack 1 to 3 channels of rack 0 will be described. To do. As can be seen from the figure, in these examples, the switches are assigned stage numbers 4 and 5 from the input side. The cell in the switch input from the left side in FIG. 17 (a) is switched by referring to the routing ST4 shown in FIG. 17 (b) in the first stage, and is referred to in the routing ST5 in the next stage to be guided to the right side. Get burned.

Therefore, in the first example of FIG. 17, ST4 =
3 and ST5 = 3, the cells in the switch are switched through the solid line route shown in FIG. 17A, and in the second example, ST4 = 7 and ST5 = 3 are set.
The cells in the switch are switched along the route of the dotted line shown in (a). Generalizing the method for obtaining this routing tag is as follows.

The rack number and channel number on the input side are AI
Din, output rack number and channel number AIDo
ut, and these are represented by binary numbers, and these are represented by AIDin = abc def AIDout = ghi jkl. However, abc and ghi are rack numbers, def and jkl are channel numbers, respectively, "a",
"G", "d", and "j" are MSBs. Then ST
4 and ST5 are represented by ST4 = Pef (1) P = c (+) i (2) ST5 = jkl (3). However, (+) represents an exclusive OR operator.

Next, a 32 × 32 configuration will be described with reference to FIG. 18 as an example. In FIG. 18A, as a third example, the case where the in-switch cell is routed from 7 channels of rack 1 to 3 channels of rack 1, and as a fourth example, 7 channels of rack 1 to 3 channels of rack 0 are used. The case of routing a cell in a switch to a channel will be described.

As shown in the figure, in these examples, each switch is assigned a stage number 1, 2, 4, 5 from the input side. Therefore, the cells in the switch input from the left side in FIG.
Referring to ST1 of the routing shown in (b), ST2 of the routing in the next stage, and ST4 and ST5 in the stages 4 and 5, switching is performed and the light is guided to the right side.

Therefore, in the third example, ST1 = 7, ST1
By setting 2 = 7, ST4 = 7, and ST5 = 3, FIG.
The cells in the switch are switched along the solid line route shown in (a), and in the fourth example, ST1 = 7, ST2 = 7, S
By setting T4 = 3 and ST5 = 3, the in-switch cell is switched along the dotted line route shown in FIG. Generalizing the method for obtaining this routing tag is as follows.

As in the case of 16 × 16, AIDin, A
It is assumed that IDout is represented by a binary number, and these are represented by AIDin = abc def AIDout = ghi jkl. However, abc and ghi are rack numbers, def and jkl are channel numbers, respectively, "a",
"G", "d", and "j" are MSBs. Then ST
1, ST2, ST4 and ST5 are calculated by ST1 = Pef (4) P = c (+) h (5) ST2 = ST1 (6) ST4 = Qkl (7) Q = b + i (8) ST5 = jkl (9) To be done. However, (+) represents an exclusive OR operator.

As described above, the conventional ATM switch is realized by dividing the switch circuit so as to be suitable for the case structure shown in FIGS.
Three types of W1, SW2, and SW3 were prepared, and SW1 was used for 16 × 16, and SW2 and SW3 were used for 32 × 32. Therefore, the 16 × 16 scale and the 32 × 32 scale have different casings and different switch cards are used. Therefore, in order to expand the function from the 16 × 16 scale exchange to the 32 × 32 scale exchange, the chassis and the switch card must be exchanged.

Further, there is a problem that the routing tag attachment method is different between the two, and the software provided to the control unit for calculating the routing tag is different. For example, in the example of FIG. 17 and FIG. 18, the same input is also applied to the first and third examples in which the cells in the switch are routed from 7 channels of the same rack 1 to 3 channels of the rack 1. Even when the switch cell is routed to the same output using, the switching direction in the stage 4 is reversed due to the difference between the above formulas (1) and (4).

As described above, the conventional ATM switch configuration has a problem in that there is no expandability in terms of hardware and inconsistency in the method of calculating routing tags in terms of software.

In order to solve these problems, for example, there is a method in which the unit switches are connected in a cross-point connection, or there are three stages as shown in FIG. This allows
In terms of hardware, it is possible to maintain extensibility, and also in terms of software, the method of calculating routing tags can be unified. However, as can be seen from FIG. 19, the connection between the unit switches becomes complicated, and therefore a large number of cables for connecting the switches are required,
The maintenance advantage is lost.

[0026]

As described above, the conventional AT is used.
In the configuration of the M switch, the switch structure of the 16-line scale and the switch of the 32-line scale are different from each other in terms of the housing structure and the switch card used, and there is no expandability in terms of hardware, and a routing tag calculation method in terms of software. There is a problem that the processing programs for the routing cannot be shared because there is no uniformity.

In order to solve these problems, if the unit switches are cross-point connected or have three stages, the connection between the unit switches becomes complicated and the advantage in maintenance is impaired. The problem arises that

The present invention solves these problems and considers expansion, which has expandability in terms of hardware, can share processing programs in terms of software, is easy to connect between unit switches, and is easy to maintain. The purpose is to realize an ATM switch.

[0029]

A first mounting unit for accommodating an ATM switch, which is connected to interface means, and which accommodates a first partial ATM switch which realizes a closed switching function in itself, and a first mounting unit. A second mounting unit for accommodating a second partial ATM switch that realizes a switching function across mounting units.

Further increasing the interface means A
When increasing the capacity of the TM switch, the first part A
The TM switch increases the number of first mounting units without changing the contents, and the second partial ATM switch changes according to the accommodation scale.

The first partial ATM switch and the second partial ATM switch are composed of a plurality of small-scale unit ATM switches.

Furthermore, the bit position to be referred to in the route information written in the header portion of the cell in the switch is the first portion AT.
Allocation is made to each of the unit ATM switches mounted on the M switch and the second partial ATM switch, and the calculation formula for calculating the allocation information of the route information is made the same regardless of the accommodation scale of the ATM switch.

[0033]

As a result, it is possible to realize an ATM switch which is highly expandable, the processing program can be used as it is even if expanded, connection between unit switches is simple, and maintenance is easy.

[0034]

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

FIGS. 1, 2 and 3 respectively show a 16 × 16-scale ATM switch according to an embodiment of the present invention, and 32.
FIG. 3 is an external view of a × 32 scale ATM switch and an 8 × 8 scale ATM switch.

As shown in the figure, the enclosures of these exchanges adopt a building block system in which they are stacked in units of mounting units (interface rack 1) capable of accommodating eight cards. Therefore, the racks 0 to 3 shown in the figure are the same, and the housing has a building block structure. Therefore, as shown in the figure, the interface rack 1 has one stage, and the case of 16 × 16 has this interface rack 1. Interface rack 1 in 2 tiers, 32x3
In case of 2 scales, it is constructed by stacking 4 stages.

In the case of 8 × 8 scale, one interface rack (INF RACK) 1 as shown in FIG.
Implement. The interface rack has an 8x8 switch card (SW8 PK) that realizes the exchange of cells in the switch between the interface cards housed in the rack.
G) 2 is implemented.

FIG. 4 shows an 8 × 8 switch card (SW8 PKG) used in the embodiment of the present invention shown in FIGS.
2 is a block diagram of FIG. In the figure, 3 is an ATM switch LSI having an 8 × 8 exchange capability, which is the same LSI as the USW used in a conventional exchange.

In the case of 8 × 8 scale, the 8 × 8 switch card (SW8 PKG) of FIG. 4 is housed in the rack 0 shown in FIG. 3, and its input terminals i1 to i7 are wired to the interface housing position of the rack. It is connected to the output terminal of the interface card. The output terminals o0 to o7 of the 8 × 8 switch card are also wired to the interface accommodating position of the rack and connected to the input terminals of the interface card.

Switching in the rack of this 8 × 8 ATM switch is performed as follows. In FIG. 4, the 8 × 8 switch card (SW8 PKG) 2 outputs to the output terminals o1 to o7 while referring to the switching information indicated by the routing tag shown in the header part of the in-switch cell of the input in-switch cell. Have the ability to

In the example of FIG. 4, the in-switch cells input to the input terminals i1 to i7 are first input to the switch LSI3 having the stage number 1, where the routing tag ST1 is referred to, and the output terminal of the LSI3. o0-o3
Be led to one of. And these output terminals o0-o3
Is connected to the input terminals i0 to i3 of another switch LSI3 whose stage number is 4, and then the routing tag ST4 is referred to here, and the in-switch cell outputs the output terminal of this LSI3 accordingly. It is led to any of o0 to o7.

In FIG. 4, it is not connected to anything indicated by NC. Output terminal o4 of LSI3 of stage number 1
o7 and the output terminals i4 to i7 of the LSI 3 having the stage number 4
8x8 switch card of rack 1 other than the rack in which its own 8x8 switch card (SW8 PKG) 2 is mounted when expanded to 16x16 scale or 32x32 scale as described later. Lead the cell in the switch to 2,
Alternatively, terminals for guiding the in-switch cells to the own rack 1 from the 8 × 8 switch card 2 of the rack 1 other than the rack in which the own 8 × 8 switch card 2 is mounted, and these terminals are called expansion terminals. .

In this case, any of the racks 1 can be configured to be closed in the rack 1 by appropriately arranging the function of the switch in the housing of the building block structure. That is, it is convenient to use the above-mentioned configuration of the 8 × 8 switch card 2 as it is for exchanging the cells in the switch between the interface cards 2 accommodated in the same rack 1.

On the other hand, in order to exchange the cells in the switch between the interface cards extending over the upper and lower racks 1, it is necessary to newly provide a switch card for switching between racks. This switch card is provided with a switch rack (SW RACK) 4 as shown in FIGS. Then, the wiring to the switch rack (SW RACK) 4
This is performed from the expansion terminal of the 8 × 8 switch card 2 accommodated in the SW8 PKG accommodation slot of the interface rack 1. Inter-rack switch cards of different types are prepared for the switch 16 of 16 × 16 scale and the switch 6 of 32 × 32 scale, and are mounted on the switch rack 4.

As described above, when the exchange is composed of the interface rack 1, the switch rack 4, the 8 × 8 switch card 2 and the inter-rack switch card 5 or 6, the expansion of the exchange function, that is, the addition of the exchange function is far more than the conventional one. To be easier. For example, when expanding from the 8 × 8 exchange shown in FIG. 3 to the 16 × 16 exchange shown in FIG. 1, one interface rack 1 and one switch rack 4 and a 16 × 16 inter-rack switch card 5 are added. If this 16 × 16 switch is expanded to a 32 × 32 switch as shown in FIG. 2, two interface racks 1 and one switch rack 4 are added.
This can be realized by adding two 32 × 32 inter-rack switch cards 6 instead of the 16 × 16 inter-rack switch card 5. Figure 5 shows the units for such an expansion.
It was shown to.

Next, the inter-rack switch card will be described.

FIG. 6 shows the configuration of a switch used in a 16 × 16 scale exchange. The switch of the 16x16 scale switch has two 8x8 switch cards (SW8 PKG) 2 stored in two interface racks 1 and a switch card for 16x16 racks (SW16 PKG) stored in the switch rack 4. ) 5 is composed of one sheet.

As can be seen from the figure, the expansion terminal of the 8 × 8 switch card (SW8 PKG) 2 is connected to the 16 × 16 inter-rack switch card (SW16 PKG) 5. In the case of a 16 × 16 scale exchange, the in-switch cells input / output at the expansion terminals of the 8 × 8 switch card (SW8 PKG) 2 are outputs to other than the own rack or inputs from other than the own rack. In the case of a 16 × 16 scale exchange, since there are only two interface racks 1, the expansion output terminal of rack 0 is connected to the expansion input terminal of rack 1 and the expansion output terminal of rack 1 is connected to the expansion input terminal of rack 0 in a pair. It is clear that connection with 1 is sufficient. Therefore,
16x16 switch card between racks (SW16 PK
G) 5 does not need a switch function, only wiring exists.

FIG. 7 shows the configuration of a switch used in a 32 × 32 scale exchange. The switch of the 32 × 32 scale switch is a switch card between racks for 4 × 8 switch cards (SW8 PKG) 2 stored in four interface racks 1 and for 32 × 32 racks stored in two switch racks 4. (SW32 PKG) 6 is composed of two sheets.

8 stored in the interface rack 1
The extension terminal of the × 8 switch card (SW8 PKG) 2 is a 32 × 32 inter-rack switch card (SW32
PKG) 6 terminal. In this case as well, the in-switch cells input / output at the expansion terminals of the 8 × 8 switch card (SW8 PKG) 2 are outputs to other than the own rack or inputs from other than the own rack. However, since there are four interface racks 1, a 32 × 32 inter-rack switch card (SW32) is provided so that the expansion output terminal of any rack can be connected to the expansion input terminal of any rack.
The PKG) 6 is provided with eight switch LSIs 3 of two stages of the stage 2 and the stage 3 and arranged as shown.

In this case, since a plurality of (two in this case) switch cards between racks are used, wiring between switch cards between racks is required.
As shown by the thick line in, the connection of the cable is extremely simple and there are few connection points, so it is superior in terms of maintenance.

Next, let us consider the routing of cells in a switch in the case of such a switch arrangement.

First, the case of a 16 × 16 scale exchange will be described with reference to FIG. In FIG. 8A, as a first example, a case where a cell in a switch is routed from 7 channels of rack 1 to 3 channels of rack 1, and as a second example, 7 channels of rack 1 to 3 channels of rack 0 are used. Consider the case of routing a cell in a switch to a channel.

In the case of a 16 × 16 exchange, the switch LSI 3 in the 8 × 8 switch card 2 is assigned stage numbers 1 and 4 from the input side. Therefore, FIG.
In (a), the in-switch cell input from the left side of the figure is
At the first stage, referring to ST1 of the routing tag shown in FIG. 8B, and at the next stage, referring to ST4, switching is performed to lead to the right side.

Therefore, in the case of the first example, the routing tags are ST1 = 3 and ST4 = as shown in (1) of FIG. 8B.
By setting 3, the cells in the switch are switched along the route of the solid line in FIG. 8A, and in the case of the second example, the routing tag is ST1 = 7, as shown in (2) of FIG. 8B. ST
By setting 4 = 3, the in-switch cell is switched along the route of the dotted line in FIG.

The calculation method of this routing tag is generalized as follows.

The rack number and channel number on the input side are AI
Din, output rack number and channel number AIDo
If ut is expressed as a binary number, AIDin = abc def AIDout = ghi jkl where abc and ghi are rack numbers, and def and jkl are channel numbers, respectively, “a”, “g”, “d”, and “j”.
Is the MSB. Then, ST4 and ST5 are calculated by ST4 = Pef (10) P = c (+) i (11) ST5 = jkl (12). However, (+) represents an exclusive OR operator.

Next, a case of a 32 × 32 scale exchange will be described with reference to FIG. In FIG. 9A, as a third example, the case where the in-switch cell is routed from 7 channels of rack 1 to 3 channels of rack 1, and as a fourth example, 7 channels of rack 1 to 3 channels of rack 0 are used. Consider the case of routing a cell in a switch to a channel.

As shown in the figure, in the case of a 32 × 32 scale exchange, the switch LSI 3 in the 8 × 8 switch card 2
Are assigned stage numbers 1, 2, 3, 4 from the input side. Therefore, in FIG. 9A, the in-switch cell input from the left side of the figure is shown in FIG. 8B at the first stage.
Reference is made to ST1 of the routing tag shown in (1), ST2 is referred to in the next stage, and ST3 and ST4 are further referred to, so that switching is performed and the light is guided to the right side. However, the third
32 x 3 when the input and output racks are the same as in the example
Since the in-switch cell is not output to the switch card (SW32 PKG) 6 between racks for two, the values of ST2 and ST3 are not referred to, and therefore any value may be used for ST2 and ST3.

Therefore, in the case of the third example, the routing tags are ST1 = 3 and ST4 = as shown in (3) of FIG. 9B.
By setting 3, the cells in the switch are switched along the route of the dotted line in FIG. 9A, and in the case of the fourth example, the routing tag is ST1 = 7, as shown in (4) of FIG. 9B. ST
By setting 2 = 3, ST3 = 1, and ST4 = 3, FIG.
The cells in the switch are switched along the route indicated by the solid line in (a).

The method of calculating the routing tag is generalized as follows.

As in the case of 16 × 16, if the rack number and channel number on the input side are AIDin, the rack number and channel number on the output side are AIDout, and expressed in binary numbers, AIDin = abc def AIDout = ghi jkl , Abc, ghi are rack numbers, and def, jkl are channel numbers, respectively, “a”, “g”, “d”, “j”.
Is the MSB. Then, ST1, ST2, ST3 and ST4 are: ST1 = Pef (13) P: abc = ghi P = 0 abc ≠ ghi P = 1 (14) ST2 = Qef (15) Q = b (+) h (16) ST3 = ikl (17) ST4 = jkl (18) However, (+) represents an exclusive OR operator.

As described above, in the present invention, the ATM switch is composed of the interface rack 1, the switch rack 4, the intra-rack switch card 2 and the inter-rack switch cards 5 and 6, so that the size of 8 × 8 to 16 × 16.
When expanding to scale, interface rack 1, switch rack 4 and switch card between racks (SW16P
KG) 5 only, and when expanding to a 32 × 32 scale, interface rack 1 and switch rack 4 are added, and switch rack switch switch SW 16 is added.
It can be realized by replacing and adding PKG5 to SW32 PKG6. Therefore, it is possible to easily expand the exchange capacity, that is, add the capacity as compared with the conventional one. Further, since the switch function is properly arranged in the building block structure of the housing, the switch card of the 8 × 8 scale (SW8 PKG), which is the smallest configuration regardless of the scale, is used when expanding the replacement scale. ) 2 can be used, and the switch card (SW16 PKG, SW32 PK) that switches between racks is used.
G) Since the switch rack 4 accommodating 5 and 6 and the interface rack 1 are separated, the interface rack 1 having the same structure can be used regardless of the scale, and it is highly compatible in this respect. In addition, the number of connection cables between switch cards can be realized at the same level as in the past, which is advantageous in terms of maintenance.

On the other hand, regarding the calculation method of the routing tag, the allocation of the bit position of the routing tag referred to in each stage of the switch network is performed by the switch card (SW8).
The reference position of the switch LSI mounted on the PKG),
By allocating and distinguishing from the reference position of the switch LSI mounted on the switch card (SW32PKG) exclusively for each stage, the switching operation differs depending on the scale even though the same input / output combination on the same stage as in the past is used. Such a defect can be eliminated. Therefore, in terms of software, the software provided to the control unit for calculating the routing tag can be the same regardless of the scale, and does not need to be changed even if the switch scale changes.

[0065]

As described above, according to the present invention, the ATM
Switchboard, interface rack, switch rack,
Since the switch card is configured with the intra-rack switch card and the inter-rack switch card, expansion to the scale can be realized by replacing the inter-rack switch card and adding other parts. Also, because the switch functions are properly arranged in the building block structure of the chassis, the same structure of interface rack can be used regardless of the scale, and the number of connecting cables between switch cards is the same as before. Absent. Furthermore,
Regarding the method of calculating the routing tag, assigning the bit position of the routing tag to the reference position of the switch LSI mounted on the switch card and the reference position of the switch LSI mounted on the switch card for each stage are distinguished. As a result, the software provided to the control unit can be the same regardless of the scale. Therefore, it is possible to easily expand the exchange capacity, that is, to add the capacity as compared with the conventional one.

[Brief description of drawings]

FIG. 1 is an external view of a 16 × 16 ATM switch that is an embodiment of the present invention.

FIG. 2 is an external view of a 32 × 32 ATM switch that is an embodiment of the present invention.

FIG. 3 is an external view of an 8 × 8 ATM switch that is an embodiment of the present invention.

FIG. 4 is a block diagram of an 8 × 8 switch card used in an embodiment of the present invention.

FIG. 5 is a block diagram of a 16 × 16 rack-to-rack switch card used in an embodiment of the present invention.

FIG. 6 is a block diagram of a 32 × 32 rack-to-rack switch card used in an embodiment of the present invention.

FIG. 7 is an external view of an extension unit according to the embodiment of the present invention.

FIG. 8 is a diagram showing an example of routing of cells in a switch of the 16 × 16 ATM switch shown in FIG. 1.

FIG. 9 is a diagram showing an example of routing of cells in a switch of the 32 × 32 ATM switch shown in FIG.

FIG. 10 is a block diagram showing an outline of an ATM switch.

FIG. 11 is a unit switch LS that constitutes an ATM switch.
Block diagram of I.

FIG. 12 is a cell format diagram of cells in a switch handled inside the ATM switch.

FIG. 13 is a block diagram of a 16 × 16 switch card of a conventional ATM switch.

FIG. 14 is a block diagram of a 32 × 32 switch card of a conventional ATM switch.

FIG. 15 is an external view of a conventional 16 × 16 ATM switch.

FIG. 16 is an external view of a conventional 32 × 32 ATM exchange.

FIG. 17 is a diagram showing an example of routing of cells in a switch in a conventional 16 × 16 ATM switch.

FIG. 18 is a diagram showing an example of routing of cells in a switch in a conventional 32 × 32 ATM switch.

FIG. 19 is a block diagram of switches in a 32 × 32 ATM switch configured by using a conventional 3-stage cross network.

[Explanation of symbols]

1 Interface rack 2 Switch card in rack (SW8 PKG) 3 8x8 ATM switch LSI 4 Switch rack 5 16x16 Switch card between racks (SW16 PKG)
KG) 6 32 x 32 rack switch card (SW32 P
KG)

Claims (5)

[Claims] 1. An ATM switch which accommodates a plurality of lines and a plurality of terminals and performs a cell switching operation between the interface means for the lines and the terminals, wherein the ATM switch is connected to the interface means and within itself. A first exchange circuit that realizes a closed exchange function, a first mounting unit that implements the first exchange circuit, and a second exchange circuit that realizes the exchange function across the first exchange circuit, An ATM exchange comprising: a second mounting unit for accommodating the second exchange circuit. 2. The number of interface means is increased to A
When increasing the accommodation scale of the TM exchange, the first exchange circuit increases the number of the first mounting units without changing the contents, and the second exchange circuit changes according to the accommodation scale. An ATM switch according to claim 1, wherein: 3. The A according to claim 1, wherein the first switching circuit and the second switching circuit are composed of a plurality of small-scale unit ATM switches.
TM exchange. 4. The unit ATM in which the bit position to be referred to in the route information written in the header portion of the in-switch cell is mounted in the first switching circuit and the second switching circuit.
4. The ATM exchange according to claim 3, wherein the ATM exchange is assigned to each switch. 5. The ATM exchange according to claim 4, wherein the formula for calculating the allocation information of the route information is the same regardless of the accommodation scale of the ATM exchange.