JPH06252940A - Connection network - Google Patents

Abstract

PURPOSE:To provide a connection network which can facilitate the high speed cell switching, can attain a building block by simple topology, can easily set a call without traffic concentration caused on the internal routes, can easily increase the scale, and also can easily realize a copying function respectively. CONSTITUTION:A stage includes N/2 pieces of 2-input/2-output unit switches which are arranged in parallel to each other, and such stages are concatenated in N stages. Then the input port numbers of the unit switches of each stage are defined as I(0), I(1)...I(N-1) together with the output numbers defined as O(0), O(1)...O(N-1) respectively. Under such conditions, the output port O(j) of the (i*2-1)-th stage is connected to the input port I((j+1)modN) of the (i*2)-th stage. Then the output port O((j+1)modN) of the (k*2)-th stage is connected to the input port I(j) of the (k*2-1)-th stage respectively (1>=k>=N/2-1). In such a constitution, each unit switch outputs the input information based on a prescribed rule as well as the route information added to the input information.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a connection network for transferring information input from a plurality of input ports to a desired output port, and particularly to information communication using fixed-length short packets called cells. The present invention relates to a coupling network suitable for a hardware-implemented cell switch used in an ATM communication system.

[0002]

2. Description of the Related Art In recent years, an S that is used in existing telephone networks has been used to secure the information transmission capacity required for communication at the time of call setup.
In place of the transfer mode called TM (Synchronous Transfer Mode), expectations and interest in the transfer mode called ATM (Asynchronous Transfer Mode) are increasing.

The ATM transmits information using short packets of fixed length called cells, and each communication terminal side passes the cell to the communication network as needed, that is, information of the communication network when the communication terminal needs it. It is a transfer mode characterized by using transmission capacity. Compared to STM, ATM is capable of providing a communication terminal with any information transmission capacity required by the communication terminal, and improves communication efficiency by using the information transmission capacity of the communication network only when the communication terminal requires it. The advantage is that Therefore, ATM is a communication network that can handle voice, data, and images in a unified manner, that is, B-ISDN (Broadband Integrated Services Digi).
tal network) has been in the limelight as a basic technology for configuring the

In order to construct an ATM network, it is necessary to realize an operation of switching a cell delivered from a plurality of input communication channels to a desired output communication channel according to the route information of the cells, that is, cell switching. . Since the communication terminal uses the information transmission capacity of the communication network when necessary, a plurality of cells are simultaneously directed to the same output communication path, that is, a conflict occurs when cell switching is performed. If a conflict occurs, transfer one of the conflicting cells to the output channel,
The remaining cells are temporarily stored until there is a vacancy in the output communication path, and output when the vacancy occurs. In order to realize this function, the cell switch is composed of a buffer that temporarily stores cells and a coupling network that transfers the cells to a desired output communication path.

As is well known, in the ATM network, communication is performed based on a kind of connection called a virtual channel (VC). When communication is started between communication terminals, a VC is set between those communication terminals. At this time, for each transmission line in the ATM network through which the above-mentioned VC passes, the VCs set up to that time on the transmission line
It is necessary to confirm that the sum of the capacities occupied by and the capacities of the VCs to be newly set do not exceed the transmission capacity of the transmission line. As a result, it is possible to prevent the traffic of more than the transmission capacity from flowing into each transmission line and the deterioration of the cell transfer quality (cell loss rate, cell delay).
If the sum of the capacities of the VCs set on the transmission path exceeds the transmission capacity of the transmission path by newly setting the VC, the VC to be newly set is not set and a call loss occurs. deal with. The operation of preventing the VC from being set on the transmission line beyond the transmission capacity of the transmission line is called call admission control. Here, the capacity of the VC is generally defined by a statistic defined in a relatively long time, such as the number of cells belonging to the VC that arrive within a fixed time of a predetermined length. It should be noted that in an extremely short period of time, cells having a capacity larger than the transmission capacity may concentrate on the transmission path (that is, a conflict). As described above, the conflict is avoided by temporarily storing the cell in the buffer in the cell switch.

The VC is also set on the cell switch. In this case, V at the input port due to cell switching
The combination of C and the combination of VC at the output port are different. Therefore, in the call admission control, it is necessary to grasp the total sum of the capacities of VCs set in the respective input / output ports of at least the cell switch.

Further, it may be necessary to keep track of the total capacity of the VCs in the internal path of the cell switch.
When a multi-stage connection network known as a banyan network is used as a cell switch, traffic is concentrated on internal routes in the banyan network, as is well known. Therefore, the total capacity of VCs set at input / output ports is summed up. Is not exceeding the transmission capacity at the I / O port,
There is a possibility that VC may be set beyond the capacity of the internal path. In such a case, it is necessary to manage the transmission capacity not only for the input / output port but also for the internal path of the cell switch by call admission control. However, if the transmission capacity is managed also on the internal path of the cell switch in this way, the calculation processing capacity required for call admission control becomes enormous, especially when a large-scale cell switching system is realized. If you try to realize it, the cost will increase.

Therefore, it is desirable that the cell switch has a configuration in which traffic concentration does not occur in the internal route. If a cell switch that does not cause traffic congestion on the internal route is used, call admission control requires only the input port and output port of the cell switch to cover the transmission capacity range, reducing the amount of calculation required for call admission control. can do.

The traffic concentration on the internal route of the cell switch appears as a conflict on the internal route of the connection network in a short time. If it can be guaranteed that no conflict occurs in the internal route of the connection network, it corresponds to the assurance that no traffic concentration occurs inside the cell switch. Cell switches having a configuration in which traffic concentration does not occur on the internal route have been studied so far so as to ensure that no conflict occurs on the internal route of the connection network. Such cell switches can be roughly classified into two classes.

The cell switches belonging to the first class are as follows. In order to ensure that no conflict occurs in the internal route of the connection network, the cell transfer speed is increased in the internal route of the connection network. Specifically, in a N-input / N-output connection network, the transfer rate of cells from a certain input port to a certain output port is set to N times the transfer rate of cells at an input / output port, and the cells are input to the connection network. Make all cells transferable. Since the cell transfer rate inside the connection network is set to N times as high as that at the input / output port, a situation occurs where cells are transferred from a plurality of input ports at the output port at the same time, that is, a conflict occurs. To avoid this conflict, a buffer is provided at the output port of the connection network.

A cell switch of this class is called an output buffer type cell switch. As a cell switch belonging to this class, the knockout switch proposed by Bell Labs in the United States is particularly famous. In this class of cell switch, a bus structure is generally used as a connection network. Therefore, in the cell switch of this class, it is easy to realize the copy function of transferring a cell input from one input port to a plurality of output ports. However, in the cell switch belonging to this class, the cell transfer rate of the connection network and the cell input rate of the output buffer are N times the cell transfer rate at the input / output ports, where N is the number of input / output ports. Therefore, even if a cell switch of a slightly larger scale (for example, 16 inputs and 16 outputs) is considered, the cell input of the connection network and the output buffer used in the cell switch becomes very high speed, which is very difficult to realize.

The cell switches belonging to the second class are as follows. A buffer (this is called an input buffer) is provided at each input port of the coupling network, and the cell to be output is selected from the cells stored in each input buffer before starting the output of the cell from each input buffer. Take action. The functional element that performs this selection operation is called a competition control unit. The contention control unit receives the route information of the cells stored in the input buffer from each input buffer, and for each output port, the cell heading for the output port is selected from the cells stored in the input buffer at most. Select one by one. At this time, at most one cell is selected in each input buffer. Each input buffer is notified of which cell has been selected, and the selected cell is output from the input buffer. Here, if the connection network is so-called non-blocking, the cells output from the respective input buffers are always transferred to different output ports, so that no conflict occurs in the internal path of the connection network. As a result, traffic concentration does not occur in the internal route of the combined network, and it becomes unnecessary to manage the VC capacity of the internal route of the combined network during call admission control.

A cell switch of such a class is called an input buffer type cell switch. As a switch belonging to this class, the 3-phase algorithm proposed by Bellcore in the United States is particularly famous. As the non-blocking connection network known so far, a crossbar switch,
There is a batcher-banyan net.

In the input buffer type cell switch, the cell output rate of the input buffer and the cell transfer rate of the connection network may be the same as or about twice as high as the cell input rate of the input buffer, and the cell transfer rate of 600 Mbps at the input / output port is required. This is advantageous when realizing a high-speed cell switch that exceeds the limit.
However, the input buffer type cell switch known so far has the following drawbacks, and it is very difficult to apply it to a high speed cell switch.

First, it can be mentioned that the non-blocking connection networks known so far are not suitable for high-speed operation. Due to its configuration, the crossbar switch needs to broadcast a signal input from one input port to all output ports. Generally, when the broadcast operation is performed, the fanout of each logic element becomes large, which makes it unsuitable for high-speed operation. Therefore, the crossbar switch is not suitable for the fast cell switch.

In the Batcher-Banyan network, the fanout of the logic element can be suppressed to a small level. But,
Its structure is very complex as is well known.
Further, from the operating principle, it is necessary to match the timing to start inputting cells from each input port in all input ports. This corresponds to the need to operate the entire Batcher-Banyan network with the bit phases aligned. In order to operate the entire Batcher-Banyan network with matching bit phases, the bit clock must be distributed to the entire Batcher-Banyan network while controlling the phase. From these points, it is very difficult to implement a batcher-banyan network that operates at high speed.

Next, the competition control unit is complicated. In the contention control unit, from among the cells accumulated in the input buffer at a certain time, at most 1 is input from each input buffer.
A complicated function of selecting cells to be output one by one to each output port is executed. The amount of calculation required for this function increases exponentially as the number of input buffers handled increases, that is, as the scale of the cell switch increases. Where the copy function,
That is, considering the function of outputting a cell input from one input port to a plurality of output ports, the amount of calculation is further increased. Therefore, it is unlikely that the competition control unit will be large-scaled, which makes it difficult to increase the size of the input buffer type cell switch. Further, the contention control unit realized by hardware is very complicated as well known. Therefore, the competition control unit is not suitable for high-speed mounting.

By the way, conventionally, a multi-stage connection network called a cross network is also known. According to this method, a relatively small non-blocking connection network is connected in three stages or more to obtain a large-scale connection network. You can Therefore,
It is also conceivable to construct a non-blocking connection network by connecting relatively small crossbar switches or batcher-banyan networks by this method. However, in the non-blocking connection network by this method, the wiring between the stages is complicated and the number of wirings is larger than the number of input / output ports. Therefore, it is difficult to implement a high-speed cross network. Also,
Since the connections between stages are complicated and the number of connections is large, it is difficult to expand the scale. Further, since there are a plurality of paths connecting the arbitrary input / output ports inside the connection network, it becomes necessary to perform routing for each VC inside the connection network, and eventually the VCs set on the internal path of the cell switch are It becomes necessary to manage the capacity.

[0019]

As described above, the cell switch can easily realize high-speed cell switching such as cell switching at a speed of 600 Mbps or more, can be formed into a building block with a simple topology, and There are technical demands for easy call setup without traffic concentration on the route, easy scale expansion, and easy copy function. There was no cell switch that could satisfy the above requirements at the same time.

The present invention facilitates high speed cell switching,
The purpose of the present invention is to provide a connection network that can satisfy the technical requirements that building blocks can be realized with a simple topology, call setup can be easily performed without traffic concentration on internal routes, scale can be easily expanded, and copy functions can be easily realized. To do.

[0021]

In order to solve the above problems, the present invention provides N input ports (N is a natural number and an even number).
In the connection network that outputs the input information input from any one of the input ports to at least one predetermined output port among the N output ports according to the route information added to the input information. Input / output unit switch N /
N stages of two stages arranged in parallel are arranged in cascade, and the input port numbers of the unit switches included in each stage are I (0), I in the order arranged on the stage.
, (1), ..., I (N-1), and the output port numbers of the unit switches included in each stage are arranged in the order of O (0), O (1), ..., O (N-). 1), the output port O of the i * 2-1 (i is a natural number, 1 ≦ i ≦ N / 2) th stage from the input port side of the connection network
(J) (j is 0 or a natural number, 0 ≦ j ≦ N−1) is connected to the input port I ((j + 1) modN) of the i * 2nd stage from the input port side of the connection network, K * 2nd from the input port side of the connection network (1 ≧ k ≧ N / 2−1)
Output port O ((j + 1) modN) of the stage of
It is configured by connecting the input port I (j) of the k * 2 + 1th stage from the input port side of the connection network.

The unit switch refers to the route information added to the input information and outputs the input information to either the straight side, the cross side, or both the straight side and the cross side according to a predetermined rule. It is characterized by The cross side is the direction toward the output port that crosses the input port where the input information of the unit switch is input, and the straight side is the side that does not cross the input port where the input information of the unit switch is input. Direction toward the output port of.

Here, according to a preferred embodiment, the route information is, for example, the number of at least one output port from which the input information is output is A, and the number of the input port from which the input information is input is B. The route information added to the input information input to the even-numbered input ports is AB-
1 or A-B-1 + N (when A-B-1 is negative), the route information added to the input information input to the odd-numbered input port is B-A + 1 or B-A + 1 + N (B-A + 1).
Is negative) or 0 (when B-A + 1 + N equals N). Further, the predetermined rule is, for example, if the route information added to the input information input to the unit switch is not 0, the input information is transferred to the cross side while decrementing the route information, If the route information added to the input information is 0, the rule is to hold the value of the route information and transfer the input information to the straight side.

Further, according to another embodiment, the route information includes a bit provided corresponding to each stage and a routing end bit reset at the time of input to the connection network. The route information added to the input information input to the even-numbered input ports is A, where A is the number of at least one output port from which information is output and B is the number of the input port to which the input information is input. -B-1 or A-B-1 + N (when A-B-1 is negative), the route information added to the input information input to the odd-numbered input port is B-A + 1 or B-A + 1 + N.
(When B-A + 1 is negative) or 0 (B-A + 1 + N is N)
The difference information is defined so that the bit corresponding to each stage having the same stage number as the difference information + 1 is set, and the other bits are reset. In this case, if the routing end bit of the route information added to the input information input to the unit switch is set, the predetermined rule resets the routing end bit of the route information to the straight side. If it is set, the routing end bit value is output to the cross side while holding the value of the routing end bit if it is not the final stage, and if the bit corresponding to the stage is set, the routing end bit is set and output to the straight side. If it is the final stage, the information is discarded if the bit corresponding to the stage is reset, and the bit corresponding to the stage is set and output to the straight side.

[0025]

The connection network according to the present invention has a self-routing function, that is, each unit switch analyzes the route information contained in the input information independently received and switches the input information to a desired output port naturally. High-speed implementation and large-scale implementation that has the function of guiding the stages, the wiring between stages is equal to the number N of input / output ports of the coupling network, and is simple, and that the number of stages is equal to the number of input / output ports, which makes building blocks easy It has favorable properties for

Further, in the connection network according to the present invention, the information transfer speed of the unit switch and the information transfer speed in the wiring between the stages are set to be twice the information transfer speed in the input / output ports regardless of the number of input / output ports. Then, it becomes a non-blocking connection network. Therefore, it is possible to realize a cell switch having characteristics that it is suitable for high-speed operation as compared with a bus or a cross network, can be easily scaled up, and has no traffic concentration on an internal route. Further, the connection network according to the present invention has a fanout number of 1 at the output port of the unit switch, and therefore is suitable for high-speed operation as compared with the crossbar switch.

Moreover, when the connection network according to the present invention is used for the cell switch, the contention control unit can be distributed and arranged in each unit switch, and the contention control is performed as if there was a problem in the input buffer type cell switch. The department does not become a bottleneck for speeding up / upsizing. Further, the cell switch using the connection network according to the present invention can easily realize the copy function.

[0028]

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

FIG. 1 shows 8 inputs 8 according to an embodiment of the present invention.
The structure of the output connection network is shown. This connection network has unit switches of 2 inputs and 2 outputs 11-14, 21-24, ..., 81-
As shown in FIG. 84, eight stages, which are configured by arranging four in parallel, are arranged in cascade, and the stages are connected as follows. The connection relation of the connection network of the present invention is generally described as follows.

The N-input / N-output coupling network in which the number of stages is N is configured by cascading N stages including N / 2 2-input 2-output unit switches. In order to describe the connection between each unit switch, the input ports of the unit switches belonging to each stage are I (0), I in order from the top.
Numbers are given as (1), I (2), ..., I (N-1).
Similarly, for output ports, from the top, O (0), O
Numbers are given as (1), O (2), ..., O (N-1).

Using this number, the connection between each stage can be described as follows. That is, i * 2-1 (1 ≧ i
≧ N / 2) output port O (j) (0 ≧
j ≧ N) and the input port I of the i * 2nd stage
((J + 1) modN) and k * 2nd (1 ≧
The output port O ((j
+1) modN) and the input port I (j) of the k * 2 + 1th stage. (J + 1) modN is
Represents the remainder when (j + 1) is divided by N. The difference between the output port number and the input port number, which are the route information, in the connection network of N inputs and N outputs where the number of stages is N is defined as follows. That is, the route information is the port number A from which the input information to be switched in the connection network is output,
Let B be the input port number to which the input information is input. (1) Even input port: A-B-1, add N when negative (2) Odd input port: B-A + 1, N when negative When the result (B-A + 1 + N) is negative, it is set to 0 when N becomes equal to N,

Is defined as The bit length of the route information defined in this way is logN (base is 2). Table 1 shows the route information defined as described above in the connection network of 8 inputs and 8 outputs according to the present embodiment.

[0033]

[Table 1]

When the route information is defined as described above, each unit switch observes the route information included in the input information, and (a)
If the route information is not "0", the route information is decremented and the input information is set to the cross side, that is, if the input information input to the upper input port in the figure is to the lower port, the lower side. If the input information input to the input port of the above is transferred to the upper port, (b) If the route information is "0", the input information is sent to the straight side, that is, to the upper input port in the figure. Input from any input port to any output port by an algorithm that transfers input input information to the upper port and input information input to the lower input port to the lower port Information can be transferred. That is, point-to-point connection can be realized. This will be briefly described with reference to FIG.

In FIG. 2, the path of the input information input from the input port 2 and output to the output port 7 on the connection network according to this embodiment is shown by a broken line. Further, a change in the route information added to the input information is also shown.

From the above-described definition (a) and (b) of the route information, the route information added to the input information input from the input port 2 and output to the output port 7 is "4". Stage 1
Then, since the route information is not “0”, the route information is decremented and the input information is output to the cross side. Similarly, the stages 2, 3 and 4 also decrement the route information and output the input information to the cross side. As a result, the input information is output to the straight side from stage 5 to stage 8. As a result, the input information is transferred to the output port 7.

Further, in FIG. 2, a path of input information input from the input port 2 and directed to other than the output port 7 is shown by a chain line. As is clear from the path of the input information indicated by the one-dot chain line, other output ports can also be transferred to the desired output port by the above-described definition of the route information and the operation of the unit switch.

Further, it is clear from the definition of the topology and route information of the connection network according to the present embodiment that the above-mentioned property holds for any input port. Therefore, the connection network according to the present embodiment can transfer input information from any input port to any output port by the above-described route information and the operation of the unit switch. Next, a method of realizing the copy function in the connection network according to the embodiment of the present invention will be described.

The route information for realizing the copy function in the connection network according to this embodiment has a format in which one bit of the routing end bit is attached to the route information of the bitmap method shown in FIG. Of this route information, the bits assigned to each stage are set as follows. That is, with respect to the input information transferred from an arbitrary input port to one or a plurality of output ports, the above-mentioned difference information is calculated, and the bit assigned to the stage equal to (difference information + 1) is set for each output port to be output. . Bits that are not set by this shall be reset. In order to realize the copy function in the connection network according to the present embodiment by using this route information, each unit switch operates as follows.

First, the unit switches belonging to stages other than the final stage perform the following operations as in the previous embodiment. That is, when the routing end bit of the input information sent to the unit switch is reset, the switching is performed so that the input information is always output to the cross side. At this time, the routing end bit remains reset. Whether or not the switching is performed so as to output to the straight side is determined by the bit assigned to each stage included in the route information included in the input information. That is, if the bit is set, it is also output to the straight side. Then, for the input information output to the straight side, the routing end bit is set. If the bit in the route information assigned to each stage is reset, it is output only to the cross side and the input information is not output to the straight side.

When the routing end bit of the input information sent to the unit switch is set, the input information is not output to the cross side,
Output only on the straight side. At this time, the routing end bit added to the input information remains set.

Next, the operation of the unit switch belonging to the final stage will be described. The operation of the final stage is different from that of the previous embodiment. If the routing end bit of the input information sent to the unit switch is reset and the route information bit assigned to the stage is set, the routing end bit is set and the input information is set. Output to the straight side. If the routing end bit is reset and the assigned route information bit is also reset, the input information is discarded.

On the other hand, if the routing end bit of the input information sent to the unit switch is set, the input information is output only to the straight side. At this time, the routing end bit for the input information remains set.

FIG. 4 shows a path along which the input information input from the input port 2 is transferred according to the above routing. In the same figure, the broken line shows the route of the input information in which the routing end bit is reset, and the dashed line shows the route of the input information in which the routing end bit is set. As is clear from the figure, the realization of the copy function in the present embodiment is performed in the method of realizing the point-to-point connection by using the route of the cell whose route information is not "0" as the cell whose routing end bit is not set. The route is obtained by setting the route of the cell whose route information is “0” as the route of the cell in which the routing end bit is set.

Next, the non-blocking property of the connection network according to the embodiment of the present invention will be described. In Figure 5, points-
The transfer paths of cells in the case of point connection and the case of realizing the copy function are shown. In the figure, the route of the cell in which the route information in the case of point-to-point connection is not "0" or the routing accommodation bit is not set in the case of realizing the copy function is shown by a broken line. Further, in the figure, the route of the cell for which the route information is "0" in the case of point-to-point connection or the routing end bit is set in the case of realizing the copy function is shown by a solid line. Further, in the figure, the points where the route information becomes "0" in the case of point-to-point connection, and where the routing end bit is set in the case of realizing the copy function are indicated by dots. As shown in the figure, if the points are regarded as cross points, it can be seen that the crossbar switch is realized by being folded inside the connection network according to this embodiment.

Further, since only one input line (broken line) and one output line (solid line) of the crossbar switch are mapped to each connection between the unit switches, the internal transfer rate is doubled, and the input port and the output port are further output. If the VC (Virtual Connection) capacity is managed at the port, it can be seen that the VC is not set on the route inside the connection network beyond the capacity of the route, that is, there is no traffic concentration. Next, an embodiment of the method for mounting the connection network according to the present invention will be described.

In the connection network according to the present invention, the number of stages is equal to the number of input / output ports.
Using the same module as the module forming the 4-input 4-output coupling network, the hardware that was operating as the output coupling network was added to operate it as an 8-input 8-output coupling network. It can be realized. This will be described with reference to FIGS.

FIG. 6 shows a 4 * 4 module as a unit of expansion. As shown in the figure, the 4 * 4 module exposes part of the connections between the stages forming the connection network to the outside.
It is possible to combine these freely.

4 using the 4 * 4 module shown in FIG.
Fig. 7 shows an implementation example when a connection network with four inputs and four outputs is created.
FIG. 8 shows an implementation example when a connection network with 8 inputs and 8 outputs is created.
FIG. 9 shows an example of implementation when a 16-input 16-output coupled network is created. As shown in these figures, the connection network according to the present invention can follow almost the same connection at any scale as compared with the connection between stages of another connection network such as a Banyan network. By defining two modules, it is possible to construct a connection network of any size.

When the copy function is realized in this embodiment, the operation of the unit switch in the final stage is different. However, this makes it possible to externally specify 1-bit information indicating whether or not the unit switch belongs to the final stage to the unit switch at the final stage of the 4 * 4 module, for example, by a component called a pin header. It is clear that the above properties can be inherited by implementing a 4 * 4 module.

Furthermore, the bit position of the routing tag to which each unit switch refers differs depending on the stage number to which the unit switch belongs. For this, the maintainer sets the stage number to which the unit switch belongs. For example, a dip switch may be added to each unit switch so that the unit switch recognizes the information set in the dip switch as the stage number and operates.

Next, FIG. 10 shows an example of a method of mounting the connection network shown in FIG. 9 in a bookshelf type casing. The bookshelf type housing is a housing in which a plurality of sub racks are stacked and modules can be mounted side by side in each sub rack. In FIG. 10, the reference numerals assigned to the respective 4 * 4 modules correspond to the reference numerals assigned to the respective modules in FIG. 9. Further, reference numerals 801 to 804 are necessary for using the present embodiment as a non-blocking network, and are speed doubling modules for doubling the internal transfer speed of the connection network. Furthermore, the broken line indicates wiring arranged on the back panel or on the back surface of the back panel, and the solid line indicates wiring arranged on the front side of the substrate.

The mounting example of FIG. 10 shows an example of housing mounting when the connection network of this embodiment is applied to an ATM switch. The 4 * 4 modules arranged side by side and the speed doubling module are mounted in one subrack. Four of these mounted subracks are stacked as shown in the figure,
By connecting each 4 * 4 module and the speed doubling module as shown in the figure, a 16 × 16 ATM switch is realized.

Further, an IN having a function of accommodating various interface points such as UNI and NNI of STM-1
The F (interface) board is supposed to be mounted on the left side of each subrack shown in FIG. The cell flow inputted from each interface point is first given to the speed doubling modules 801-804, where the cell transfer rate is made twice as fast as the interface point and then inputted to the connection network. On the other hand, the cell flow output from the connection network upon receiving the switching is the speed doubling modules 801-804.
Is input to each INF board after being returned to the cell transfer rate at the interface point.

Here, the speed doubling modules 801 to 8
The configuration of 09 is a dual buffer capable of accumulating two cells for the path handling the cell flow from the INF board to the connection network, and the cell handling for the path handling the cell flow from the connection network to the INF board. It may be a FIFO memory capable of accumulating a relatively large number, for example, 32 cells. Further, backpressure, which will be described later, may be applied in the path from the speed doubling module to the speed doubling module via the connection network to improve the cell discard rate of the entire ATM switch. In this case, from the connection network to INF
If the FIFO on the path handling the cell flow to the substrate can store a large number of cells, the cell discard rate can be improved accordingly.

In FIG. 10, since the number of unit switches to which the unit switches of the connection network are connected is fixed at 3, there is almost no wiring congestion when the connection network is mounted in a bookshelf type housing. There is.

The architecture of the unit switch forming the connection network according to the present invention may be any of an input buffer type ATM switch, an output buffer type ATM switch and a common buffer type ATM switch. For these, if the input unit has a function to match the internal operating state of the unit switch and the phase of the input cell, the clock given to the unit switch forming the coupling network does not have to be phase-controlled, The cell switch is suitable for high-speed operation.

Further, a function of observing whether or not the buffer to which the output port is connected has a space for receiving a cell at the output port of the unit switch and outputting only when the buffer can receive the cell, a so-called back pressure function By setting the above, it is possible to improve the cell discard property of the connection network.

[0059]

As described above, according to the present invention, it is possible to easily provide a non-blocking coupling network by a very simple connection between unit switches. Therefore, it is suitable for high-speed operation and is a modular building block configuration. It is also easy to implement, suitable for large-scale, has no traffic concentration on the internal route, and can realize the copy function. To provide a connection network having desirable properties as a cell switch for ATM exchanges. You can

[Brief description of drawings]

FIG. 1 is a diagram showing a connection network according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a routing method on a connection network according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating a format of route information used in a connection network that realizes a copy function according to an embodiment of the present invention.

FIG. 4 is a diagram showing a transfer route of input information in a connection network realizing a copy function according to an embodiment of the present invention.

FIG. 5 is a diagram for explaining non-blocking property of a connection network according to an embodiment of the present invention.

FIG. 6 is a diagram illustrating an implementation method of a connection network according to an embodiment of the present invention.

FIG. 7 is a diagram illustrating an implementation method of a connection network according to an embodiment of the present invention.

FIG. 8 is a diagram for explaining an implementation method of a connection network according to an embodiment of the present invention.

FIG. 9 is a diagram for explaining a mounting method of a connection network according to an embodiment of the present invention.

FIG. 10 is a diagram for explaining a method of mounting the connection network according to the embodiment of the present invention on a bookshelf type housing.

[Explanation of symbols]

11-14, 21-24, ..., 83, 84 ... Unit switch

Claims (3)

[Claims] 1. N input ports (N is a natural number and an even number)
A connection network that outputs input information input from any one of the input ports to at least one predetermined output port among the N output ports according to the route information added to the input information; N / 2 stages of input / output 2 output unit switches are arranged in parallel and N stages are arranged in cascade, and the input port numbers of the unit switches included in each stage are I (0 ), I (1), ..., I (N-1)
And the output port numbers of the unit switches included in each stage are O (0), O in the order arranged on the stage.
(1), ..., O (N−1), i * 2-1 (i is a natural number, 1 ≦ i ≦ N / from the input port side of the connection network.
2) The output port O (j) of the second stage (j is 0 or a natural number, 0 ≦ j ≦ N−1) and the input port I ((j + of the second stage) i * from the input port side of the connection network.
1) modN) and the output port O ((j + 1) modN) of the k * 2nd (1 ≧ k ≧ N / 2−1) stage from the input port side of the connection network and the connection network of the connection network. Input port I of the k * 2 + 1st stage from the input port side
(J) is connected to the unit switch, the unit switch refers to the route information added to the input information, and outputs the input information to the straight side and the cross side and to the straight side and the cross side according to a predetermined rule. A connection network characterized by outputting to any one of. 2. The route information is input to even-numbered input ports, where the number of at least one output port from which the input information is output is A and the number of the input port to which the input information is input is B. The route information added to the added input information is added to the input information input to the odd-numbered input ports A-B-1 or A-B-1 + N (when A-B-1 is negative). The route information is B-A + 1 or B-A + 1 + N
(When B-A + 1 is negative) or 0 (B-A + 1 + N is N)
When the route information added to the input information input to the unit switch is not 0, the predetermined rule is that the input is performed while decrementing the route information. The rule is that information is transferred to the cross side, and if the route information added to the input information is 0, the value of the route information is held and the input information is transferred to the straight side. The connection network according to claim 1. 3. The route information comprises a bit provided corresponding to each stage and a routing end bit reset at the time of input to the connection network, and at least one of the input information is output. The route information added to the input information input to the even-numbered input ports is A-B-1 or A-, where A is the number of one output port and B is the number of the input port to which the input information is input. B-1 + N
(When A-B-1 is negative), the route information added to the input information input to the odd-numbered input ports is B-A + 1 or B-A + 1 + N (when B-A + 1 is negative) or 0 ( B
-A + 1 + N is defined as (when N becomes equal to N), the bit corresponding to each stage having the same stage number as the difference information + 1 is set, and the other bits are reset. The predetermined rule is that if the routing end bit of the route information added to the input information input to the unit switch is set, the routing end bit of the route information is reset to the straight side. For example, if it is not the final stage, the value of the routing end bit is held and is output to the cross side, and if the bit corresponding to that stage is set, the routing end bit is set and output to the straight side, and the final If it is a stage, the information is discarded if the bit corresponding to the stage is reset, 2. The connection network according to claim 1, wherein the rule is that a bit corresponding to a bit is set and is output on the straight side.