JPS62237890A - Lattice lottery type non-blocking switch - Google Patents

Abstract

PURPOSE:To attain the connection between input and output terminals not in use without changing the connection path set already by increasing the number of switch elements more than a number nC2 (n is the number of input and output terminals) by a prescribed number. CONSTITUTION:A non-blocking switch Sn-1 640 is arranged between links up to the (n-l)th link from the uppermost link among n-set links to form the stage 1. Then (n-l)-set of paths (switch elements) 630 is arranged upward sequentially stepwise among upper/lower adjacent links started from the n-th link (the lowermost link) as the stage 2 adjacent to the right of the stage l. Finally, the non- blocking switch Sn-1 640 is arranged from the 2nd link from the upper part toward the n-th (lowermost) link as the stage 3 adjacent to the right side of the stage 2. Thus, the input/output terminals not in use are connected without changing the connection path set already.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a switch in an exchange, and more specifically, a switch that connects a plurality of input terminals and a plurality of output terminals without any internal width difference. It is related to a replaceable switch that can be used.

[Conventional technology]

As an amida type switch, in n links that connect n (where n is any integer) input terminals and n output terminals one-to-one and transmit signals in one direction, Between two links that are vertically adjacent to each other, as a whole. The inventors of the present invention have already proposed a device in which n large terminals and n output terminals can be connected without internal interference (blocking) by providing two switch elements, and the patent application 61-30
A patent application is pending as No. 546.

However, when making a new connection between the input terminal and the output terminal of the conventional amida-type switch according to such an existing proposal, if there is no empty path, the already connected path must be rearranged. There were times when I couldn't connect. In other words, in order to prevent internal comparison from occurring, it may be necessary to change the connection path between the preset large terminal and the output terminal even if the connection relationship between the person and the output does not change. In addition to being complicated, if the terminals are in use, there is a problem that the route changes during communication.

Figure 5 shows n (4) input terminals connected one-to-one between n (n = 4) input terminals and n output terminals to transmit signals in one direction.
total n C only between adjacent links.
An explanatory diagram showing an example of a conventional cross-shaped switch according to an existing proposal that can prevent internal comparison from occurring by connecting two (six) switch elements and rearranging the paths. It is.

In the figure, 101, 102, 103, 104 are input terminals, 111, 112, 113, 114 are output terminals, 121,
122,123.124 is a link, 131.132,1
33, 134, 135, 136 are switch elements (hereinafter,
(called a path).

This path can be realized, for example, by using a switch element as shown in Fig. 6; in Fig. 6, when the switch element is OFF, (a
) As shown in (b), if the link connecting the input terminal 211 and the output terminal 221 and the link connecting the input terminal 212 and the output terminal 222 are connected through, and the link is ON, as shown in (b). It will be understood that the above two tree links are cross-connected as shown in FIG.

By the way, now - as an example, only path 132 in Fig. 5 is OF
Consider the state shown in Fig. 7 where F is applied. At this time, the large terminal 103 is connected to the output terminal 112 through paths 134, 135, and 133, and the input terminal 104 is connected to the output terminal 113 through paths 134, 135, and 136.

Here, the Amidakuji moves according to the rule that it always changes direction if it hits an intersection, but the same rule applies to the Amida-shaped switch as well, that if it starts from the input terminal and hits a path, it always changes direction. A connection path is formed between the two.

Now, if we change the connection destination of the input terminal 103 from 112 to 111 from the state shown in FIG.
The FF state must be as shown in FIG. As can be seen from Figures 7 and 8, input terminal 104 is connected to connection destination 1.
Regardless of whether or not 13 is changed, the route along the way must be changed.

[Problem that the invention seeks to solve]

The present invention enables the connection of unused large terminals in amida-type switches without changing the preset connection routes.

The problem to be solved is to enable connection between the output terminals. SUMMARY OF THE INVENTION It is therefore an object of the present invention to provide a cross-shaped non-blocking switch which makes the above possible.

[Means and actions for solving the problem] In order to solve the above problem, the number of switch elements is increased from 2 nCs, and the connection between unused input terminals and output terminals is established in any state. settings without changing the connection route. Then, i pieces (to change the expression, (2”-n-1)
) and arrange them according to a certain rule.

In the case of Figures 7 and 8, for example, as shown in Figure 9, if you place a path 137 in addition to the 02 bus (Figure 7 shows the state in which this path is OFF). ), by turning on this path, the input terminal 10 can be accessed without changing the route from the input terminal 104 to the output terminal 113.
3 can be switched from the output terminal 112 to the output terminal 111.

The non-blocking switch according to the present invention eliminates the need to rearrange previously set paths in any state and prevents internal comparison from occurring.

〔Example〕

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

FIG. 1 is an explanatory diagram showing one embodiment of the present invention.

In the figure, 600 is n large terminals, 610 is n output terminals, 620 is 0 links, 630 is (n-1) paths, and 640 is the number of terminals (n-1). It is a non-blocking switch.

A switch Sn-+ 640 is placed between the (n-1)th links from the top among the n links, and this is set as stage 1. Here, the switch SI%-1 has (n-1) input terminals and (n-1) output terminals (n
-1) It is a non-blocking switch of the X (n-1) switch. Here, the configuration of the non-blocking switch will be understood later as you read the following description.

Next, stage 2 is placed to the right of stage 1 (
n-1) paths are sequentially arranged in a stepwise manner between the upper and lower adjacent links, starting from the n-th link (the lowest link). Finally, to the right of the stage 3, a non-blocking switch S, 640 is arranged from the second link from the top to the nth (bottom) link.

The non-blocking switch 5n-1 itself also recursively includes the non-blocking switch 5n-2 as stage 1, the stepped arrangement of paths as stage 2, and the non-blocking switch 57 as stage 3.
-2, and hereinafter, the non-blocking switch 57-2 will also have a similar configuration in a recursive manner.

The fact that the cross-shaped switch shown in FIG. 1 is non-blocking can be explained as follows.

For each path of the amida-shaped switch, when it is ON, the input terminal number to be used from the top to the bottom of the diagram and the input terminal number to be used from the bottom to the top in the diagram, and when it is OFF, it is to be turned ON. When doing so, input terminal numbers to be used from the top to the bottom of the diagram and input terminal numbers to be used from the bottom to the top can be assigned respectively.

If it is desired to change the output light of two input terminals in a certain state, it is sufficient to turn ON or OFF the paths assigned the two input terminal numbers.

At that time, the routes of terminals other than those two input terminals are not affected at all. Any state can be realized by combining such exchanges between two terminals.

Therefore, if switching can be performed so that a path given all the combinations of selecting two terminals from n terminals always exists, the switch will be non-blocking. That is, it is sufficient to explain that in the configuration of FIG. 1, there always exists a path to which n CZ sets of input terminals are assigned.

It is clear that if switching is performed using the algorithm below, there will be a path to which all combinations of input terminals are assigned in any state.

(Switching between terminals from the 11th terminal to (n-1) is performed within the switch of stage l.

(2) To switch large terminal n, if all stages 2 paths are in the ON state, switch the stage 2 path assigned the input terminal number to be replaced with input terminal n.
Make it FF. In the subsequent state, a path with a combination of n and other terminals other than the terminal pair assigned to the OFF path always exists in stage 3.

(3) When one of the paths in stage 2 is in the OFF state, when switching the terminal pair attached to the OFF path, turn that path ON. When performing switching between other n and other terminals, it is performed within the switch of stage 3.

In the case of n=2, the situation is clearly as shown in FIG.

When n=3, the switch shown in FIG. 2 may be placed in place of the switch 5n-1 shown in FIG. 1 (see FIG. 3). Similarly, n
In the case of =4, the configuration as shown in FIG. 4 is sufficient, and in this way, the configuration in the case of n is generally obtained.

Furthermore, a configuration in which the structure shown in FIG. 1 is symmetrically transformed vertically and horizontally also becomes non-blocking.

At this time, the required number of paths P□ is determined by using the number of paths P, %-1 when the number of terminals is (n-1), P, 1 = 2p, -, +
It can be expressed as fi 1. Since P2=1, solving this recurrence formula yields =2
”-n-1.

〔Effect of the invention〕

As explained above, in the amida-type non-blocking switch according to the present invention, by performing switching according to the following algorithm, unused input terminals and output terminals can be It has the advantage of being able to freely switch between terminals.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram showing one embodiment of the present invention, FIG.
4, n=2. 3. 4 is an explanatory diagram showing an embodiment of the present invention in each case, FIG. 5 is an explanatory diagram showing an example of a conventional amida-type switch based on the existing proposal, and FIG. 6 is a switch element used in the amida-shape switch. Figures 7 and 8 are explanatory diagrams illustrating that paths may need to be rearranged when switching the terminals formed by the cross-shaped switch shown in Figure 5. The figure is an explanatory diagram showing a solution example in which switching is realized without requiring path relocation in that case. Explanation of symbols 101-104...Input terminal, 111-114...Output terminal, 121-124...Link, 131-137...
...Path (switch element), 211-212...Input terminal of switch element, 221-222...Output terminal of switch element, 600...n input terminals, 610...n
Output terminals, 620...0 links, 630...
Non-blocking switch with (n-1) paths and 640--number of terminals (n-1). Agent Patent attorney Akio Namiki Agent Patent attorney Seiyu Matsuzaki 4 Fig. 6 Fig. 6 ■ (α) 0FF (71 case π γ Fig. plt 8 Ta light 0 Fig.

Claims (1)

[Claims] 1) In n links that connect n input terminals and n output terminals one-to-one to transmit signals in one direction, first, the n links Regarding the (n-1)th link from the top, the number of input terminals is (n-1) on the input terminal side.
In 1), the number of output terminals is the same (n-1) (n-1)×
(n-1) is placed as the first stage, and on the right side, between the bottom link and the adjacent link immediately above it (second link from the bottom), When connected between two matching links, in the off state it connects the two links through, in the on state it connects the switch element that connects the two links cross, and on the right side , Similarly, connect the same switch element as above between the second link from the bottom and the adjacent link immediately above it (third link from the bottom), and then connect the same switch element as above to the top link. A total of (n-1) switching elements as described above are connected to the , and further, on the right side, between the (n-1)th links counting from the bottom of the n links, the above-mentioned switch elements are connected. (n-1)×(n-1) grid-shaped switches similar to the above are arranged as the second stage, and the (n-1)×( n-1
) is arranged as the first stage in the same way as above, that is, on the input terminal side (n-2)
×(n-2) grid-shaped switches, a total of (n-2) switching elements as described above arranged in the same manner as above on the right side thereof, and further on the right side, as a second stage, the above-mentioned switch elements. and (n-2)×(n-2) grid-shaped switches arranged in the same manner, and similarly, the above-mentioned (n-2) constitutes the first stage and the second stage, respectively. It consists of a connection configuration in which ×(n-2) amida type switches are configured recursively and this is repeated sequentially, or it consists of a connection configuration in which the connection configuration described above is converted to point symmetry or line symmetry, and the resulting A cross-shaped non-blocking switch (where n is an integer) characterized in that (2^n-n-1) switching elements as described above are connected to the wafer.