DE1121655B - Method and circuit arrangement for searching for and selecting free connection paths in a multi-stage field of crosspoints - Google Patents

Description

The subject of the main patent 1048 956 is a Method according to which free connection paths in a field of crosspoints, which can be as many as desired Can have switching stages, can be searched for and after that, if several free connection paths are available, one can be selected. In such an array of crosspoints, in short Called the switching matrix, the crosspoints in the individual stages of the switching matrix are arranged like a cross-field. Crosspoints that are switched multiple times in rows and columns form a switching matrix, which is implemented in each case by a coordinate switch. Crossbar selectors, Cross-coil selectors or relay couplers can be used. At each crosspoint there is a coupling element which is used when a connection path is established via this coupling point enters its working state. It can, for. B. set several so-called crosspoint contacts will. The individual switching stages of the switching network can contain several switching matrices and are via intermediate lines, each with several wires, z. B. Speech cores and assignment cores, connected in a certain way. The arrangement of these intermediate lines is determined by the grouping plan, which results from considerations of traffic theory. the Intermediate lines are primarily arranged in such a way that each of a switching matrix of the a switching stage at least one intermediate line to each switching matrix of the neighboring switching stage leads.

to search and select

of free communication routes

in a multilevel field

of crosspoints

Addendum to patent 1,048,956

If several connection requests to be routed via this switching matrix arise at the same time, so they are expediently dispatched one after the other, so that the dispatching process is unambiguous is not endangered. The connection requests are processed by means of a central Facility, which is called a marker here. The marker is below information lines other things in connection with the switching matrix.

In the circuit arrangements for carrying out the known methods mentioned, a so-called path search network is superimposed on the switching network, the wires of which are assigned to the intermediate lines and which are directly connected to one another at the points of the switching network where switching matrices are located. The connection points are referred to here as marking nodes. Various markings are to be applied to such a route search network, which must not interfere with one another. With this route search network, the following are therefore required for each applicant:
Siemens & Halske Aktiengesellschaft,

Berlin and Munich,
Munich 2, Wittelsbacherplatz 2

Herbert Knoblau, Krailling near Munich,
Dipl.-Ing. Rainer Rodrian, Munich-Solln,

Dipl.-Ing. Ulrich Körber, Munich,
and Dipl.-Ing. Hans-Joachim Jabczynsky,

Munich-Solln,
have been named as inventors

Intermediate line two separate path search cores provided. You can then connect to these veins in the course of the route search, different markings independently of one another with the help of certain Potentials are applied. The potentials are at interfaces that cut across the route search network are placed parallel to coupling steps, evaluated for the selection of individual sections of the route. the The selected route sections then determine the connection route to be used.

The invention now specifies a method in which the route search network to be used has to have only one instead of two wires per intermediate line. Such a route search network is therefore much simpler and therefore requires less effort. This is particularly advantageous if changes are to be made in the grouping of the switching matrix, as they are z. B. incurred, if, as a result of expansions, certain outputs of the switching matrix are particularly strong

109 759/99

have to take over. Changes to the grouping of the switching matrix also have corresponding Changes in the route search network result. These changes are less so Workload linked, the fewer wires the route search network has.

The method according to the invention is therefore a method for searching and selecting free connection paths in a switching network containing any number of switching stages, which is a Has route search network, the wires of which are assigned to its intermediate lines and marker nodes assigned to one another via switching matrices are connected and in which the input and output belonging to the desired connection are connected at the same time are marked, whereupon at an interface running across the route search network a section of the route marked from the entrance and exit is selected and then marked again This renewed marking runs in the opposite direction to the original marking in the route search network is transmitted to other interfaces on which with the help of the now coinciding Markings each time a further suitable section of the route is selected until through the selected ones Routes a route is defined in the route search network, according to which a connection route is to be switched through, according to patent 1048 956. This method is characterized in that after the selection of a section of the route at an intersection, the renewed marking of this section of the route due to a restriction of the markings previously present at this interface is brought, in such a way that only the selected section of the road is marked here.

Since in the method according to the invention in the course of the route search, the previously created Markings are more and more restricted, are those in the mentioned known methods The second route search wires used are not required and can therefore be saved.

The essence of the invention is explained in more detail in the description with reference to the drawings.

Fig. 1 shows a simple grouping plan for a seven-stage switching matrix, which here as an example a switching matrix is used;

Fig. 2 shows the course of the speech wires α and b for a connection path between an input and an output of this switching network;

3, 4, 5 to 7 and 8 and 9 show examples of circuit arrangements for performing the method according to the invention;

Fig. 10 shows how Figs. 1 to 7 are to be put together. The marks MA and MB drawn in must meet so that the associated coupling steps in the various figures are each in the same alignment.

In order to facilitate understanding of the circuit arrangement according to the invention, the structure of the switching matrix shown in FIG. 1 and the representation of the speech wires shown in FIG. 2 will first be explained. Fig. 1 shows a seven-stage switching network with the switching stages A to G. Each switching stage contains several switching matrices, each switching matrix being implemented by a crossbar. So contains z. B. the switching stage A, the switching matrices Al to Al, the switching stage B , the switching matrices B 1 to Bk , etc. In each stage, the switching matrices are the same among themselves in this example. The inputs of the switching matrix are at switching stage A and are at the same time the inputs of the switching matrix of this switching stage. Each switching matrix of switching stage A has / inputs and k outputs. Each switching matrix of switching stage B therefore has k switching matrices each with / inputs. Each input of the switching stage A can reach every switching matrix of the switching stage B. In the same way

ίο the structure of the switching network continues over the others Coupling levels continued. The outputs of the switching matrix of the switching stage G are also the outputs of the switching matrix. The coupling matrices are only indicated schematically in the illustration and the Intermediate lines only partially drawn. At the crossing points of the rows, i.e. the rows (horizontally) and the columns (vertically) of the coupling multiples, i.e. at the coupling points, are located in the crossbar contacts that have this

so crosspoints are assigned and therefore also called crosspoint contacts. So is z. B. in the switching matrix A 1 at the intersection of the / th column and the first row of the crosspoint a 1/1, to which the crosspoint contact 1 ka 1/1 and others are assigned. Accordingly, the coupling point lying in the switching matrix B 1 are b 11 m of the coupling point contact LKB 11 m and other associated, etc. Such cross-point contacts are inserted into the network of the speech wires and not shown here networks of other veins. The grouping plan shown in FIG. 1 represents the scheme according to which the switching matrices or marking nodes assigned to them are connected to one another via the link wires in the various networks. The speech wires α and b of the switching matrix are routed via the crosspoint contacts that are located at the crossing points in the switching matrices.
Fig. 2 shows the course of the speech wires a and b between an input and an output, namely a very specific one has been picked out of the many possible connection paths in the network. It is produced in that, in the course of the route search, the crosspoint contacts located in it are set, that is to say closed. In FIG. 2, however, these crosspoint contacts are shown in the idle state and therefore as open. This connection path leads, for example, from the switching matrix input Ti j to the switching matrix output Z11.

The switching matrix input TIj is at the / th input of the switching matrix A 1 of the switching stage A, and the switching matrix output Z11 is at the first output of the switching matrix G1 of the switching stage G. From the switching matrix input TIj , the connection path in this example leads via the coupling point contact 1 ka 1/1 to Output 1 of the coupling multiple A 1. This crosspoint contact is therefore at the intersection of the column / and the line 1 of the coupling multiple A 1. The multiple circuit symbols drawn to the right and left of the crosspoint contact 1 ka 1/1 indicate that in the columns and rows of the coupling multiple several crosspoint contacts are connected at the same time. The left multiple circuit symbol indicates the k crosspoint contacts connected to one column and the right multiple circuit symbol indicates the / crosspoint contacts connected to a respective row. From the crosspoint contact lkaljl an intermediate

5 6

wire to input 1 of the coupling matrix B 1, the path search wires that correspond to the

of the switching stage B. Here belongs the switching point-related switching matrix ending intermediate lines

contact lkbllm to the connection path. Also assigned here are connected to each other. In the

two multiple circuit symbols are shown. Fig. 3 lead the path search veins shown, the

5 / wires then lead from the output m of the switching matrix B 1, via the marker nodes jA 1, jB 1, / Cw,

a connecting wire to the first input of the fDn, / El, fFp and / Eq.

Coupling multiple Cm of the coupling stage C. About Further, in the path search veins Belegungskon crosspoint contact 1 extends the core connects contacts inserted. They are in each case with route search arteries, manure path, and that it still goes over which the free intermediate lines belong, closed crosspoint contacts lkdnml, lkelp, lkfp 11 and io and with route search arteries, the intermediate lkglpl to be occupied up to output Z11. lines belong, open. In the path search artery. On the basis of FIGS. 3 to 10, various circuitry belongs to the intermediate line, which are illustrated by the coupling point examples that work according to the path search method according to the invention contact 1 / approx. 1/1 for the coupling point contact lkbllm. They pass under it, the occupancy contact bab 11 is inserted, differ from one another in various notices. First of all, they differ in the additional occupancy contacts bbcml in search cores, and in which way the restriction bcdnm, bdeln, befpl and bfglp is inserted. Besides being brought about by markings. There are also certain differences in the path search arteries, depending on the ladder Gab 11, Gbcml, Gcdnm, Gdeln, Gefpl and whether the interfaces are inserted by switching matrices 20 Gfglp. The directional decoupling conductors are in this way or intermediate lines are laid. Polished to detect that markings of entrances or Ausder at the relevant sections of the path went to the first intersection markings, various switching means can be put, but not in the opposite direction, overused, such. B. Relay. The design will carry. The first interface is here with the various circuit arrangements also 25 of the coupling stage D. Now those shown in FIG. 3 are influenced by these means. After all, only part of the route search wires placed on it is also of influence, whether in the switching matrix on both networks of these wires. In the entire network pages of the first interface there are further interfaces as many marking nodes as there are switching multiplexes or whether this is the case on only one side in the switching matrix according to FIG. 1. Between all these features there are now number 30 between which the intermediate lines run. A wealth of combinations, which can come together in each case at a switch, the marking node jA 1, across all the examples. From these closed free intermediate line arenas of these many possible examples, some special marking potentials for several marking nodes are selected in the characteristic and shown in the relevant coupling stage B. The decoupling directional ladder is shown in the path figures. 35 search veins prevent this marking potential in the circuit shown in FIGS. 3, 4 and 8 in an inadmissible manner from the marking nodes of the exemplary embodiments, the restriction of the respective coupling stage B from also going backwards to other markings that were previously present with the assistance of the switching matrix is transmitted, where additional, for the application of marking potentials by occupied and therefore by open occupancy serving switching means, such as. B. of contacts, to 40 contacts interrupted path search arteries brought into failure. These switching means can be bypassed in a manner desired. The presence has the effect of being closed in a kierknoten or in a corresponding way on the intermediate line. Accordingly, either under the decoupling directional conductor in the other path search switching matrices or marking nodes or under heels.

Intermediate lines selected. In the case of the one in FIGS

4, 5 and 8 shown and different from the circuit example shown

Restriction of the markings that are present in each case.

with the participation of the already existing occupancy works, the marking nodes described above

contacts made. These are in intermediate lines instead of switching matrices. Further

connection wires inserted. It is therefore appropriate there- 50 are also required in these route search networks

moderately under intermediate lines laying contacts and decoupling directional conductors in appropriate

chooses. inserted accordingly.

First, some circuit examples in the circuit example shown in FIG

individually described, in which the mentioned are additional on both sides of the first interface

special switching means are provided to limit Mar- 55 interfaces. The interfaces are

markings are provided. This includes that in Fig. 3 placed in such a way that they are switching matrices or mar-

circuit example shown. Cut the kier knots shown there. As parts of the way are therefore

Cable cores belong to the route search network. each switching matrix selected. In the first

The route search network is here in a similar way among the switching matrices of the interface

as in FIG. 2 from the network of speech cores, 60 coupling stage D selected. The centrally located

ie the a- or 6-cores, just an extract from the KVD multiple switch. With the help of

posed. The network of the route search cores is the relay, which is at the marking nodes of this coupling stage

The switching matrix is superimposed and is inserted after that in FIG. 1, the marking state becomes this

established grouping plan. However, it has marking nodes for the switching matrix selector KVD but some special features, which are also reported in 65. At the marker node fDn they are in series

Fig. 3 are indicated. First the coupled relays lSdn and 2Sdn are inserted. Of the

Replaced many times by so-called marking nodes, the center tap of the series connection is via the idle

which each consist of a connection point in which contact nwd is connected to ground. The other mar-

kierknoten this coupling stage, inserted relays are provided in a corresponding manner. The switching matrix selector KVD has inputs that are individually assigned to the individual marking nodes. The series circuit of the two normally open contacts Isdn and 2sdn is connected to the input assigned to the marking node fDn. The normally open contact Isdn belongs to relay 1 Sdn, and normally open contact 2 .sdn belongs to relay 2 Sdn. When these two relays are energized, the normally open contacts Isdn and 2sdn are closed, whereby ground potential is applied to the relevant input of the switching matrix selector KVD . As a result, a certain marking state of the marking node fDn is reported to the switching matrix selector KVD. This is the marking state that is present when this marking node or the associated switching matrix comes into question when searching for the connection path to be searched for. The relays provided in the other marking nodes of this coupling stage, which are not shown in FIG. 3, also have working contacts via which they may apply ground potential to the inputs of the switching matrix selector K VD assigned to them. If ground potential is applied to one or more of the inputs of the switching matrix selector KVD while searching for a route, this triggers a selection process in this selector through which one of the associated switching matrices is selected.

The switching matrix selector KVD has pairs of normally open contacts which are assigned to the switching matrices or marking nodes. The contacts of the pair belonging to the selected switching matrix or marking node are closed to identify the election result. The Markierknoten Fdn are vdn the operating contacts 1 and 2 assigned VDN. One connection of one of these working contacts is connected to one end of the series circuit of the relays 1 Sdn and 2 Sdn. The other connections are each connected to ground via the normally closed contacts lwcb and 1 wef and via the relay WD . The function of the contacts lwcb and lwef will be given later. When the switching matrix Dn is through, as indicated by the multiple circuit symbols η, it is connected to all the working contacts of the switching matrix selector KVD assigned to the marking node. If any of the switching matrices assigned to this marking node is selected, the marking with ground potential is restricted.

It was assumed above that the relays lSdn and 2Sdn had been energized in the course of the route search. At the interface located at the coupling stage D , ground potential was used for marking. In addition, input and output markings must be made to mark inputs and outputs for which a connection path is to be found. For this purpose, a different marking potential is used than at the interfaces. In the present circuit example, positive potential is used for this purpose. The connection path to be searched should lead from the input TIj ao to the output Z11 of the switching matrix. The contacts tlj and ζ 11 are therefore closed for input and output marking. If there is still a free connection path via the switching matrix Dn , a current flows from the input Tlj via the contact tlj and the marker nodes fA 1, fB 1, fCm, the relay lSdn and via the contact nwd. This current can also flow branch via further switching matrices of the switching stages B and C, provided they are in free connection paths. In addition, a current initially flows from the output ZIl via the contact ζ 11, the marker nodes / Gl, fFp, / El, the relay 2Sdn and the contact nwd. This current can also branch, via other marker nodes of the coupling stages F and E. as long as these are in free connecting paths. As a result of the selection process already described, contact nwd is opened and contacts 1 vdn and 2 vdn are closed. The relays 1 Sdn and 2 Sdn are therefore de-energized again. Relays are also inserted at the marking nodes of the other coupling stages. Those of these relays that belong to marker nodes in connection paths that lead via marker node fDn , however, continue to be energized. this will

the coupling multiple selector KVD is selected and enables 45, since the contacts 1 vdn and 2 vdn are

therefore the contacts 1 vdn and 2 vdn are closed, the marking node fDn is connected to ground via these contacts and via the relay WD . As a result, the relay WD is energized, as will be explained in more detail later. The already mentioned break contact nwd now belongs to the relay WD. Corresponding break contacts are provided for all marking nodes of the coupling stage D. All these normally closed contacts are opened when the relay WD is energized. The marking nodes are thereby separated from the ground. However, the marking node fDn belonging to the selected switching matrix is an exception. This is now connected to ground via contacts lvdn and 2 vdn of the switching matrix selector KVD. Ground potential is used here as marking potential. The selection of the switching matrix Dn has therefore led to the mass Markierpotential is invested exclusively in the companies belonging to this switching matrix Markierknoten fDn. The relay WD is now not only energized when the switching matrix Dn is selected, but also when another switching matrix belonging to the switching stage D is selected. It had been closed.

Individual switching matrices or marking nodes must now be selected in other switching stages in order to define a connection path. This can be started after the contacts of the previously de-energized relays have come to rest. The relay WD comes after the completion of the selection process with the coupling multiple selector KVD in the course of the restriction of the marking at the first interface under power. It has the working contact wd, which is used to trigger the dialing processes to be carried out at the neighboring interfaces. Initially, the triggering of the dialing process for coupling stage C is considered.

Via the contact wd , ground is applied to an input of the switching matrix selector KVC , via which the triggering of a selection process is effected. The switching matrix selector KVC is assigned to switching stage C and corresponds to the switching matrix selector KVD in structure and function. As already mentioned, relays are also inserted in the marking nodes of the coupling stage C. The relay Scm belongs to the marking node jCm. These relays have normally open contacts,

which are each connected to the inputs of the switching matrix selector KVC . Each of these inputs is assigned to a marking node. When activated, a normally open contact is used to apply ground potential to the relevant input. Because of the marking of the input TIj with the help of the contact AIj and the marking node jDn with the help of the contact 1 van some of these relays are energized, so that ground potential is also applied to some inputs of the switching matrix selector KVC. In the course of the selected selection process, a switching matrix belonging to these inputs is selected. The switching matrix selector KVC has contacts, one of which applies ground potential to the marking node belonging to the selected switching matrix. The coupling matrix Cm may be selected. There is then the Markierknoten JCM by belonging to the switching matrix selector KVC working contact vcm connected to ground. The WCB relay is also inserted into this connection, which is then energized. The multiple circuit symbol m in the relay WCB indicates that this relay is inserted in all connections belonging to the marker node. The WCB relay has contacts lwcb and 2wcb. The normally closed contact lwcb is in series with the contact lvdn and interrupts the connection to ground there when it is actuated. If the contact 1 we / is actuated, the marking node jDn in the coupling stage D is no longer under the influence of ground potential. Therefore, in the coupling stage C, only that marking node is still under the influence of ground potential which was connected to ground by a contact of the switching multiple selector KVC . As already described, this is the marking node jCm. So only this marking node is marked here. The associated switching matrix Cm belongs to the selected connection path. With the help of the contact 2 web , a selection process is then triggered at the switching four-way selector KVB , by means of which a switching matrix is now selected for the switching stage B in a manner analogous to that previously for the switching stage C. Let this be the switching matrix B 1.

On the basis of the grouping plan according to FIG. 1, only a single connection path now leads from a switching matrix of switching stage B to an input of the switching matrix. Therefore, there is even a single connection path between the switching matrix Bl and input tij. A connection path between the input T1 j and the switching matrix Dn is thus established on the basis of the selection processes described so far. In the functional example described, it leads over the switching matrices Al, Bl and Cm.

Simultaneously with the selection of switching matrices in switching stages, which are located between the first interface and the inputs of the switching matrix, there is also a selection of switching matrices in switching stages, which are located between the first interface and the outputs of the switching matrix. The facilities provided there for this purpose correspond completely to the facilities that are provided for the coupling stages C and B.

An input of the coupling multiple selector KVE, which belongs to the coupling stage E , is also connected to the normally open contact wd of the relay WO , via which a selection process is triggered in the coupling multiple selector KVE when the contact wd is actuated. As with the marking nodes of coupling stage C, relays are inserted in the marking nodes of the coupling stage E , with the aid of which the marking state of this marking node is reported to the coupling multiple selector KVE. After selecting a coupling multiple for coupling stage E , the relay WEF is energized and opens its already mentioned break contact lwef. In addition, it closes its ίο normally open contact 2 wef, which triggers the selection of a coupling multiple in the coupling stage F, which is carried out by the coupling multiple selector KVF. After selecting this switching matrix, a connection path between the switching matrix Dn and the output Z11 of the switching matrix is established. Let it run over the coupling multiples El, Fp and Gl. This means that the entire connection path to be searched for has been found.

If several outputs had been marked in the switching matrix G1 instead of the output ZIl, this would have had no effect on the route search described above. After completing this route search, all that remains to be done is to select one of the various marked exits. If the various marked outputs do not all belong to the same switching matrix, but are distributed over several switching matrices of switching stage G, a switching matrix of this switching stage must also be selected in the course of the route search. A switching matrix must therefore also be provided in this coupling stage. After a switching matrix has been selected there, a specific one can then also be selected from the marked outputs of this switching matrix. An example of how such a selection can be made will be described later. In the same way that several outputs of the switching matrix can be taken into account when searching for a route, several inputs can also be taken into account at the same time. The necessary dialing devices must then be provided at coupling stage A. Several exits as well as several entrances can be taken into account at the same time without difficulties in finding a route.

When searching for a route, the WD, WCB and WEF relays were used to control the sequence of the various processes. It should be noted here that this sequence control should only serve as an example to show the possibility that the processes in question can be controlled. When searching for a route, you can also choose intermediate lines instead of switching matrices. The interfaces must then be relocated accordingly. The route search is otherwise carried out using the same procedure. In order to restrict the markings, the original marking paths are interrupted at the first intersection for all sections of the road, except for the selected section, and then the rest of the marking path is also interrupted after selecting one section at the adjacent intersections at the first intersection. Instead, marking paths are closed over the last selected sections of the path. The last existing marking paths are then displayed at any other adjacent interfaces that may be present, depending on the selection of a section of the path

109 759/99

interrupted and only closed again over the last selected sections of the route. These procedural steps are repeated until the route you are looking for is set via the switching matrix.

In the grouping plan according to FIG. 1, the number of inputs of a coupling stage is always the same the number of outputs of the previous coupling stage. At the connections of the switching matrix Therefore only one intermediate line is connected to coupling stages. There can also be grouping plans for switching fields in which these Restrictions for their structure are not observed. This can also be done with such switching networks Searching and selecting connection paths made according to the method according to the invention will. Let us consider the case that between the selected switching matrices of adjacent switching stages there is more than one intermediate line. FaUs determine the selection processes switching matrices is then the one to be used of these intermediate lines has not yet been determined. To set them down can therefore be reached here with the help of an additional dialing device under these intermediate lines to choose one. An example of this will be described later. But if the election processes affect intermediate lines anyway, additional dialing devices are not required.

In the circuit example shown in Fig. 3, the first interface is at the switching stage D, which is the middle switching stage of a switching network. Instead, the first interface can also be located at other coupling stages. The number of further interfaces lying sideways from it is then different on the one hand than on the other. The case can then also arise that on the one hand no further interfaces are required at all. It may then be possible to simplify the design of the sequence control.

In Fig. 4 a circuit example is shown, in which the first interface is placed in such a way that none in the direction of the outputs of the switching matrix further interfaces are necessary. This is a restriction on the first interface existing markings removed the mark that was attached to the interface on The next side of the switching matrix, i.e. the side where the outputs are located, is created, and through replaces a marking on the section of the road selected at the first intersection. This is followed by Selection of a section of the path at the neighboring interface is the one created at the first interface Marking is canceled again and selected by one at its neighboring interface Replaced markings that were laid out for a stretch of the way. These procedural steps are from interface to interface continued until a routing in the switching matrix is fixed.

In the circuit example shown in FIG. 4, switching matrices are selected as path sections. First, the case is considered where a connection path between an input and an output of the switching network is to be found, namely between the input Π / at the switching stage A and the output IZ11 at the switching stage E. Two input and output markings are used different marking potentials applied via resistors are used. The contact ti j belonging to the input TIj applies positive potential to the path search network via the resistor RtIj. The contact Iz 11 belonging to the output IZ11 applies a negative potential to the path search network via the resistor IRzIl. As has already been described in detail, the route search network itself has marker nodes, decoupling guides and occupancy contacts. At the interfaces there are centrally arranged switching multiple selectors, the inputs of which are each connected to the marking nodes belonging to the relevant interface. If they are in free connecting paths, an intermediate potential between the two marking potentials occurs at these marking nodes. A switching matrix selector now selects one from each of the marker nodes which have the mentioned intermediate potential, whereby a switching matrix is determined in the relevant switching stage.

In the circuit example shown in FIG. 4, the first interface is at the switching stage D. The switching matrix selector KVD provided there is the first to carry out a selection process. It has individually assigned outputs to its inputs. One each of these inputs and outputs is shown. It is connected to the input of the selector assigned to it via the resistor Rdn and the normally closed contact wen. Let the mentioned intermediate potential occur here, inter alia, at the marking node jDn as a result of the input and output marking. If this intermediate potential is applied to at least one input of the switching matrix selector KVD , a selection process is triggered on it, on the basis of which it applies a negative potential to one of the outputs that are assigned to the marked inputs. Let this negative potential be applied here at the output shown with the resistor Rdn . As a result, this negative potential has an effect on the marking node jDn via the resistor Rdn and the contact wen . The outputs of the switching matrix selector are still connected to the relay Wd via the mixing gate OD . This relay is energized when there is a negative potential at at least one selector output. That is to say, with its other connection it is connected to ground. The relay Wd has a normally closed contact, which is in series with the contacts that are connected to the outputs of the switching matrix, so z. B. also in series with the contact Iz 11. This is the normally closed contact lwd. As long as the relay Wd is de-energized, the presence of its contact lwd has no effect. When the relay Wd is energized, the normally closed contact lwd is opened and the previous supply of the negative marking potential is thereby interrupted. However, the marking node jDn is now marked directly by the negative marking potential supplied to it from the switching matrix selector KVD.

In the coupling stage C , the intermediate potential is now applied to those marking nodes which are in connection paths between the input TIj and the marking node jDn or the switching matrix Dn . The relay Wd still has the normally open contact lwd. This is connected to an input of the switching matrix selector KVC , which belongs to the switching stage C. When this contact is activated, a voting process is triggered for this voter. He

takes place in the same way as with the multiplexer KVD . Here, too, negative marking potential is applied to the marking node belonging to the selected switching matrix via a resistor. The coupling matrix Cm should be selected here. The negative marking potential is then applied to the marking node fCm via the resistor Rcm. The relay Wc is connected to the outputs of the switching matrix selector KVC via the mixing gate OC. After choosing a coupling multiple, it comes under power and activates its contacts, including its normally closed contact wen. The supply of the negative marking potential is therefore interrupted at the marking node jDn. Now a switching matrix has to be selected for coupling stage B. This is done in an analogous manner to the switching stage C, this time with the help of the switching matrix selector KVB. This completely defines a connection path between the input TIj and the output IZ11.

The route search method used in the circuit example according to FIG. 4 can also be used with switching matrices are used, which are structured in any way. Then only, if necessary, further elective facilities to be provided. In principle, the switching matrix can also have any number of switching stages have, since the devices provided for the individual coupling stages depend on their number are independent. This also applies to the method described with reference to FIG. 3 and also to the procedures to be described later.

In Fig. 8, a further circuit example is shown in which only on one side of the first Interface further interfaces are provided. Again, the limitation of the one Interface existing markings each brought about by previously applied markings canceled and replaced by a marker created for the last selected section of the route will. Coupling multiples are chosen as the path sections. In the path search arteries are, as with the already Circuit examples described, decoupling directional conductor and occupancy contacts inserted.

In this circuit example, ferromagnetic toroidal cores with an approximately rectangular hysteresis loop are used to detect markings. Such toroidal cores have two magnetic saturation states, which are used here, among other things, as operating positions of the toroidal cores. Each toroid has a first, a second and a third winding. The first windings of the toroidal cores are looped into the marking nodes assigned to the coupling matrices. The toroidal cores are therefore assigned to the marking nodes. Two different marking potentials are used here for input and output marking. The input side of the switching network is at switching stage A, where positive potential is applied in the event of a marking. The output side of the switching matrix is at the switching stage E, where negative marking potential is applied in the event of a marking. A current therefore flows over the route search arteries belonging to the free connecting routes. In order to limit this current, current limiting resistors are provided at the supply points of the marking potentials, which are intended to determine the maximum strength of the resulting current. With the help of the contact / 1, the positive marking potential can be applied to the marking knee.ifA 1 . It is fed in via the limiting resistor RtI. With the help of the contact ζ 1, the negative marking potential can be applied to the marking node / El. It is fed in via the limiting resistor Rz 1. The current that may flow in the network of path search cores is so strong that at most the simple magnetizing current flows through the first windings of the toroidal cores, i.e. a current that is just as strong as it is necessary to convert one toroidal core from one to the other To bring saturation. This current depends on the occupancy of the switching network. It can branch out in a coupling stage via the windings 1 of several toroidal cores. With a single winding, it is then smaller than the simple magnetizing current. If, however, only one coupling matrix is possible for the connection path there, the entire current flows through the same winding, which is therefore traversed by the simple magnetizing current. The current limiting resistors prevent this current intensity from being exceeded. The second windings of the toroidal cores are each grouped together at the same interface, e.g. B. connected in series, and are initially traversed by twice the magnetizing current, however, in the opposite direction. The third windings of the toroidal cores are individually connected to inputs of the matrix switch assigned to the interfaces. The first interface is at coupling stage B. The switching matrix selector KVB is located there. The winding 3 of the toroidal core RB 1 is connected to one of its inputs. The winding 2 of this toroidal core is connected in series with the second windings of the other toroidal cores located at this interface, but not shown. The series circuit is then connected to the flip-flop circuit MB , the function of which will be explained later. The winding 1 of the toroidal core RB 1 is looped into the marking node / Sl.

To select a coupling multiple, the interface in question is used via the second Windings of the toroidal cores flowing double magnetizing current interrupted, causing in the third windings of the toroidal cores belonging to the connection paths in question, a voltage surge arises. These voltage surges are applied to the inputs of the relevant switching matrix fed. In the third windings of those toroidal cores that are not among those in question Connection paths, there are no voltage surges. This fact is now even more precise explained. As already mentioned, in the case of the toroidal cores, the one flowing through their second windings Power interrupted. The toroidal cores belonging to connection paths that are not in question then retain their state of magnetization. It is therefore not a cause for arising a voltage surge present. In the case of the toroidal cores, the connection paths to be considered belong, the situation is different. At first they were simultaneously under the influence of the The magnetizing current flowing in one direction and the magnetizing current flowing in the other direction double magnetizing current. The magnetization state that had set itself results from the difference in the magnetizing currents. The direction of the associated magnetic

15 16

ization is determined by the direction of double the gnetisierungsstromes brought into working position before starting the path search, as this is always done by closing the contact st. In the star working position because of the limitation of the other current, it supplies the coincidence gate VA 1 kere current. After switching off this current also positive potential. This potential is then affected by the Ma inputs of the coincidence gate UA 1 flowing in the opposite direction 5 after the contact ί 1 has been closed at both of the Ma inputs of the coincidence gate UA 1. The bistable gnetisierungsstrom. It causes a flip-flop circuit KB is now also via the mixed magnetization of the toroidal cores concerned. As a result, the gate OB to the outputs of the switching matrix selector is connected in the third windings of this KVB . And at an output of the ring cores, a surge voltage, which, as already io switching matrix selector sprinkled auferwähnt positive potential occurs to the inputs of the associated Koppelviel-, the flip-flop KB is supplied to tray selector. brought their rest position, so that at the coincidence if at least one gate UA 1 no longer supplies a positive potential to a switching matrix. If such a voltage surge is supplied, a selection process is consequently also disappeared at the marking node, in which the input 15 fA 1 the positive potential. Gears are taken into account at the flip-flop circuit KB in which a span is also fed to the other marking nodes of the coupling joint. Connected under the associated coincidence gates assigned to level A. One of the switching matrices is selected. Apart from what the multiple circuit symbol / indicates, the coincidence gates corresponding to the first interface on the 20 are also present in the marking nodes of the coupling stages B to restrict the previously existing markings. On the next side of the coupling matrix previously connected to marking node jB 1, the coincidence applied marking potential is switched off and instead of gate UBl. One of its two inputs is connected to the marker node connected to the selected switching matrix KC at the switching stage C. closed.

In the illustrated in Fig. 8 circuit EXAMPLE 25 As shown in Fig. 8 can be seen, are at each of the first interface is located at the coupling stage as switching stages B, C and D are the same devices The aforementioned Kippschalutng MB is used for the implementation of the Electoral and marking operations interrupting the through the second windings of the present. In the coupling stage C, the toroidal cores flowing double magnetization coupling multiple selector KVC, the monostable breakover current, are used. The coupling circuit MB is monostable and circuit MC, which responds to the output of the mixer with a delay. In the quiescent state, it causes the mixing gate OC , gatters OB is connected, that the connected windings 2 of the ring and the bistable flip-flop circuit KC. Furthermore, the cores are live. The winding of the toroidal cores OAB is looped in with the marking knot of the coupling via the mixing gate, the marking knot. This includes the toroidal core RCm. Out of level A, that is to say also connected to the marking node jA 1, there are coincidence gates, such as the coincidence bond. If an input gate UCm. If the switching matrix selector KVB and an output marking has been carried out at one of its outputs, then a certain potential is applied to the marking node of the coupling stage A via the mixing gate OB, the flip-flop MC. This potential has a temporary effect in the working position, so that the mixing gate OAB to the flip-flop circuit MB a 40 current that through the windings 2 of the toroidal cores and brings them temporarily in the working position. As a result, flows that belong to the coupling stage C are interrupted, as already explained, in the windings 3. The bistable multivibrator MC responds hesitantly to the inputs of the switching matrix selector KVB , so that it is ensured that voltage surges are connected beforehand, which in turn switching stage B, the marking processes are ended. It trigger a voting process. The switching matrix 45 then finds outputs in the same way as with the switching selector KVB , which are individually assigned to its level B, the selection of a connection input for the sought-after connection. After expiry of the coupling multiple belonging to the manure path. Thereupon the selection process is marked with a previously existing marking at the output that is subject to the corresponding restriction that is assigned to the selected switching matrix. It can then put indicative potential in this circumstance. For this purpose, the coupling stage D uses the positive potential for the connection path here. To this output switching matrix used is to be selected, which belongs to the selected switch matrix, where the switching matrix selector KVD, the tilting Markierknoten are connected so that the positive step-MD and KD and the mixing gate OD for the potential at the same time at these Markierknoten Toggle implementation of the election and marking process is presented, namely via a delimitation 55. If a switching matrix is selected there, stood. If the switching matrix Bi is selected, a connection path between the head is also determined, the positive potential is determined via the non-pelmultiple A 1 and the switching matrix El, the limiting resistor shown on the marking The path search is thus ended,
node jB 1 created. In the example described above for In addition, the marking previously carried out at the marking 60 at the switching network node jA 1 must be switched off and one marking node must be marked on the output side. This is what the flip-flop KB and that have been used for. The connection coincidence gate UA 1. found by the route search. The marking of the marking path can therefore be used to connect the node fA 1 via an input and associated switching matrices and thus two outputs of the coincidence gate UA 1. The one 65 see one of the inputs that are connected to an A 1 at the switching matrix input of the coincidence gate UA 1, and one of the outputs that are connected to the coupling output of the flip-flop KB . Often E 1 are used. Given this is a bistable multivibrator, which is one of the inputs and outputs mentioned.

went to the coupling multiple to make a selection.

Here, too, outputs located at different coupling matrices can be taken into account at the same time when searching for a route. One of the switching matrices at which these outputs are located must then first be selected. Then you have to choose the output to be used yourself. The selection among these switching matrices can be done in exactly the same way as previously described. If one disregards the presence of the marker node JE 1 in the circuit according to FIG. 8, then in the switching matrices or marker nodes of the switching stage D, outgoing lines can be seen as belonging to the outputs of the switching matrix. If negative marking potential has been applied to several of these lines and a switching matrix or marking node of the coupling stage D has been selected, as already described, a specific output itself must also be selected. The toroidal cores provided for these lines, such as the toroidal core R1 and the selector PZDE and the monostable multivibrator MDE , can be used for this purpose. The PZDE voter can be referred to here as a place voter. It is used to select the place at which the output of the switching matrix to be used is located on one of the switching matrices of switching stage D. The second windings of all ring cores at the outputs are connected in series and connected to the monostable multivibrator MDE . The third windings of the toroidal cores are connected in series and each connected to an input of the PZDE location selector . If a switching matrix has been selected for coupling stage D , the associated marking node is marked. Let it be z. B. the marking node jDn. In addition, the monostable multivibrator MDE is brought into the working position via the mixing gate OD , whereby the double magnetizing current flowing through the windings 2 is interrupted. In the case of those toroidal cores that belong to the outputs of the selected coupling multiple, a magnetizing current also flows through the first windings. This also includes the toroidal core R 1. In its third windings, a voltage surge is generated when the double magnetizing current is interrupted. These toroidal cores are taken into account in the subsequent location selection by the PZDE location selector. If the location of the output to be used is selected, the output itself is fixed, since the switching matrix on which it is reading has already been selected. This means that a clear connection route is also found in this route search. In an analogous manner, inputs of the switching matrix that are distributed in any way can also be taken into account at the same time when searching for a route.

In the route search described with reference to the circuit according to FIG. 8, switching matrices were selected in the relevant switching stages, which have to define the resulting connection route. If two coupling multiples of adjacent switching stages are only connected by a single intermediate line, as is also the case with the grouping plan shown in FIG. 1, a connection path is uniquely determined by the selection of switching stages. If the number of intermediate lines running between two switching stages is not limited in this way, after the selection of the switching matrices to be used, one must optionally be selected from the several intermediate lines connecting two switching matrices. In FIG. 9, among other things, a selection device is shown with the aid of which this selection is made. In this FIG. 9, the two coupling stages B and C with their selection devices, etc., as they are also present in the circuit according to FIG. 8, are shown first. The windings of toroidal cores, to which the illustrated toroidal core RBCmV belongs, are inserted into the path search cores connecting the marker nodes that are assigned to the coupling multiples of these coupling stages. This toroidal core belongs to one of those path search cores that connect the marker nodes jB 1 and fCm . It may be the switching multiples of the switching stages B and C z. B. each be connected by three intermediate lines. Accordingly, the associated marker nodes are each connected by three route search wires. Therefore, in the Markierknoten three times as much routing conductors are each connected to one another than in the Markierknoten in the circuit of Fig. 8. This is indicated in Fig. 9 in that there m in place of the multiple circuit symbols and k is the multiple circuit numeral 3 ■ m and 3 ■ k are shown. The toroidal cores assigned to the mentioned path search cores, like the toroidal cores used in the circuit according to FIG. 8, have three windings 1, 2 and 3. As already mentioned, the windings 1 are each looped into the path search cores. The windings 2 are all connected in series and connected to the monostable and on-delayed flip-flop circuit MBC , from which they are fed with twice the magnetizing current in the idle state. The windings 3 of these toroidal cores are only partially connected in series here, and their series connections are connected to different inputs of the place selector PZBC . This location selector has as many inputs as a maximum of intermediate lines run between two switching matrices. The series-connected windings 3 of those toroidal cores are connected to the first input of the place selector PZBC.

which belong to path search cores and thus to intermediate lines, each of which connects as the first two switching matrices. The series-connected windings 3 of those toroidal cores are connected to the second input of the place selector PZBC.

which belong to route search cores and thus to intermediate lines, which each connect as a second two coupling multiples. Additional windings connected in series and correspondingly belonging together are connected to the other inputs of the place selector PZBC. The location selector PZBC , like the switching matrix, has outputs assigned to its inputs, one of which is designated by a certain potential after a selection process. The bistable multivibrator I KC is connected to the outputs via the mixing gate OBC. This flip-flop takes over the task of the flip-flop KC otherwise present in the coupling stage C.

The sequence of an intermediate line selection will now be briefly described. The switching matrix B1 may first have been selected in the switching stage B. Then the monostable multivibrator MC is made to respond, whereupon in already

109 759/99

Written way in the switching stage C a switching matrix is selected with the help of the switching matrix selector KVC . Let this be the coupling matrix Cm again. In the mixing gates OC also the monostable multivibrator is connected MBC here other than the members of the switching stage D monostable multivibrator, which is part of the provided for the intermediate line choice selection means. This intermediate line selection must be completed before the switching matrix selection in switching stage D takes place. The monostable multivibrator of the coupling stage D must therefore have a sufficiently long response delay. Before it responds, the monostable multivibrator MBC must respond, whereupon the double magnetizing current supplied by it is interrupted. As a result, voltage surges occur on the windings which are inserted in the path search cores which extend from the previously marked marking node jB 1 and which belong to free intermediate lines. As a result, a dialing process is triggered in the place selector PZBC , by means of which one of these intermediate lines is selected. The result of the election is indicated by the fact that the place selector PZBC applies a certain potential to one of its outputs. Devices (not shown here) for evaluating the election result are connected to its outputs. In addition, the bistable flip-flop 1 KG, which had previously been brought into the working position when the contact st was closed, is now brought back into the rest position via the mixing gate OBC. As a result, the positive potential previously supplied there by the flip-flop IKC disappears at one input of the coincidence gate UB1 , whereby the marking of the marking node fB 1 is switched off. A switching matrix can now be selected in coupling stage D. It was shown above how the route search is to be carried out when switching matrices of neighboring switching stages are connected to one another by more than one link. The route search method according to the invention can therefore also be used with switching matrices configured in any way.

In the circuit examples discussed so far, special aids were provided for restricting the marking in the switching matrix. In the circuit example dealt with below, switching means which are already present in any case are also used for this purpose, which are inserted into the route search wires assigned to the intermediate lines and identify their occupancy status. For this purpose, i.e. if they belong to non-selected intermediate lines, they are brought into the operating position that they have when the intermediate lines in question are occupied. This circuit example is shown in FIGS. FIG. 5 shows an extract from the route search network and the selection devices provided for the individual interfaces. These are essentially intermediate line selectors. The switching means used to restrict markings are assignment contacts that are inserted into the route search wires. They are closed when the associated intermediate line is free and open when it is occupied. It is useful here to select Intermediate lines in order to define a connection route. Accordingly, intermediate line selectors are provided on the intermediate lines that run between the coupling stages B and C, C and D, D and E and E and F, namely the intermediate line selectors ZLBC, ZLCD, ZLDE and ZLEF. Here, too, decoupling directional conductors are inserted into the path search cores, which are polarized in such a way that they do not hinder the transmission of the markings to be made. Two different marking potentials are used here for input and output marking. B. positive potential for input marking

ίο and z. B. negative potential for output marking.

At the interfaces are in the route search veins

Pairs of changeover contacts inserted. With their help, the inputs of the link selector are connected to the route search wires via the coincidence circuits individually assigned to the route search wires, provided that one of the marked route search wires is to be selected at the relevant interface. If there is no dialing process to take place, the changeover contacts have such an operating position that the associated path search vein is switched through. In leading from Markierknoten fDn to Markierknoten fE 1 intermediate line the changeover are 1 rde 1 η and 2 rde 1 n. If they are brought into working position, so this Wegesuch- be separated ädern and the resulting ends to the two inputs of the coincidence circuit Vdein connected. An input of the intermediate line selector ZLDE is connected to the output of this coincidence circuit.

Since a total of η ■ ο intermediate lines run between the switching matrices of the switching stage D and the switching matrices of the switching stage E (see grouping plan Fig. 1), the intermediate line selector ZLDE also has η ■ ο inputs. There are also no coincidence circuits. The intermediate line selector ZLDE has its inputs individually assigned outputs to which relays are connected. This includes the relay ZdeIn. These relays take over the evaluation of the selection process carried out by the intermediate line selector ZLDE. After completing a selection process, the relevant relay is energized via an output.

The route search sequence will now be described using the circuit example shown in FIG. 5. A connection path between the input TIj and the output Z11 should be sought. The required marking is made by closing the contacts ti j and ζ 11. The first interface is here z. B. between the coupling stages D and E. There is the intermediate line selector ZLDE. First of all, at the first interface, the pairs of changeover contacts inserted in the route search wires are brought into the working position. Negative marking potentials are fed from the output Z11 to one of the inputs of the coincidence circuits present at these interfaces, which belong to intermediate lines that can be reached from the marked output. The other inputs of coincidence circuits, which belong to free intermediate lines and can be reached from the marked input, are supplied with the positive marking potential from the input TIj. The coincidence circuits used here at the interfaces have the property of only delivering a certain potential at their output if a negative potential is fed to their one input and a positive potential to their other input.

This means that only the inputs of the intermediate line selector ZLDE assigned to the free and suitable intermediate lines are supplied with this specific potential. When the dialing process then takes place, one of the intermediate lines assigned to these inputs is selected. Let this be the intermediate line that connects the switching matrices Dn and El and whose associated path search artery is shown in FIG.

The relay Zdeln then also comes under power via the associated output of the link selector ZLDE. As a result, as will be explained later, all occupancy contacts that are located in path search veins that run between marker nodes of the coupling stages D and E are opened. However, this does not include the occupancy contact bdeln, which is located in the path search artery belonging to the selected intermediate line. In addition, at the one adjacent interface, namely at the interface between the coupling stages C and D, all pairs of changeover contacts are then brought into the working position for the purpose of selecting an intermediate line. The analog processes then take place, as described above, in the intermediate line selector ZLCD . The intermediate line leading from the switching matrix Cm to the switching matrix Dn may be selected here. The relay Zcdnm connected to the link selector ZLCD then comes under power. This relay also causes the marking to be restricted by opening occupancy contacts that were previously closed because they belong to free intermediate lines. The occupancy contact belonging to the selected link , i.e. contact bcdmn, remains closed.

The analog dialing processes now take place at the neighboring interface, which is located between the switching stages B and C. It is not necessary to restrict the marking here, since this interface has no neighboring interface at which a choice still has to be made.

However, a choice must also be made for the second interface adjacent to the first interface, that is to say for the interface between the coupling stages E and F. It is triggered with the aid of the relay Zbc. This relay is connected to the outputs of the link selector ZLBC via the OBC mixing gate . With this choice, the intermediate line selector ZLEF has to select an intermediate line. The intermediate line leading from the switching matrix E 1 to the switching matrix Fp may result.

If only one intermediate line runs between the switching matrices of the switching stages F and G as well as B and A , the entire connection path to be searched for has now been found. A switching matrix is determined by the choice of intermediate line in the switching stages B and F , namely the switching matrices B 1 and Fp at which the selected intermediate lines end. In addition, a switching matrix is also defined in each of the switching stages A and G , namely the switching matrix A 1 through the input TIj to be used and the switching matrix Gl through the output ZIl to be used. Since intermediate lines were chosen between the other switching stages, more than one intermediate line can also run between two switching matrices of adjacent switching stages there. If this is not the case, the interfaces can also be placed in such a way that one coupling stage is skipped at a time. This is already described in the main patent in connection with a method aimed at the selection of intermediate lines. It should also be noted that the first interface can easily be placed anywhere in the switching network. This method can also have multiple inputs

ίο and / or exits taken into account when searching for a route will.

In FIGS. 6 and 7, the devices are shown which are used to actuate occupancy contacts in the routing cores when intermediate lines are occupied and with the help of which restrictions can be made to the marking, as well as devices which are used to control the sequence of dialing and marking processes to serve. These devices are related to the circuit example according to FIG. 5. FIG. 6 shows an excerpt from the network of assignment cores, the c cores, namely those cores which lead from input T1 to output Z11 and to the selected intermediate lines belong. The occupancy wheels run exactly like the speech cores α and b (see Fig. 2), in the case of the switching multiples via crosspoint contacts, of which the crosspoint contacts 2kaljl, Ikbllm, Ikcmln, Ikdnml, Ikelp, 2 kfρ 11 and Ikglpl are shown.

Occupancy relays assigned to the intermediate lines are connected to the occupancy wires. For example B. to the intermediate line that leads from the switching matrix A 1 to the switching matrix B 1, the occupancy relay BAB 11. This occupancy relay has the occupancy contact bab 11. The occupancy relay BBCm 1 belongs to the intermediate line that leads from the switching matrix B 1 to the switching matrix Cm his occupancy contact bbcml. If, for example, this latter intermediate line is occupied, a path leads from there via closed crosspoint contacts from at least one input to at least one output of the switching matrix. There, a positive potential is applied to the occupancy wires. This therefore has an effect on the occupancy relay BBCm 1, so that it is energized, and its occupancy contact bbcmi is therefore open. The associated route search artery is therefore not taken into account when searching for a route, since it is interrupted by the open occupancy contact. It should be noted that the occupancy veins c yet to hold actuated crosspoint contacts suitable auxiliaries, namely retaining coils of cheese constituents as the holding coil HA Iy, HBIl etc. are connected.

The occupancy relays, which belong to intermediate lines at interfaces where the marking must be restricted, each have two windings I and II instead of one winding. These include the occupancy relays BCDnm and BDEXn. The winding parts are additional windings. They are each in series with a normally closed contact that is actuated when the associated intermediate line has been selected by an intermediate line selector. So is z. B. the winding II of the occupancy relay BDEIn in series with the normally closed contact 1 zdeln. This normally closed contact belongs to the relay Zdeln, which is energized via an output of the link selector ZLDE, which supplies the link between the coupling

multiple Dn and £ 1 is assigned, i.e. precisely the intermediate line to which the occupancy relay BDEIn also belongs.

The windings II of the occupancy relays belonging to the same interface are also in series with a contact of a control relay, which is energized when the marking is to be restricted. The windings II of occupancy relays, which belong to the interface between the coupling stages D and E , are connected in this way to the contact 2 red of the control relay RCD (see FIG. 7). When this control relay is energized, all of these windings II are energized, with the exception of the winding in which the individually assigned and in series break contact has been opened. This is winding II that belongs to the occupancy relay of the selected link. All occupancy contacts are therefore opened, with the exception of the one belonging to the selected intermediate line. In this way, the desired restriction of a marking can be made. For example, the normally closed contact lzdeln can be opened before contact 2 red is closed. The occupancy contacts are then opened at the interface between the switching stages D and E for all intermediate lines, except for the one leading from the switching matrix Dn to the switching matrix E 1, unless they were already open because the relevant intermediate lines were busy.

The control relays RDE, RCD, RBC and REF (see FIG. 7) are used to control the sequence of the selection processes taking place at the various interfaces and the marking restrictions. Each interface has one of these relays. Among other things, they operate the pairs of changeover contacts inserted in the route search wires of this interface. The control relay RDE z. B. belongs to the interface between the coupling stages D and E and actuates, among other things, the changeover contacts 1 rde 1 η and 2 rde In. It can be energized by actuating the control contact st . It is de-energized again by actuating the normally closed contact lrcd, which is actuated when the control relay RCD belonging to the adjacent interface between the coupling stages C and D is energized. The relay RCD is energized when one of the relays connected to the intermediate line selector ZLDE is energized, as these relays have contacts provided for this purpose. The normally open contact 2 zde 1 n belongs to them. The other control relays are each temporarily energized in a corresponding manner.

The sequence of the route search already described will now be briefly repeated in connection with this, in particular with regard to its control by the control relays etc. First, after the marking potentials have been applied to the input to be used and to the output to be used, the control contact st is closed. As a result, the control relay RDE is energized and operates its changeover contacts, whereby the intermediate line selector ZLDE is connected to the path search wires via the coincidence circuits Vdein etc. After the dialing process in the intermediate line selector ZLDE comes z. B. the relay Zde In is energized. With its normally closed contact lzdeln, it prevents winding II of the occupancy relay BDE1 η from being energized. With its normally open contact 2 zde In, it energizes the RCD control relay. Its contact 2 red energizes the windings II of the occupancy relays at the first interface, with the exception of the occupancy relay BDEIn. The previous marking is therefore restricted there. Its break contact lrcd again interrupts the circuit of the control relay RDE. In addition, due to its switchover contacts, to which

ίο the changeover contacts 1 rcdnm and 2 rcdnm belong, the intermediate line selector ZLCD is now connected to the relevant intermediate lines. The same processes as with the first interface are repeated here. Then the control relay RBC is energized via the contact 2zcdnm. After the selection process in the intermediate line selector ZLBC , the relay Zbc is energized and, with its contact zbc, energizes the control relay REF , which also initiates the selection process at the interface between the coupling stages E and F. After its completion, the path search is over.

The actuation of the contacts ti j and ζ 11 and, if applicable, the control contact st takes place with the aid of a central control device, which is referred to as a marker. Such a marker can be used in all of the circuit examples described above. The mode of action of such markers is already known, so that it does not need to be discussed further. The marker to be used here also has to ensure that any connection requests that arise are processed one after the other, for which purpose it temporarily applies marking potential to the inputs and outputs in question. Then the crosspoint contacts belonging to the connection path found are to be actuated, whereby this connection path is set, ie switched through. As already mentioned, the coupling matrices can be implemented by means of coordinate switches of various types, such as crossbar selectors, cross-coil selectors or relay couplers. The way in which the selected connection path is set depends on the type of crossbar switch used and also depends on whether the connection path is determined by selected intermediate lines or switching matrices. In the main patent several examples are described in detail how, after the selection of the relevant route sections, the setting of the connection route found is to be carried out when using crossbar switches. Switching means are provided here for setting certain connection paths, which are used to actuate crosspoint contacts and are looped into a network of setting cores superimposed on the coupling matrix. To hold the actuated crosspoint contacts, when using cross-coil selectors, special holding coils are provided which are connected to the network of the occupancy wires.

Claims (17)

PATENT CLAIMS: 1. A method for searching for and selecting free connection paths in a switching network containing any number of switching stages, which has a path search network, the wires of which are assigned to its intermediate lines and over Marking nodes assigned to switching matrices are connected to one another and in which the input and output belonging to the desired connection are marked at the same time, whereupon a section of the path marked from the input and output is selected at an interface running across the path search network and then marked again, this renewed marking opposite to the original marking in the route search network is transmitted to further interfaces at which a further suitable section of the route is selected with the help of the markings that now coincide here, until the selected route sections define a route in the route search network, according to which a connection route is to be switched through, according to the patent 1 048 956, characterized in that in each case after the selection of a route section (switching matrix or intermediate line) at an interface, the renewed marking of this route section by restricting the section at this interface le previously existing markings is brought about in such a way that only the selected section of the road is marked here. 2. The method according to claim 1, in which at least one further interface is placed in the switching matrix on both sides of the first interface, characterized in that to restrict the markings first at the first interface (coupling stage D) for all sections of the road except for the selected section of the road (Coupling multiple Dn or marking node fDn) interrupted the original marking paths and then, after selecting one section of the path at the adjacent interfaces (coupling stages C and E) at the first interface (coupling stage D), the remaining marking path (via fDn) also interrupted and instead marking paths via the last selected path sections (coupling multiples Cm and E 1 or marking nodes fCm and / El), whereupon the last existing marking paths are interrupted at any other adjacent interfaces (coupling levels B and F) after selecting a path section and only over the last selected Sections of the route are closed again, and that these process steps are repeated until the routing sought is established via the switching matrix (Fig. 3). 3. Circuit arrangement for carrying out the method according to claim 2, in which coupling matrices are selected as path sections, characterized in that a first potential (+) for input and output marking and a second potential (ground) deviating therefrom for marking marking nodes at the interfaces be used that decoupling directional conductors (Gabll, Gcdnm, Gdeln, Gefpl, Gfglp,... ), which are polarized in such a way that they do not hinder the transmission of markings, and occupancy contacts (bab 11, bbcm 1, bcdnm, bdeln are befpl, bfglp,...), which are closed with freedom of the respective intermediate pipes and open at their Belegtsein inserted, that at the first interface (switching stage D) by means of relays (lSDN / lSDN,...), which are inserted into the routing network, the marking status of the marking nodes (fDn, ...) is reported to a central switching matrix selector (KVD) , which is used to select a switching matrix and to Mar kierung of the associated Markierknotens is represented by exclusive application of the second Markierpotentials (ground) is carried out at the belonging to this switching matrix Markierknoten (fDn), and that at the adjacent interfaces (coupling steps C, E, B, F) corresponding central switching matrix selector (/ CFC, KVE, KVB, KVF) are provided, which are also used to select a coupling multiple and mark the associated marker node (Fig. 3). 4. The method according to claim 1, in which further interfaces are placed in the switching matrix only on one side of the first interface, characterized in that the marking (-) is canceled to restrict the markings present at the first interface (coupling stage D) the side of the switching matrix closest to the interface (side with exits) is created, and is replaced by a marking ^ -) on the section of the path selected at the first interface (coupling manifold Dm or marking node fDn) , whereupon after selecting a path section at the adjacent interface (coupling stage C) the marking (-) applied at the first interface (coupling stage D) is canceled again and by a marking (-) applied to its neighboring interface (coupling stage C) at the selected path section (coupling manifold Cm or marking node fCtri ) is replaced, and that these procedural steps are continued from interface to interface until routing is defined in the switching matrix (Fig. 4). 5. A circuit arrangement for performing the method of claim 4, are selected in the switching matrices as paths pieces, characterized in that different for the input and output labeling two resistors (RtIj;... RzIl, ■ ■ ■) applied Markierpotentiale (+, -) are used that decoupling directional conductors (Gabll, Gbcml, Gcdnm, Gdeln), which are polarized in such a way that they do not hinder the transmission of the markings, and occupancy contacts (bab 11, bbcml, bcdnm, bde 1 n), which When the relevant intermediate lines are free and open when they are occupied, it is inserted that at the interfaces (switching stages D, C, B) centrally arranged switching multiple selectors (KVD, KVC, KVB) are provided, the inputs of which are connected to the relevant interface Marking nodes (fDn..., FCm..., FBl...) Are connected and which, under the marking nodes, have an intermediate potential between the marking potentials l, select one in each case, whereby a switching matrix (Dn, Cm, B 1) is determined, and there apply the marking potential (-) which was previously on the side of the switching matrix (side with Outputs) and at the same time interrupt the previous supply of this marking potential (-) (Fig. 4). 109 759/99 6. Circuit arrangement for performing the method according to claim 4, in which coupling matrices are selected as path pieces and in which ferromagnetic toroidal cores with approximately rectangular hysteresis loops are used, characterized in that two different marking potentials (+, -) are used for input and output marking, that in the path search arteries decoupling directional conductors (GabU, Gbcml, Gdeln, Gefpi), which are polarized in such a way that they do not hinder the transmission of the markings and occupancy contacts (babll, bbcml, bdeln, bejpV), which are closed when the relevant intermediate lines are free and when they Occupied are open, are inserted that at the interfaces (coupling stage B, C, D) in the marker nodes QBl ..., FCm ..., fDn ...) Assigned to the switching matrices, first windings (1) of toroidal cores (RDn. .., RCm..., RB 1) are looped in via which, under the influence of the marking potentials (+, -), if they are in connection paths in question , a maximum of the simple magnetizing current flows that in each case at the interfaces (coupling stage B, C, D) second windings (2) of the toroidal cores (RB 1 ..., RCm. . ., RDn) are combined and are traversed jointly by double the magnetizing current in the opposite direction and third windings (3) are individually connected to the inputs of the interfaces (coupling stage B, CD) assigned central switching matrix selectors (KVB, KVC, KVD) that for The purpose of choosing a coupling multiple at an interface is to interrupt the double magnetizing current flowing there, which means that the connection paths in question belong to the third windings (3) of the toroidal cores (RB 1 ..., RCm ..., RDn ...) , a voltage surge occurs that the relevant switching matrix selects one of the switching matrices belonging to the latter toroidal cores and the marking potential (+) that was previously applied to the first interface (switching stage B) on the side (side with inputs) of the switching matrix that is closest to the the marker node (/ Bl, fCm, fDn) belonging to the selected coupling array (B 1, Cm, Dn) (Fig . 8th). 7. Circuit arrangement according to claim 6, characterized in that in each case with the connection of the marking potential (+) to a marker node (JBl, jCm, fDn) belonging to a selected switching matrix (B 1, Cm, Dn ) with the aid of a delayed responsive monostable multivibrator (MB, MC, MD) twice the magnetizing current at the adjacent interface, at which the next selection is to be made, is temporarily switched off (Fig. 8). 8. Circuit arrangement according to claim 6 or 7, characterized in that the marking potential (-) which is only supplied up to one selection process is supplied via an AND circuit (UB 1, UCm) , one input of which is connected to the relevant output of the switching matrix selector (KVB , KVC) of the associated interface and the other input of which is connected to a bistable multivibrator (KC, KD) , which is connected via a mixer (OC, OD) to the outputs of the switching matrix selector (KVC, KVD) that comes into operation at the next election is (Fig. 8). 9. Circuit arrangement according to one of claims 6 to 8, characterized in that limiting resistors (Rt 1, RzI) are provided at its feed points to limit the current flowing when marking potential (+, -) is applied (Fig. 8). 10. The method according to claim 1, characterized in that in order to restrict the markings present at the interfaces (intermediate coupling stages DE, CD, BC, EF) in the path search cores (/ cores) inserted and the occupancy state of the associated intermediate lines characterizing switching means (bdeln .. ., bcdnm..., bbcml ..., befpl...) are also used by bringing them to the operating position that they have when the intermediate line in question is occupied (open) if they belong to non-selected intermediate lines (Figures 5 to 7). 11. Circuit arrangement for carrying out the method according to claim 10, in which intermediate lines are selected as path sections, characterized in that two different marking potentials (+, -) are used for input and output marking in the path search network, that decoupling directional conductors (Gab 11. .., Gbcml..., Gcdnm..., Gdeln ..., Gfglp. ..), which are polarized so that they do not hinder the transfer of the markings and occupancy contacts (bah 11 ..., bbcm 1 .. ., bcdnm ..., bdeln ..., befpl ..., bfglp. ..), which are closed when the relevant intermediate lines are free and open when they are occupied, are inserted that at the interfaces with the help of pairs of changeover contacts individually (Irdelnllrdeln..., Ircdnmllrcdnm ..., lrbcmlllrbcml ... lrefpl / lrefpl) which are inserted into the routing wires, centrally arranged sequentially intermediate line selector (ZLDE, ZLCD, ZLBC, ZLEF) via the routing wires (/ veinlets) assigned co incidence circuits (Udein. .., Ucdnm ..., Ubcm 1 ..., Uefp 1 ...) are connected to the route search wires (/ wires) and that these intermediate line selectors (ZLDE, ZLCD, ZLBC, ZLEF) select one of the marked route search wires (Figures 5 to 7). 12. Circuit arrangement according to claim 11, characterized in that in each case after the Selection of a route search vein the occupancy contacts leading to non-selected intermediate lines belong to be opened (Fig. 5 to 7). 13. Circuit arrangement for performing the method according to one of claims 1, 2, 4 and 10 or according to one of Claims 3, 5 to 11, characterized in that for consideration several outputs and / or several inputs of the switching matrix in the determination of a connection path through the relevant inputs or outputs on the switching matrix Marking potential are designated at the same time and that after determining a connection path between a switching matrix to which inputs are connected and a switching matrix to which outputs are connected, each with at least one free input or output through one selection process each for the such specified switching matrices a single free input or output for the connection path is determined (according to patent 1,062,761). 14. A circuit arrangement according to claim 13, characterized in that a place selector {PZDE in FIG relevant switching matrix determines an input or output. 15. Circuit arrangement according to one of claims 3, 5, 6 to 9, 13 and 14, in which is selected under switching matrices and in the case of the present switching network, the switching matrices of two adjacent switching stages are each connected via several intermediate lines, characterized in that additional intermediate line selector (PZBC in Fig. 9) are provided, which each have as many inputs as a maximum of intermediate lines run between the switching matrices of the respective switching stages (B, C) and with the aid of which a selection is made among these intermediate lines. 16. Circuit arrangement for performing the method according to one of claims 1, 2, 4 and 10 or according to one of claims 3, 5 to 9, 11 to 15, characterized in that switching means for actuating crosspoint contacts for setting of connection paths determined by the route search (1 / ta 1/1, lkbllm, lkcmln, lkdnml, lkelp, lkfpll, lkgpl in Fig. 2) are looped into a network of setting cores superimposed on the coupling field. 17. Circuit arrangement according to claim 16, characterized in that the holding of the actuated Crosspoint contacts is made with the help of switching means that are connected to the Coupling matrix superimposed network of assignment cores are connected. For this purpose 2 sheets of drawings © 109 759/99 1.62