AU8203898A - Ring circuit with transport loop and card protection - Google Patents

Description

1 Ring Circuit with Transport Loop and Card Protection The present invention relates to a ring circuit in accordance with the preamble to patent claim 1. To protect the transmission of data within a telecommunications network, dual ring circuits are often used that have a working ring and a protection ring which connect the applicable network nodes with one another. Such systems are explained in such references as [5] and [6]. The applicable ITU recommendations for Synchronous Digital Hierarchy transmission networks are specified in Section 5 of [3] (see also Section 2.7.2 of [4]). Synchronous Digital Hierarchy transmission networks (SDH (SONET in the USA)) are characterized in that all participating network elements work in the normal case with one single, centrally generated clock frequency that is extracted from the data stream of a selectable input. In this way a tree-like frequency distribution network that comprises several network nodes can be constructed through a data network with any desired structure. On the basis of a piece of information supplied with the signal that identifies the quality of the frequency of the relevant data stream, automatic switching to the best alternate frequency source can be performed in case of malfunctions. At the transitions from SDH-transmission networks to Plesiochronous Digital Hierarchy transmission networks (PDH), multiplexers are provided (terminal multiplexers, add/drop multiplexers or frequency division multiplexers) that under certain circumstances are supplied with the system clock only over a single data connection. The Plesiochronous Digital Hierarchy (PDH) does not allow an individual channel to be taken directly from a data stream; all hierarchy levels of the multiplex system in which the channels are combined into systems with an ever increasing channel count must -rT 2 always be traversed. On the receiver side, these same hierarchy levels are traversed in the reverse order in order to be able to then further distribute the individual channels. The Synchronous Digital Hierarchy (SDH), in contrast, permits direct access to signals of specific bandwidth within a many-channeled system, in order to be able to route these signals to a subscriber or an exchange. It is also possible to access the broad-band signal stream in order to replace specific signals by others without having to traverse the entire multiplex hierarchy. This access is achieved through a computer-controlled switching network. The structure of the data streams transmitted in the Synchronous Digital Hierarchy (SDH) is described in detail in [2] and [4]. The Synchronous Digital Hierarchy (SDH) is based on the synchronous transmission of transport modules (STM-n) in which data are inserted. A base transport module STM-1, which consists of a frame with 9 lines and 270 columns or 2430 fields with 8 bit data capacity, has a maximum data content of 19440 bits. The STM- 1 modules are transmitted at a clock frequency of 8000 Hz, by which means a transmission channel with a capacity of 155.52 Mbit/s is created. As is shown in Fig. 1, rows 1 - 3 (Regeneration Section Overhead) and 5 - 9 (Multiplex Section Overhead) form the first 9 columns of a synchronous transport module STM-1 of the section head (Section Overhead - SOH). Row 4 of the first 9 columns contains an administrative unit AU-4, in which is provided an administrative unit pointer (Pointer) AU-4 PTR, which identifies the field in which a signal accommodated by the administrative unit AU-4, or the first field (JI) of a virtual container (e.g. VC-4), begins. The remaining 261 columns, which are provided to accommodate the virtual container VC-4, form the data field (payload), which is structured differently depending on the structure and the transmission rate of the data to be transmitted. For example, three 34 Mbit/s channels or 63 2 Mbit/s channels or even a continuous sequence of ATM cells can A4 be contained in one virtual container VC-4 that has a path frame head (Path Overhead 0, 3 POH). The defined multiplex structure is shown in Fig. 6-1 and Fig. 6-2 of [2]. A virtual container VC-4 can, in addition to the frame head POH, contain one container C-4, three virtual containers VC-3 or 63 virtual containers VC-12, which each have one container C 3 or C-12 as well as a frame head POH. The virtual containers VC-3 and VC-12 are relocatably contained in transport frames known as Tributary Units TU-3 or TU-12, which have, time-multiplexed in the first byte, a tracked administrative unit pointer (Pointer) that points to the first field of the virtual containers VC-3 and VC-12. The transport frames TU-3 and TU-12 are combined in transport groups TUG-3 or TUG-2 and TUG-3. One transport group TUG-3 contains three transport frames TU-3 or seven transport groups TUG-2, of which each contains three transport frames TU-12. Through the head fields POH, the data are identifiable down to the container level. Individual payload channels can therefore be taken from or added to a transport module STM-1 without disassembling the entire synchronous module STM-1. The beginning of a transmission in the data field is specified through the pointer PTR contained in the relevant frame structures (AU-4, TU 3, TU-12). Individual containers are thus identifiable and can be combined in various ways by the elements of the Synchronous Digital Hierarchy (SDH) and transmitted through the network. In accordance with section 2.7 and section 2.11 of [1], switching matrices (higher order path connection function (HPC-n) or lower order path connection function (LPC-n)) are provided for virtual containers VC-n with higher order numbers (n = 3 or 4) and for virtual containers VC-m with lower order numbers (m = 11, 12 or 2). The structure of a device for transferring data streams for the Plesiochronous Digital Hierarchy into data streams for the Synchronous Digital Hierarchy is described in [1] and shown therein at the module level in Figure 2-1. The functions of the individual modules are described in detail in [1] - [4]. The network architecture to which the device corresponds is described in detail in [3] and [4]. In that context, the network essentially consists of three types of components: adaptation modules (each shown in the form of a trapezoid), termination modules (each shown as a triangle), and connection units (see for 4 example Fig. 4-1 of [3] or Fig. 2.7 of [4]). The functions of the modules are explained briefly on the basis of Fig. 9, which already shows a device in accordance with the invention. A synchronous interface SPI is provided for conversion and synchronization of the optical signals received over a glass fiber. Only the termination module OST is shown (the conversion and synchronization must always be provided in an adaptation module). If a fault is detected in the synchronous interface SPI, an error message LOS (loss of signal) is sent to the regenerator section RS. The frame synchronization of the transport modules STM-n as well as the scrambling and descrambling of the data take place in the adaptation module RS/OS of the regenerator section RS. Frame synchronization is performed using the six data bytes Al, A2 (see Fig. 10) contained in the first line of the section overhead (SOH or RSOH). If frame synchronization cannot be established during a specific period, an error message LOF (loss of frame) is generated. In the termination module RST (regenerator section termination) of the regenerator section RS, moreover, an error control procedure (bit interleaved parity check BIP-8) takes place using the data byte B 1 contained in the second line of the section overhead (SOH or RSOH). If an error is detected or an error message LOS or LOF is already present, an alarm signal AIS (alarm indication signal) is sent to the multiplex section MS. In another adaptation module MS/RS of the regeneration section RS, adaptation to auxiliary layers takes place, through which a communications channel between the regenerators (using the RSOH bytes Dl, D2 and D3), a voice channel for service purposes (using RSOH byte E1) and a user channel (using the RSOH byte F1) are created. In the termination module MST (multiplex section termination) of the multiplex 4Z 5 section MS, checking of the signal quality is performed using the three bytes B2 contained in the fifth line of the section overhead (SOH or the first line of the MSOH), whereupon in case of reduced quality, the error message "signal poor" (signal degrade SD), or in case of bad quality, the error message "signal failed" (signal fail SF) is sent to the multiplex section protection module MSP, which is essentially an expansion of the termination module MST. Moreover, the three bytes KI and K2 contained in the fifth line of the section overhead (SOH or the first line of the MSOH) are sent to the multiplex section protection module MSP. If the bit pattern 111 occurs in bits 6, 7 and 8 of byte K2, an alarm signal AIS is detected. If the bit pattern 110 occurs in bits 6, 7 and 8 of byte K2, a remote receiver failure FERF (far end receiver fail) is detected. In the adaptation module MSA (multiplex section adaptation), the position of the virtual container VC-4 within the payload is determined through pointer processing. A switching matrix for higher order virtual containers (higher order path connection) is not shown. In the termination module HPT (higher order path termination), the evaluation of the path frame head (path overhead POH) of the virtual container VC-4 takes place, whereupon the position of the virtual containers VC-12 or VC-3 within the payload is determined in the adaptation module HPA (higher order path adaptation) through pointer processing. In the switching matrix LPC (lower order path connection), the lower order virtual containers VC-12 or VC-3 are switched in accordance with the traffic paths provided. A monitoring module LPOM (lower order path monitoring), which evaluates parts of the path frame head (path overhead POH) of the virtual containers VC-12 or VC-3, serves to establish alarm and quality information, which is provided for protection measures (sub network protection). Particularly when a switching matrix HPC is used for higher order virtual containers, an appropriate monitoring unit HPOM is provided. z' 6 In the termination module LPT (lower path termination), the evaluation of the path frame head (path overhead POH) of the virtual containers VC-12 or VC-3 takes place, whereupon the localized containers C12 or C3 (see Fig. 1) are sent to the adjacent adaptation module LPA (lower order path adaptation), where it is transferred synchronized into a data stream of the Plesiochronous Digital Hierarchy and delivered to the Plesiochronous Interface PPI (interface to the Plesiochronous Digital Hierarchy). The modules described above thus permit fault recognition in the multiplex section (multiplex section protection) or in deeper layers of the network (sub-network path connection protection). Measures to protect the transport functions of the network are described in Section 5 of [3] (see also Section 2.3 of [7]). In this process, faulty (failed or degraded) transport units are replaced by protection units, which are present in the ratio m:n (m=protection, n=working; normally, m=n= 1; more rarely, m= 1, n> 1). Sub-network connection protection is accomplished through switching of (lower or higher order) virtual containers VC-n to protection channels by means of the switching matrices (LPC or HPC). Ring structures, as are described in Section 5 of [3], are preferably used for protection of transport paths. In this context, as shown in Fig. 2, several network nodes A, B, C, D, which are provided as transfer components between the Plesiochronous and the Synchronous Digital Hierarchy, are connected to one another through two opposing transport loops tsw (working loop) and tsp (protection loop). The working loop tsw (working) is protected in this context by the protection loop tsp (protection). Data transmission from network node A to network node C takes place through the working loop tsw via network node D and through the protection loop tsp via network node B. If one of the transport loops tsw or tsp fails (e.g. through a defect occurring at the network node B or D), transmission takes place through the remaining transport loop tsw or tsp. If one of the two network nodes A or C fails, however, no data transmission is possible. r O 7 For this reason it is customary to additionally protect the network nodes A, B, C, D with a protection unit. Thus in Fig. 3, two network cards AW, AP; BW, BP; CW, CP and DW, DP are present at each network node. In the event of failure of the first network card AW (working card) at network node A, data transmission takes place over the second network card AP (protection card). Fig. 4 shows a network card, known from [6], connected to two transport loops (2-fiber ring), that could be used in the ring circuit of Fig. 3. Fig. 5 shows two network cards per Fig. 4, incorporated in the two transport loops tsw and tsp. In this case, the full data traffic of both transport loops tsw and tsp passes through the two network cards, for which reason the full processing capacity, i.e. a total of four add/drop multiplexers, must be available in each of these network cards for each transport loop tsw and tsp. This results in significant cost. Further, in case of a complete failure of one of the two network cards shown in Figure 5, both transport loops tsw and tsp are interrupted, which results in a severe limitation of the opportunities for protection. In particular, the network cards that are still functional can only be supplied with the data of one transport loop tsw (working) or tsp (protection). The object of the present invention is therefore to specify a ring circuit with transport loop protection and card protection that can be implemented at low cost. This object is achieved through the measures specified in the characterizing portion of patent claim 1. Advantageous embodiments of the invention are specified in additional claims. The ring circuit in accordance with the invention can be implemented with simple to-produce network cards, since the two transport loops tsw and tsp do not pass through the two neighboring network cards (e.g. AW and AP). Only the data of one transport loop 8 tsw or tsp are processed in each network card. Since the two transport loops tsw and tsp do not pass through both neighboring network cards, only one transport loop tsw or tsp is interrupted when a network card (e.g. AW or AP) fails, for which reason the data of both transport loops tsw or tsp are still present at the neighboring card. This is significant not only for reasons of protection, but also because the protection channels can be used in normal operation for the transmission of non-redundant data. Upon failure of a working card, the neighboring protection card assumes its tasks and can thus forward the non redundant data of both transport loops tsw and tsw e.g. to a PDH network. With the reduction in cost in accordance with the invention, therefore, an improvement of the protection function is additionally achieved. The ring circuit in accordance with the invention is preferably employed in Synchronous Digital Hierarchy networks. However, an advantageous application of the solution in accordance with the invention is also possible in other networks e.g. Plesiochronous Digital Hierarchy networks. The invention is described below in greater detail by way of example with the aid of a drawing. Shown in: Fig. 1 are possibilities for construction of an STM-1 frame, Fig. 2 is a known ring circuit with four network nodes A, B, C and D, which are connected to one another through two transport loops tsw, tsp, Fig. 3 is a known ring circuit with four network nodes, at each of which are provided two network cards AW, AP; BW, BP; CW, CP and DW, DP, Fig. 4 is a known network card with add/drop functionality for two transport loops tsw, tsp, Fig. 5 are two network cards in accordance with Fig. 4 connected in series, Fig. 6 is a ring circuit in accordance with the invention with two network cards El, WI; E2, W2; E3, W3; E4, W4, that are passed through by two transport loops tsw, tsp, in the normal operating state,

--L>

9 Fig. 7 is the ring circuit in accordance with Fig. 6 after a failure of the network card W3 and an interruption of the transport loop tsw, Fig. 8 is one possible construction of a network card El, ..., E4; WI, ..., W4 in accordance with the invention, Fig. 9 is the ring circuit in accordance with Fig. 6 with additional switchover options, Fig. 10 are the columns 1 through 9 of a Synchronous Transport Module STM-n in detail, Fig. 11 is the network card E3 and W3 in the same configuration (north/south) although shown differently, and Fig. 12 is the ring circuit in accordance with Fig. 9 in case of a fault, Fig. 13 is an additional construction of a network card W3 * in accordance with the invention with a unit #s, which is suitable for selective through-switching of working channels and protection channels, and Fig. 14 is a network card W3 * in accordance with Fig. 13 as well as a network card E3* with a changeover switch s located outside of the card E3* for the working channel and the protection channel The object of Figures 1 through 5 and 10 was explained at the outset. Fig. 6 shows the structure of a ring circuit in accordance with the invention with four network nodes NK1, NK2, NK3 and NK4, each consisting of one working card and one protection card El, WI; E2, W2; E3, W3; E4, W4, that are connected to one another through one working loop and one protection loop tsw (working) and tsp (protection), upon which data from the same source are transmitted in opposite directions (the number of connected network nodes NK can of course differ significantly from the ring circuit described by way of example). The network cards El, W1, E2, W2, E3, W3, E4, W4 are constructed identically and have two inputs liI and li2 (see Fig. 6, E3 and W3), a ring-side output lo and at least one interface tio (PPI) to a network separate from the ring, for example a Plesiochronous Digital Hierarchy network. The first input liI of each network card E3 and )OkA 10 W3 is connected to the working loop tsw (working) and the second input li2 of each network card E3 and W3 is connected to the protection ioop tsp (protection). This situation is represented visually in Figure 11. Instead of coupling the transport loops tsw and tsp to the network cards E3 and W3 in an east-west direction in the known manner (the working loop tsw from the east and the protection loop tsp from the west), the working loop tsw is connected from the north, and the protection loop tsp (protection) is connected from the south to both network cards E3 and W3. The circuit of Figure 11 is identical here to the circuit of Figure 6. It must be noted that in contrast to the known arrangement of Figure 5, the output lo of the working card W3 is not connected to an input liI or li2 of the associated protection card E3, and the output lo of the protection card E3 is not connected to an input liI or li2 of the associated working card W3. On every network card E3 and W3, the data from input liI or li2 are routed through a switch s to a switching unit # (for example, an add/drop multiplexer), which in normal operation transmits the data through the input lo back to this same transport loop tsp or tsw as well as through an interface to a network separate from the ring. Therefore, in normal operation or in protection operation, every network card E3 or W3 terminates only the data flow of one transport loop tsw or tsp. In contrast to known arrangements (see Fig. 5) in which the data of both transport loops tsw and tsp are terminated and forwarded in the working card and the protection card, the solution in accordance with the invention results in a significantly reduced cost (only two add/drop multiplexers are required instead of the four used in the arrangement in Fig. 5). The specific network card W3 in which the data of the first transport loop tsw (working) are read in through switch s in normal operation is the working card. The specific network card E3 in which the data of the second transport loop tsp (protection) are read in through switch s in normal operation is the protection card.

11 As described, the data of both transport loops tsw and tsp are both present at both network cards E3 and W3 such that in case of a failure ofone network card E3 or W3 the data of both transport loops tsw and tsp can still be read in through switch s at the second network card W3 or E3. Through the solution in accordance with the invention, the network can therefore be protected against the failure of a network card (e.g. the working card W3) and simultaneously against a line interruption of the protection loop tsp in the east (the simultaneous failure of a west card and a line in the west is also noncritical in known arrangements since supply would take place from the east direction). As described with the aid of Fig. 7, the protection functionality is significantly improved by the measures in accordance with the invention as compared to known systems. In Fig. 6, data transmission takes place in normal operation from network node NK1 with the network cards El and W1 to the network node NK3 with the network cards E3 and W3. Data, for example from the Plesiochronous Digital Hierarchy, are carried by the working loop tsw via the working cards WI and W4 to the working card W3 where they are sent back to the Plesiochronous Digital Hierarchy network. On the other side, data from the Plesiochronous Digital Hierarchy are carried through the protection loop tsp via the two protection cards E1, E2 to the working card W3. Thus identical data are present at the inputs liI and li2 of the working card W3 insofar as the data are redundantly transmitted in normal operation as well, so that in case of a failure of the working loop tsw switchover to the input li2 can take place (reception of data from the protection loop tsp). Preferably, however, non-redundant data are transmitted through the protection loop tsp in normal operation. These data can therefore be transmitted to the protection card E3 and terminated there. With the aid of Fig. 7, the protection mechanisms are described that in case of failure of a working card (failure of the working card W3) and/or in case of interruption of 12 the working loop tsw. In known ring circuits with one working loop and one protection loop, the combination of these failures is no longer controllable. It can be seen from Fig. 5 that, with a failure of the west card, the data supply from the west is interrupted, and with a line interruption in the east, the data supply from the east is interrupted, whereupon the data no longer reach the subscribers connected to the relevant network node NK. As a result of the interruption shown in Fig. 7 at the output of working card W4 at network node NK4, the working loop tsw is interrupted before reaching the working card W3 at the network node NK3. An error (LOS / loss of signal) is therefore detected at the input liI of the working card W3. As a result of switching the switch s to the input li2, therefore, the data from the protection loop tsp are supplied to the working card W3, by which means data transfer to the subscriber is ensured. As a result of the failure of the working card W3 shown in Fig. 7, data transfer to the subscriber is interrupted. After detection of the failure, a control unit activates the protection card E3, to which the data transmitted through the protection loop tsp are supplied through the switch s. In case of a failure of the working card W3 and an interruption of the working loop tsw, the same measures are taken (activation of the protection card E3 and data transfer over the protection loop tsp. In case of a failure of the working card W3 and an interruption of the protection loop tsp, protection card E3 is again activated, whose switch s is changed over for reception of the data of the working loop at the input li1. In accordance with the invention, two critical faults can thus be controlled with simplified measures (use of two add/drop multiplexers instead of four). The faults occurring in the arrangement according to Fig. 7 have the consequence RIA4 that the working loop tsw does not forward any data to the network cards E2 and W2 of az~ 13 the next network node NK2. It would therefore be necessary to change to receiving data carried over the protection loop tsp, by changing over the switch s on the working card W2. A fault would thus propagate itself over one or more network nodes NKx, ..., NKy. In order to avoid this problem, in the ring circuit in accordance with Fig. 9 the outputs lo of the network cards E, W of every network node NK are connected on the east and west sides to a switch slE, ..., s4E or s1W, ..., s4W, from whose moving contact the working loop and protection loop are continued. In the event of a failure of one network card, the transport loops tsw and tsp can thus always be supplied with data from the second network card. The situation in the case of failure of the working card W3 at network node NK3 is shown in Figure 12. As described, the protection card E3 is activated, which takes over the exchange of data with the Plesiochronous Digital Hierarchy network (add/drop functionality) and transmits the appropriately modified data at output lo. Since no data are present at the output 1o of the failed working card W3 of network node NK3, the working loop tsw remains interrupted (cf. Fig. 7) if no further measures are taken. By changing the switch s3W, the input of the continued working loop tsw can be switched to the output lo of the protection card, by which means error-free operation at the interfaces of the network node NK3 can be ensured. Measures to correct the fault thus need only be taken within the network node NK3. The fault that has occurred does not have any effects on other network nodes (NK1, NK2, and NK4). Additionally shown in Fig. 12 is a control unit CTRL that receives the error messages from the working and protection cards and performs the necessary switching. In systems of the sd that receives error messages from the working and protection cards and performs the necessary switching. In systems of the Synchronous Digital Hierarchy, the "Synchronous Equipment Management Function" (SEMF) is provided for this purpose. The ring circuit in accordance with the invention can advantageously be employed

)/

14 in Synchronous Digital Hierarchy networks. One possible structure of a network card for SDH operation in accordance with the invention is shown in Fig. 8. It can be seen from this that separate test paths and processing or input paths, implemented with the adaptation and termination modules described at the outset, are provided therein for the working and protection loops. In this context the input paths are preferably carried in parallel to such a degree that complete checking of the data of both transport loops tsw and tsp is carried out separately, by which means the criteria for changeover between the transport loops tsw and tsp are obtained. Subsequently the two input paths are combined via the switch s, after which the add/drop functionality for the selected transport loop tsw and tsp is realized with the known adaptation and termination modules. In the present case the switch s is provided before the adaptation module HPA (higher order path adaptation) by means of which the position of the virtual containers VC-12 and VC-3 within the payload is determined through pointer processing. Moreover, the adaptation module HPA provided for the input paths is connected in parallel to the plesiochronous interfaces PPI with switching matrices LPC (lower order path connection) as well as termination and adaptation modules LPT, LPA. The data exchange with the Plesiochronous Digital Hierarchy network can take place in parallel within the deeper hierarchy levels. Proceeding from one of the switching matrices LPC, there is provided an output path leading to the output lo that is equipped with adaptation and termination modules of the higher SDH hierarchy levels. Between the relevant switching matrix LPC and the subsequent adaptation module HPA of the output path is interposed a monitoring module LPOM (lower order path monitoring) which provides the changeover criteria for the sub-network protection (lower order path protection) and the protection for the paths of deeper hierarchy levels. The circuit arrangement of Fig. 8 has the advantage that a comparison of the signal quality for the data delivered through the working loop and protection loop tsw; tsp can be performed. In relatively simple systems the signal checking can also be carried out after 15 the changeover switch s. Here, in the event of a signal failure a changeover to reception of the data from the second input 11 or li2 is performed without its quality being known. The described network card permits protection against errors occurring within higher or lower hierarchy levels (multiplex section protection and sub-network protection) within the ring circuit in accordance with the invention. As already described, protection is further provided against card failures (card protection). The solution in accordance with the invention can be used in any desired networks. It is essential that the error detection mechanisms necessary for implementing protection measures be provided for the network cards and the transport loops. The invention permits various solutions with regard to the through switching of working channels and protection channels. Fig. 13 [verb missing] the additional construction of an network card W3 * in accordance with the invention with a unit #s, which is suitable for the selective through-switching of working and protection channels. When a fault occurs, the channel in which a fault has occurred is determined through monitoring functions. If, for example, one of n working channels is detected as faulty, it can be replaced by a protection channel with the aid of the switching matrix #s. The switching matrix #s is therefore provided for the purpose of switching through the working channels and if applicable the protection channels that are functioning error-free. Preferably only one switching matrix #s, which is provided for lower order path connection function LPC, is used for all incoming and outgoing channels. Faults can therefore occur in the working channel and in the protection channel at the same time, without the a system failure occurring. A through-switching on the level of the higher order path connection function HPC is also possible. Advantageously, the hierarchy level upon which a changeover occurs is selectable. L1 Fig. 14 shows the network card W3* in accordance with Fig. 13 as well as a ONr 16 network card E3* with a changeover switch s located outside of the card E3* for the working channel and the protection channel. Instead of through-switching on the level of the higher or lower order path connection function HPC or LPC, a changeover of the working and protection channel takes place outside of the network card E3*. Bibliography: [1] ITU-T Recommendation G.783 (version 01/94) [2] ITU-T Recommendation G.707 (version 03/96, replaces the previous Recommendations G.707, G.708 and G.709) [3] ITU-T Recommendation G.803 (version 03/93) [4] M. Sexton, A. Reid, Transmission Networking - SONET and the Synchronous Digital Hierarchy, Artech House 1992 [5] WO 95/22860 [6] U.S. 5,517,489 [7] ITU-T Recommendation G.782 (version 01/94)

Claims (10)

1. Ring circuit with one working loop and one protection loop (tsw, tsp) through which several network nodes (NK1, NK2, NK3, NK4) of a first network (SDH) are connected to one another, each of which has one working card and one protection card (W 1, El; W2, E2; W3, E3; W4, E4) that are suitable for the exchange of data between the loops (tsw, tsp) and a second network (PDH) via an interface (PPI, tio), characterized in that the working cards and protection cards (WI, W2, W3, W4, El, E2, E3, E4) have on the loop side a first and a second input (lil, li2) that are capable of connection to an output (lo) through a first changeover switch (s) and a switching matrix (LPC; HPC, #, #s), in that the working loop (tsw) is connected through the first input and the protection loop (tsp) is connected through the second input (liI; li2) to the working card and the protection card (WI, El; W2, E2; W3, E3; W4, E4) of a network node (NK1; NK2; NK3; NK4), and in that the working loop (tsw) is continued from the output (lo) of the working card (W1; W2, W3, W4) and the protection loop (tsp) is continued from the output (lo) of the protection card (El; E2, E3, E4) directly or through a second or third changeover switch (s1W, s2W, s3W, s4W or, respectively, slE, s2E, s3E, s4E), wherein the changeover of channels takes place on a predetermined or selectable hierarchy level by means of a changeover switch (s) provided within or outside of the working cards and protection cards (W1, W2, W3, W4, E1, E2, E3, E4) or through the switching matrix (LPC; HPC, #, #s).

2. Ring circuit in accordance with claim 1, characterized in that the second and the third changeover switch (s1W, s2W; s3W, s4W or, respectively, slE, s2E, s3E, s4E) are connected to the outputs (lo) of the working card (WI; W2, W3, W4) and the protection card (E1; E2, E3, E4) so that in case of failure of a working card or protection card (WI or E1; W2 or E2; W3 or E3; W4 or E4) of a network node (NK1; NK2; NK3; NK4) the data can be taken from the output (lo) of the neighboring card (El or W1; E2 or W2; E3 18 or W3; E4 or W4) and sent to the associated loop (tsw; tsp).

3. Ring circuit in accordance with claim 1 or 2, characterized in that in normal operation the processing and routing of the data of the working loop (tsw) takes place on the working card (WI; W2, W3, W4) and the processing and routing of the data of the protection loop (tsp) takes place on the protection card (El; E2, E3, E4).

4. Ring circuit in accordance with claim 1, 2 or 3, characterized in that the first changeover switch (s) can be switched to receive data from the protection loop (tsp) or, respectively, the working loop (tsw) when a fault is detected in the data of the working loop (tsw) or, respectively, the protection loop (tsp).

5. Ring circuit in accordance with claim 4, characterized in that the first and the second input (lii; li2) of every working and protection card (W1, El; W2, E2; W3, E3; W4, E4) are each connected to the first changeover switch (s) through an input circuit that is provided for testing and processing of the data supplied by the working loop (tsw) or by the protection loop (tsp).

6. Ring circuit in accordance with claim 5, characterized in that the input circuits for Synchronous Digital Hierarchy networks include the modules of the regenerator section and if applicable the modules of the multiplex section.

7. Ring circuit in accordance with claim 6, characterized in that the output (lo) of every working and protection card (WI, W2, W3, W4, El, E2, E3, E4) is connected to an output circuit that corresponds to the structure of the input circuits.

8. Ring circuit in accordance with claim 6 or 7, characterized in that the changeover switch (s) supplies the received data to at least one switching matrix (LPC; 19 HPC, #), whence they are carried through the output circuit to the output (lo) or to the interface (tio; PPI), which connect the working or protection card (WI, W2, W3, W4, El, E2, E3, E4) to the second network.

9. Ring circuit in accordance with claim 8, characterized in that the switching matrix (LPC; HPC, #) is connected to the output circuit via a monitoring module (HPOM; LPOM).

10. Ring circuit in accordance with one of claims 1 - 9, characterized in that every network node (NKI; NK2; NK3; NK4) has at least one control unit (CTRL) which is provided to receive the error messages from the input circuits and from the monitoring module (HPOM; LPOM) as well as to carry out the necessary protection measures.