The information stored in the management section of data blocks in a certain transmission section of a telecommunication network is used to check the quality of the data transmission in this transmission section. The management section of each data block contains parity bits for checking the data transmission quality in a transmission path comprising the transmission section.

The parity bit values are set depending on the data of the preceding data block including its parity bits and the bits used for information. The source and sink of the transmission section are limited and suitable to change the bits used for the tandem information and the parity bits.

The invention also concerns a source and a sink of a tandem connection in a telecommunication network, especially a SDH network in which parity bits in the management sections of data blocks are provided to check the quality of the data transmission in the transmission paths of the data block. The source to check the quality of the data transmission in a transmission section that forms a part of the transmission path has means to insert additional information for the tandem connection in suitable bits in the respective management section of the data blocks, and has means to compensate this change (and potential changes in the parity bits in the management section of the same data block) in the parity bits in the- management section of the following data blocks. The sink itself has means to reset bits in the respective management section of the data blocks in which additional information was stored for the tandem connection to check the quality of the data transmission in a transmission section that forms a part of a transmission path. It also has means to compensate this change and potential changes in the parity bits in the management section of the same data block in the parity bits in the management section of the following data block.

In telecommunication networks, the quality of the transmission of data signals is usually checked using additional transmitted values. For example, according to the SDH standard, virtual containers stored in the transport frames are used as data blocks for data transmission in which parity bits are also transmitted in additional to the payload. These parity bits can be used to determine at a monitoring point whether a bit has been put into the prior container along the prior transmission path.

As an illustration, the path U1 of a virtual container from the network of a first network operator B to the network of a second network operator C via the network of a third network operator A is represented in Fig. 1.

At an end point E1 in the network of the network operator B, the virtual container consists of the useful load and a management section (POH = path overhead). At an endpoint E2 in the network of the network operator C, the virtual container is divided into the individual data streams. The information in the management section of the virtual container is required to ensure the transport of the virtual container between the endpoints. Among other- things, parity bits in the management section, especially a parity byte (B3 in the case of VC4 or VC3) or two parity bytes (2 MSBs in the V5 byte in the case of VC12 or VC20), are set. For each of the used parity bits, a XOR link of predetermined data bits of a preceding virtual container is established, and the result is stored in the parity bit. In this manner, it can be determined whether a data error has arisen before an end point E2 in the network of the network operator C or before an observation point in the transmission section U1 by renewed XOR linkage of the data bit of the prior virtual container that was used to determine the parity bit, and by a subsequent comparison with the parity bit.

Network operator A can also evaluate the used parity bits in the management section of the virtual container as described. If an error is determined, the operator only knows that the error arose before the monitoring site and not whether it occurred in its own network or beforehand.

A network operator whose network is used for a part of the transmission path U1 is frequently interested in specially monitoring the quality of the sections U2 from a transmission path U1 in his own network.

So-called tandem connection monitoring is used additionally for this purpose. Tandem connections are known in practice and are specified in the ETSI standard ETS 300417-4-1 for VC12, VC2, VC3 and VC4. Certain bits are used in the management section of a virtual container for this procedure just to transmit information in a transmission section U2 to be evaluated. For such bits, the N1 bytes in the management section of the virtual containers of SDH data flows are suitable for VC3 and VC4, and the N2 bytes are suitable for VC12 and VC2.

The tandem connection begins at a source Q in which information for the tandem connection monitoring is entered into the bits for monitoring. The tandem connection ends with a sink S that terminates the bits used for information transmission (preferable at zero).

The modification of the management section of the virtual container by a tandem connection should not influence the quality information via the overall path of the virtual container between endpoints E1, E2. In particular, the change of the used bits should not cause the information transmitted in the parity bits to become useless.

Since the parity bits are calculated via a complete prior virtual container including the bits in the management section of the container, the data in the transmission section U2 with activated tandem monitoring when bits in the source Q are changed no longer correspond to the parity byte of the following virtual container. If the source Q and the sink S of the tandem connection are also not turned on/off for the identical virtual container, non-corresponding data arise outside of the transmission section U2 if no further measures are taken when the tandem connection is activated and deactivated.

It is therefore necessary for the parity values to be compensated at the source Q and the sink S after activating a tandem connection when bits in the prior virtual container are modified by storing information in them or removing information from them for a quality check in the transmission section U2. Compensation ensures that a source/sink of a tandem connection is transparent for existing data errors that may have arisen- beforehand in the transmission path. For the next virtual container, the changed parity bits must also be taken into consideration in the compensation.

This compensation is successful until the tandem connection is interrupted and the compensation is stopped. Even when the last changes in the bits used for the information for the tandem connection are compensated in the parity bits of the next virtual container, it is probable that disagreement will arise in the uncompensated parity bits of the following virtual container since the values of the parity bits of a virtual container are continuously used for calculating the parity bits of the next virtual container. Each time a tandem connection is deactivated, a virtual container can arise whose parity bits do not correspond to the data of the prior virtual container and hence show nonexistent bit errors at a monitoring site.

Since the network operator at which the transmission path Ul of a virtual container terminates has no information on the tandem monitoring in the transmission path in most cases, he will interpret such detected errors as actual data errors.

If many segments of the transmission path U1 of a virtual container are monitored and if the monitoring functions are frequently turned on an off, this always leads to the detection of supposed errors at down-stream observation points or at the termination site of the transmission path of the virtual container without an error actually existing.

The invention is based on the problem of creating a procedure that allows the source or sink of a tandem connection to be turned on and off so that no supposed errors are displayed in the parity bits of the data blocks. Likewise, a source and a sink of a tandem connection in a transmission path should be created in a telecommunication network that allows the process to be used.

This problem is solved according to the invention by a process to activate and deactivate the source and sink of a tandem connection. The information stored in the management section of data blocks in a certain transmission section of a telecommunication network is used to test the quality of the transmission of the data in this transmission section. The management section of each data block also contains parity bits for checking the data transmission quality in a transmission path that comprises the transmission section. The parity bit values are set depending on the data of the prior data block including its parity bits and its bits used for the information. The source and sink delimit the transmission section and are suitable for changing the bits used for the tandem information and the parity bits. The process comprises the following steps: a) Activation of the source and sink of the tandem connection, b) Setting the bits provided for the tandem information of the arriving data block corresponding to the tandem information to be transmitted in the source, or resetting the bits in the sink, c) Compensating changes of bits for the tandem information and/or parity bits in the data block in the parity bits of the following data block, d) Repetition of steps b) and c) as long as the monitoring is to be carried out, e) Signaling that the tandem monitoring should be deactivated, and f) Modifying the bits provided for the tandem information, or the bits of another unused byte of the current data block so that the change in the parity bits of the same data block is compensated (based on the adaptation to the values of the prior data block) in regard to the calculation of the parity bits in the following data block.

In addition, the problem is solved with a source according to the preamble of the claim 5 or with a sink according to the preamble of claim 6 in that the source or sink has additional means to compensate the change of parity bits in the management section of a data block in the bits that are used for the tandem connection for transmitting additional information, or in bits of another unused byte in regard to the calculation of the parity bits of the following data block after the tandem connection was ended.

The advantage of the process according to the invention and the sources and sinks of a tandem connection according to the invention is that, when a tandem connection is deactivated, compensation occurs that prevents supposed errors from being displayed.

Normally, as described above, the change in the bits used for tandem information is compensated by adapting the parity bits in the following data block as long as the tandem connection is activated. Each time the parity bits are adapted, any adaptation in the parity bits of prior containers is also taken into account. This compensation is also provided in the procedure according to the invention.

In addition when the tandem connection is deactivated, the last change of the parity bits is compensated in the same data block by adapting the tandem bits that are no longer necessary for transmitting information in the tandem connection. A compensation can also be carried out using bits of another unused byte in the data block. In this manner, the parity bits in the following data block agree with all the data of the data block even if the values in the parity bits of the prior data block are changed by the last compensation (assuming there are no actual transmission errors).

Preferred embodiments of the process according to the invention and the source and sink according to the invention of a tandem connection are found in the subclaims.

The process according to the invention will be further explained in the following using an exemplary embodiment that refers to figures.

Fig. 1: A tandem connection in a section of a data transmission path in a telecommunication network, Fig. 2: The change of a parity byte (B3) and a byte (N1) used for a tandem connection in VC4s at the source of a tandem connection after the start of the connection, Fig. 3: The change of a parity byte (B3) and a byte (N1) used for a tandem connection in VC4s at the source of a tandem connection after ending the connection, Fig. 4: The change of a parity byte (B3) and a byte (N1) used for a tandem connection in VC4s at the sink of a tandem connection after the start of the connection, Fig. 5: The change of a parity byte (B3) and a byte (N1) used for a tandem connection in VC4s at the sink of a tandem connection after ending the connection.

The data of VC4s were assumed to be 00hex up to the parity byte (B3) and the tandem byte (N1) for reasons of simplification.

Fig. 1 has already been described in reference to the state of the art.

Fig. 2 illustrates the procedure at source Q of this tandem connection for the transition from an inactive to an active tandem connection in a transmission section U2 from Fig. 1.

On the left in Fig. 2 are portrayed sequential virtual containers VC#1 IN-VC4#4 IN at the source. Each of the virtual containers VC#1 IN-VC4#4 IN has a parity byte B3 in its management section and a byte N1, designated "tandem byte"in the following, that is used in the transmission section U2 for transmitting the information for the tandem connection.

On the right side of Fig. 2 are the virtual containers VC#1 OUT-VC4#4 OUT leaving the source of the tandem connection. To allow differentiation, the associated parity bytes are identifie as B3'and the associated tandem bytes as N1'.

While the first virtual container VC4#1 passes through the source, the tandem connection is not activated in the transmission section. Neither the parity byte B3 nor the tandem byte N1 of the first virtual container VC4#1 are changed in the source.

The tandem connection is activated between the arrival of tandem byte N1 of the first virtual container VC4#1 IN and the tandem byte N1 of the second virtual container VC#2 IN.

The parity byte B3 of the second virtual container VC4#2 that contains a checksum for the prior virtual container VC$#1 is not changed at the source since the data of the virtual container VC4#1 remained unchanged and hence agree with the arriving parity byte B3. The tandem byte N1 of the virtual container VC4#2 whose bits were all set to"0"contains at the source a value corresponding to the information to be transmitted for the tandem connection. A value of N1'= 10010100"was determined for the information to be transmitted and set 1).

In the directly following third virtual container VC4#3 that passes through the source, the parity byte B3 is adapted since the values of the tandem byte N1'of the prior virtual container VC4#2 OUT are included in it.

Below is a calculation of the new parity byte B3'using XOR links according to the equation from ETSI standard ETS 300417-4-1: B3'(t) = B3 (t-1) = B3'(t-1) Nl (t-1) Nl'(t-1)(t) = B3 (t-1) = B3'(t-1) Nl (t-1) Nl'(t-1) B3 (t) (1) With X (t): bytes of the current VC4 IN frame X' (t): byte of the current VC4 OUT frame X (t-1): byte of the prior VC4 IN frame X'(t-1) (t-1) byte of the prior VC4 OUT frame This yields a value of"10110000"that is assigned to the outgoing B3'byte of the third virtual container VC4#3 OUT 2). The parity bytes B3, B3'of the prior arriving or outgoing or virtual container VC4#2 IN and VC4#2 OUT in equation (1) do not influence the result due to their identity.

The tandem byte N1'of the outgoing virtual container VC4#3 OUT is also assigned a value corresponding to the information 3) to be transmitted for the tandem connection;"01100000"in this case.

The calculation of the parity byte B3'of the following virtual container VC4#4 OUT is analogous to that of the parity byte B3'of the third virtual container VC4#3 OUT.

In equation 1, the differences between the incoming and outgoing parity bytes B3, B3'of the prior virtual container VC4#3 are relevant. The value of"11010000" results for the parity byte B3'of the outgoing fourth virtual container VC4#4 OUT 4).

The described procedure of changing the tandem byte of a virtual container and compensating this change and the change of the parity byte of the same virtual container in the parity byte of the following virtual container is continued for all virtual containers arriving at the source 5) until the tandem connection is to be interrupted.

The process at the source Q in Fig. 1 when the tandem connection ends is shown in Fig. 3. As in Fig. 2, the virtual containers VC4#1 IN-VC4#4 IN arriving at the source are shown on the left side, the virtual containers VC4#1 OUT-VC4#4 OUT leaving the source are shown on the right side.

The tandem connection is still active for the first virtual container VC4#1 passing through the source, and the procedure carried out for the third and fourth virtual container in Fig. 2 is applied 1), 2).

After the tandem byte Nl, Nl' of the first virtual container VC4#1 passes and before the tandem byte N1 of the second virtual container VC4#2 arrives, there is a report that the tandem connection is to be interrupted.

In the parity bytes B3, B3' of the second virtual container VC4#2, a parity compensation is carried out analog to the parity compensation in Fig. 2 independent of the time of the report 3) since there must be an adaptation to the values in the outgoing first virtual container VC4#1 OUT so that a supposed error will not be detected when evaluating the parity bit B3'in the transmission section after the tandem source.

As a result of the report that the tandem connection is to be interrupted, the tandem byte N1'of the second virtual container VC4#2 OUT is no longer assigned an information value. Instead, the value of the tandem byte N1 of the second virtual container VC4#2 is adapted so that the parity compensation in the parity byte B3'of the same virtual container VC4#2 is compensated 4).

The value for the tandem byte N1'of the second virtual container VC4#2 OUT is determined according to the equation: Nl'(t)(t) (t) = B3 (t) B3' (t) Nl (t), With X (t): byte of the current VC4 IN frame X' (t): byte of the current VC4 OUT frame Which produces a tandem byte N1'of Nl' (t) =,"00100100" "10100100""00000000"="10000000".

For the third virtual container VC4#3, the result of this compensation in the tandem byte N1'of the second virtual container VC4#2 OUT is that a parity compensation in the parity byte B3 can end 5) since the value that would result when the compensation for B3'is calculated is the same as the existing value of the arriving parity byte B3 if no bit errors have arisen in the transmission section beforehand in the VC4#2.

All subsequently arriving tandem bytes N1 and parity bytes B3 can now pass through the source 6), 7), 8) without a change, and the deactivation of the tandem connection at the source is terminated.

According to the process described with reference to Fig.

2 and 3 for activating and deactivating the source of a tandem connection, there cannot be a value that would display a supposed error when comparing the data of the prior virtual container in any of the priority bytes B3' of the virtual containers leaving the source.

The procedure at the sink S of the tandem connection at the end of the transmission section U2 in Fig. 1 corresponds in principle to the procedure at the source Q and will be described with reference to Fig. 4 and 5. In- both figures, the structure with the incoming and outgoing virtual containers corresponds to that in Fig.

2-4.

Fig. 4 shows that in the sink of the tandem connection the tandem bytes N1 are terminated at zero starting at the activation of the sink that is between the arrival of the tandem byte N1 of the first virtual container VC4#1 IN and the arrival of the tandem byte N1 of the second virtual container VC4#2 IN. The first tandem byte N1'set to zero is hence that of the second virtual container VC4#2 1).

In the next arriving parity byte B3, that of the third virtual container VC#3, the data change is compensated in the second virtual container VC4#2 OUT corresponding to the above equation (1) 2).

As long as the sink is not deactivated again, the tandem byte N1 of the arriving virtual container is alternately terminated at zero 3), 5) and the changes in the tandem byte N1 and parity byte B3 I are compensated in the parity byte B3 of the following virtual container 4).

The procedure to deactivate the sink S in Fig. 2 is illustrated in Fig. 5.

The sink is still activated while the first virtual container VC4#1 passes through the sink. Hence, analogous to the description with reference to Fig. 4, the parity byte B3 in the first virtual container VC4#1 is compensated corresponding to the changes in the prior virtual container 1), and the tandem byte N1 is terminated at zero 2).

The report that the sink should be deactivated occurs between the arrival of the tandem byte N1 of the first virtual container VC4#1 IN and of the tandem byte N1 of the second virtual container VC4#2 IN.

The parity byte B3 of the second virtual container VC4#2 is first compensated corresponding to the changes of the parity bytes B3, B3' and the tandem bytes N1, N1' in the first virtual container VC4#1 3).

The tandem byte N1 of the second virtual container VC4#2 is no longer terminated at zero. Instead, the compensation in the parity byte of the second virtual container VC4#2 from B3'"10110000"to B3'="10000000"is compensated by the tandem byte N1'leaving the sink 4).

The calculation of the resulting tandem byte N1'of the second virtual container VC4#2 OUT is again calculated based on the incoming and outgoing parity bytes G3, B3' and the incoming tandem byte N1 of the second virtual container VC4#2 according to the above equation (2), and the following is obtained: Nl' (t) =,"10110000","10000000","01100000"-"01010000".

In calculating the parity value using the data of the second virtual container VC4#2 OUT leaving the sink, the same value results as from calculating the parity value via the second virtual container VC4#2 IN arriving at the sink. The parity byte B3 of the next arriving virtual container VC4#3 IN hence does not have to be compensated to prevent a supposed error from being detected at a later site 5).

All following tandem bytes and parity bytes can pass through the source without being changed 6), 7). The tandem bytes are not terminated at zero since otherwise a- change of the parity byte in the following virtual containers would be necessary, or a parity byte that displays a supposed error would be passed on.