WO1995019082A1

WO1995019082A1 - Synchronous transmission systems

Info

Publication number: WO1995019082A1
Application number: PCT/GB1995/000014
Authority: WO
Inventors: Andrew James Barker
Original assignee: Gpt Limited
Priority date: 1994-01-05
Filing date: 1995-01-05
Publication date: 1995-07-13
Also published as: DE69532882T2; CN1141704A; CN1080505C; GB2285554B; GB9400057D0; ATE264588T1; GB9500141D0; AU1324295A; CA2180550A1; FI962752A; AU686718B2; JPH09507357A; GB2285554A; US5905731A; ZA9565B; FI962752A0; EP0738444B1; EP0738444A1; DE69532882D1

Abstract

In a digital telecommunication system having data packages carrying identification data, the identification data is characterised in that it is unique to the package within a station and does not depend on the identity of the originating station.

Description

SYNCHRONOUS TRANSMISSION SYSTEMS

The present invention relates to telecommunication systems and more particularly to

digital hierarchy systems (SDH). In such systems data is switched within an individual

station to the appropriate output port. Also alternate routes are provided between

stations by which data can be transmitted from an originating stations port to a receiver

stations port so that failure of one route, e.g. due to accidental damage, will not prevent the data from reaching the receiver stations port via the other route.

The data is sent in the form of two packages or so-called virtual containers which essentially are made up of two sections, one being the actual data which is transmitted, i.e. the so-called payload, and the other being data which is concerned with the integrity of the data, referred to hereinafter as overhead or ancillary data.

One of the requirements for a system of the kind outlined above is that there must be some means for checking whether the package is correctly switched (or routed) at each individual station along the overall path of the data. Essentially this is achieved by use of the ancillary or overhead data area at each station in order to include an identity, and other monitoring information, to ensure correct switching.

In our co-pending application no. 9301575 we disclose an arrangement where the package is identified in terms of the channel number of the data, and the receiving stations card number. However there is a potential problem with such an arrangement in situations where the package travels by the alternative route referred to earlier. In the event of failure of the first route the package will acquire a new channel and card number. This means that the destination port of the station will not recognise the package because it will be looking for a package which has an identification which is characteristic of the first route. This problem is a direct result of the use of the alternate route protection identified earlier.

The present invention is concerned with overcoming this problem.

According to the present invention in a digital telecommunications system of the kind referred to, data packages carry identification data which is unique to the package within a station, and does not depend on channel or card numbers.

In designing systems of the kind referred to there is always the conflicting requirement of on the one hand making maximum use of the transmission capacity to transmit payload data and on the other hand ensuring the correct transmission of that payload data. In other words in each data package a balance has to be struck between the ratio of payload capacity and overhead or ancillary capacity.

Further design constraints are the internationally agreed standards and protocols which

are necessary to ensure that the various national telecommunications systems can communicate with one another, e.g. the so-called CCITT standards and protocols.

Included in the ITU (formally CCITT) standards concerning the above mentioned overhead or ancillary data is one which calls for data whose purpose is to provide monitoring of the actual data, referred to as the V4 byte. This is only valid between stations, allowing it separate use within a station. In addition some data packages (or payloads) do not have such an overhead that may be used, in which case a part of the Section Overhead of the traffic is used in its place. For clarity only the V4 byte is

described below, but the same principle applies to the Section Overhead.

According to one aspect of the present invention overhead or ancillary data characteristic

of the identification of a data package is substituted for the so-called V4 byte to ensure the correct routing of the data package whilst at the same time not creating an increase in the overhead or ancillary data contained in the data package. In other words the inclusion of such identification data does not result in a requirement for increased bandwidth.

More specifically where an add-drop multiplexer (ADMUX) is employed to switch data from a data transmission channel into a so-called tributary channel the package identification data is substituted for the V4 byte as the package enters the ADMUX and is re-substituted as the package leaves the ADMUX.

The package identification data may have the characteristics referred to in our co- pending application number 9301575 or the characteristics referred to above in connection with the present invention.

How the invention may be carried out will now be described by way of example only and with reference to the accompanying drawings in which: Figure 1 is a diagrammatic representation of a typical transmission loop connected to a main transmission path; and

Figure 2 shows diagrammatically and in more detail one of the ADMUX shown in Figure 1.

A transmission loop 2 is connected to a main transmission path 1 through a switch in ADMUX 3. There is included in the transmission loop 2 ADMUX switches 4, 5 and 6 each of which serves tributaries 7, 8 and 9 respectively.

Data destined for tributary 9 can reach the ADMUX 6 either through the path 3,4 or through the path 3,5. Data is sent on both paths, with the terminating ADMUX (ADMUX 6) selecting which to use. This provides protection against failure of either path. When the packet of data reaches its destination at 9, there must be means to check that the correct packet of data has been switched down to it by the switch in ADMUX 6. The present invention is concerned with ensuring that this is the case in arrangements where the so-called path protection is provided, i.e. the data can reach its destination by one of at least two routes.

Referring to Figure 2, the ADMUX 6 (shown in dotted lines) consists essentially of four cards 10, 11, 12 and 13. Card 11 is the switch itself. Cards 10 and 12 are so-called synchronous transport modules (STM) which can be designed to multiplex at different levels e.g. 1, 4 or 16 which equate to bit rates of 155.52 Mbit/s, 622.08 Mbit/s and 2488.32 Mbit/s respectively would thus be designated STM1, STM4 or STM16 tributary card.

Data packages entering the ADMUX 6 at A and B will have inserted into the ancillary or overhead data in place of the V4 byte, data which uniquely identifies that data

package within that ADMUX. This contrasts with the arrangement disclosed in our co-

pending application no. 9301575 where the data package is identified in terms of its

entry point into the ADMUX, for example by means of its channel number and card

number. This would result in the packages at A and B having different identifications.

This identity is distinct from the other data which forms the overhead and which is designed to enable various checks to be made in order to monitor the integrity of the data being transmitted.

Having described the inventive concept illustrated in broad terms in relation to the accompanying drawings there will now be given a more detailed description of the embodiment of the present invention.

It is necessary to ensure that switch protection takes place automatically a failure condition at a destination port is detected or alternatively on demand from the ADMUX controller.

There are a number of conditions which can be used to initiate switching and these are: - mismatch on path identification number - failure of parity checks

- failure of alignment

- loss-of system-clock transitions

The first two- of the above involve the monitoring of diagnostic messages inserted by

sender ports which are compared with data which is downloaded from a Mux Controller as a 'comparison message' or which is calculated internally (parity errors). For TU-1 and TU-2 SDH signals, where the diagnostic messages are contained in the V4 bytes and are completely updated only every 16 frames (4 V4 bytes), the comparison is carried out every four frames on a partial V4 message. For TU-3, AU-3 and AU-4 signals, the diagnostic message is fully updated every frame.

According to the invention a path is first set up, and a controller is arranged to generate a 16 bit number which uniquely identifies that path, this number is inserted at each sender port into the diagnostic message for that channel, a separate number being used for each unidirectional path (i.e. a bidirectional path utilises 2 numbers). It will be appreciated that the 16 bit number generated, allows up to 65536 unique path numbers the ADMX4 being capable of having up to 1008 bi-directional or 2016 unidirectional cross connections.

At each destination port, the path identity number of each channel, inserted by the sender port, is arranged to be compared with the expected address (downloaded from the Mux Controller configuration-data). The address check is arranged to be carried out on traffic from both Switch units simultaneously, to detect failures of either switch. This method of failure detection is

preferably subject to a persistence check and protection switching is in accordance with

established priorities.

In carrying the invention to effect parity checks are carried-out on the data of each channel. Parity bits are calculated at the sender port and inserted into the diagnostic

messages such that even parity is maintained in each channel. At the destination port, the parity of each channel is again calculated and any parity errors generate an alarm signal. As with the source address checks, the parity checks are carried out on traffic from the working switch and the standby switch simultaneously. This method of failure detection is subject to a persistence check and protection switching is in accordance with the allocated priorities.

For TU- 1 or TU-2 signals, V4 bytes occur every fourth frame (i.e. every multiframe) and a complete V4 diagnostic message requires four such bytes. Each parity bit for a channel is relevant to all bits from that channel within the previous multiframe (four frames), including the V4 byte. The four parity bits of the V4 message are therefore completely independent of each other and are related to the data of the previous four frames only.

For TU-3, AU-3 and AU-4 signals only eight parity bits are contained within the diagnostic message within the Signal path overhead (SOH). Since a complete diagnostic message is transmitted every frame, these are used to indicate the parity of the data within the previous frame. Each peripheral card incorporates transition-loss detectors which the internal clock signals from each switch. A failure of the (38.88 MHz) clock shall be defined as a loss- of clock transitions for a period exceeding a nominal threshold of 125-500nS. There need not be persistency check on clock failures.

A check is necessary to identify any alignment failure. If the alignment bits contained within the diagnostic messages are incorrect, the detecting destination port is arranged to generate a loss-of-frame alarm. This check is subject to a persistence check as described later.

A request to switch is arranged to be issued to all destination ports on receipt of a switch- failure indication by the Mux. controller. A manual request may also be issued when demanded by an operator via the NMI. These requests are treated as switch-failure conditions and acted upon in accordance with prescribed routines. No persistence check is carried out on such conditions.

In cases of diagnostic message failure, a persistence check is performed to prevent spurious switching during transient conditions (e.g. MSP switching). This check takes the form of a threshold on the number of permissible consecutive mismatches of the diagnostic messages, with the messages being partially updated and compared every multiframe. A hardware persistence check is therefore applied to the channel identification number and the parity check independently. This persistence is programmable from between 1 and 256 successive failures of the check. The enabling and disabling of protection is carried out according to the following

cπtena: a. Unconnected - no checks performed b. Connected - all checks performed c. - Protected - all checks performed

The V4-bvte diagnostic message format is as shown in the following table.

MSB LSB

Parity Alignment bits bits.

V4 byte 1 PI 0 Al 0 D3 D2 Dl DO

V4 byte 2 P2 0 A2 0 D7 D6 D5 D4

V4byte 3 P3 0 A3 0 Dll D10 D9 D8

V4 byte 4 P4 0 A4 0 D15 D14 D13 D12

These bits are defined as follows :- Data Parity

PI P2 P3 P4

Alignment Bits

Al A2 A3 A4

Path Identification Number

DO to D15

The various Section Overhead diagnostic message configurations are treated as follows: TU-3 and AU-3 The diagnostic message is contained in the first three columns of the first three rows of the SOH, as follows. Section Overhead Bytes

Row Column

•

1 2 3

1 AMI AM2 AM3

2 BM1 BM2 BM3

3 CM1 CM2 CM3

Where the bytes shown have the format:

MSB LSB

AMN = PI P2 P3 P4 P5 P6 P7 P8

BMN = D7 D6 D5 D4 D3 D2 Dl DO

CMN = D15 D14 D13 D12 Dl l D10 D9 D8

(where N = 1 to 3)

AU-4

The diagnostic message is contained in the first column of the first three rows of the SOH.

The three bytes 'A', 'B' and 'C have the same format as bytes 'AMN', 'BMN' and 'CMN' for TU-3 and AU-3 signals.

Comparison messages sent from the mux. controller and stored in an ASIC at the destination ports have the following format:

MSB LSB

Byte l = D7 D6 D5 D4 D3 D2 Dl DO Byte 2 = • D15 D14 D13 D12 Dll D10 D9 D8

The system described above gives an effective protection arrangement for ADMX switch plane protection without any need for disabling a card when protection is applied to a traffic path.

Claims

1. In a digital telecommunication system having data packages carrying identification data, the identification data is characterised in that it is unique to the package within a station and does not depend on the identity of the originating station.

2. A system as claimed in Claim 1 which comprises:

a) means to set up a uni-directional route or path through the switch;

b) a controller to generate an identification number which uniquely identifies that route or path;

c) means to insert that identification number at each sender port into a diagnostic message for that channel ; and

d) means at each of a plurality of destination ports to compare the actual identification number received with that expected.

3. A system as claimed in Claim 2 in which the unique identification number is applied to the data package as it enters an ADMUX, in place of the V4 byte normally carried by the data package, and is replaced by the V4 byte as the data package leaves the ADMUX.

4. A system substantially as hereinbefore described with reference to and as shown in the accompanying drawings.