EP0850517A1

EP0850517A1 - Network element and input-output unit for a synchronous transmission system

Info

Publication number: EP0850517A1
Application number: EP97934546A
Authority: EP
Inventors: David Scheer; Jürgen Kasper; Werner Beisel; Wolfgang Krips
Original assignee: Alcatel SA; Alcatel Alsthom Compagnie Generale dElectricite
Current assignee: Alcatel Lucent SAS
Priority date: 1996-07-10
Filing date: 1997-07-10
Publication date: 1998-07-01
Also published as: WO1998001971A1; DE19627728A1; US6208648B1

Abstract

The invention concerns a network element and an input-output unit (1) for a synchronous transmission system according to the standard for Synchronous Digital Hierarchy. The input-output unit (1) contains a signal processing device (2) that has a reception data memory (5), a transmission data memory (7), a control unit (6) and a processor (3). The processor (3) is a digital signal processor which essentially carries out the processing of a STM-N signal.

Description

Network element and input / output unit for a synchronous transmission system

The invention relates to a synchronous transmission system according to the preamble of claim 1. The invention also relates to a network element according to the preamble of claim 6 and an input / output unit according to the preamble of claim 7.

A synchronous transmission system is e.g. B. a transmission system for the synchronous digital hierarchy (SDH system). In an SDH system, signals to be transmitted are combined according to a certain pattern and structured according to a frame. Such a frame is referred to as a synchronous transport module STM-N; he is e.g. B. in the ITU-T recommendation "Recomendation G.707 (Draft) (11/95)", e.g. B. Chapter 7 "Multiplexing method". Within the scope is an area for tax data, i. H. for "Section Overhead SOH" and "AU-n pointer", and for user data, d. H. for "payload".

The SDH system is made up of a number of network nodes that are connected to one another by physical transmission media (e.g. optical fibers, coaxial cables). The network nodes are usually made up of groups of individual network elements (e.g. add / drop multiplexer, cross-connect) for which various functions are defined. The CCITT recommendation "Recommendation G.783", Chapter 2 "Transport Terminal Functions", defines the network elements according to elementary functions, which include interface, monitoring and connection functions. The interface function to the physical transmission medium is provided by an interface device. An interface device (SPI, SDH Physical Interface) is used in the receiving direction for clock recovery from the received signal and for detecting a signal loss (LOS, loss of signal); it supplies the signals LOS, DATA and TIMING (see G.783, Figure 2.2). In the sending direction, the

Interface device u. a. the task of sending out a signal to be transmitted with the system clock.

The interface devices are usually implemented by a combination of optical transmitter and receiver modules and standard components (eg TDC2302C from Texas Instruments) or ASICs. Such a standard component has u. a. following functions: It sends and receives STM-1 signals with a bit repetition rate of 155.52 Mbit / s. It recognizes the frame of the incoming signal and sends a frame indication signal. It also provides markings for the states of loss of signal (LOS) and loss of frame (LOF).

Subsequently, the received STM-1 signal is further processed in a signal processing device, which is also a standard component (eg TDC3003 from Texas Instrument) or an ASIC. This standard component has the following functions, among others: It is responsible for the entire processing of the overhead. Depending on an external clock, it generates pointers for received and to be transmitted signals and executes pointer actions. It also carries out monitoring functions for the Bl, B2 and B3 coding, the error display (Far End Block Error, FEBE) and the counting of pointer actions. The interface device and the

Signal processing device form an input / output unit, which connects to other components such. B. a switching matrix in a cross-connect, the network element.

Increasing demands on the network elements with regard to complexity and integration density lead to ever more complex circuits and to ASIC developments with increasing number of gates. This requires complex simulations and tests that result in ever longer simulation and test times.

The invention is based on the object of specifying a synchronous transmission system and a network element therefor, in which the increasing requirements are dealt with in a simple manner. A synchronous transmission system is the subject of claim 1 and a network element is the subject of claim 6. An input / output unit for a network element is the subject of claim 7.

An advantage of the invention is that the network elements can be flexibly adapted to evolving (ITU-T / ETSI) standards without extensive circuit changes; necessary adjustments can be made quickly. There is no need for complex circuit developments.

The invention is explained below by way of example with reference to drawings. Show it:

1 is a block diagram of an input / output unit for a network element,

2 is an illustration for explaining group formation,

Fig. 3 is an illustration for explaining a transfer of memory contents. 1 shows an exemplary block diagram of an input / output unit 1 for a network element with the components required for understanding the invention. The network element is part of a transmission system for the synchronous digital hierarchy, an SDH system. Several network elements can be combined at a network node which is connected to further network nodes by one or more physical transmission media. Network elements are e.g. B. Cross-Connect, Add / Drop Multiplexer and line systems. The input / output unit 1 according to the invention is described using a cross-connect for an SDH system which has a switching matrix connected to the input / output unit 1.

The input / output unit 1 has an interface device 4 and a signal processing device 2, which has a received data memory 5, a transmitted data memory 7, a control unit 6 and a processor 3. In Fig. 1 is also an interface 10, which establishes a connection to the switching matrix, a program memory 8 and a central control circuit 9, which are connected to the input / output unit 1, but u. U. can be arranged spatially away from it in the network element. The program memory 8 is also used to store data and to forward messages.

At the interface device 4, an STM-N signal enters and exits the network element; The STM-N signal is sent and received serially, i. H. the individual bytes of a frame are received and sent serially. The

Interface device 4 has the function of the SDH Physical Interface SPI known from "Recommendation G.783"; For the further description, the interface device 4 is therefore referred to as SPI. The transmit data memory 5 and the receive data memory 7 are random access memories (RAM), which are also referred to as read-write memories. The program memory 8 is also a read-write memory, for. B. a dynamically programmable read-write memory (DPRAM). The structure of the input / output unit 1 is explained below, and then its function is explained. The SPI 4 has a data output 11, an output 12 for a clock derived from the received STM-N signal, an output 13 for a frame identification signal (AI byte), a data input 15 for data to be sent and an input 16 for a fixed in the network element System clock. The data output 11 is connected to a data bus 26 (8-bit parallel bus), which supplies user data (payload) and control data (SOH) to a data input 17 of the received data memory 5 bytes in series. The outputs 12 and 13 are connected to inputs of the control unit 6. The system clock can be fed to a further input 14 of the control unit 6. The control unit 6 is connected by a control bus to an input 18 of the received data memory 5 and by a control bus to an input 19 of the transmitted data memory 7; Memory addresses generated by the control unit 6 can be fed via these inputs 18, 19: a write address is fed to the input 18 and a read address is fed to the input 19. The term write pointer is also used for the write address and the term read pointer for the read address.

The received data memory 5 has a data output 20 to which a data bus 23 is connected. A data connection 21 of the processor 3 and a data input 22 of the transmission data memory 7 are also connected to the data bus 23. The data connection 21 of the processor 3 is a data input and output, so that bidirectional data transmission is possible. An address bus 24 is connected to the receive data memory 5, the transmit data memory 7 and the processor 3; Via this address bus 24, the receive data memory 5 and the send data memory 7 are addressed by the processor 3. The data bus 23 and the address bus 24 are also connected to the interface 10, which, as already mentioned, establishes the connection to the switching matrix. For example, the data bus 23 is 64 bits wide (8 bytes) and the address bus 24 is 32 bits wide (4 bytes). The control unit 6 is connected to an interrupt connection 25 of the processor 3, whereby a synchronization of the processor 3 and the control unit 6 is possible. The processor 3 is preferably a digital signal processor DSP, e.g. B. a TMS320C80 from Texas Instruments, details can be found in the product description. A general description of the function and programming of digital signal processors is e.g. B. from M. Kappelan et al, "Digital Signal Processors", Funkschau 16/1993 (part 1, pages 66 - 69), Funkschau 17/1993 (part 2, pages 66 - 69) and Funkschau 18/1993 (part 3, Pages 136-141).

For the following description, an STM-1 multiplex signal with a VC-4 payload (3 TUG3) and the DSP TMS320C80 and its structure are assumed. In principle, however, another programmable microprocessor or DSP can also be used.

The DSP TMS320C80 has a central processor (master processor), four parallel processors and a transfer processor (transfer controller), which is responsible for data transfer between external and internal memories.

The function of the input / output unit 1 in the receiving direction is described below with the aid of FIGS. 2 and 3. The SPI 4 receives the frames of an STM-1 signal, which has a frequency of 155.52 MHz, and performs the functions already mentioned: e.g. B. clock derivation, and frame detection (AI, A2 bytes). The SPI 4 feeds data, ie the overhead and payload bytes of the STM-1 signal, to the data input 17 of the received data memory 5. In the received data memory 5, the overhead and payload bytes are logically organized so that four groups are formed, each containing the same type of data: three groups for the three tributary units TU and one group for the overhead OH (81 bytes). This group formation is shown schematically in FIG. 2 using an exemplary data stream. From left to right they are individual bytes arranged as follows: TU3 # 1, TU3 # 2, TU3 # 3, 9 OH, TU3 # 1, TU3 # 2, TU3 # 3, etc. After grouping, there is a group for TU3 # 1, a group for TU3 # 2, a group for TU3 # 3 and a group for OH. Each group can be processed individually without the need for data from another group. As a result, the signal processing task is divided into four sub-tasks and the processor 3 can access similar data in a coherent manner. The organization of the overhead and payload bytes transmitted via the data bus 26 into the four groups is controlled by the control unit 6, which generates the memory addresses. For this purpose, a counter is present in the control unit 6, which has a range from 0 to 2429 and is automatically reset.

Access to the data stored in the receive data memory 5 is controlled by the transfer processor present in the processor 3, which accesses the data according to a transfer list which is transferred to it by the central processor also present in the processor 3. The transfer list specifies the address of an internal processor memory to which the content of an address of the received data memory 5 is to be transferred; the transfer list thus contains all the information that is necessary to transfer data from a source to a sink. The transfer list is continuous, e.g. B. recalculated for each frame.

The central processor is responsible for the actual signal processing of the received STM-1 signal, it monitors the processing of the received data by the four parallel processors. The signal processing takes place in accordance with a program which is fed to the processor 3 from the program memory 8. With the aid of the program, all the procedures that are necessary to process the received data and then to feed it to the interface 10 can be carried out. About these procedures include z. B. the entire processing of the overhead, the generation of pointers, function monitoring of the Bl, B2 and B3 coding and the counting of pointer actions.

The (external) clock that an incoming STM-1 signal has can be different from the (internal) system clock, i. H. the incoming STM-1 signal and an STM-1 signal to be sent are not synchronized. The incoming STM-1 signal is processed with the system clock, which means that the address (write pointer), under which a data byte (payload or overhead) is to be saved, from the address (read pointer), from which a data byte is read is different. It follows that the data bytes stored in the receive data memory 5 belong to two different frames of the received STM-1 signal; the AI byte is therefore not the "oldest" byte stored in the receive data memory 5. Since the data must be processed in accordance with the order of the incoming STM-1 signal, the transfer processor reads out the data from the received data memory 5 for each group (FIG. 2) in two blocks and rearranges them in such a way that the data is in the internal memory of processor 3 are in the correct order. When all the required data is in the internal memory of the processor 3, the mentioned signal processing procedures can be applied to the data of the four blocks.

In Fig. 3, the temporal rearrangement described is shown schematically. Fig. 3a. shows the linear arrangement of the TU3 # 1 bytes in the receive data memory 5, in which the write pointer points to the beginning of the fourth byte. To the right of the write pointer is the first block, which has 779 bytes. The second block drawn to the left of the write pointer has 3 bytes. The bytes of the first block belong to a frame that was stored first and the bytes of the second block belong to a subsequent frame. The situation that arises after the time re-sorting by the transfer processor is in Fig. 3b shown. In this likewise linear arrangement, the 779 bytes of the first block are arranged on the left and then the 3 bytes of the second block on the right.

The transfer processor also performs a clock adaptation (multiplexer section adaptation) of the incoming STM-1 signal to the system clock. For each frame, it calculates the address (write pointer) under which the data is to be stored in the internal memory and specifies the size of the blocks. The size of the blocks can change by +/- 3 bytes, depending on the direction in which the write pointer is moving. If a change in the block size is required, such a need for change is included in the continuous calculation of the transfer list. The next time the frame arrives, the change in block size is taken into account.

The fill level of the internal memory of the processor 3 is queried after each frame has arrived; if the fill level has exceeded an upper limit value or fallen below a lower limit value, a pointer action is initiated in the corresponding tributary unit TU. A pointer action is used to read more or less from the internal memory of the processor 3 in a known manner. If the columns are rearranged after such a pointer action, an outgoing frame is formed which consists of four separate data groups in four separate memory areas. This frame is fed to the transmission data memory 7. The control unit 6, the transmission data memory 7 and the SPI 4 then form an STM-1 signal to be transmitted; the STM-1 signal to be transmitted is formed in the reverse order to the signal processing in the receiving direction.

The description of the invention was based on an STM-1 signal, but the invention is not restricted to this. In the event that an STM-N signal is to be processed, the number of blocks must be increased by a factor of N. From the preceding description it can be seen that the use of the processor 3 for processing an STM-1 signal creates a flexible processing concept. The processor 3 present in the signal processing device 4 can be supplied with a changed program by the program memory 8, whereby changes in e.g. . B. in the ETSI or ITU standard without extensive circuit changes.

Claims

MPatents

1. Synchronous transmission system for digital signals combined into a multiplex signal, in which network elements are present which are connected to one another by one or more transmission media, in which each network element has an input / output unit (1) for receiving and transmitting the multiplex signal, which has an interface device (4) and a signal processing device (2) which processes data coming from the interface device (4), characterized in that the signal processing device (2) has means (5, 6, 7) for storing the data and a programmable processor (3) that the means (5, 6, 7) for storing arrange the data according to a fixed scheme so that the processor (3) can access and process them, and that the means (5, 6, 7) for storing the processor (3) incoming data of the interface device (2) can supply.

2. Synchronous transmission system according to claim 1, characterized in that the means for storing (5, 6, 7) store the data coming from the interface device (4) and from the programmable processor (3) so that they can be serially fed, that they are arranged in groups of similar data.

3. Synchronous transmission system according to claim 2, characterized in that the processor (3) can access each of the groups block by block, and that the size of the blocks is variable.

4. Synchronous transmission system according to claim 3, characterized in that the processor (3) can re-sort the blocks in time and store them in an internal memory.

5. Synchronous transmission system according to claim 3 or 4, characterized in that the processor (3) accesses the data arranged in groups with the aid of a transfer list in which information is defined in order to transmit data from a data source to a data sink. and that the processor (3) recalculates the transfer list after each change in the block size.

6. Network element for a synchronous transmission system, with an input / output unit (1), which has an interface device (4) that can transmit a multiplex signal and process a received multiplex signal, and a signal processing device (2) that is provided by the interface device (4 ) processing incoming data, characterized in that the signal processing device (2) has means (5, 6, 7) for storing the data and a processor (3) in that the means (5, 6, 7) for storing the data in accordance with a predetermined Arrange the scheme so that the processor (3) can access and process it and that the means (5, 6, 7) for storing data coming from the processor (3) can feed the interface device (2).

7. input / output unit (1) for a network element, which is part of a synchronous transmission system, with an interface device (4) that can transmit a multiplex signal and process a received multiplex signal, and a signal processing device (2) that the interface device ( 4) processes incoming data, characterized in that the signal processing device (2) has means (5, 6, 7) for storing the data and a processor (3) in that the means (5, 6, 7) for storing the data according to a Arrange the defined scheme so that the processor (3) can access and process it and that the means (5, 6, 7) for storing data coming from the processor (3) can feed the interface device (2).