AU2071400A - Controlled slip in the transmission of a synchronous data stream over an asynchronous communications network - Google Patents

Description

P/00/011 Regulation 3.2

AUSTRALIA

Patents Act 1990

ORIGINAL

COMPLETE SPECIFICATION STANDARD PATENT Invention Title: ft f 'Controlled slip in the transmission of a synchronous data stream over an asynchronous communications network' The following statement is a full description of this invention, including the best method of performing it known to us: FHPSYDrFNATP1T2n1Rii S0n S CEO0368594.5 2 Controlled Slip in the Transmission of a Synchronous Data Stream Over an Asynchronous Communications Network Field of the invention This invention relates to a method of transmitting an asynchronous data stream over a transmission path in an asynchronous communications network, and to a receiver for receiving such a data stream.

Background of the invention Broadband communications networks are known which use an asynchronous transmission method. The most widely known transmission method of this kind is referred to as 10 ATM (asynchronous transfer mode). The fundamentals of this method are described in an Sarticle by G. Koch et al, "ATM: Schlissel zum "Information-Highway"", Telekom Unterrichtsblitter, Vol. 48, 4/1995, pp. 196-205. According to this method, all incoming :*information is packed in cells of equal length, provided with a cell header, and placed via a .0 multiplexer on the transmission line.

Besides time-uncritical information such as text and data transmission, it must be possible to transmit time-critical information, such as speech, video signals, and isochronous bit streams, over such asynchronous communications networks. For this purpose, the function AAL Type 1 (AAL1) is provided in the third protocol layer, the ATM Adaptation Layer (ALL), for continuous-time services with fixed transmission rate, such as speech.

During the transmission of the above-mentioned time-critical information, it is necessary for the transmitter and receiver to operate synchronously with each other. Otherwise data may be lost during transmission. When the receiver is synchronized to the average clock rate of the received data stream, the resulting phase variation may exceed a maximum permissible limit.

An alternative is the external synchronization of the terminals. This requires, however, that a uniform network clock, e.g. a 2-MHz clock, is available in the communications network.

That is not always the case. For instance, a network clock failure may occur, which in the case of time-critical services results in loss of data during transmission. Also, there may be no uniform network clock available, for example in the case of international connections and in the CE00368594.5 3 case of ATM connections over nonsynchronous transmission systems, such as local area networks (LANs).

For that case it has been proposed to compensate for frequency differences in the event of synchronization failures or in international data traffic using controlled slip as in conventional pulse code modulation (PCM) systems. This, however, entails transmission impairments. Particularly during transmission over nonsynchronous transmission systems, the resulting slip rate would be too high.

Summary of the invention It is therefore an cbject of the invention to provide a method whereby synchronous or approximately synchronous (plesiochronous or isochronous) data streams can be transmitted S:i over an asynchronous communications network with less perceptible disturbance occurring than in the prior art. Another object of the invention is to provide a receiver which can receive such data streams and in which the information contents of received data streams are less affected by variations in transmission rate than in prior-art receivers.

According to a first aspect of the present invention there is provided a method of transmitting an at least approximately synchronous data stream from a transmitter to a receiver, each of which uses its own clock signal, over a transmission path in an asynchronous communications network, with a slip operation being performed to compensate for a frequency difference between the clock signals of the transmitter and the receiver, wherein the data stream arriving at the receiver is monitored to detect time periods in which the effect of a slip on the transmitted data contents is small, and that a slip operation is preferably performed in such a detected time period.

According to a second aspect of the present invention there is provided a receiver for receiving an at least approximately synchronous data stream over a transmission path in an asynchronous communications network comprising a buffer for temporarily storing the received data stream, a clock source for generating a clock signal for reading the buffer and means for performing a slip operation when the filling level of the buffer rises above an upper threshold or falls below a lower threshold, a monitoring device for monitoring the data stream and for detecting time periods in which the effect of a slip on the transmitted data contents is CE00368594.5 4 small, with the monitoring device controlling the means for performing the slip operation in such a way that the slip is preferably performed in such a detected time period.

Particular advantages of the invention are that all transmission rates can be used, that uniform network timing is not necessary, and that the method according to the invention can be carried out over existing communications networks.

The invention is particularly suited to the transmission of PCM-coded speech (PCM pulse code modulation) over an ATM link.

Brief description of the drawings The invention will become more apparent by reference to the following description of two embodiments of the invention taken in conjunction with the accompanying drawings, in which: Fig. 1 shows a first embodiment of the invention, in which two terminals, namely a transmitter and a receiver of a synchronous data stream, are interconnected via an asynchronous communications network; 15 Fig. 2 is a block diagram of a receiver in accordance with the invention; Fig. 3 is a flowchart of the method in accordance with the invention; Fig. 4 shows a second embodiment of the invention, in which two terminals are •interconnected via an ATM network and a local area network; and Fig. 5 is a block diagram of a network element between the ATM network and the local area network.

Detailed description of the embodiments Referring to Fig. 1, there are shown two terminals 11 and 12 which are connected to multiplexers 13 and 14, respectively. The two multiplexers 13, 14 are interconnected by a communications network that is based on the ATM concept. The two terminals are ISDN telephones (ISDN integrated services digital network). The multiplexers 13, 14 are ATM multiplexers of a conventional switching center. Via the ATM network 15, a bidirectional ISDN link with a transmission rate of 64 kb/s for digital speech transmission is established, over which PCM-coded speech information is transmitted. Over this link, a data stream containing the coded speech information is transmitted between the two terminals. Each of the terminals CE00368594.5 thus operates both as a transmitter and as a receiver. For a better understanding, however, only one direction of transmission, namely that from the terminal 11 to the terminal 12, will be considered in the following. Therefore, terminal 11 will be referred to as a transmitter, and terminal 12 as a receiver.

For error-free speech transmission, it is necessary that transmitter 11 and receiver 12 operate synchronously with each other. However, synchronization of the two terminals 11, 12 to an identical external timing reference cannot be accomplished in any case, as was explained above.

Transmitter 11 and receiver 12 each have their own internal clock sources, which generate respective clock signals. The transmitter generates the synchronous data stream, which contains the coded speech information, at its local clock rate. The data stream is transported over the ATM network to the receiver. There the received data stream is written into a buffer and read at the receiver clock rate. The receiver clock thus serves as a read clock for reading the buffer. A frequency difference between transmitter and receiver would result in an overflow or underflow of the buffer. "Underflow" means that the buffer is too empty because it was read more rapidly than it was filled, so that further reading is no longer possible for lack of stored data. To avoid this overflow or underflow, a slip operation is performed when an upper or lower threshold for the filling level of the buffer is reached. Either a part of the buffer is read doubly (positive slip) or a part of the buffer is skipped during reading (negative slip). In this way, the filling level of the buffer is brought back to a region in which no overflow or underflow can occur. Such a slip entails a loss of data, which during voice calls appears as a disturbance, for instance as a click.

To minimize the effect of a slip on the contents of the data stream, a fundamental idea of the invention is to "listen into" the data stream in order to perform the slip operation precisely when its effect on the data contents is minimal. During voice calls, that is the case in silent intervals, for example. Thus, according to the invention, a slip operation is performed at instants at which it does not result in perceptible disturbances.

An embodiment of a receiver 20 that operates in the manner described is shown in Fig.

2. The receiver 20 includes a buffer 21 for storing the received data stream and a clock source 22 for generating a clock signal for reading the buffer 21. The writing of the data stream is done in the usual manner, which is not discussed here.

CE00368594.5 6 The clock source 22 is a free-running crystal oscillator that generates a frequency of 2 MHz. A read address generator 23, in this embodiment a simple counter, generates read addresses for reading the buffer 21. A monitoring device 24 monitors the read data stream and detects time periods in which the effect of a slip on the transmitted data contents would be small. In this embodiment, silent intervals, non-speech intervals of the subscriber at the transmitting end, are detected. A control circuit 25 monitors the filling level of the buffer 21, and performs a slip operation when the filling level has risen above an upper threshold or fallen below a lower threshold and if, at the same time, a silent interval is present. For a slip operation, the current address value of the read address generator 24 is simply increased (negative slip) or decreased (positive slip) by the control circuit 25 by a predetermined value, e.g. a value corresponding to one frame length. If necessary, a slip operation can also be performed outside fe.. a silent interval if the filling level of the buffer rises above or falls below the respective oo i threshold by more than a predetermined value without a silent interval having occurred. In this way, overflow or underflow of the buffer can be avoided even if no silent intervals occur.

15 Preferably, however, a slip operation is performed in a silent interval detected by the monitoring device 24. This minimizes the effect of the slip on the information content of the transmitted data stream.

The above-described modules 21-25 are preferably incorporated in an interface circuit of the receiver 20. The data read from the buffer are fed to a digital-to-analog converter 26 in the 20 receiver 20. The analog signal generated by the D/A converter 26 is reproduced by an external loudspeaker 27.

By performing a slip operation, it is also possible to skip two or more frames at a time.

Preferably, the filling level of the buffer is centered by performing the slip operation, so that the buffer is then approximately half full. In the case of unframed signals, an arbitrary portion of the data stream can be skipped or repeated by a slip during reading, in which case the length is preferably chosen so that the buffer is subsequently centered again. If a continuous, i.e. a permanent and approximately constant, frequency offset between the transmitter clock, at whose pulse rate the data stream is received on an average, and the receiver clock is detected, it is advantageous to bring the filling level of the buffer by a slip operation to the opposite boundary of the buffer. This increases the time period until the next slip operation. In the event of irregular phase variations, however, centering of the buffer filling level is the better solution.

CE00668594.5 7 The detection of silent intervals is known per se, from statistical multiplexers or satellite communications, and will not be described here in detail. The invention can also be used with data contents other than coded speech, e.g. with video data. In that case, the slip operation is performed in idle intervals between data packets. The monitoring device will therefore detect idle intervals between data packets.

Silent intervals and idle intervals between data packets are detected heuristically. It may therefore happen that a slip operation is performed at a wrong place because either no silent or idle interval occurs or because a silent or idle interval was detected incorrectly. This results in a residual error rate, which, however, is lower than if the decision on the slip operation were made exclusively on the basis of the buffer filling level.

In the embodiment, the method comprises the steps shown in the flowchart of Fig. 3: SStep 30: At the transmitter, a data stream is generated at a first clock rate.

Step 31: The data stream is transmitted over a transmission path in an asynchronous communications network.

Step 32: The data stream is received at a receiver.

Step 33: The received data stream is written into a buffer of the receiver.

**Step 34: The buffer is read at a second clock rate, which is generated at the receiver.

Step 35: By means of a monitoring device, the data stream is monitored for silent intervals.

Step 36: A check is made to determine whether the filling level of the buffer has reached an upper or lower threshold. If not, no slip operation will be performed and the data stream continues to be received normally, (step 32).

Step 37: When either of the thresholds is reached, a check is made to see if there is a silent interval.

Step 39: If there is a silent interval (step 37), a slip operation will be performed in order to recenter the filling level of the buffer. After that, the data stream continues to be received normally (step 32).

CE00368594.5 8 Step 38: If there is no silent interval (step 37), a check is made to see if the filling level of the buffer has already risen or fallen below the threshold to the point that an overflow or underflow is impending. If that is the case, the slip operation will be performed nevertheless (step 39). Otherwise, nc slip operation will be performed and the data stream continues to be received normally in order to wait for a silent interval (step 32).

The process ends with the end of the data stream, when the transmitter stops sending data.

In a preferred embodiment of the method according to the invention, the size of the buffer used to temporarily store the received data stream is negotiated and fixed during establishment of a call between transmitter and receiver. The available buffer capacity thus represents a configuration parameter of the AAL1 function during call establishment. This ••go S parameter, together with the frequency accuracy of the clock sources of the transmitter and receiver, determine the slip rate. The latter is adapted to meet the requirements of the data contents to be transmitted.

For the transmission of coded speech, at a frequency accuracy of the clock source in the receiver of 20 ppm, a buffer capacity of, 5 ms may be fixed. This corresponds to a buffer size of 40 bytes at a transmission rate of 64 kb/s. This gives a slip rate of one slip per 4 min. A slip can therefore be placed in a silent interval with very high probability.

For the transmission of coded video signals, e.g. for a videoconference, at a receive-side frequency accuracy of 10 ppm, a buffer capacity of 100 ms may be fixed. This corresponds to a buffer size of 256 kbytes at a transmission rate of 2 Mb/s. A slip then occurs after about 2.8 hours, the slip operation will most probably be performed only after the end of the videoconference, and thus not at all.

The above-described first embodiment assumes that a voice call is established solely via the ATM network with the aid of the AALI function. In this configuration, the invention is of advantage, for example, in the case of international connections, for which no common network timing is available, or in the event of a network clock failure. The following describes a second preferred embodiment of the invention.

Fig. 4, like Fig. 1, shows two terminals 41 and 42. The first terminal 41 is connected directly to a first communications network 45, which is an ATM network. The second terminal CE00368594.5 9 42 is connected to a multiplexer 44. The multiplexer 44 is connected to a communications network 46 for transmitting conventional pulse-code-modulated signals using time-division multiplexing (PCM network). The two networks 45 and 46 are interconnected via a network element IWU 47. The network element 47 provides the connection between the two networks 45, 46 and, because of its function, can also be referred to as a gateway or an interworking unit (IWU). The two terminals are ISDN telephones. Between the two terminals, a 64-kb voice link is to be established. To that end, the network element 47 must ensure interoperation between the ATM network and the lccal area network (interworking function).

First, only the "upstream" direction from the first terminal 41 to the second terminal 42 will be considered. The first terminal operates as a transmitter and is transmitting a synchronous data stream over the ATM network 45. The data stream is transmitted through the ATM o oo network by "circuit emulation" at 64 kb/s using the AAL1 function. The network element 47 receives the data stream and must clock the latter anew, since clock signal transmission over the ATM network is not possible with sufficient accuracy.

The clock rate used for the network element 47 is the clock rate of the PCM network 46.

A difference between the clock rates of the first terminal 41 and the network element 47 is eeoc compensated for by slip. operations if an overflow or underflow of the buffer used for temporarily storing the data stream is impending. The network element 47 sends the newly clocked data stream at a transmission rate of 64 kb/s as a time slot of a PCM signal over the PCM network 46 to the terminal 42.

A block diagram of the network element 47 is shown in Fig. 5. Only the modules necessary to provide timing for the received data stream are illustrated. Other modules, such as interface circuits, are not shown for the sake of clarity. The construction is analogous to that shown for a terminal in Fig. 2. The network element 47 has a buffer 51 for temporarily storing the received data stream and a read address generator 53, which is supplied with an external clock signal PCM-CLK 52 from the PCM network. Thus, the clock source in this case is a clock recovery circuit which derives the clock signal from the data stream received from the PCM network.

The data stream read from the buffer 51 is monitored for silent intervals by a monitoring device 54. When a silent interval is detected and the buffer filling level has risen above the upper threshold or fallen below the lower threshold, the read address generator 53 is controlled CE00368594.5 via a control circuit 55 in such a way that a slip operation is performed. The data stream, which has thus been clocked anew, is then transmitted as a time slot of a PCM signal over the PCM network to the second terminal 42. The network element 47 thus operates simultaneously as a transmitter and a receiver.

In the opposite direction, in the "downstream" direction, synchronization is effected in the network element 47 to the data stream received from the terminal 42.

It will be understood that the invention disclosed and defined herein extends to all alternative combinations of two or more of the individual features mentioned or evident from the text or drawings. All of these different combinations constitute various alternative aspects of the invention.

The foregoing describes embodiments of the present invention and modifications, obvious to those skilled in the art can be made thereto, without departing from the scope of the p inve *present invention.

*oo* .o

Claims (9)

1. A method of transmitting an at least approximately synchronous data stream from a transmitter to a receiver each of which uses its own clock signal, over a transmission path in an asynchronous communications network with a slip operation being performed to compensate for a frequency difference between the clock signals of the transmitter and the receiver, wherein the data stream arriving at the receiver is monitored to detect time periods in which the effect of a slip on the transmitted data contents is small, and that a slip operation is preferably performed in such a detected time period.

2. A method as claimed in claim 1 wherein coded speech is transmitted as the data contents of the data stream, and wherein as the time periods in which the effect of a slip is minimal, silent intervals are detected.

3. A method as claimed in claim 1 wherein coded video signals are transmitted as the data contents of the data stream, and wherein as the time periods in which the effect of a slip is minimal, idle intervals between packets of the video signals are detected.

4. A method as claimed in claim 1 wherein at the receiver the data stream is temporarily stored in a buffer and the filling level of the buffer is centered by each slip operation performed.

A method as claimed in claim 1 wherein at the receiver the data stream is temporarily stored in a buffer and the filling level of the buffer is brought by each slip operation performed to the opposite boundary of the buffer in order to compensate for an approximately constant frequency difference between the clock signal of the transmitter and the receiver.

6. A method as claimed in claim 1 wherein during call establishment between transmitter and receiver, a capacity of a buffer to be used for temporarily storing the received data stream is fixed.

7. A receiver for ;eceiving an at least approximately synchronous data stream over a transmission path in an asynchronous communications network comprising a buffer for temporarily storing the received data stream, a clock source for generating a clock signal for SCEOe368594.5 12 reading the buffer and means for performing a slip operation when the filling level of the buffer rises above an upper threshold or falls below a lower threshold, a monitoring device for monitoring the data stream and for detecting time periods in which the effect of a slip on the transmitted data contents is small, with the monitoring device controlling the means for performing the slip operation in such a way that the slip is preferably performed in such a detected time period.

8. A method of transmitting a data stream substantially as hereinbefore described with reference to the accompanying drawings.

9. A receiver substantially as hereinbefore described with reference to the accompanying drawings. *0* SDated this 7th day of March 2000 e. ALCATEL by its attorneys Freehills Patent Attorneys o*o« oo ft eo