Method of bi-directional data transmission over a two-wire line

The invention relates to a method for bidirectional data transmission over a two-wire line, digital data for sending or receiving being modulated or demodulated by means of discrete multi-tone modulation (DMT) and the data to be sent and received being separated.

In order to eliminate disruptive influences on the data to be transmitted, known methods of this type perform the separation of the modulated data in frequency division multiplexing (FDM), with different frequency ranges being defined for the two transmission directions. Another possibility for separation consists in the use of the echo cancellation method (EC), in which the influence of the transmitting part on the receiver is suppressed through the use of adaptive filters.

To carry out these separation methods, a relatively high computing power is required, which greatly increases the costs for data transmission. Especially when used in cases in which, such as ADSL (Asymmetry Digital Subscriber Line), high data rates are transmitted in one transmission direction (downstream) from a central data system to a subscriber located on the periphery and comparatively low data rates are transmitted in the other transmission direction (upstream) should, the effort driven by these known data transmission methods is only subject to poor use.

The FDM method generates a lower and an upper frequency band during transmission in accordance with the two transmission directions. However, since the cable attenuation is frequency-dependent, it is very difficult to achieve the same transmission quality for both transmission channels; in the majority of cases the transmission quality is better in one direction than in the other. In general, however, it is desirable to be able to offer the same quality as possible for both channels. Furthermore, in FDM, the variation of the transmission capacity is associated with considerable effort, since it requires an adaptation of the band filters used in each case, so that the channel bandwidth can be increased or decreased accordingly.

The echo cancellation method, which is the second method known from the prior art in connection with DMT modulation, also has disadvantages, albeit of a different type. With this method, near-end crosstalk is a major technical problem, since the signal distance between the transmitted and received signals is very large. Very high requirements must therefore be met for the A / D converters provided in the transmitting and receiving parts, since the transmitted and received signals occur simultaneously and these must be separated accordingly. The high level differences between the transmitted and received signals require a correspondingly high resolution of the A / D converter, which is therefore very expensive.

Since the DMT signal requires a relatively long settling and settling time (prefix), only FDM and EC separation methods have always been used in the previous solutions.

Overall, FDM and echo cancellation have considerable disadvantages, which, however, have always been accepted because no better solution for the DMT process has been offered.

In the two-wire time separation method described in DE-A1-36 32 710 for full duplex transmission between a main station and a slave station, the directional separation between transmitted and received signals is achieved by alternately sending and receiving data packets on both sides of the transmission path. The time division multiplex transmission used in this case takes place through the formation of time frames which each contain exactly two data packets that are sent in opposite directions. A burst of information is periodically transmitted from a master station, which is received by the slave station after the signal time. The secondary station then begins sending out an information burst of the same duration. The cycle duration essentially results from twice the duration of the information burst and the signal propagation times.

Both the method disclosed in EP-A1-0 426 961 and that in the article IEEE Transactions on communication Vol. Com. 29, No. 11, 1981, the time-shared two-wire digital subscriber transmission system corresponds to such a time-division method, but there is no mention of a signal modulated by means of DMT.

The object of the invention is to achieve a very good transmission quality with relatively little technical effort in a method of the type mentioned at the outset, it being possible to change the transmission capacity easily and inexpensively.

Another object of the invention is to provide a method of the type mentioned at the outset, the implementation of which requires the least possible bandwidth.

AT 406 533 B

A further aim of the invention is to provide a method which is characterized by low complexity with regard to the use of hardware or computing power, so that it can be carried out in a simple and inexpensive manner.

A further aim of the invention is to create a method with which transmissions which to a large extent only take place in one of the two transmission directions can be carried out at a high transmission speed.

According to the invention this is achieved in that - in a manner known per se - the data to be sent and received are separated by time division multiplexing (TDM), the associated multiplex time frame being divided into a predeterminable number N of time slots, and a predeterminable number thereof K of time slots excluding one transmission direction, e.g. transmission, and the remaining number (NK) of time slots excluding the other transmission direction, e.g. Received, assigned.

As a result, in the case of multi-tone (DMT) -modulated signals, the separation of data to be sent and received is carried out using the time division multiplexing (TDM) which is known per se. This combination of DMT modulation and TDM method is not described in the aforementioned publications, from which only the time division multiplex transmission of data packets emerges, and there is in no way any suggestion for a person skilled in the art to proceed in the manner according to the invention.

The present invention fulfills all of the above-mentioned objects by combining DMT modulation and TDM methods with one another. However, since the DMT signal requires a relatively long settling or settling time (prefix), the solution according to the invention has never before been considered by those skilled in the art. The prejudice that has arisen in this regard that the combination of DMT and TDM processes is not efficient is therefore overcome by the invention, although longer prefix times result, the advantages of high transmission quality at relatively low costs and the easy-to-implement variability of the transmission capacity The disadvantage of the slightly longer transmission time not only outweighs but far surpasses it.

The TDM method used according to the invention offers the advantage of the same transmission quality in both transmission directions, since sending and receiving take place with the same line attenuation. As a result, both transmission directions can take place with the least possible reduction in quality, which naturally results in an improvement over the FDM method, since with this only the frequency band that is poorly transmitted can be viewed as a quality reference value. Another advantage of the method according to the invention is the very simple change in the transmission capacity, which is made possible by the appropriate choice of the number of time slots for the respective transmission direction.

Since either only transmitter or only receiver functions are active in the method according to the invention, less processor power is required than in conventional methods, since the latter have to cope with a very high internal data traffic. This makes it possible to implement a transmission carried out according to the method according to the invention in a very cost-effective manner. Another advantage of the temporally separated sending and receiving results in a low bandwidth requirement, which allows the method according to the invention to be used in a wide variety of ways.

The method according to the invention also makes it possible to use low-resolution A / D converters, so that a substantial reduction in costs can be achieved in this way.

In a development of the invention, it can be particularly advantageous in the case of asymmetrical data transmission if the majority of the data is transmitted in one transmission direction and only a small remainder is transmitted in the other. This is the case when the number N of time slots is selected to be much larger than the number K. This condition is preferably fulfilled when N is equal to 30 and K is equal to 1.

Since the method according to the invention can be used for data transmission over telephone lines, it can e.g. The dialing of the number on the line leads to impulsive disturbances, which cause a transmission error that must be corrected. However, the data transmission does not have to take place via telephone lines; within the scope of the invention it can take place via any two-wire line suitable for this purpose. In the same way, a wide variety of electromagnetic disturbances, including those external to the system, can have an influence on data transmission.

AT 406 533 B

The well-known ARQ (Automatic Repeat Request) method is usually used for error correction in such a way that the data transmission remains error-free even in the event of any disturbances on the line, whereby the data throughput can, however, drop sharply, since an incorrectly transmitted data packet is repeated until it is error-free Will be received.

In a further embodiment of the invention it can therefore be provided that in the multiplex time frame of the data transmission a predeterminable number of time slots for ARQ (Automatic Repeat Request) transmission repetitions are provided on average.

In this embodiment, transmission overcapacity is therefore always available. If a data block is received incorrectly, the receiver only requests a repetition as often as is possible within the overcapacity available in the time average, so that the nominal data throughput can be kept constant, unaffected by the repetitions of transmission. If the transmission is error-free, a more redundant signal is transmitted. The duration of the time span over which the time averaging takes place is essentially limited by the storage capacity of the ARQ buffer used.

According to another variant of the invention, it can be provided that, in the event of a faulty transmission, the data are transmitted in modified form, for example by means of a computational algorithm.

In this way, the error that occurs during transmission, which is caused by cutting off part of the amplitude in the event of transmission overload, can be corrected.

In a particularly preferred manner, it can be provided that the data are modified by logical inversion.

This inversion operation represents an algorithm that is very easy to calculate and can be implemented without great effort.

Furthermore, it can be provided that the switching frequency of an interference source, for example a power supply unit, is synchronized with one of the carrier frequencies of the discrete multi-tone modulation.

In this way, the DMT process, which is sensitive to frequency-selective interference, can be secured against known sources of interference. When the switching frequency of the interference source is synchronized to one of the carrier frequencies of the DMT modulation, the interference only affects this carrier frequency and its multiples, so that it can be compensated for by an adaptive algorithm.

In the case of several two-wire lines routed next to one another, on each of which data is transmitted, there is usually crosstalk, which naturally has a disruptive effect on the transmission.

According to another embodiment of the method according to the invention, in which data is transmitted over two or more two-wire lines, which are at least partially stirred in crosstalk distance, it can be provided that the time division multiplex operation (TDM) is carried out synchronously on all two-wire lines, so that on all Two-wire lines are either sent or received at the same time.

This means that either transmission or reception is always carried out at the same time, so that the individual receivers can be prevented from being influenced by transmitters that are not directly connected.

The invention is explained in more detail below with reference to an exemplary embodiment shown in the drawings.

It shows:

1 shows a block diagram for carrying out an embodiment of the invention

Procedure and

2 shows a schematic representation of a tent frame according to the invention.

A bidirectional data transmission of digital data according to the block diagram shown in Fig. 1 is carried out by converting the digital data coming from a data source 1, 4 in the transmitting part 50 to an analog transmission signal during transmission and via a line transformer 13 of a two-wire line 100 to one at the end this line 100 located participants are transmitted. In contrast, a signal arriving on the two-wire line 100 is fed via the line transformer 13 as a received signal to the input of a receiving part 51, where it is converted into digital data. Since the method according to the invention never sends and receives at the same time, the line transformer 13 can be used instead of an otherwise customary hybrid circuit, whereby the often problematic adaptation of the hybrid circuit to the line impedance is eliminated from the outset. An annoying crosstalk caused by a hybrid circuit, through which signal residues from the transmitter to the

AT 406 533 B

Recipients from the same subscriber side can therefore be ruled out as a source of interference for this method.

In the embodiment shown in Fig. 1, the transmitting and receiving part 50, 51 both a central data point C (CENTRAL) and a peripheral data point R (REMOTE) is shown in a single block diagram, which is to be understood as the central data point C is connected to the data point R via the transmitter 13, the two-wire line 100 and a further transmitter 13. R are marked with ATU-C only ”or ATU-R only.

Without restricting the general applicability of the method according to the invention, a home video system is described as an exemplary embodiment of asymmetrical data transmission, in which the video information of various videos is stored in a mainframe as data in compressed form in the central data point C and can be called up via a peripheral data point R. The control information is exchanged between data points C and R via a bidirectional control channel, with a data rate of 64 kbit / s being specified. This control information can relate to various commands to be issued by the participant, such as PLAY, REWIND or the like. ä, as they are known from a video recorder and refer to internal control commands and is comparatively small in its amount compared to the broadband information sent out by the central data point C, which essentially contains the video information that is only available at a data transfer rate of 2.048 Mbit / s one direction from C to R.

However, the data rates mentioned can also be chosen completely different for the method according to the invention, for example much higher, with a data rate of approximately 50 Mbit / s to 150 Mbit / s also being available for the broadband information to be transmitted in only one direction. The transmitted information can represent any type of voice, image or data information. A different rate can also be implemented for the bidirectional control channel, which, however, can fulfill not only control functions but all possible data transmission functions.

At the input-side part of the transmitting part 50, two different data inputs are formed for the data point C and only one data input for the data point R. The data stream from data source 1 arrives at the first input, which is the same for C and R, e.g. essentially sends out control commands which arrive via a subsequent scrambler 2 in a send buffer 3 following this, the data coming from the data source 1 being converted in the scrambler 2 according to a predeterminable algorithm. This prevents a long-lasting, constant logical state and achieves a balanced statistical distribution of the binary states. Subsequently, the scrambled signals are temporarily stored in the transmit buffer 3. In the data location R, the data emerging from the transmit buffer 3 are multiplexed via a device MUX with other data which are generated in the ARQ buffer 24 and contain repeat instructions.

At the second input of the transmission part 50, which is designed only for data point C, the data stream comes from the data source 4, which generates the broadband information, via a subsequent scrambler 5 and via an ARQ (Automatic Request) buffer 6, which contains a CRC generator , via which error correction coding takes place, to the second input of the transmitting part 50. The data converted in the scrambler 5 are temporarily stored in the ARQ buffer 6 and repeated in the event of an incorrect transmission. A special ARQ transmission technique according to the invention is described below.

The data arriving serially via the inputs of the transmitting part 50 are combined in the encoder 7 to reduce the data rate in a predetermined length and assigned to a corresponding symbol for further processing on the basis of a coding table. Furthermore, this coded signal is modulated in the subsequent DMT (Discrete Multi Tone) modulator 8 according to this known method and passed through a high-pass filter 9 which essentially suppresses the voice frequency band to avoid interference. The digital output signal of this high-pass filter 9 is converted via a digital-to-analog converter 10 into an analog signal, which reaches the converter 13 via a band-pass filter 11 and then via an amplifier 12. On the one hand, the bandpass filter 11 once again fulfills the function of the high-pass filter 11 and, on the other hand, it cuts off the high-frequency voltage peaks caused by the analog-digital converter 10. To fulfill the sampling theorem, the frequency of the analog-to-digital conversion is selected so that the analog-to-digital converter 10 scans at least twice for the highest frequencies that occur

AT 406 533 B

The transmitting part 50 and the receiving part 51 are controlled by a TDM (Time Division Multiplex) unit 30 so that, according to the invention, the data to be transmitted and received are separated by time division multiplexing, the associated multiplex time frame being divided into a predeterminable number N of time slots , and thereof a number K of time slots of the time frame excluding one transmission direction, e.g. Sending, and the remaining number NK of time slots is assigned exclusively to the other transmission direction, e.g. receiving. For this purpose, the TDM unit controls the transmitting part 50 and the receiving part 51 by activating them at the given time. The transmitting part 50 and the receiving part 51 are never in operation at the same time, as a result of which the processor power required for the control can be designed correspondingly low. Since this also rules out any influencing of the own transmitter on the receiver, only a low resolution is required for the analog-to-digital converter 16 of the receiver part. This advantage has a very cost-effective effect due to the direct proportionality of resolution and price with analog-digital converters.

The method according to the invention has the advantage of a relatively low bandwidth requirement and very low complexity, which is evident in the hardware or in the computing power required. With conventional methods for the separation of sending and receiving, a considerable part of the computer power for internal communication is lost, while with the method according to the invention this auxiliary computer capacity can be kept very low.

The method according to the invention has its limit where the proportion of sending and receiving approaches the 50% percentage limit, since other methods such as echo canceling or the like can then be carried out with the same or less effort.

FIG. 2 shows the time frame which is subdivided into time slots and is used in the method according to the invention. The two directions of transmission are identified by the terms upstream and downstream. In this example, the entire time frame is 20.625 ms long and divided into different slots of 625 ps, with the majority of the data being transmitted in the downstream direction. This division is particularly advantageous when a bidirectional channel with a low and a unidirectional channel with a high data rate is required in one transmission direction. In the illustrated embodiment, control commands are transmitted via the bidirectional channel through the time slots labeled CONTROL in the downstream and upstream direction and via the unidirectional channel through the 30 downstream time slots labeled VIDEO with an auxiliary slot in the time average of video information. This type of transfer can take place for any information.

The distribution of the transmission and reception capacities can be adapted to the respective conditions by selecting the number of upstream or downstream time slots. If the workload changes, this ratio can be adjusted automatically according to current requirements. The specified transmission and reception times have the advantage over frequency division multiplex transmission that data received and data to be transmitted do not have to be processed at the same time, which reduces the computer performance or the hardware effort can be designed to be correspondingly low. A coded and DMT-modulated data unit is transmitted in each DMT slot.

For ARQ retransmissions, according to an embodiment according to the invention, a predeterminable number of time slots for ARQ retransmissions is provided in the multiplex time frame of the data transmission in the time average. For this purpose, when the data are sent, they are constantly written into the ARQ send buffer 6 and passed on from this to the encoder 7 again. The outgoing data from the buffer 6 is transmitted faster than it is filled. The last data block is entered again in the resulting gap, but this is recognized as a repeated block on the receiver side and automatically eliminated. Thus, if the transmission is error-free, transmission is always carried out with overcapacity, without the transmitted information content being greater.

As soon as a transmission error occurs, the receiver in the peripheral data point R detects the error by means of its CRC error detection in the ARQ unit 24 and then passes on the command via the multiplexer of the send buffer 3 for data repetition, which is then sent as control information over the bidirectional channel will. In the central data point C, this information is demultiplexed after passing through the receiver part 51 in the receiver buffer 27 and a control command is given to the ARQ buffer 6 to repeat the faulty transmission.

AT 406 533 B

In this exemplary embodiment, only one auxiliary slot is available in the time average, which corresponds to an overcapacity of 3.33%. The duration and number of auxiliary slots are not subject to any restriction in this context and can be adapted to the respective conditions within what is technically feasible.

After an incorrect transmission, the retransmission is carried out in the following time frame, which can extend over several successive time slots. Averaged over time, only one time slot per frame should be used for the repetitions in this example.

The time span over which the time average is calculated is determined by the size of the ARQ buffer memory. As soon as this is filled with information, no further repetitions can be carried out and the faulty data block must be output as transparent.

Compared to a conventional ARQ method, the time span specified for the data repetitions is fixed in the time average. As a result, it cannot happen that the transmission is repeated until it is error-free and the transmission time is thus greatly increased due to a longer-lasting disturbance. With the well-known ARQ method, the data transmission is repeated even in the event of any faults until it is received without errors, which, however, greatly reduces the data throughput. On the other hand, due to the fixed overcapacity, which is between 2 and 10%, but preferably between 3 and 5%, the transmission in the method according to the invention is only repeated as often as is possible within the scope of the overcapacity in order to maintain the nominal data throughput. If one of the incorrect data blocks in succession cannot be received repeatedly and correctly, it is output transparently.

In the case of a signal modulated by discrete multi-tone modulation (DMT), the ratio of peak value to mean value is very large, so that clipping of the signal peak is a frequent source of error. In order to correct this error in a simple manner, after a faulty data transmission, the digital bit sequence can be modified during the repetition process in the transmitter, for example by a computational algorithm, and then transmitted again. In the receiver, the calculation algorithm used is applied in reverse and the data is recovered. This means that this transmission error can be eliminated very effectively. In particular, it can be carried out in a simple manner in terms of circuitry or computation to transmit the faulty data in inverted form

Another source of interference in the DMT process results from the switching frequency of the voltage supply used, e.g. the power supply unit, since this switching frequency is in the transmission range and thus shows its effect as a frequency-selective interference. In addition, these disturbances are dependent on other influencing variables, such as the load currently on the power supply unit. This type of interference can be reduced by synchronizing the switching frequency of the power supply to one of the carrier frequencies of the DMT modulation. This disturbance only affects this carrier frequency and its multiple, so that it can be compensated very easily by an adaptive algorithm.

In FIG. 1, the receiving part 51 corresponding to the transmitting part 50 is also shown. The signals arriving from the other side of the subscriber via the two-wire line 100 and the transmitter 13 are sent to the input via a bandpass filter 14 and an AGC (Automatic Gain Control) unit, which generates an approximately constant-amplitude signal regardless of the current signal conditions on the line an analog-digital converter 16 belonging to the receiving part 51, the output of which is connected to a high-pass filter 17. The signal present at the input of the high-pass filter 17 is fed back to the AGC unit 15 as a manipulated variable via an AGC control circuit 18.

After the high-pass filter 17, the demodulation of the signal takes place, from which the transmitted pilot tone is fed to a pilot AGC unit 20 only in the peripheral data point R, from which a reference signal for the clock generation unit 31 of the peripheral data point R is obtained in the clock generation unit 21. This clock generation unit 31 generates the time base for the TDM unit 30 and for the system clock. The data location C does not require a clock generation unit, since an independent time base is provided here.

The linear distortions caused by the transmission path are eliminated in an equalizer 22 with an update function connected to the DMT demodulator 19. Subsequently, the encoding takes place in a decoder 23 in accordance with a decoding table, whereupon a serial bit stream is again present at the output of the decoder 23, which has two

AT 406 533 B

Outputs is performed. The first output, which is the same for data locations C and R, consists of a receive buffer 27 for control information, a subsequent descrambler 28 in which the data are restored in their correct order and the data sink 29 which receives the control data sent. The second output of the receiving part 51, which is only provided for the data point R, is connected to an ARQ buffer 24, which temporarily stores the broadband information transmitted from the data point C, verifies and, if necessary, the command via a control unit integrated in the ARQ buffer 24 to send the incorrectly transmitted data again to the multiplex input of the send buffer 3, which is transmitted back to data point C. A descrambler 25 is connected to the output of the ARQ buffer 24, followed by a data sink 26 for taking over the broadband information.

If data is transmitted over two or more two-wire lines, which are at least partially routed at a crosstalk distance, it can happen that crosstalk occurs due to the mutual inductive influence of the two-wire lines. This undesirable disruption can occur particularly in a central data system in which many outgoing two-wire lines are routed next to one another.

In one embodiment of the method according to the invention, this type of disturbance is avoided in that the time-division multiplex operation is carried out synchronously on all two-wire lines. This means that all two-wire lines are either sent or received at the same time, so that no further influence is possible.