DE4301701C1 - Verifying digital signal at output of noisy transmission path - using cross-correlation between received signal and processed transmission signal - Google Patents

Abstract

The signal recognition method uses a cross-correlator (2) supplied directly with the received signal at one input and with a processed transmission signal at its other input, to provide a cross-correlation function. The received signal is sampled by the cross-correlator using a sampling clock signal (TA) with a period which is less than half the data clock period (TM) of the transmitted data. Pref. the cross-correlation function is proportional to the peak-to-peak amplitude of the output signal from the transmission path in the absence of noise during transmission of individual rectangular pulses with a length corresponding to the data clock period. USE - For detecting data from transmission path with high attenuation and noise level.

Description

The invention relates to a method for the detection of distorted and heavily disturbed digital signals with the help of a digital cross correlator, the undistorted and undisturbed digital signal must be available as a reference and a circuit arrangement for carrying out the method.

The use of a cross correlator is a well known and suitable means to get uncorrelated interference signals eliminate. It takes advantage of the fact that uncorrelated interference signals the result of the correlation measurement with a long (theoretically infinitely long) correlation time do not affect.

A method for the detection of an analog signal is described in JP 03- 103 772 A2 known. A signal is branched and in one first way over a detector and a low pass to one first input of a comparator. On the second way the signal goes to a signal processing, here one Directional coupler. The at the two outputs of the directional coupler pending signals are opened via a further branch a detector and from there on another low pass passed the second input of the comparator. So it will the presence of a signal using a cross correlator proven.  

Relating to a digital transmission signal m (t) on the one hand and the superposition e (t) from a component g (t) correlated to the transmission signal and an uncorrelated component n (t) on the other hand (see FIG. 1) this means

d. H.

Φ _{m, e} (τ) ≡ Φ _{m, g} (τ) [2]

The result of the cross-correlation measurement is therefore the same correlation function that you would have measured if that Interference signal n (t) would not have been present.

The cross-correlation function Φ _{m, g} (τ) can be used to draw conclusions about the presence or absence of correlated signals, but it is not possible to make any statements about the shape or the amplitude of the correlated signal component.

Another known method to determine distortion is the consideration of the so-called eye diagram.

The eye diagram is obtained by applying the disturbed and distorted signal e (t) at the output of the channel to the vertical deflection of an oscilloscope and at the same time synchronizing the time deflection with the period 2T _m . If the binary data sequence sent is a random sequence, the so-called eye diagram results on the oscillograph screen, from which one can draw conclusions about the channel distortions. The disadvantage of this method is that it only provides evaluable results when the signal-to-noise ratio is very favorable.

A third known method is the synchronous one Averaging with periodic transmission signal.  

This method assumes that the binary data sequence sent is periodic. The channel output signal e (t) is written into a digital memory starting at the time T _s and ending at the time T _s + T ₁ , so that a section of the signal, namely e (T _s ) to e (T _s + T ₁ ) is saved. Subsequently, further time segments of the signal e (T _s + nT _m ) to e (T _s + nT _m + T ₁ ) are added to the old memory content. This process can be repeated until an overflow occurs in one of the memory registers. After the process is complete, the contents of the memory registers must be divided by the number of additions. The result is the presentation of a section of e (t), free from interference. The disadvantage of this method is that the binary data sequence sent must be periodic. If this cannot be achieved, the method cannot be used.

The object of the invention is to provide methods and associated circuit arrangements for carrying out the methods with which the time function at the output of the channel can be determined, which would be measured if a defined transmission signal, hereinafter referred to as the 'test function' m _T (t), would be fed into the channel and the channel would be trouble-free.

It is important that the digital signal m (t) actually transmitted over the channel is assumed to be given and cannot be influenced. The test function m _T (t) can therefore not be fed in in real terms and is merely simulated by the formation of the cross-correlation function. To do this, this method must be able to eliminate the influence of the interference.

The problem is solved as in the claims specified.  

In short, there is the object of the invention in doing this to a cross correlator at its first input investigating signal e (t) and at its second input from the transmission signal m (t) processed in a special way Signal m '(t) to be supplied and, if necessary, the thus formed Cross correlation function using a Scale proportionality factor.

The present invention is described in more detail below.

Information generated by a news source as an electrical digital signal via a severely disturbed and distorting channel. The undistorted and undisturbed Signal (hereinafter referred to as 'transmission signal' m (t)) available for reference.

Now the task is to provide the evidence with what amplitude and with what curve shape the distorted There is a digital signal at the output of the transmission channel.

The transmission signal is first described.

A message source generates a binary value a _n at the discrete times nT _m . The sequence of the binary values is random and cannot be influenced.

With the help of a modulator, these binary values are to be transmitted in the form of "amplitude shift keying" in the form of analog voltages ( FIG. 1).

For this purpose, the binary values with a sequence of rectangular voltage pulses also generated at intervals of the cycle time T _m

_Tr U _(t) = U _0/2 rect (t / T _m)

linked, such that the transmission signal at the output of the Modulators the form

assumes ( Fig. 2), wherein

The following must distinguish between random and non-random Transmission signals are distinguished. Broadcast signals are called "Randomly" designated if its autocorrelation function only has a single main maximum at position 0 or the Secondary maxima of the autocorrelation function outside the Measuring range of the cross correlator used. In addition, it is assumed that the States logic 1 and logic 0 equal on average over time occur frequently (1-0 uniform distribution). Are these Conditions are not met, the transmission signal is called "non- randomly ".

This broadcast signal would be distorted and disturbed Transmission channel.  

The channel distortions are to be regarded here exclusively as linear distortions, ie they can be described by the shock response h (t) of a linear, time-invariant system (LTI system). Thus the signal at the output of a disturbed transmission channel is (see Fig. 1):

e (t) = g (t) + n (t) = m (t) * h (t) + n (t) [3]

With
e (t) = signal at the output of the disturbed transmission channel,
g (t) = signal at the output of the undisturbed transmission channel,
h (t) = shock response of the transmission channel,
n (t) = interference signal,
* Symbol for the convolution operation defined by equation [4]

The interference signal n (t) is by definition the useful signal g (t) uncorrelated, d. H. with

Defined cross-correlation function of useful and interference signals is identical to zero.  

According to the invention, a cross correlator at its first Input the signal to be examined e (t) and at its second Input on from the transmission signal m (t) in a special way processed signal m '(t) supplied and the thus formed Cross correlation function using a Proportionality factor scaled.

In the case of random transmission signals in the form of a curve and amplitude, the scaled cross-correlation function is identical to the signal that would be measured at the output of the transmission channel if a defined test function m _T (t) were fed into the channel and the channel were free of interference. This interference-free channel response to the test function is called g _T (t) in the following. Fig. 3 shows the four test functions that are used in the following.

Both the message actually sent and the Disruptions must be seen as given and not influenced will.

There are possible solutions for various test functions specified, depending on the application, which Test function can be implemented or is appropriate.

Three test functions are shown in FIG. 3:

Fig. 3a rectangular pulse,

FIG. 3b Double rectangular pulse,

FIG. 3c rising edge,

Fig. 3d falling edge.

The type of transmission signal processing depends on the Available signals. There will be four different ones Types of transmission signal processing described. Of the selected signal processing in turn depends on the for Scaling the cross-correlation function needed Proportionality factor.

As mentioned above, must be between random and non-random transmission signals in the sense of the described A distinction can be made between:

In the case of non-random transmission signals, it is only possible in special cases to determine the curve shape of the channel response g _T. However, it is always possible to determine the peak-to-peak amplitude of g _T. A further proportionality factor is required for this, which can be determined from the autocorrelation function Φ _{m, m} (τ) of the transmitted signal.

The auto-correlation function of the transmission signal can be at stationary transmission signals once before starting the measurements the cross correlator or in the case of non-stationary transmission signals during the measurement with a separate autocorrelator be determined. (Such are stationary transmission signals Signals referred to, the autocorrelation function in the Not changed over several correlation measurements.) Im In the case of periodically repeated bit patterns, the Autocorrelation function also calculated from the bit pattern will. These methods of determining Autocorrelation function of a binary signal are state of the art Technology and not the subject of the invention.  

In the following explanations, it is assumed that the transmit signal is in digital form and the receive signal in analog form. In the cross-correlator, the analog signal is sampled and quantized with the sampling clock period T _A. The ratio of the data clock period T _m to the sampling clock period T _A is R. So that the sampling theorem is fulfilled, R <2 must be selected. The transmission signal m (t) can assume the logical states '0' and '1'. The processed transmission signal m '(i) can assume the three numerical values -1.0 and 1 (ternary signal).

A cross correlator that has an analog and a ternary signal correlated, is described in DE 39 11 155 A1.

The following are the four types of broadcast signal conditioning and the resulting proportionality factors for describes the scaling of the cross-correlation function:

The first type of broadcast signal conditioning is the To supply a signal to the reference input of the cross-correlator, which one

• - assumes the value +1 for the duration of a sampling period T _A if the transmission signal has the logical state '1', ie if the binary value a _n = 1 is transmitted,
• - assumes the value -1 for the duration of a sampling period T _A if the transmission signal has the logic state '0', ie if the binary value a _n = 0 is transmitted and
• - assumes the value 0 for all remaining R-1 samples that occur within the data clock period T _m .

This processed signal is m ₁ '(i). It is generated by a circuit arrangement of transmit signal processing 1 ( FIG. 4) which requires the transmit signal m (t), the associated data clock T _m and the sampling clock T _{A of} the cross-correlator as input signals.

Assuming a randomly transmitted binary data sequence a _n with 1-0 uniform distribution, the following applies to the cross-correlation function Φ _{m₁ ′, g} (τ)

R is the ratio of the data clock period T _m to the sampling clock period T _A and g _T1 (τ) the channel response to the test function m _T = rect (t / T).

This equation can be solved for the sought function g _T1 (τ):

g _T1 (T) = R Φ _{m₁ ′, g} (τ) [7].

The scaling of the cross correlation function according to equation 7 can e.g. B. from a computer connected to the output of Cross correlator is connected to be adopted.

The circuit arrangement for calculation of g _T1 (τ) shows Fig. 4.

In the case of non-random transmission signals, the cross-correlation function has, in addition to the main maximum described by g _T1 (τ), further secondary maxima c _n g _T1 (τ-nR). These are shifted by integer multiples of R and weighted with coefficients c _n <1.

If these secondary maxima overlap each other, the shape of g _T1 (τ) can no longer be uniquely reconstructed from the cross-correlation function and the scaling according to equation 7 produces incorrect results.

However, it is possible to calculate the peak-to-peak amplitude of the sought signal g _T1 (τ) if the autocorrelation function Φ _{m, m} (τ) of the transmitted signal is known. The following applies:

If the data clock signal T _m required for the circuit arrangement according to FIG. 4 described above is not available, another form of transmission signal processing can be selected:

This consists of the reference input of the cross correlator to supply a signal which

• - assumes the value +1 for the duration of a sampling period T _A when the transmission signal changes from '0' to '1' (rising edge),
• - assumes the value -1 for the duration of a sampling period T _A when the transmission signal changes from '1' to '0' (falling edge) and
• - assumes the value 0 if the transmission signal does not change.

This signal is generated by a circuit arrangement of transmit signal conditioning 2 in FIG. 5, which requires the transmit signal m (t) and the sampling clock T _{A of} the cross-correlator as input signals.

This signal m ₂ ′ (i) prepared in this way is the difference quotient of the transmission signal apart from the factor ½:

m ₂ ′ (i) = ½ (m (i) -m (i-1)).

The factor ½ is used for reasons of simpler implementation not in the processing of the broadcast signal, but in the subsequent scaling of the cross correlation function considered.

Assuming a randomly transmitted binary data sequence a _n with 1-0 uniform distribution, the cross-correlation function formed with the transmitted signal prepared in this way can be performed on the test function according to the channel response g _T2

dissolve. This signal is in the following double rectangular pulse called. The following applies:

g _T2 (τ) = 2R Φ _{m₂ ′, g} (τ) [9].

The scaling of the cross correlation function according to equation 9 can e.g. B. from a computer.  

The entire circuit arrangement for calculation of g _T2 (τ) Fig. 5 shows.

With this type of transmission signal processing, in the case of non-random transmission signals, there will generally be a superposition of several shifted and differently weighted functions c _n g _T2 (τ-nR), which is why the shape of the function g _T2 sought cannot be clearly reconstructed from the cross-correlation function. Analogous to equation 8, the peak-to-peak amplitude of the sought signal g _T2 (τ) can also be calculated here if the autocorrelation function Φ _{m, m} (τ) of the transmitted signal is known. It applies

If the channel transfer function has a pronounced high-pass character (lower cut-off frequency »1 / T _m ), the channel output signal consists of pulses which are correlated to the edges of the transmission signal and whose pulse duration is much shorter than the data period T _m .

Real transmit signals have finite slope, which can be different on the rising and falling edge. As a result, the channel responses to the rising and falling edges are not symmetrical. In this case, it is expedient to channel the response g _T3 to the "rising edge" test function (ε _s (t)) and the channel response g _T4 to the "falling edge" test function (ε _f (t)) in two separate correlation measurements with different Determine types of transmission signal processing:

In a first measurement, the reference input of the Cross-correlator fed a signal which

• - assumes the value +1 for the duration of a sampling period T _A when the transmission signal changes from '0' to '1' (rising edge) and
• - assumes the value 0 for all other samples.

This signal is m ₃ '(i). It is generated by a circuit arrangement of transmit signal conditioning 3 in FIG. 6, which requires the sampling clock of the cross-correlator and the transmit signal as input signals.

The cross correlation function formed in this way can be resolved according to g _T3 :

where / m ₃ '/ is the time average of the processed transmission signal m ₃ ' (i).

In a second measurement, the reference input of the Cross-correlator fed a signal which

• - assumes the value +1 for the duration of a sampling period T _A when the transmission signal changes from '1' to '0' (falling edge) and
• - assumes the value 0 for all other samples.

Let this signal be m ₄ '(i). It is generated by a circuit arrangement of transmit signal processing 4 in FIG. 7, which requires the sampling clock of the cross-correlator and the transmit signal as input signals.

The cross-correlation function thus formed can be resolved after g _T4 .

where / m ₄ '/ the time average of the processed transmission signal m ₄ ' (i).

The scaling according to equation [11] or [12] can e.g. B. from a computer to which the input parameters Cross-correlation function and the mean of the processed Transmit signal to be entered.

Equations [11] and [12] apply to both random and also for non-random transmission signals.

The entire circuit arrangement for calculating g _T3 for a rising edge is shown in FIG. 6, and the circuit arrangement for calculating g _T4 for a falling edge is shown in FIG. 7.

In the case of the transmit signal preparations 3 ( FIG. 6) and 4 ( FIG. 7), the mean / m ′ / of the processed transmit signal m ′ (t) is required for the calculation of the proportionality factor. Therefore, a circuit arrangement must also be provided which determines the mean value of a signal supplied to it.

The following are circuit arrangements for the previously described signal processing 1 to 4 described.

Transmit signal processing 1 ( FIGS. 4 and 8)

In order to obtain a processed transmission signal m ₁ '(i), as is required for carrying out the method according to claim 1 or 2, namely

• - assumes the value +1 for the duration of a sampling period T _A if the transmission signal has the logical state '1', ie if the binary value a _n = 1 is transmitted,
• - assumes the value -1 for the duration of a sampling period T _A if the transmission signal has the logic state '0', ie if the binary value a _n = 0 is transmitted and
• assumes the value 0 for all remaining R-1 samples which occur within the data clock period T _m ,

a circuit arrangement is proposed according to the invention, which consists of an inverter I1, two flip-flop circuits FF1 and FF2, two sampling circuits AT1 and AT2 and two NOR gates NOR1 and NOR2 exist.

The flip-flop circuits FF1 and FF2 are each acted upon by the data clock T _m inverted in the inverter I1 at the clock input C and thus clocked. The data signal m (t) is fed to the flip-flop circuit FF2 at the data input D, and the data input D is connected to the potential high H in the flip-flop circuit FF1.

The output signal from FF1 appearing at the output Q of FF1 is sampled by the sampling circuit AT1 with the sampling clock T _A. At the output Q of AT1, the mantissa M of the signal m ₁ '(i) appears.

The inverted output signal of FF2 appearing at the output of FF2 is sampled by the sampling circuit AT2 with the sampling clock T _A. At the output Q of AT2, the sign V of the signal m ₁ '(i) appears.

The signals appearing at the outputs of AT1 and AT2 are each supplied to one of the NOR gates NOR1 and NOR2. The NOR gate NOR1 also receives the signal from the output of FF1 supplied to the NOR gate NOR2 from the output Q of FF2. The output signal from NOR1 is used to reset FF1 and is fed to this at the R input. The output signal from  NOR2 is used to set FF2 and becomes this at the S input fed.

Transmit signal conditioning 2 ( FIGS. 5 and 9)

In order to obtain a processed transmission signal m ₂ '(i), as is required for carrying out the method according to claim 3 or 4, which is namely

• - assumes the value +1 for the duration of a sampling period T _A when the transmission signal changes from '0' to '1' (rising edge),
• - assumes the value -1 for the duration of a sampling period T _A when the transmission signal changes from '1' to '0' (falling edge) and
• - assumes the value 0 if the transmission signal does not change,

a circuit arrangement is proposed according to the invention, which consists of an inverter I3, two flip-flop circuits FF3 and FF4, two sampling circuits AT3 and AT4, two NOR gates NOR3 and NOR4 and an OR gate OR.

The data inputs D of the flip-flop circuits FF3 and FF4 are each connected to the high H potential. The clock input C is connected to the transmission signal m (t) inverted in the inverter I3 in the flip-flop circuit FF3 and to the transmission signal m (t) in the flip-flop circuit FF4. The output signal present at the output Q of FF3 is sampled by the sampling circuit AT3, the output signal of FF4 by the sampling circuit AT4 with the sampling clock T _A. The value appearing at the output Q of AT3 forms the sign V of the processed transmission signal m ₂ '(i) and is supplied to the OR gate OR. The signal appearing at the output Q of AT4 is also fed to the OR gate OR. The signal appearing at the output of the OR gate forms the mantissa M of the processed transmission signal m ₂ '(i).

The inverted outputs of FF3 and AT3 or FF4 and AT4 a NOR gate NOR3 or NOR4 is fed in each case, whose output signals each have a flip-flop circuit FF3 or reset FF4.

Transmit signal conditioning 3 ( FIGS. 6 and 10)

In order to obtain a processed transmission signal m ₃ '(i), as is required for carrying out the method according to claim 5, which is namely

• - assumes the value +1 for the duration of a sampling period T _A when the transmission signal '0' to '1' changes (rising edge) and
• - assumes the value 0 for all other samples,

a circuit arrangement is proposed according to the invention, which consists of a flip-flop FF5, a sampling circuit AT5 and a NOR gate NOR5.

The data input of the flip-flop FF5 is connected to the potential high H, the clock input C to the transmission signal m (t). The output signal of the flip-flop FF5 is sampled by the scanner AT5 with the sampling clock T _A. The inverted output signal of the flip-flop FF5 and the inverted output signal of the scanner AT5 control the two inputs of a NOR gate NOR5, the output signal of which resets the flip-flop FF5. The processed transmission signal m ₃ '(i) appears at the output Q of AT5.

Transmit signal processing 4 ( FIGS. 7 and 11)

In order to obtain a processed transmission signal m ₄ '(i), as is required for carrying out the method according to claim 6, which is namely

• - assumes the value +1 for the duration of a sampling period T _A when the transmission signal changes from '1' to '0' (falling edge) and
• - assumes the value 0 for all other samples,

a circuit arrangement is proposed according to the invention, which consists of an inverter I6, a flip-flop FF6, one Scanning circuit AT6 and a NOR gate NOR6 exists.

The data input of the flip-flop FF6 is connected to the potential high H, the clock input C to the transmission signal m (t) inverted in the inverter I6. The output signal of the flip-flop FF6 is sampled by the sampler AT6 with the sampling clock T _A. The inverted output signal of the flip-flop FF5 and the inverted output signal of the scanner AT6 control the two inputs of the NOR gate NOR6, the output signal of which resets the flip-flop FF6. The processed transmission signal m ₄ '(i) appears at the output Q of AT6.

Claims (10)

1. Method for the detection of digital signals at the output of a distorting and disturbed transmission channel with the aid of a cross correlator and a transmission signal processing with the following features:

• a) the cross-correlator is designed so that under the condition of the transmission of a random digital signal with 1-0 equal distribution, the cross-correlation function at a first output proportional to the output signal of the transmission link with interference-free transmission of a single square-wave pulse with the length of the data clock T _m and that Proportionality factor is equal to the ratio of data clock period T _m to sampling clock period T _A ;
• b) at the output of a transmission signal conditioning for the duration of a sampling clock period T _A , when the binary value 1 is transmitted over the transmission link, the value 1, when the binary value 0 is transmitted, the value -1 and for all other sampling values within the data clock period T _m occur, the value 0 generates;
• c) the received signal is given to a first input of the cross-correlator;
• d) the transmission signal is given to the input of the transmission signal conditioning;
• e) the processed transmission signal is given to a second input of the cross-correlator;
• f) the cross-correlator samples the received signal with a sampling clock of a period T _A which is less than half the data clock period T _m .

2. Method for the detection of digital signals at the output of a distorting and disturbed transmission channel with the help of a cross correlator and a transmission signal processing with the following features:

• a) the cross-correlator is designed so that the cross-correlation function at a first output is proportional to the peak-to-peak amplitude of the output signal of the transmission link with interference-free transmission of a single rectangular pulse with the length of the data clock T _m and that the proportionality factor is equal to the ratio of data clock period to sampling clock period divided by the peak-to-peak amplitude of the auto-correlation function of the transmit signal;
• b) at the output of a transmission signal conditioning, the value 1 is generated for the duration of the sampling clock period when the binary value 1 is transmitted over the transmission link, the value -1 when the binary value 0 is transmitted and the value 0 for all other sampling clock periods;
• c) the received signal is given to a first input of the cross-correlator;
• d) the transmission signal is given to the input of the transmission signal conditioning;
• e) the processed transmission signal is given to a second input of the cross-correlator;
• f) the cross-correlator samples the received signal with a sampling clock of a period T _A which is less than half the data clock period T _m .

3. Method for the detection of digital signals at the output of a distorting and disturbed transmission channel with the help of a cross correlator and a transmission signal conditioning with the following features:

• a) the cross-correlator is designed so that under the condition of the transmission of a random digital signal with 1-0 equal distribution, the cross-correlation function at a first output proportional to the output signal of the transmission link with interference-free transmission of a double rectangular pulse with the length of the data clock T _m and that Proportionality factor is equal to the ratio of data clock to sampling clock times 2;
• b) at the output of a transmission signal conditioning, for the duration of the sampling clock period, when the binary value of the signal transmitted over the transmission link changes from 0 to 1, the value +1, when the binary value changes from 1 to 0, the value -1 and for all others Sample clock periods generated the value 0;
• c) the received signal is given to a first input of the cross-correlator;
• d) the transmission signal is given to the input of the transmission signal conditioning;
• e) the processed transmission signal is given to a second input of the cross-correlator;
• f) the cross-correlator samples the received signal with a sampling clock of a period T _A which is less than half the data clock period T _m .

4. Method for the detection of digital signals at the output of a distorting and disturbed transmission channel with the aid of a cross correlator and a transmission signal processing with the following features

• a) the cross-correlator is designed so that the cross-correlation function at a first output is proportional to the peak-to-peak amplitude of the output signal of the transmission path with interference-free transmission of a double rectangular pulse with the length of the data clock T _m and that the proportionality factor is equal to the ratio of data clock to sampling clock times 2 divided by the peak-to-peak amplitude of the autocorrelation function of the transmitted signal;
• b) at the output of a transmission signal conditioning device, the value +1 for the duration of the sampling clock period when the binary value of the signal transmitted over the transmission link changes from 0 to 1, the value -1 when the binary value changes from 1 to 0 and for all other sampling clock periods generates the value 0;
• c) the received signal is given to a first input of the cross-correlator;
• d) the transmission signal is given to the input of the transmission signal conditioning;
• e) the processed transmission signal is given to a second input of the cross-correlator;
• f) the cross-correlator samples the received signal with a sampling clock of a period T _A which is less than half the data clock period T _m .

5. Method for the detection of data signals at the output of a distorting and disturbed transmission channel with the aid of a cross-correlator and a transmission signal preparation with the following features:

• a) the cross-correlator is designed so that the cross-correlation function at a first output is proportional to the output signal of the transmission link with interference-free transmission of a rising edge and that the proportionality factor is equal to the reciprocal of the mean value of the output signal of the transmission signal processing;
• b) at the output of a transmission signal conditioning device, the value +1 is generated for the duration of the sampling clock period, when the binary value of the signal transmitted over the transmission link changes from 0 to 1, and the value 0 for all other sampling clock periods;
• c) the received signal is given to a first input of the cross-correlator;
• d) the transmission signal is given to the input of the transmission signal conditioning;
• e) the processed transmission signal is given to a second input of the cross-correlator;
• f) the cross-correlator samples the received signal with a sampling clock of a period T _A which is less than half the data clock period T _m .

6. Method for the detection of digital signals at the output of a distorting and disturbed transmission channel with the help of a cross correlator and a transmission signal conditioning with the following features:

• a) the cross-correlator is designed so that the cross-correlation function at a first output is proportional to the output signal of the transmission link with interference-free transmission of a falling edge and that the proportionality factor is equal to the reciprocal of the mean value of the output signal of the signal conditioning;
• b) at the output of a transmission signal conditioning device, the value +1 is generated for the duration of the sampling clock period when the binary value of the signal transmitted over the transmission link changes from 1 to 0, and the value 0 for all other sampling clock periods;
• c) the received signal is given to a first input of the cross-correlator;
• d) the transmission signal is given to the input of the transmission signal conditioning;
• e) the processed transmission signal is given to a second input of the cross-correlator;
• f) the cross-correlator samples the received signal with a sampling clock of a period T _A which is less than half the data clock period T _m .

7. Circuit arrangement for a transmission signal processing for performing the method according to claim 1 or 2 with an inverter (I1), a first flip-flop circuit (FF1), a second flip-flop circuit (FF2), a first sampling circuit ( AT1), a second sampling circuit (AT2), a first NOR gate (NOR1) and a second NOR gate (NOR2), which are connected as follows:
Clock input (C) of flip-flop circuits (FF1) and (FF2) with the data clock T _m inverted in the inverter (I1),
Data input (D) from flip-flop circuit (FF1) with the potential high (H),
Data input (D) from flip-flop circuit (FF2) with the transmission signal m (t),
Reset input (R) of flip-flop circuit (FF1) with the output of the NOR gate (NOR1),
Set input (S) of flip-flop circuit (FF2) with the output of the NOR gate (NOR2),
Output signal from flip-flop circuit (FF1) to the data input (D) of the scanner (AT1),
inverted output signal from flip-flop circuit (FF1) with input of the NOR gate (NOR1),
inverted output signal from flip-flop circuit (FF2) to the data input (D) of the scanner (AT2),
Output signal from flip-flop circuit (FF2) with input of the NOR gate (NOR2),
Clock inputs (C) of the samplers (AT1) and (AT2) with sampling clock T _A ,
inverted output signal of the scanner (AT1) with input of the NOR gate (NOR1),
inverted output signal of the scanner (AT2) with input of the NOR gate (NOR2),
then the mantissa (M) appears as the output signal of the scanner (AT1) and the sign (V) of the processed transmission signal m ₁ '(i) as that of the scanner (AT2). 8. Circuit arrangement for a transmission signal processing for performing the method according to claim 3 or 4 with an inverter (I3), a first flip-flop circuit (FF3), a second flip-flop circuit (FF4), a first sampling circuit ( AT3), a second sampling circuit (AT4), a first NOR gate (NOR3) and a second NOR gate (NOR4) and an OR gate (OR), which are connected as follows:
Clock input (C) of flip-flop circuit (FF3) with the transmission signal inverted in the inverter (I3),
Clock input (C) of flip-flop circuit (FF4) with the transmission signal,
Data input (D) from flip-flop circuit (FF3) and (FF4) with the potential high (H),
Output signal at the output (Q) of the flip-flop circuit (FF3) with the data input (D) of the sampling circuit (AT3),
Output signal at the output (Q) of the flip-flop circuit (FF4) with the data input (D) of the sampling circuit (AT4),
Reset input (R) of flip-flop circuit (FF3) with the output of the NOR gate (NOR3),
Reset input (R) of flip-flop circuit (FF4) with the output of the NOR gate (NOR4),
inverted output signal at the output () of flip-flop circuit (FF3) with the input of the NOR gate (NOR3),
inverted output signal at the output () of flip-flop circuit (FF4) with the input of the NOR gate (NOR4),
Clock input (C) from sampling circuit (AT3) with sampling clock T _A ,
Clock input (C) from sampling circuit (AT4) with sampling clock T _A ,
Output signal at the output (Q) of the sampling circuit (AT3) and output signal at the output (Q) of the sampling circuit (AT4) with the inputs of the OR gate (OR),
inverted output signal at the output () of the sampling circuit (AT3) with the input of the NOR gate (NOR3),
inverted output signal at the output () of the sampling circuit (AT4) with the input of the NOR gate (NOR4),
then the mantissa (M) appears at the output of the OR gate (OR) and the sign (V) of the processed transmission signal m ₂ '(i) at the output of the sampling circuit (AT3). 9. Circuit arrangement for a transmission signal processing for performing the method according to claim 5 with a flip-flop (FF5), a sampling circuit (AT5) and a NOR gate (NOR5), which are connected as follows:
Clock input (C) of flip-flop (FF5) with the transmission signal m (t),
Data input (D) from flip-flop (FF5) with potential high (H),
Output (Q) of flip-flop (FF5) with data input (D) of the sampling circuit (AT5),
Reset input (R) of flip-flop (FF5) with the output of the NOR gate (NOR5),
Inputs of the NOR gate (NOR5) with the inverted output signals from flip-flop (FF5) and sampling circuit (AT5), clock input (C) from sampling circuit (AT5) with sampling clock T _A , then appears at the output (Q) of the sampling circuit (AT5) the processed transmission signal m ₃ '(i). 10. Circuit arrangement for a transmission signal processing for performing the method according to claim 6 with an inverter (I6), a flip-flop (FF6), a sampling circuit (AT6) and a NOR gate (NOR6), which are connected as follows:
Clock input (C) of flip-flop (FF6) with the transmission signal m (t) inverted in the inverter (I6),
Data input (D) from flip-flop (FF6) with potential high (H),
Output (Q) of flip-flop (FF6) with data input (D) of the sampling circuit (AT6),
Reset input (R) of flip-flop (FF6) with the output of the NOR gate (NOR6),
Inputs of the NOR gate (NOR6) with the inverted output signals from flip-flop (FF6) and sampling circuit (AT6), clock input (C) from sampling circuit (AT6) with sampling clock T _A , then appears at the output (Q) of the sampling circuit (AT6) the processed transmission signal m ₄ '(i).