Method for the correction of phase and amplitude errors in COFDM signals.

ES 2 184 745 T3

DESCRIPTION

Procedure for the correction of phase and amplitude errors in COFDM signals.

The invention relates to a method for the phase and amplitude correction of a broadband reception signal with the aid of reference signals according to the preamble of claim 1 of the patent and to a circuit arrangement for carrying out the method.

It is known for the publication “Entwicklungstendenzen im Fernsehrundfunk”, published in telekom praxis, edition 12/92, pages 32 to 38, which is used worldwide on a digital terrestrial television system. For digital transmission, the COFDM (Coded Orthogonal Frequency Division Multiplex) procedure must be used in this television system. In this way, a high transmission quality is achieved with a more efficient use of the frequency ranges used for transmission.

The COFDM transmission method is known from WO 88/00417. The information to be transmitted is converted in the first place in an analog-digital way, if the information was present in an analog form. Next, the digital information is encoded with a channel encoder, with which an error protection code is added to the information. Finally, the channel-coded information of a plurality of individual carrier frequencies is modulated, the temporal sequence of the information being encrypted. In this way, a high protection against time and frequency selective interference is achieved.

A method for regulating the frequency of local oscillators of a receiver is known from WO 92/10043. The received digital television data is transmitted in this case according to the OFDM (Orthogonal Frequency Division Multiplex) procedure. In addition to the digital television data, two pilot signals are still transmitted in each transmission frame, which have a different constant frequency. To regulate the frequency of the local oscillators in the receiver, the beginning of an emission time division is detected, after which the position of the two pilot signals in the transmission box is determined and the difference between the signals is then calculated. two pilot signals as a function of time. From this, a regulation quantity can be derived for the local oscillators.

A procedure for correcting the frequency response is known from the previously unpublished application P 42 40 609.9 of the Applicant Firm. In the case of terrestrial transmission, digital signals (DAB) are used. In this case, distortions of the amplitudes and of the selective phases of the frequency of a transmission signal COFDM modulated in the differential phases are detected from a control signal, which calculates a corresponding correction characteristic curve and weights the spectrum of the signal from transmission with the characteristic curve of the correction. In this case, it does not matter whether the detected distortions have appeared on the transmission path or in the structural groups of the emitter or receiver. The procedure also makes it possible to estimate the reliability of the reception signals.

In this procedure, it is a drawback that for the calculation of the characteristic curve of the correction, in the case of distortions that fluctuate in time and frequency to a great extent, a high cost of calculation is necessary.

Another drawback is that the method can only correct a part of the distortions possible in the case of signals not modulated in the differential phases.

A method for the phase and amplitude correction of a broadband reception signal is known from EP 0 365 431 A1. In this method, a transmission of the radio signal is carried out by means of symbols. A protection interval is added to the signal transmission.

Therefore, the object of the invention is to indicate a procedure for the correction of a COFDM transmission signal not modulated in the differential phases. The cost of calculation necessary for this must be as low as possible.

This task is solved according to the invention through the characteristic features of claim 1 of patent 1.

The method according to claim 1 has the advantage that, by virtue of the reference signal transmitted in each transmission frame, the corrections of the amplitudes and phases of the data signals can be interpolated as a function of time and time. frequency for the entire transmission frame. This calculation can be performed for the individual carrier frequency or for a group of several individual carrier frequencies in a transmission frame. Another window is that a maximum probability decision is made regarding the amplitude and phase of each individual carrier frequency. In this way, the reduced corrections of the phases or of the amplitudes are carried out immediately, recognizing in the case of large corrections that the reliability of the reception signal is reduced.

Furthermore, the procedure has the advantage that the amplitude correction is adapted to the modified reception ratios. Through the consideration of a number of signals received or lately, a continuous curve of the amplitude of the corrected reception signal is obtained.

The method according to claim 2 has the advantage that by virtue of the known time course of the distortion of the amplitudes and of the phases, a correction of the amplitudes and of the phases can also be calculated according to a characteristic correction curve of higher order.

The procedure according to claim 3 of the patent has the advantage that through a single multiplication of the spectrum of short duration of the reception signal by the characteristic curve of the correction, the correction of the amplitudes and the correction of the phases.

The procedure according to claim 4 has the advantage that no transformation or retransformation of Fou2 is necessary.

ES 2 184 745 T3 rier itself for the selective correction of the frequency of the amplitude and the phase of the reception signal. In this way, considerable computational and circuit technical expense can be avoided.

The use of the method according to claim 5 of the patent has the advantage that a correction of television transmissions is made possible. Furthermore, it is advantageous that the correction values ​​for the correction of the amplitudes can be easily calculated, since the amplitude of the spectrum during a transmission of digital television data according to the COFDM procedure adopts in an ideal way discrete values.

The circuit arrangement according to patent claim 6 has the advantage that the calculation of the correction characteristic curve, the correction of the individual carrier frequencies and the maximum probability decision can be carried out through a control unit, which it must be provided anyway for the control of the receiver. The additional circuit expense for correcting the reception signal is thus limited to a relatively inexpensive and highly integrated memory.

The invention is described and explained in detail below. In this case:

Figure 1 shows a possible curve of the individual carrier frequencies used for transmission according to the COFDM procedure as a complex indicator.

Figure 2 shows a possible phase correction of an individual carrier frequency.

Figure 3 shows possible phase differences of the received individual carrier phases and of the originally modulated phases on the individual carrier frequencies.

Figure 4 shows possible phase differences of the received individual carrier phases and of the originally modulated phases on the individual carrier frequencies after phase correction, and

Figure 5 shows a circuit arrangement for the realization of the correction.

In the following example, the COFDM procedure is used for the transmission of digital television data. In this case, the individual carrier frequencies are modulated with the 4-PSK television data (Phase - Shift - Keying = Phase Shift Modulation). In the receiver is received in each transmission frame after a synchronization signal SYNC for the time synchronization of the receiver, in which the amplitudes of all the individual carrier frequencies are zero, a reference signal REF with a reference phase PR already A data block followed with n data signals D1 to Dn, which are modulated in phase with respect to the reference phase PR. Through the 4-PSK modulation of each individual carrier frequency, two binary states, for example 01, ie one dibit, can be transmitted with each individual carrier frequency.

Figure 1 shows the transmission signal, where the individual transmitted carrier frequencies are represented as complex indicators with amplitude and phase as a function of time t and frequency f. The frequencies f1 to fm are the individual carrier frequencies of the COFDM transmission procedure, which are arranged equidistant on the frequency axis f. The signals SYNC for the synchronization of the receiver, REF for the transmission of the reference phase PR and D1 to Dn for the transmission of the television data are arranged equidistant on the axis of time t and form a transmission frame. By virtue of the modulation of the phases of the data signals D1 to Dn, the individual carrier frequencies can be arranged out of phase with respect to the reference phase PR free of modulation of the reference signal REF, which is expressed through corresponding turns of complex indicators. The reference phase PR is set again for each frame of transmission through the reference signal REF. Therefore, the reference signal REF cannot have any phase distortion. In this case, it is not necessary that all the individual carrier frequencies of the reference signal REF have the same reference phase PR. The data signals D1 to Dn show phase distortions most of the time according to figure 1. E^orThese are caused through the transmission path and / or through structural groups of the emitter and receiver. For greater clarity, the representation of the possible distortions of the amplitudes is dispensed with.

For demodulation in the receiver, the phase difference between the transmitted reference phase PR and the phases of the individual carrier frequencies of the same transmission frame is calculated. Next, this phase difference is decoded into binary values, so that the transmitted information is obtained back.

Due to the distortions, the phase difference in the case of a 4-PSK modulation is not exactly a multiple of an integer number of 90 °, but for example 1000 To be able to decode, also in the case of distortions, which are very small , still binary states from the phase differences, a decision of maximum probability is carried out.

In this way insignificant phase deviations are corrected and the transmission system has a certain tolerance to errors. For example, phase values ​​in an interval between 45^° and 135^° on the receiver at a phase of 90^°. Figure 2 illustrates this graphically. The ideal phase Pi, which is recognized in the case of the maximum probability decision, does not correspond to the position of the phases Pe of the individual carrier frequency received. The possible limits for the maximum probability decision are represented with dashed lines.

Figure 3 shows the phase deviation P

ES 2 184 745 T3 of the individual carrier frequencies received from one of the four possible phase positions in the case of 4-PSK modulation. By definition, the reference signal REF of a first transmission frame, which is received at time t1, does not present any phase distortion. The phase distortions of the data signals D1 to Dn are represented as a function of frequency. Since during SYNC sync signals all individual carrier frequencies have zero amplitude, no signal can be determined. The reference signal REF of the second transmission frame defines again the reference phase PR for the following data signals D1 to Dn, assuming that the reference phase PR is invariable with time. However, the reference phases PR (t1) and PR (t2) received in different transmission frames may have a phase difference P due to phase distortions. This phase difference is represented in FIG. 3 as a function of the frequency for the reference signal REF, which is received at time t2 in the second transmission frame.

Especially by virtue of the time-dependent oscillations of the paraometers, there are deviations ΔΡ of the received phases, as shown in figure 3, which are not conditioned by the modulation and which are not within the described correctable error tolerance. above, and despite the error tolerance, errors may appear during decoding. To avoid this, a phase correction is carried out on the reception signal. In this phase correction, the control unit calculates the phase difference dPR between at least two reference phases PR (t1) and PR (t2) of the same individual carrier frequency. In general, for this purpose, reference signals REF of consecutive transmission frames are evaluated in the time followed (1).

dPR = PR (t1) - PR (t2) (1)

If this phase difference dPR is, in terms of the absolute value, greater than a threshold value, it is recognized that a correction of the phases is necessary, and a quotient Q is formed from the phase difference dPR and the time dTR, that is between the reception of the two reference signals REF (t1) and REF (t2), following (2).

^Q = dTR ⁽²⁾

This Q ratio now indicates the time-dependent linear modification of the PR reference phase. In this way, a predictable linear phase error Pfi can be calculated for each data signal Di as a function of the instant of its emission according to (3).

Pfi = Q · (T (REF, t1) - T (Di)) (3)

By subtracting the phase errors Pfi calculated in this way from the phases of the individual carrier frequencies in FIG. 3, time-dependent linear phase errors can be eliminated.

Figure 4 shows the phase distortions Pn, adjusted according to the calculated portion, which can be corrected by virtue of the following maximum probability decision, so that no transmission error appears. In this case, the correction of the phases can have, in general, different repercussions. As an example, consider the individual carrier frequency f1. Following the representation shown in figure 3, this individual carrier frequency for the data signal D1 does not present any phase distortion Δp and presents a reduced phase distortion ΔΡ for the data signal Dn. The distortion of the phases ΔΡ of the reference signal of the second transmission frame REF (t2) exhibits, in comparison with the reference signal of the first transmission frame REF (t1), already an average distortion of the phases. The distortion of the phases ΔΡ of the individual carrier frequency f1 therefore increases with time t, but not linearly. Therefore, by virtue of the phase correction Pfi calculated linearly, for the individual carrier frequency f1, also after phase correction, a distortion of the phases Pn results, which is, however, more reduced, as represented in figure 4. For the data signal D1 a distortion of the phases Pn occurs, which is within the error tolerance and therefore does not lead to a transmission error. For the data signal Dn there is no distortion of the phases Pn of the individual carrier frequency f1, since through the linear correction the distortion of the phases P that was originally present can be eliminated.

The time dependence of the distortion of the phases P can be analyzed through a control unit MP, so that an optimal mathematical model can be found for the calculation of the correction values ​​of the phases Pfi for the interpolation of the phases of the data signals D1 to Dn, which do not, of course, have to be linear.

There is the possibility of incorporating at the same time in the calculation also PR reference phases of previously received transmission frames and carry out a weighting of these PR reference phases as a function of time. In this way a consideration of temporally non-linear phase distortions is possible. Equations (1) to (3) must then be modified accordingly.

In order not to have to perform the calculation described above for each individual carrier frequency f1 to fm, it is conceivable to perform the above calculation, for example, only for one out of every two or for one out of every 16 individual carrier frequencies and interpolate the missing Q ratios to starting from the Q ratios calculated as a function of the individual carrier frequency. In this way, channel capacity can be released in the reference signal REF, which can be used for the transmission of additional data.

According to the phase correction described above, the amplitudes of the individual carrier frequencies are also corrected. The

ES 2 184 745 T3 amplitudes of the reference signals are temporarily stored in a memory SPE, the amplitudes of the individual carrier frequencies are set again by virtue of the interpolated correction values ​​of the amplitudes. The interpolation of the amplitude can also be carried out in this case both as a function of time and also as a function of frequency.

From the correction of the amplitudes and the phases as a function of the frequency, a complete correction characteristic curve is then calculated, by which the short-term spectrum SP of the reception signal is multiplied. The correction characteristic curve is composed of the amplification value for the amplitude and the correction value of the phases Pfi for each individual carrier frequency. The multiplication can be performed at a place in the signal path, where the short duration spectrum is already present anyway, for example after Fourier FET transformation.

As an alternative to this, a correction indicator can also be calculated for an individual carrier frequency, which results from the indicator of the difference between two reference indicators of an individual carrier frequency. AND^orThis is multiplied as a function of the reception time of the signal to be corrected by a scale factor and added to the indicator originally received, to obtain the indicator corrected in phase and amplitude for the individual carrier frequency.

In order not to have to recalculate this characteristic correction curve for each transmission frame, it is stored in a SPE memory and can be used for the correction of data signals from several transmission frames. A new calculation of the correction characteristic curve can be started through the user or through the MP control unit. To this end, the MP control unit analyzes the corrected amplitudes and the corrected phases of the individual carrier frequencies. As soon as the amplitudes or phases of the reference signals REF oscillate strongly, a new calculation of the correction characteristic curve is necessary. Figure 5 shows a circuit arrangement for the realization of the phase correction and the amplitude of the reception signal. After the digital reception signal of a first transmission frame was present as a short duration spectrum through a fast Fourier transform FFT in the spectral range, the control unit MP analyzes this spectrum SP.

The reference signal REF and the data signals D1 to Dn are temporarily stored in a memory SPE. In the same way, the signals of a second transmission frame are temporarily stored in the SPE memory, which is transmitted after the first. Next, from the reference signals REF of the two transmission frames, the correction characteristic curve is calculated through the MP control unit, which is also stored in the SPE memory. After the individual carrier frequencies of the first transmission frame have been multiplied by the correction characteristic curve, the individual carrier frequencies are fed to the highest probability decision through the MP control unit and, finally, are further processed. on the DMOD demodulator. If it is checked during the maximum probability decision that the phases of the received individual carrier frequencies differ greatly from the possible modulation phases (in 4-PSK modulation: 90 °, 180 °, 270 °, 360 °), it is recognized that the information is not reliable.

Then, in the SPE memory, the data received from the first transmission frame is overwritten by the data received from the third next transmission frame. The reception data of the second transmission frame are then multiplied by the correction characteristic curve already calculated previously and are fed to the decision of maximum probability. If it is found in this decision via the MP control unit that the oscillations of the amplitudes or the phase distortions are large, then a new correction characteristic curve is calculated via the MP control unit. This new correction characteristic curve is calculated as a function of the reference signals REF of the second and third transmission frame. The new correction characteristic curve replaces the one stored up to now in the SPR memory and is used to correct the reception data after the first transmission frame. As soon as necessary, a new correction characteristic curve is calculated again.

If non-linear phase distortions must also be corrected, the previously transmitted REF reference signals must also be stored in the SPE memory and taken into account by the MP control unit during the calculation, especially of the correction values. of the phases.