Broadcast receiver

The present invention relates to a receiver for broadcast signals, e.g. DAB, FM, DVB, and digital or analog short-wave RF broadcast signals, e.g. DRM or AM signals, in particular to receivers for digitally modulated signals and/or multi-transmission system receivers for at least two different broadcast signals.

State of the art multi-transmission system receivers that can be switched between different RF broadcast signals usually comprise separate processing blocks each of which is directly related to a specific type of RF broadcast signal in-between the input stages of a RF receiver, like antenna, front-end block, A/D converter and IQ generator, and the output stage, like D/A converter. Such individual circuits respectively perform e.g. the digital frequency demodulation which is needed for the reception of a frequency modulated signal, the digital frequency adjustment of the receiver in case of reception of AM or short-wave broadcast signals and the digital frequency adjustment and digital gain control of the receiver in case of DAB reception. In these solutions, the AFC (Automatic Frequency Control) and the AGC (Automatic Gain Control) as well as the digital filtering of a complex baseband signal, the sampling rate decimation and the frequency demodulation are separate blocks which leads to high realization costs.

Therefore, it is the object underlying the present invention to provide a receiver for broadcast signals that has lower realization costs than comparable state of the art receivers.

According to the present invention this object is solved with a receiver according to independent claim 1. Preferred embodiments thereof are defined in dependent claims 2 to 19.

According to the present invention an efficient realization of a receiver uses a polyfunctional circuit that receives the complex baseband signal generated from a respective RF broadcast signal which can be switched into different modes dependent on the broadcast signals that should be receivable. In a first mode (mode A) the polyfunctional circuit performs the digital frequency demodulation which is needed for the reception of a frequency modulated (FM) signal. According to a preferred embodiment (mode A- 1 ) additionally a digital neighbor channel suppression and sampling rate decimation is performed by the polyfunctional circuit. In a second mode (mode B) the polyfunctional circuit performs the digital frequency adjustment (AFC=automatic frequency control) of the receiver which can be used for the reception of digitally or analog modulated signals, like DAB, DVB, digital short-wave (DRM) or AM reception. In a third mode (mode C) the polyfunctional circuit performs the digital frequency adjustment and the digital gain control (AGC=automatic gain control) of the receiver which is needed in case of DRM, DAB or DVB reception. The polyfunctional circuit has at least the functionality to perform either mode C, or mode C and mode A and/or mode B, or mode A and mode B.

According to the present invention the digital AGC and the digital AFC are realized in one circuit so that the same hardware of the AFC can also be used for the AGC. Furtheron, the digital filtering of complex signals and a sampling rate decimation are included in the polyfunctional circuit to perform a linear phase neighbor channel suppression and noise shaping. Therefore, no additional adders or multipliers are needed for the digital filtering and sampling rate decimation. Since all functionalities needed according to the prior art are combined in one polyfunctional circuit that can be switched into different modes according to the present invention, hardware parts that have to be used for reception of all or several of the RF broadcast signals, i.e. that are not used in parallel, but only during reception of one kind of broadcast signal are shared so that basically the same elements do not have to be included twice within the receiver, but only once.

In a preferred embodiment the circuit combines all three modes for the reception of a digital signal like DAB, DRM or DVB and the reception of analog signals like FM or AM. Furtheron, this polyfunctional circuit can be realized with only three adders and without multipliers which is a quite big advantage in comparison with state of the art multi-transmission system receivers.

The circuit according to the present invention is based and works with the CORDIC-algorithm. The CORDIC-algorithm was published by Jack E. Volder, “The CORDIC Trigonometric Computing Technique”, Institute of Radio Engineers, IRE Transactions on Electronic Computers, Vol. EC-8, pp. 330-334, 1959. To get a better understanding of the functionality of the polyfunctional circuit according to the present invention, in the following a short overview of the CORDIC-algorithm is given.

The CORDIC-algorithm is an iterative algorithm to rotate a complex vector. The algorithm can be realized very simple in hardware, because only shift and add operations are needed.

The CORDIC-algorithm rotates a complex input vector z  ( k ) = x  ( k ) + jy  ( k ) = x ( 0 )  ( k ) + jy ( 0 )  ( k ) = z ( 0 )  ( k ) ( 1 )

in a first iteration:

x (1) ( k )=σ (0) ( k )· y (0) ( k )

y (1) ( k )=−σ (0) ( k )· x (0) ( k )  (2)

with σ (i) ( k )=±1: micro rotation direction.

Beginning from the second to the N th iteration, another formula is used

x (i+1) ( k )= x (i) ( k )+σ (i) ( k )·δ (i) ·y (i) ( k )

y (i+1) ( k )= y (i) ( k )−σ (i) ( k )·δ (i) ·x (i) ( k )  (3)

with δ (i) =2 1−i : step with of the i th iteration.

The factor δ (i) =2 1−i can be realized simply by a shift operation.

The CORDIC-algorithm output signal after the N th iteration w(k)=z (N) (k) is described by the following formula: w  ( k ) = z ( N )  ( k ) = z ( 0 )  ( k ) ·  - j  ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) · ∏ i = 1 N - 1     1  ( δ ( i ) ) 2 = z ( 0 )  ( k ) · K ·  - j  ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) ( 4 )

By splitting this formula into amplitude and phase follows:  w  ( k )  =  z ( N )  ( k )  =  z ( 0 )  ( k )  · ∏ i = 1 N - 1     1 + ( δ ( i ) ) 2 = K ·  z ( 0 )  ( k )  ( 5 )

with ϕ rot ( i ) = { π 2 arctan  ( δ ( i ) )     for     i = 0 i > 0 :

micro rotation angle of the i th iteration step and φ(z (0) (k)) angle of the input vector z (0) (k).

The rotating angles φ rot (i) are constant and can therefore be written in a ROM.

The amplitude is amplified by a constant factor K = ∏ i = 1 N - 1     1 + ( δ ( i ) ) 2 ,

which is independent on the rotation itself.

The CORDIC-algorithm can be switched between two modes: rotation mode and vector mode.

In rotation mode, the micro rotation direction σ (i) is generated from the phase error φ error (i) ϕ error ( 0 )  ( k ) = ϕ  ( k ) 

 ϕ error ( i + 1 )  ( k ) = ϕ  ( k ) + ∑ l = 0 i  σ ( l )  ( k ) · ϕ rot ( l ) ( 6 )

by the following formula:

σ (i) ( k )=−sign(φ error (i) ( k ))  (7)

The following example describes the CORDIC-algorithm in rotation mode: Lets suppose for the example, the input vector is z ( 0 )  ( k ) =  j     π 3

and the rotation phase is ϕ  ( k ) = 2     π 3 ≡ 120

degrees. The CORDIC-algorithm has N=7 iteration steps. The following table shows the phase error φ error (i) after each iteration in degrees:

FIG. 12 shows the vector z (i) after each iteration. The input vector z ( 0 )  ( k ) =  j     π 3

is shown by a small circle. The large circle with the amplitude 1 is only for orientation purpose.

In rotation mode the algorithm rotates the input vector z (0) (k) by a phase value ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) .

With increasing number of iterations N the factor φ error (N) (k) can be neglected and therefore with ϕ error ( N )  ( k ) = ϕ  ( k ) + ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) ≈ 0 ( 8 )

follows ϕ     ( k ) ≈ ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) . ( 9 )

Together with equation (5) ϕ     ( z ( N )  ( k ) ) = ϕ  ( z ( 0 )  ( k ) ) - ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) ( 10 )

follows:

φ( z (N) ( k ))=φ( z (0) ( k ))+φ( k )  (11)

and finally with equation (4):

w ( k )= z (N) ( k )= z (0) ( k )· K·e jφ(k)   (12)

The vector w(k) is generated from the input vector z(k) by an amplification with a constant K and a phase rotation by the phase φ(k).

In vector mode the angle of each input vector φ(z (0) (k)) is calculated. Therefore the input vector z (0) is rotated to the x-axis and the micro rotation angles φ rot (i) are added to φ(z (0) ). For the rotation of the input vector z (0) to the x-axis, the micro rotation direction σ (i) is generated from the imaginary component y (i) (k) using the following formula:

σ (i) =sign( y (i) )=sign(imag( z (i) ))  (13)

The angle φ sum (i) (k) of the vector z (i) is calculated using the following equation:

φ sum (0) =0

The following example describes the CORDIC algorithm in rotation mode: Lets suppose for an example, the input vector is again z (0) (k)=e jπ/3 and the CORDIC-algorithm has N=7 iteration steps. The following table shows the phase sum φ sum (i) (k) in degrees:

FIG. 13 shows the vector z (i) after each iteration. The input vector z (0) (k)=e jπ/3 is shown by a small circle. The large circle with the amplitude l is only for orientation purpose.

With an increasing number of iterations N the phase ϕ sum ( N )  ( k ) = ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) ( 15 )

of the output vector z (N) is getting neglectable: ϕ sum ( N )  ( k ) = ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) = ϕ  ( z ( N )  ( k ) ) ≈ 0 ( 16 )

Together with equation (5) ϕ  ( z ( N )  ( k ) ) = ϕ  ( z ( 0 )  ( k ) ) - ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) ≈ 0 ( 17 )

follows: ϕ  ( z  ( k ) ) = ϕ  ( z ( 0 )  ( k ) ) ≈ ∑ i = 0 N - 1  σ ( i )  ( k ) · ϕ rot ( i ) ( 18 )

Furtheron, the digital AFC and the digital frequency demodulation are both state of the art. They are, for example, described in EP 0 486 095 B1 “Digital Receiver”. Also the digital filtering of real signals is known, e.g. from EP 0 7412 478 A2 “Circuit for Performing Arithmetic Operations in a Demodulator”.

The present invention and its embodiments will be better understood from the following detailed description thereof taken in conjunction with the accompanying drawings, wherein:

FIG. 1 shows a function block of a receiver according to the present invention that works in mode A, i.e. that performs a frequency demodulation,

FIG. 2 shows the function blocks of the polyfunctional circuit according to the present invention in the embodiment shown in FIG. 1,

FIG. 3 shows the function blocks of a receiver according to the present invention that works in mode A- 1 , i.e. that performs a digital linear phase filtering for noise shaping and neighbor channel suppression, sampling rate decimation with a decimation factor of 2 and the digital frequency demodulation,

FIG. 4 shows the function blocks of the polyfunctional circuit according to the present invention in the embodiment shown in FIG. 3,

FIG. 5 shows the function blocks of a receiver according to the present invention that works in mode B, i.e. that performs the digital frequency adjustment,

FIG. 6 shows the function blocks of the polyfunctional circuit according to the present invention in the embodiment shown in FIG. 5,

FIG. 7 shows the function blocks of a receiver according to the present invention that works in mode C, i.e. that performs the digital frequency adjustment and the digital gain control of the receiver,

FIG. 8 shows the function blocks of the polyfunctional circuit according to the present invention in the embodiment shown in FIG. 7,

FIG. 9 shows the realization of the averaging low pass filter shown in FIG. 8,

FIG. 10 shows the realization of the polyfunctional circuit according to the present invention that can perform modes A, B and C,

FIG. 11 shows the realization of the polyfunctional circuit according to the present invention to perform modes A- 1 , B and C,

FIG. 12 shows a chart for explaining the rotation mode of the CORDIC-algorithm,

FIG. 13 shows a chart for explaining the vector mode of the CORDIC-algorithm,

FIG. 14 is a table showing the states of switches, adders/subtracters and shift registers for a mode A of the circuit depicted in FIG. 10,

FIG. 15 is a table showing the states of switches, adders/subtracters and shift registers for a mode B of the circuit depicted in FIG. 10,

FIG. 16 is a table showing the states of switches, adders/subtracters and shift registers for a mode C of the circuit depicted in FIG. 10,

FIG. 17 is a table showing shift and add operations as described in connection with FIGS. 8 and 9, and

FIG. 18 is a table showing the states of switches, adders/subtracters and shift registers for a mode A- 1 of the circuit depicted in FIG. 11 .

FIG. 1 shows the functionality of a receiver according to the present invention for mode A in which the polyfunctional circuit 5 performs a frequency demodulation. A FM broadcast signal is received with an antenna 1 and converted with a front-end block 2 , an A/D converter 3 and an IQ generator 4 into a complex baseband signal z(k). This complex baseband signal z(k) might also undergo a neighbor channel suppression within the IQ generator 4 before it is fed to the polyfunctional circuit 5 according to the present invention that performs the digital frequency demodulation and outputs a digital multiplex stereo signal MPX(k) which gets demultiplexed by a stereo demultiplexer 6 A before it gets converted into an analog audio signal by a D/A converter 7 .

FIG. 2 shows that the frequency demodulation within the polyfunctional circuit 5 is done by a differentiation of the phase of the frequency modulated carrier. Therefore, according to the present invention, the phase value φ(z(k)) for every sample of the complex baseband signal z(k) gets calculated by the CORDIC-algorithm in vector mode with N iteration steps in a phase calculator 5 a before the phase value φ(z(k)) differentiated by a differentiator 5 b to perform the frequency demodulation. The differentiator then outputs the stereo multiplex signal MPX (k) or STMUX (k), as it is also shown in FIG. 2 which shows the two function blocks of the polyfunctional circuit 5 representative for the FM demodulation.

According to the present invention, the frequency demodulation is done by a differentiation of the phase value φ(z(k)) of the frequency modulated complex baseband signal z(k) using the following approximation: MPX  ( K ) =   t   kT  ϕ z  ( t ) ≈    ϕ  ( z  ( k ) ) - ϕ  ( z  ( k - 1 ) ) T ( 19 )

FIG. 3 shows the functionality of a receiver according to the present invention in mode A- 1 in which the complex baseband signal does not undergo a neighbor channel suppression within the IQ generator 4 , but in which the neighbor channel suppression is formed within the polyfunctional circuit 5 . Therefore, as it is shown in FIGS. 3 and 4 that shows the functionality of the polyfunctional circuit 5 shown in FIG. 3, the complex baseband signal generated by the IQ generator 4 is fed to the polyfunctional circuit 5 in which it first passes through a FIR filter and down conversion stage 5 c before the signal processing is performed as described in connection with FIGS. 1 and 2. The FIR filter performs the neighbor channel suppression and consists of two real FIR lowpass filters, one for the inphase component and one for the quadrature component of the complex baseband signal. After the filtering with e.g. lowpass filters comprising nine taps a sampling rate decimation with a decimation factor of 2 is performed before the frequency demodulation is again done by a differentiation of the phase of the frequency modulated carrier as it is described above.

FIG. 5 shows the different functionality of a multi-transmission system receiver according to the present invention in case an AIM, DRM, DAB or DVB, i.e. a digitally or analog modulated broadcast signal is received by the antenna 1 and processed by the front-end block 2 , the A/D converter 3 and the IQ generator 4 which again might include a neighbor channel suppression before the so generated complex baseband signal z(k) is fed to the polyfunctional circuit 5 according to the present invention which then feeds its output signal w(k) to a short-wave processing circuit 6 B that includes a neighbor channel suppression, a demodulation, an adjustment and additional processings or a DAB processing circuit 6 C which includes a FFT, de-interleaving, a Viterbi-decoder, an adjustment, and a MPEG-decoding before the analog audio signal is finally generated by the D/A converter 7 . For the automatic frequency control the frequency offset Δf(k) is fed from the short-wave processing circuit 6 B or the DAB processing circuit 6 C to the polyfunctional circuit 5 according to the present invention. As it is shown in FIG. 5, the polyfunctional circuit 5 performs an automatic frequency control to synchronize the receiver with the transmitter in frequency. The automatic frequency control corrects the frequency offset between the transmitter and the receiver that is calculated by the synchronization algorithm within the short wave processing circuit 6 B or the DAB processing circuit 6 C

FIG. 6 shows the functionality of the polyfunctional circuit 5 performing the automatic frequency correction according to the present invention. The digital baseband signal z(k) is fed to the CORDIC-algorithm block 5 a which again comprises N iteration steps and which this time works in rotation mode and rotates every input sample z(k) by a phase value φ(k) which is generated from the frequency offset Δf(k). The phase value φ(k) is calculated from the frequency Δf(k) using the well known formula (21) ϕ  ( t ) = ∫ - ∞ t  Δ     ω  ( τ )      τ = 2  π  ∫ - ∞ t  Δ     f  ( τ )      τ ( 21 )

Therefore, the frequency offset Δf(k) received from the short-wave processing circuit 6 B is fed into an integrator 5 d that calculates the phase φ(k) according to the above formula (21) and feeds it to the CORDIC-algorithm block 5 a.

In a discrete system this formula can be written as follows: ϕ  ( k ) = T  ∑ 1 = - ∞ k  Δ     ω  ( 1 ) = 2  π     T  ∑ 1 = - ∞ k  Δ     f  ( 1 ) ( 22 )

From equation (12)

w ( k )= z (N) ( k )= z (0) ( k )· K·e jφ(k)

follows together with equation (22) the following equation (23): z ( N )  ( k ) = K · z ( O )  ( k ) · e jT  ∑ l = - ∞ k     Δω  ( l )    ( 23 )

With Δω(l)= { 0 Δω     for  1 < 0 1 ≥ 0

follows:

z (N) ( k )= K·z (0) ( k )· e jΔω·k·T =K·z (0) ( k )·e j2πΔf·k·T   (24)

So the input signal of the CORDIC-algorithm is amplified with a constant K and mixed by a multiplication with a complex carrier e j2πΔf·k·T .

FIG. 7 shows the functionality of a receiver according to the present invention in case of reception of digitally modulated broadcast signals, e.g. of DRM, DAB or DVB reception, in which the circuit performs the digital frequency adjustment and the digital gain control of the receiver after complex baseband signal z(k) has been generated. In this case the polyfunctional circuit 5 passes its output signal to the short-wave processing circuit 6 B or the DAB processing circuit 6 C. The synchronization again passes the frequency offset Δω(k) to the polyfunctional circuit 5 according to the present invention. The DAB processing circuit 6 C sends its output signal to the D/A converter 7 that again produces the analog audio signal.

The AGFC (Automatic Gain and Frequency Control) is needed to again correct the frequency offset between the transmitter and the receiver and to adjust the amplitude of the wanted DAB signal optimal to the following processing, since the neighbor channel suppression of an unwanted DAB neighbor channel can be done in the front-end, in the digital part or a combination of both and the amplitude of the wanted DAB signal has to suite the following processing.

FIG. 8 shows the functionality of the polyfunctional circuit 5 performing the automatic gain and frequency control. The automatic gain control is integrated in the circuit that also performs the automatic frequency control. The automatic gain control uses the shift registers needed to perform the CORDIC-algorithm to adjust the amplitude. The frequency adjustment within the AGFC is performed as described above. Additionally, the average amplitude of the input complex baseband signal z(k) is calculated and an amplification of the input complex baseband signal z(k) depending on the average amplitude thereof is performed to realize the automatic gain control.

The average amplitude is calculated by filtering the absolute of the quadrature component z Q (k) of the complex baseband signal z(k). Therefore, the absolute value of the quadrature component z Q (k) is calculated in an absolute calculator 5 g which passes its result to an averaging low pass filter 5 h that in turn feeds its output signal d(k) to a pre-amplification block 5 e with a resolution of 1 bit with included saturation to avoid overflow which is situated in the signal path of the complex baseband signal z(k) in front of the CORDIC-algorithm block 5 a and to a fine-amplification block 5 f with a resolution of 0.25 bit with included saturation to avoid overflow which is situated in the signal path of the complex baseband signal after the CORDIC-algorithm block 5 a.

The averaging lowpass filter is shown in FIG. 9 . It has the transfer function H averaging  ( z ) = zA z - ( 1 - A ) ( 25 )

and the time constant of the averaging low pass filter can be calculated using the following equation: τ = - T ln  ( 1 - A ) ( 26 )

with T being the sampling rate.

Within the averaging low pass filter the input absolute of the quadrature component |z Q (k)|=|y(k)| is fed to a first multiplier 5 i that is advantageously realized by a shift register and which multiplies this input signal with a constant A before it is fed to an adder 5 j which adds the time-delayed output signal d(k) of the averaging low pass filter and which feeds its output signal to a second adder 5 k which subtracts the same time-delayed output signal d(k) of the averaging low pass filter which is multiplied by the constant A to produce the output signal d(k). The time-delayed output signal d(k) is generated by a delay circuit 5 m that receives the output signal d(k) from the output of the second adder 5 k and delays it by a time-constant T. Of course, the average amplitude can also be calculated by filtering the absolute of the inphase component |z I (k)|=|x(k)|.

FIG. 10 shows a block diagram of a polyfunctional circuit 5 according to the present invention that is able to perform modes A, B, and C. The polyfunctional circuit 5 consists of a multifunction circuit 51 that receives the IQ multiplexed complex baseband signal z(k) and the frequency offset Δf(k) and outputs one output signal D to an output circuit, a rotation circuit 52 that receives the IQ multiplexed baseband signal z(k) and outputs a first and a second output signal to said output circuit, and said output circuit that receives the output signal D of the multifunction circuit 51 and the first and second output signals of the rotation circuit 52 and outputs the processed IQ multiplexed baseband signal w(k).

The multifunction circuit 51 comprises a first switch S 1 that receives the frequency offset Δf(k) at its first input terminal (the first input terminal of a switch is always drawn as a black dot, the second input terminal of a switch is always drawn as a hollow circle and the movable output terminal of a switch is always drawn as a black dot with a line connected to one of the input terminals in the figures, having a second input terminal that runs free and having its movable output terminal connected to a first terminal of a second switch S 2 within the multifunction circuit 51 . Furtheron, the multifunction circuit 51 comprises a fourth switch S 4 having its second terminal connected to the movable output terminal of the second switch S 2 and its movable output terminal to a first input of a first adder/subtracter A 1 . The first input of the first adder/subtracter A 1 receives the first summand or diminuend depending on the function which is carried out by the first adder/subtracter A 1 . Furtheron, the multifunction circuit 51 comprises a third delay circuit 511 having a delay time T that receives the sum or the difference calculated and output by said first adder/subtracter A 1 at its first input, a second latch signal LE 2 at its second input and which outputs the output signal D of the multifunction circuit 51 at its output. The output signal D is also connected to the first terminal of the fourth switch S 4 . Furtheron, the multifunction circuit 51 comprises a fifth switch S 5 that receives the complex baseband signal z(k) at its first input and the output signal D of said multifunction circuit 51 at its second input. The movable output terminal of said fifth switch S 5 is connected to a multiplication circuit 512 also included in said multifunction circuit 51 that multiplies with a constant A, that is advantageously realized by a shift register and which output is connected to the first input of a sixth switch S 6 that is also included in said multifunction circuit 51 . Still furtheron said multifunction circuit 51 comprises a sixteenth switch S 16 having its first input connected to a constant value source 513 which is also included in said multifunction circuit 51 and which represents a constant frequency offset in mode B, its second input connected to the movable output terminal of said sixth switch S 6 and its movable output terminal connected to the second input of said first adder/subtracter A 1 that receives either the second summand or the subtrahend according to the respective functionality of said first adder/subtracter A 1 . Further components of said multifunction circuit 51 are a first delay circuit 514 , a second delay circuit 515 , a third switch 33 and a ROM 516 . The first delay circuit 511 has a delay T and has its input connected to the output of said first adder/subtracter A 1 and its output connected to the second input of said second switch S 2 . The second delay circuit 515 also has a delay T and has its first input connected to the output of said first delay circuit 514 and has its second input connected to the first latch signal LE 1 . The output of the second delay circuit 515 is connected to the first input of said third switch 33 which second input is connected to said ROM 516 and which movable output terminal is connected to the second input of the sixth switch S 6 .

Said rotation circuit 52 comprises seventh to twelfth switches 37 to S 12 , a first shift register 521 and a second shift register 522 , a second adder/subtracter A 2 and a third adder/subtracter A 3 , and fourth and fifth delay circuits 524 and 525 . The seventh switch receives the complex baseband signal z(k) at its first input terminal and has its movable output terminal connected to the second input of the eighth switch S 8 . The movable output terminal of the eighth switch S 8 is connected to the input of the first shift register 521 which has a control input connected to a first shift signal Sh 1 . The output of the first shift register 521 is connected to the input of a first amplification and saturation block 523 that is also included in said rotation circuit 52 and that performs an amplification by 2 4 with saturation to avoid overflow and to the second input terminal of the twelfth switch S 12 . Which first input terminal is connected to the output of the first amplification and saturation block 523 .

The movable output terminal of said twelfth switch S 12 is connected to the second input of the second adder/subtracter A 2 that receives either the second summand or the subtrahend according to the respective functionality of the second adder/subtracter A 2 . The output of said second adder/subtracter A 2 is connected to the input of the fourth delay circuit 524 that has a delay T and which output provides the first output signal of said rotation circuit 52 . Said first output signal of said rotation circuit 52 is also connected to the first input of the eighth switch S 8 , the second input of the ninth switch S 9 and the first input of the tenth switch S 10 . The second input of the tenth switch S 10 is running free and the movable output terminal is connected to the first input of said second adder/subtracter A 2 that receives the first summand or diminuend according to the functionality of said first adder/subtracter A 2 . The movable output terminal of the second switch 39 is connected to the input of the second shift register 522 which control input is connected to a second shift signal Sh 2 . The output of said second shift register 522 is connected to the second input of the third adder/subtracter A 3 that is receiving the second summand or the subtrahend according to the respective functionality of said third adder/subtracter A 3 . The output of said third adder/subtracter A 3 is connected to the fifth delay circuit 525 that has a delay T and which output provides the second output signal of said rotation circuit 52 . Said second output signal of said rotation circuit 52 is also connected to the second input terminal of said seventh switch S 7 , the first input terminal of said ninth switch S 9 and the first input terminal of the eleventh switch S 11 . The second input terminal of said eleventh switch S 11 is running free and the movable output terminal of said eleventh switch S 11 is connected to the first input of said third adder/subtracter A 3 that receives the first summand or diminuend according to its functionality.

Said output circuit comprises a sixth delay circuit 53 , a seventh delay circuit 54 , a second amplification and saturation block 55 , an eighth delay circuit 56 , and a thirteenth, a fourteenth, and a fifteenth switch S 13 , S 14 , and S 15 . The sixth delay circuit 53 has a delay T and receives the second output signal of the rotation circuit 52 at its first input and a third latch signal LE 3 at its second input. The output of the sixth delay circuit 53 is connected to the second input terminal of the thirteenth switch S 13 which has its first input terminal connected to the output of the seventh delay circuit 54 but has a delay T and receives the first output signal of the rotation circuit 52 at its first input and a fourth latch signal LE 4 at its second input. The movable output terminal of the thirteenth switch S 13 is connected to the input of the second amplification and saturation block 55 that calculates an amplification by 22 with saturation to avoid overflow and has its output connected to the first input terminal of the fourteenth switch S 14 which has its second input terminal connected to the movable output terminal of the thirteenth switch S 13 . The movable output terminal of the fourteenth switch S 14 is connected to the second input terminal of the fifteenth switch S 15 . The first input terminal of the fifteenth switch S 15 is connected to the output signal D of said multifunction circuit 51 . The movable output terminal of said fifteenth switch S 15 is connected to the eighth delay circuit 56 that has a delay T and which provides the output signal w(k) of the polyfunctional circuit 5 .

The states of the switches, adders/subtracters and shift registers of the circuit described above and shown in FIG. 10 for mode A are shown in FIG. 14 . The circuit has 24 different internal states ( 0 . . . 23 ) and is clocked with f c =24f s , i.e. 24 times faster than the clock frequency of the input sampling rate f s of the polyfunctional circuit 5 . In state 0 , the rotation circuit reads the inphase component x (0) (k) of the time-multiplexed complex baseband signal z(k). In state 1 , the rotation circuit reads the quadrature component x (0) (k) of the time-multiplexed complex baseband signal z(k). Beginning from state 1 to state 12 , the rotation circuit performs N=13 CORDIC-rotations of the input signal z(k).

The multifunction circuit 51 performs beginning from state 1 to state 13 the summation of the microrotation angles φ rot (i) described in equation (14). In state 14 the multifunction circuit 51 performs the frequency demodulation described in equation (19).

The states of the switches, adders/subtracters and shift registers of the circuit shown in FIG. 10 for mode B, i.e. AM-mode are shown in FIG. 15 . The circuit has 12 different internal states ( 0 . . . 11 ) and is clocked with f c =12f s , i.e. 12 times the clock frequency of the input sampling rate f s of the polyfunctional circuit 5 . In state 0 , the rotation circuit reads the inphase component x (0) (k) of the time-multiplexed complex baseband signal z(k) and in state 1 the quadrature component y (0) (k) of the time-multiplexed complex baseband signal z(k). Beginning from state 1 to state 11 , the rotation circuit performs N=11 CORDIC-rotations of the input signal z(k). The multifunction circuit 51 integrates in state 0 the frequency offset Δf(k) as described in equation (22) and beginning from state 1 to state 10 it calculates the phase error φ error (i) as it is described in equation (6).

FIG. 16 shows the states of the switches adders/subtracters and shift registers of the circuit described above and shown in FIG. 10 for mode C. Again, the circuit has 12 different internal states ( 0 . . . 11 ) and is clocked with f c =12f s , i.e. 12 times the clock frequency of the input sample rate f s of the polyfunctional circuit 5 . In state 0 , the rotation circuit reads the inphase component x (0) (k) of the time-multiplexed complex baseband signal z(k) and in state 1 the quadrature component y (0) (k) of said time-multiplexed circuit signal. During reading the inphase and quadrature components of the complex baseband signal z(k) the circuit performs a pre-amplification using the first shift register 521 that is controlled by the first shift signal Sh 1 and the first amplification and saturation block 523 which performs an amplification with 2 4 . Beginning from state 1 to state 10 , the rotation circuit performs N=10 CORDIC-rotations of the input complex baseband signal z(k) and instate 11 the rotation circuit performs a fine-amplification by multiplying the CORDIC output signal with the factors 1.5, 1.25, 1.00, or 0.875. This multiplication is controlled by the average amplitude d(k) as described in connection with FIGS. 8 and 9 and is performed using shift and add operations as shown in FIG. 17 .

The multifunction circuit 51 integrates in state 0 the frequency offset Δf(k) as described in equation (22) and beginning from state 1 to state 9 it calculates the phase error φ error (i) as it is described in equation (6). In states 10 and 11 , the multifunction circuit calculates the absolute of the quadrature component y(k) of the input complex baseband signal z(k) and performs the average filtering as described in equation (25).

In FIG. 16 the numbers 1 to 10 in brackets that are shown in connection with some of the arguments in the table have the following meaning:

( 3 )-( 8 ): depending on the average input amplitude d(k) which is identical to the signal D in mode C, the fine-amplification is performed using FIG. 17 .

( 1 )-( 2 ): Depending on the average input amplitude d(k) that is identical to the signal D in mode C, the rough pre-amplification is performed by varying the value Sh 1 of the first shift register 521 in states 0 and 1 . Values agc (0) and agc (11) have to be identical during one sample.

( 9 )-( 10 ): The averaging low pass filter can be enabled and disabled by the signal. In case of a null-symbol of the DAB-frame ( 9 ) and ( 10 ) can be set to 0 to avoid overshooting of the signal amplitude after the null-symbol.

The amplification has to be held constant during one OFDM-symbol.

FIG. 11 shows a circuit according to the present invention to perform modes A- 1 , B, and C.

In this case, the polyfunctional circuit 5 additionally comprises a ring buffer circuit 57 that receives the inphase and quadrature components g I ( 1 ) and g Q ( 1 ) of the complex baseband signal g( 1 ) and that leads only one of a quadrature component g Q and a inphase component g I to the multifunction circuit 51 and both components g Q and g I to the rotation circuit 52 . In this case, the inphase component g I is input to the rotation circuit 52 through the first fixed terminal of the seventh switch S 7 as the IQ-multiplexed complex baseband signal z(k) described in connection with the polyfunctional circuit 5 shown in FIG. 10 and the quadrature component g Q is input to the rotation circuit through the first input terminal of an additional nineteenth switch S 19 included in said rotation circuit 52 that is inserted with its path in-between the second input terminal and the movable output terminal in-between the second input terminal of the ninth switch S 9 and the first output signal of the rotation circuit 52 .

The ring buffer circuit 57 itself comprises a seventeenth switch S 17 , an eighteenth switch S 18 , a ninth delay circuit 59 , and a tenth delay circuit 58 . The first input terminal of the seventeenth switch S 17 receives the quadrature component g Q ( 1 ) of the complex baseband signal g( 1 ) and the first input terminal of the eighteenth switch S 18 receives the inphase component g I ( 1 ) of the complex baseband signal g( 1 ). The movable output terminal of the seventeenth switch S 17 is connected to the ninth delay circuit 59 that has a delay 11 T and that provides the quadrature component g Q at its output to the rotation circuit 52 . This output signal of the ninth delay circuit 59 is also connected to the second input terminal of the seventeenth switch S 17 . The movable output terminal of the eighteenth switch S 18 is connected to the input of the tenth delay circuit 58 that has a delay 11 T which output provides the inphase component g I to the multifunction circuit 51 and the rotation circuit 52 and to the second input terminal of the eighteenth switch S 18 .

FIG. 18 shows the states of the switches, adders/subtracters and shift registers of the polyfunctional circuit 5 shown in FIG. 11 in mode A- 1 . The polyfunctional circuit 5 shown in FIG. 11 has 24 different internal states ( 0 . . . 23 ) and performs a sampling rate decimation with a decimation factor of 2. Therefore, the circuit is clocked with f c =12f sg , f sg is the sampling rate of the circuits input signal g( 1 ).

In states 0 to 8 the rotation circuit performs a 9-tap FIR filtering for complex signals and therefore the rotation circuit gets the needed input samples from the ring buffer circuit 57 . Beginning from state 9 to state 21 , the rotation circuit performs N=13 CORDIC-rotations of the filtered and sampling rate decimated signal z(k).

The multifunction circuit 51 performs beginning from state 9 to state 21 the summation of the microrotation angles φ rot (i) as described in equation (14). In state 22 the multifunction circuit 51 performs the frequency demodulation described in equation (19).

As the polyfunctional circuit 5 shown in FIG. 11 is besides the switch S 19 and the ring buffer circuit 57 identical to the polyfunctional circuit 5 shown in FIG. 10, the functionalities for modes B and C can easily be derived from the description above in connection with the changes to the circuit.

It is clearly to be understood that the polyfunctional circuits 5 shown in FIGS. 10 and 11 can be modified by those skilled in the art without departing from the scope of the dependent claims and the invention is not limited to these specific circuits.

According to the present invention the shift registers that are quite costly that are included in the rotation circuit 52 are used for different purposes and therefore the receiver according to the present invention is cheaper than such receiver according to the prior art. Furtheron, the circuit is realized with only 3 adders/subtracters and no additional adders or multipliers are needed for the digital filtering in sampling rate decimation and the same hardware is used from the automatic frequency control and the automatic gain control, since the multipliers 512 and 5 i are advantageously realized by shift registers.