DE102008006428A1 - Consequence equalizer for equalizing modulated signal in modulated signal receiving circuit, has quantization error circuit for determining quantization error of quantizer, where components are designed to work in polar coordinates - Google Patents

Abstract

The equalizer (1) has a quantizer (3) arranged at an output side to measure an output signal, where an input signal is supplied to a forward filter (2) e.g. finite impulse response (FIR) filter. The output signal of the quantizer is supplied to a rearward filter (4). An adder (5) is connected between the forward filter and the quantizer, where the output signal of the rearward filter is supplied to the adder. A quantization error circuit (7) determines the quantization error of the quantizer, where the components are suitably designed to work in polar coordinates (r, phi). An independent claim is also included for a method for equalizing a modulated signal.

Description

The The invention relates to a circuit arrangement for equalizing a modulated signal according to the preamble of the claim 1.

circuitry for equalizing a modulated signal, so-called equalizer or equalizer, are known from the prior art. They are in circuits used to receive modulated signals to inadequacies of transmitter, transmission link and receiver input circuit compensate. They are used in both digital and analog transmission methods.

A circuit arrangement for equalizing a modulated signal according to the prior art is in 2 shown. The circuit arrangement 21 for equalizing a modulated signal or short the equalizer or equalizer, in the illustrated embodiment consists of an adaptive forward filter 22 , which is designed as an FIR filter with adaptive coefficients. The forward filter 22 is an adder 25 downstream. The output signal of the adder 25 is a quantizer 23 can be supplied, which in turn provides a quantized output signal. The equalizer 21 also includes an adaptive backward filter 24 , which is designed as an IIR filter with adaptive coefficients, wherein the backward filter 24 Input side, the quantized output signal is supplied. The output signal of the backward filter 24 becomes the adder 25 fed. It is further a quantization error circuit 27 provided for determining a quantization error δ. The quantization error δ is multiplied by a multiplier 29 amplified by a factor ε and is the forward filter 22 and the backward filter 24 can be fed to form the filter coefficients. The equalizer 21 is a receiver circuit 20 as well as a circuit 20a upstream, which mixes the received signal into baseband. The complex-valued signal mixed into the baseband is then the input filter to the forward filter 22 of the equalizer 21 fed.

It is a common procedure that the receiver circuit 20 the radio frequency signal first mixes to an intermediate frequency and then the signal through a circuit 20a is transformed into the baseband. So that no information is lost, the signal paths of the baseband branch are executed in a complex manner, ie that an in-phase and a quadrature-phase component I, Q are transmitted on at least two signal paths. It is irrelevant whether the circuit blocks are implemented as analog or digital circuits and from which of these blocks, if necessary, a digitization takes place.

One of the receiver circuit 20 downstream equalizer 21 , which works in the baseband, must also be executed in this case as a complex circuit with two signal paths I, Q. Usually there is such an equalizer 21 in two parts, an adaptive forward filter 22 and a decision feedback adaptive backward filter 24 , This structure is for example in the chapters 5.7 and 14 of the textbook message transmission described by KD Kammeyer and in 2 shown. The adaptive forward filter 22 is here as FIR filter and the decision-feedback adaptive backward filter 24 implemented as an IIR filter. The filters 22 . 24 are controllable by adjustable filter coefficients. Under ideal signal conditions, ie when there are no intersymbol interferences, a central filter coefficient is 1, all other coefficients being 0. The filter coefficients are calculated from a quantization error δ of the equalizer 21 downstream quantizer 23 certainly. To determine the filter coefficients, the quantization error δ is amplified by a gain factor ε and the filter coefficients are calculated from the amplified quantization error. To the stability of the equalizer 21 especially when interacting with other function blocks, such. As a sampling rate recovery, a carrier control or a gain control, the gain factor ε, which changes the adaptive coefficients, must be set very low.

From the US 4237554 It is also known to incorporate into the signal path after the amplification of the quantization error δ a leak which pulls the coefficients towards their normal position, ie that the central coefficient is 1 and all other coefficients are 0.

It is also common practice, the gain factor ε, which alters the adaptive coefficients, as well as the size of the leak to conform to the reception conditions. In particular, be during the acquisition phase uses different values than during signal tracking.

The approaches described in the prior art have several disadvantages, since the entire signal processing of the equalizer takes place in the Cartesian baseband system I, Q. On the one hand, experience has shown that the strongest echo signals are to be expected directly behind the original signal. To eliminate these echoes, one would like to have the highest possible amplification for the change of the filter coefficients, ie a large amplification factor ε. A receiver circuit arranged in front of the equalizer and the means for mixing the signal into the baseband also have phase noise. Since the phase angle Φ zeitab varies widely, the coefficients in the back of the equalizer will no longer have a correct phase reference. At this point, it would be advantageous to set a higher gain in the direction of the phase Φ. However, these opposing requirements can not be easily solved with an equalizer operating in Cartesian coordinates (I, Q).

Here uses the present invention.

It the object of the invention is a circuit arrangement for equalizing a modulated signal to provide the does not have the disadvantages of the prior art and it also allows a high gain in the direction of Phase Φ easy to adjust.

These The object is achieved by a circuit arrangement for equalizing a modulated signal with the feature of claim 1 solved.

advantageous Further developments of the invention are the subject of the dependent claims.

A inventive circuit arrangement for equalizing a modulated signal consists of a forward filter, to which an input signal can be fed, a quantizer, the output is arranged and at which an output signal can be tapped, a reverse filter that the Output of the quantizer can be fed in at the beginning is, an adder between the forward filter and the quantizer ge is switched and that the output signal of the backward filter can be supplied and a quantization error circuit for Determination of a quantization error of the quantizer. The basic Idea of the invention consists in the fact that at least one of the Components of the circuit arrangement is suitably formed in Polar coordinates to work.

It is advantageous if the forward filter, the input signal can be supplied in polar coordinates. The input signal can z. B. by a forward filter upstream IQ / rΦ coordinate converter implemented from the Cartesian coordinate system in polar coordinates become. The IQ / rΦ coordinate converter can do this this implementation, for example, by tables, so-called look-up tables, or realize a Cordic circuit.

Around also on the output side an interface to downstream circuit arrangements to create, which work in Cartesian coordinates, it is advantageous the quantizer a rΦ / IQ coordinate converter downstream. The output rΦ / IQ coordinate converter is advantageously similarly formed and the input side IQ / rΦ coordinate converter.

Around To achieve optimum performance of the equalizer, it is advantageous the forward filter as FIR filter with adaptive filter coefficients train. Likewise, it is advantageous the backward filter to provide as an IIR filter with adaptive filter coefficients.

Around optimal control of the filter coefficients of the forward and to reach the backward filter, it makes sense that at the quantization error circuit on the output side, the quantization error can be tapped in polar coordinates. The quantization error becomes calculated separately in r and Φ components and on the output side provided at the quantization error circuit. In a preferred embodiment of the invention, the filter coefficients of the forward and backward filters controllable by the quantization error, wherein a radial component the quantization error with a first factor and a phase component of the quantization error can be amplified with a second factor is. This can be achieved, for example, a high gain in Adjust phase direction while gain is kept low in the radial direction. The realization of such Gain is especially easy through multipliers possible.

It is further a method for equalizing a modulated Claims in which a forward filtering of the Input signal, quantization of the output signal, backward filtering of the quantized output signal and an addition of the backward filtered Signal to the forward filtered signal performed where a quantization error is the quantization of the output signal is calculated. The basic idea with this method lies also in that at least one of the steps is performed in polar coordinates becomes.

The Invention will be described below with reference to the attached Figures described in detail.

It demonstrate:

1 a circuit arrangement for equalizing a modulated signal according to the invention and

2 a circuit arrangement for equalizing a modulated signal according to the prior art (already discussed).

1 shows a block diagram of a circuit arrangement according to the invention 1 for equalizing a modulated signal. The Schaltungsanord voltage 1 for equalizing a modulated signal is in the sequence shortly as an equalizer 1 designated. The equalizer 1 is an IQ / rΦ coordinate converter 13 upstream of which an input signal in Cartesian coordinates I, Q, which was received for example by a receiver and then mixed into the baseband, can be fed. On the output side, the IQ / rΦ coordinate converter provides a signal in polar coordinates r, Φ. This signal in polar coordinates is a forward filter 2 , which is designed as an FIR filter with adaptive filter coefficients, fed. The output side signal of the forward filter 2 is an adder 5 can be fed, which also works in the polar coordinates r, Φ. The adder 5 is a quantizer 3 downstream, which quantizes an input-side signal and provides the output side in polar coordinates r, Φ. The output signal of the quantizer 3 is a rΦ / IQ coordinate converter 15 can be supplied, which converts the signal from polar coordinates r, Φ in Cartesian coordinates I, Q, so that subsequent circuit blocks can work again in Cartesian coordinates I, Q.

The equalizer 1 further includes an adaptive backward filter 4 which is designed as an IIR filter with adaptive filter coefficients. The backward filter 4 is the output side signal of the quantizer 3 fed. The output signal of the backward filter 4 is the adder 5 that between the forward filter 2 and the quantizer 3 is arranged to be supplied to perform addition of the forward filtered signal and the backward filtered signal and the sum to the quantizer 3 provides. The equalizer 1 further includes a quantization error circuit 7 which has a quantization error δ _r , δ _{Φ of} the quantizer 3 calculated in polar coordinates r, Φ. In the quantization error circuit 7 For example, subtractors may be arranged, one of which is the input and the output of the quantizer in the radial component r and the other the input and the output of the quantizer 3 in the phase component Φ can be supplied. On the output side, the quantization error δ _r , δ _{Φ can} then be subtracted component by component at the subtractors. Both the quantization error in the radial component δ _r and the quantization error in the phase component δ _Φ are each a multiplier 9 . 11 can be fed in and separately amplified by gain factors ε _r , ε _Φ . The amplification factors ε _r , ε _Φ may, for example, also depend on a coefficient τ. τ represents a temporal component, so that z. B. for the radial component of the quantization error δ _r in the vicinity of the central coefficient a high gain ε _{r is} adjustable and for the phase component of the quantization error δ _Φ at a later time, a high gain factor ε _Φ can be adjusted.

In conclusion, it should be noted that a complete execution of the equalizer 1 in polar coordinates r, Φ is not absolutely necessary. It can also be individual parts, such as the forward filter 2 and the backward filter 4 in polar coordinates r, Φ, the other components performing signal processing in Cartesian coordinates I, Q. It is also possible that the equalizer 1 and the upstream and downstream components in a mixed coordinate system of Cartesian coordinates I, Q and polar coordinates r, Φ work. In this embodiment, it is then possible that a quantization error δ is further processed, for example, in the in-phase component I and the phase component Φ and suitably amplified.

In the following, the function of the equalizer just described becomes briefly 1 described.

An input signal in Cartesian coordinates I, Q becomes an IQ / rΦ coordinate converter 13 supplied and converted by these in polar coordinates r, Φ. The signal in polar coordinates r, Φ is then used by the forward filter 2 filtered with adaptive filter coefficients and an adder 5 fed. The output signal of the adder 5 becomes a quantizer 3 which performs a quantization of the signal and provides an output signal in polar coordinates r, Φ. The output signal in polar coordinates r, Φ is then converted back into Cartesian coordinates I, Q by an rΦ / IQ coordinate converter. The output signal of the quantizer 3 also gets a backward filter 4 which performs infinite impulse response filtering and adaptive filter coefficients. The backward filtered signal also becomes the adder 5 which performs addition of the forward filtered signal and the backward filtered signal, and the result of this addition to the quantizer 3 provides. The adjustment of the adaptive filter coefficients for the forward filtering and the backward filtering is carried out on the basis of a quantization error δ _r , δ _Φ by a quantization error circuit 7 is determined. The individual components of the quantization error δ _r , δ _Φ, which is determined in polar coordinates r, Φ, are separately multiplied by amplification factors ε _r , ε _Φ and provided for coefficient determination for the forward filtering and the backward filtering. The amplification of the individual components δ _r , δ _Φ can also be amplified by gain factors ε _r (τ), ε _Φ (τ), which are dependent on a time coefficient τ. Thus, for example, in the vicinity of the central coefficient, a large gain of the radial component of the quantization error δ _r and, at a later time, a large amplification of the phase component of the quantization error ε _Φ can be achieved.

1

Circuitry / equalizer

2

forward filter

3

quantizer

4

backward filter

5

adder

7

Quantisierungsfehlerschaltung

9

multipliers

11

multipliers

13

IQ / rΦ coordinate converting

15

r Q / IQ coordinate converter

δ _r

radial component of the quantization error

δ _Φ

phase component of the quantization error

r

radial component

Φ

phase component

I

In-phase component

Q

Quadrature phase component

ε _r

Radial strength / first factor

ε _Φ

Phase gain / 2. Factor time coefficient

20

receiver

20a

circuit for mixing to baseband

21

Circuitry / equalizer

22

forward filter

23

quantizer

24

backward filter

25

adder

27

Quantisierungsfehlerschaltung

29

multipliers

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• - US 4237554 [0006]

Claims (17)

Circuit arrangement ( 1 ) for equalizing a modulated signal with the following components: - a forward filter ( 2 ) to which an input signal can be fed, - a quantizer ( 3 ), which is arranged on the output side and on which an output signal can be tapped off, - a backward filter ( 4 ), to which the output signal of the quantizer ( 3 ) can be fed on the input side, - an adder ( 5 ) between the forward filter and the quantizer ( 3 ) and the output signal of the backward filter ( 4 ) and - a quantization error circuit ( 7 ) for determining a quantization error δ _r , δ _{Φ of} the quantizer ( 3 ), characterized in that at least one of the components is adapted to operate in polar coordinates (r, Φ). Circuit arrangement ( 1 ) according to claim 1, characterized in that the forward filter ( 2 ) a signal in polar coordinates (r, Φ) can be fed. Circuit arrangement ( 1 ) according to one of the preceding claims, characterized in that the forward filter ( 2 ) an IQ / rΦ coordinate converter ( 13 ) is connected upstream. Circuit arrangement ( 1 ) according to one of the preceding claims, characterized in that the quantizer ( 3 ) a rΦ / IQ coordinate converter ( 15 ) is connected downstream. Circuit arrangement ( 1 ) according to one of the preceding claims, characterized in that the forward filter ( 2 ) is designed as an FIR filter with adaptive filter coefficients. Circuit arrangement ( 1 ) according to one of the preceding claims, characterized in that the backward filter ( 4 ) is designed as an IIR filter with adaptive filter coefficients. Circuit arrangement ( 1 ) according to one of the preceding claims, characterized in that on the output side at the quantization error circuit ( 7 ) the quantization error (δ _r , δ _Φ ) in polar coordinates (r, Φ) can be tapped. Circuit arrangement ( 1 ) according to one of claims 5 to 7, characterized in that the filter coefficients are controllable by the quantization error (δ _r , δ _Φ ). Circuit arrangement ( 1 ) according to one of the preceding claims, characterized in that a first multiplier ( 9 ) is provided for amplifying the radial component (δ _r ) of the quantization error (δ _r , δ _Φ ) with a first factor (ε _r ). Circuit arrangement ( 1 ) according to one of the preceding claims, characterized in that a second multiplier ( 11 ) is provided for amplifying the phase component (δ _Φ ) of the quantization error (δ _r , δ _Φ ) with a second factor (ε _Φ ). A method of equalizing a modulated signal, comprising the steps of: - forward filtering an input signal, - quantizing an output signal, - backward filtering a quantized output signal, - adding the backward filtered signal to the forward filtered signal, and - detecting a quantization error (δ _r , δ _Φ ) , characterized in that at least one of the steps is performed in polar coordinates (r, Φ). Method according to claim 11, characterized in that that the forward filtering in polar coordinates (r, Φ) he follows. Method according to one of claims 11 or 12, characterized in that the backward filtering in polar coordinates (r, Φ). Method according to one of claims 11 to 13, characterized in that the quantization error (δ _r , δ _Φ ) in polar coordinates (r, Φ) is determined. Method according to one of claims 11 to 14, characterized in that the quantization error (δ _r , δ _Φ ) in a radial component (δ _r ) and a phase component (δ _Φ ) with different factors (δ _r , δ _Φ ) is amplified. Method according to one of claims 11 to 15, characterized in that the forward filtering is adaptive with finite impulse response and filter coefficients by the quantization error (δ _r , δ _Φ ) are adjustable. Method according to one of claims 11 to 15, characterized in that the backward filtering is carried out with infinite impulse response and filter coefficients by the quantization error (δ _r , δ _Φ ) are adjustable.