JP2007318349A - Fm receiver - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an FM receiver that can effectively reduce multipath distortion by an equalization algorithm with fast convergence and is suitably mounted on a vehicle. <P>SOLUTION: The FM receiver extracts a distortion component produced in a pilot signal band from an FM composite signal obtained by demodulating an FM reception signal filtered by a variable coefficient filter, and controls the filter coefficient of the variable coefficient filter to minimize an evaluation function for the distortion component, thereby obtaining equalization characteristics having superior convergence for change in multipath environment. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an FM receiver having an equalizer for equalizing multipath distortion.

  In an FM receiver that receives FM broadcast waves, it is known that reception quality deteriorates due to multipath distortion in a so-called multipath environment in which a target broadcast wave arrives via a plurality of propagation paths. A constant envelope algorithm (Constant Modulus Algorithm: CMA) is known as one of equalizer algorithms for reducing multipath distortion. Japanese Patent No. 30111948 (Patent Document 1) discloses an example of an equalizer using CMA. According to Patent Document 1, an FM reception signal is input to a variable coefficient filter, and a filter coefficient (tap coefficient) is controlled based on CMA, whereby an equalized signal from the filter, that is, an FM in which multipath distortion is reduced. A reception signal is output.

CMA is one of so-called blind equalization algorithms that do not require a reference signal. According to CMA, the equalized output signal from the variable coefficient filter is expressed as follows.

ここでμはタップ係数の一回当たりの更新量を決定するステップ係数を表し、σは可変係数フィルタからの等化出力信号のターゲット振幅を表す。 The tap coefficient is updated according to CMA as follows.

Here, μ represents a step coefficient for determining an update amount per tap coefficient, and σ represents a target amplitude of the equalized output signal from the variable coefficient filter.

Originally, an FM broadcast wave has a constant envelope, but if a multipath exists, it does not have a constant envelope. Using this point, CMA reduces multipath distortion by updating the tap coefficient so that the FM reception signal has a constant envelope.
Japanese Patent No. 30111948

  As can be seen from Equation (2), the CMA updates the tap coefficient according to the square error of the output signal amplitude (y) of the variable coefficient filter with respect to the target amplitude (ρ). Therefore, it can be seen that even if the phase of the output signal of the variable coefficient filter is different from the phase of the desired output signal, the tap coefficient is not updated if the amplitude is constant. For example, in an inferior multipath environment where FM broadcast waves arrive via a large number of paths with different propagation distances, the phase of the output signal of the variable coefficient filter is the same as the target amplitude, but the phase is larger than the desired phase. It may be off. In such a case, the multipath distortion is not reduced because the tap coefficient should be updated but not updated correctly.

  For this reason, it has been pointed out that CMA has poor convergence characteristics compared to other algorithms, that is, tap coefficients converge slowly. Poor convergence characteristics of the tap coefficient are particularly problematic in FM receivers mounted on a moving body that moves at high speed such as an automobile. That is, when viewed from a high-speed moving body, the multipath environment changes every moment. For this reason, if the convergence is slow, the equalizer cannot follow the change in the multipath environment, and the multipath distortion is not sufficiently reduced.

  An object of the present invention is to provide an FM receiver suitable for in-vehicle use in which multipath distortion can be effectively reduced by an equalization algorithm with fast convergence.

  According to an aspect of the present invention, a receiving unit that receives FM broadcast waves and outputs a received signal; a first filter that filters the received signal according to a variable filter coefficient; and a received signal that is filtered by the first filter An FM demodulator that generates a composite signal including a main signal, a sub signal, and a pilot signal; and a distortion component generated in a target band other than the band in which the main signal and the sub signal are arranged from the composite signal. There is provided an FM receiver comprising: a second filter to be extracted; and a coefficient control unit that controls the filter coefficient so as to minimize an evaluation function for the distortion component.

  According to a second aspect of the present invention, a receiving unit that receives FM broadcast waves and outputs a received signal; a first filter that filters the received signal according to a variable filter coefficient; and a filter that is filtered by the first filter An FM demodulator that demodulates the received signal to generate a composite signal including a main signal, a sub signal, and a pilot signal; distortion generated in a target band other than the band in which the main signal and the sub signal are arranged from the composite signal A second filter for extracting a component; and a sum of a first evaluation function for the distortion component and a second evaluation function for an error with respect to a target amplitude of the received signal filtered by the first filter is minimized. An FM receiver comprising: a coefficient control unit that controls the filter coefficient.

  According to the present invention, since the filter coefficient is updated using both the distortion component included in the FM composite signal, that is, both the amplitude information and the phase information, the convergence speed, which is a drawback of the CMA that uses only the amplitude information for updating the filter coefficient. Can be improved, and the demodulation accuracy of the FM reception signal is improved.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
The FM receiver according to one embodiment of the present invention shown in FIG. 1 is suitable for mounting on a mobile object such as an automobile. According to the FM receiver of FIG. 1, an FM broadcast wave is first received by the antenna 10, and an output signal (FM reception signal) of the antenna 10 is input to the front end 11. The front end 11 includes a tuner that amplifies the FM reception signal, converts the reception signal of the desired channel to an intermediate frequency (IF), and outputs the intermediate frequency (IF). The reception signal output from the front end 11 is digitized by an analog-digital converter (ADC) 12.

  The digitized received signal 101 is input to the variable coefficient filter 13. The variable coefficient filter 13 performs equalization processing to reduce multipath distortion included in the received signal 101 by filtering the input received signal 101 according to a variable filter coefficient, and outputs the equalized output signal 102. Generate.

  The equalized output signal 102 from the variable coefficient filter 13 is input to the FM demodulator 14. In this example, the FM demodulator 14 includes an arctangent calculator 15 and a differentiator 16. In the arc tangent calculator 15, phase information is obtained by calculating the arc tangent of the equalized output signal 102 from the variable coefficient filter 13. This phase information is differentiated by the differentiator 16 to generate a so-called FM composite signal 103.

  In the case of FM stereo broadcasting in Japan, for example, in the case of FM stereo broadcasting in Japan, as shown in FIG. 2, a low (L + R) signal up to 15 kHz, that is, a stereo left (L) signal and a right (R) signal. (LR signal), that is, a difference signal (sub signal) between the L signal and the R signal is arranged in a region centered at 38 kHz. In the case of monaural broadcasting, only the main signal exists and no sub signal exists. In the composite signal, a pilot signal for identifying whether the broadcast is stereo broadcast or monaural broadcast is further multiplexed at 19 kHz.

  The FM composite signal 103 output from the FM demodulator 14 is subjected to (L + R) + (LR) and (L + R)-(LR) operations (stereo demodulation) by the stereo demodulator 17. Stereo signals (L signal and R signal) or monaural signals are reproduced.

  The FM composite signal 103 is also input to the extraction filter 17. The extraction filter 17 extracts the distortion component 104 generated in the target region other than the band in which the main signal and the sub signal are arranged from the FM composite signal 103. Specifically, for example, a distortion component generated in a band of 15 kHz to 23 kHz (hereinafter referred to as a pilot signal band) in which a pilot signal included in the FM composite signal 103 is arranged, that is, a pilot signal other than the pilot signal existing in the pilot signal band. A distortion component 104 of the signal is extracted. The distortion component 104 thus extracted by the extraction filter 17 is input to the coefficient control unit 18.

  The coefficient control unit 18 performs control while updating the filter coefficient (tap coefficient) of the variable coefficient filter 13 so that the evaluation function representing the distortion component 104 extracted by the extraction filter 17 is minimized. That is, only the pilot signal originally exists in the pilot signal band, but if there is multipath distortion, a distortion component due to multipath is mixed into the pilot signal band.

  Therefore, such a distortion component 104 is extracted by the extraction filter 17, and the filter coefficient of the variable coefficient filter 13 is set so that the distortion component 104 is minimized, in other words, the evaluation function representing the distortion component 104 is minimized. If updated, multipath distortion is effectively reduced in the equalized output signal 102 from the variable coefficient filter 13. A method for calculating the evaluation function representing the distortion component 104 will be described in detail later.

Hereinafter, the multipath distortion reduction operation in the FM receiver of FIG. 1 will be described in detail.
The equalized output signal 102 from the variable coefficient filter 13 corresponding to the digitized received signal (discrete signal) 101 output from the ADC 12 at time n is represented by the following equation y (n).

Since the FM broadcast wave has information on the frequency, the FM demodulator 14 can demodulate it by obtaining the frequency component of the equalized output signal y (n) of the variable coefficient filter 13. In order to obtain the frequency component of the equalized output signal y (n), it is necessary to obtain the phase change of y (n). Therefore, in the FM demodulator 14, first, the phase component of the equalized output signal y (n) from the variable coefficient filter 13 is extracted by the arc tangent calculator 15. This phase component extraction process can be written as follows using an arctangent function.

ここで、式（５）のｚ(n)がＦＭコンポジット信号１０３である。 The output signal of the arctangent calculator 15 is input to the differentiator 16. The differentiator 16 obtains the phase change (phase change from time t-1 to time t) of the output signal of the arctangent calculator 15 at time t, and extracts the frequency component. The processing of the differentiator 16 can be written as

Here, z (n) in the equation (5) is the FM composite signal 103.

  The solid line in FIG. 3 indicates the frequency characteristic of the original composite signal, and the dotted line indicates the frequency characteristic of the composite signal generated from the FM reception signal that has passed through the multiple propagation paths. The L signal is a single tone of 1 kHz, and the R signal is no signal. In the dotted line in FIG. 3, multipath distortion occurs in the band of the composite signal, which deteriorates the quality of the audio signal.

  As shown in FIG. 2, the composite signal should not contain any signal other than the pilot signal in the pilot signal band (15 kHz to 23 kHz). However, when the FM broadcast wave subjected to multipath distortion is FM demodulated, the dotted line in FIG. As shown, the distortion spreads over the entire area, and a distortion component is also generated in the pilot signal band.

  Since a distortion component is generated in the pilot signal band (15 kHz to 23 kHz) due to the influence of the multipath distortion, if the filtering by the variable coefficient filter 13, that is, the equalization process is performed so as to reduce the distortion component, the multipath distortion is generated. Can be reduced. The equalization processing for reducing the multipath distortion by the method of the present embodiment uses both amplitude and phase information, and therefore converges faster than the conventional CMA.

  That is, in the CMA using only the amplitude, as described above, the tap coefficient may not be updated correctly in a poor multipath environment, which causes the convergence of the tap coefficient to be delayed. On the other hand, the distortion component generated in the pilot signal band due to the multipath distortion includes both amplitude and phase distortions, so that the variable coefficient filter 13 can reduce the distortion component in the pilot signal band. If the coefficient is updated, the convergence characteristic as an equalizer is improved.

ただしＳ_j(m)は抽出フィルタ１７のフィルタ係数、Mはフィルタ１７のフィルタ長をそれぞれ表す。フィルタ係数Ｓ_j(m)は、抽出フィルタ１７によってパイロット信号帯域内の歪み成分を抽出できれば、どのような係数も使用することが可能であるが、例えば次式に示されるフーリエ係数を用いることができる。

Next, a specific method for extracting a distortion component in the pilot signal band and reducing multipath distortion will be described. There are several methods for extracting the distortion component of the pilot signal band. Here, a method using an FIR filter as the extraction filter 17 will be described. As a method for extracting a distortion component within a pilot signal band, either a method for extracting a distortion component together with a 19 kHz pilot signal or a method for extracting a distortion component without including a pilot signal can be used. The method will be described. The distortion component v (n) in the pilot signal band extracted by the extraction filter 17 can be written as follows.

However, S _j (m) represents the filter coefficient of the extraction filter 17, and M represents the filter length of the filter 17. Any coefficient can be used as the filter coefficient S _j (m) as long as the extraction component 17 can extract a distortion component in the pilot signal band. For example, a Fourier coefficient represented by the following equation is used. it can.

Here, f _j indicates the frequency of the distortion component to be extracted.

  Due to the nature of the FM composite signal, all components extracted by the extraction filter 17 are unnecessary distortion components. Therefore, in the present embodiment, the coefficient coefficient of the variable coefficient filter 13 is updated by the coefficient control unit 18 so that this distortion component is reduced. Hereinafter, some specific examples of the coefficient control unit 18 will be described.

(First specific example of the coefficient control unit 18)
In the first specific example of the coefficient control unit 18, an evaluation function J is defined for the distortion component v (n) extracted by the extraction filter 17, and the variable coefficient filter 13 is set so that the evaluation function J is minimized. Update filter coefficients. The following formula shows an example of the evaluation function J.

  Actually, the ensemble average of the power v (n) v (n) of the distortion component v (n) or the instantaneous value of v (n) v (n) is used as the evaluation function J. In this embodiment, the latter v (n ) A method using the instantaneous value of v (n) as the evaluation function J will be described as an example. Of course, it is also possible to take a method of using the ensemble average as the evaluation function J.

  In the above description, the distortion component around the pilot signal that does not include the pilot signal is extracted by one filter 17 and the evaluation function for the distortion component is used. However, a plurality of distortion components are extracted using a plurality of extraction filters. It is also possible to extract and use an evaluation function for those distortion components. That is, Equation (8) uses the sum of distortion component powers v (n) v (n) extracted by a plurality of extraction filters as an evaluation function, and updates the filtering coefficient of the variable coefficient filter 13 so as to minimize it. You may make it do.

  Further, when the extraction filter 17 extracts a distortion component together with the pilot signal from the pilot signal band, the evaluation function evaluates the power of the component obtained by subtracting the undistorted pilot signal component from the output signal of the extraction filter 17 with respect to the distortion component. It can be a function.

The filter coefficient update formula by the coefficient control unit 18 can be expressed by the following formula using the evaluation function J.

である。 Here, from the equations (6), (7) and (8)

It is.

FIG. 4 is a block diagram illustrating the configuration of the coefficient control unit 18 that performs the above-described filter coefficient updating operation. The coefficient control unit 18 receives the digitized reception signal output from the ADC 12 in FIG. 1 and the output signal from the variable coefficient filter 13. The digitized received signal is a vector, and the vector length is the same as the number of taps of the variable coefficient filter 13. In the coefficient control unit 18, first, the α calculation unit 30 calculates partial differential values α ₀ (n) to α _L (n) of the evaluation function J. The partial differential values α ₀ (n) to α _L (n) represent the gradient (including direction) of the filter coefficient update of the variable coefficient filter 13.

Now, when the vector length of the digitized received signal from the ADC12 and L, the α calculation unit 30 the signal x ₀ ~x _L of L elements of the received signal vector is input. The signals x _{0 to} x _L are divided by 1 + y (n) ² which is the sum of the square of the output signal y (n) of the variable coefficient filter 13 and the fixed value “1” by the dividers 311 to 31L. The output signals of the dividers 311 to 31L are input to the multipliers 321 to 32L. Here, the coefficients of the multipliers 321 to 32L have the same configuration as the first tap coefficient of the extraction filter 17 that extracts the distortion component in the pilot signal band in the FM composite signal 104, and the partial differential value α ₀ of the evaluation function J. (n) to α _L (n) are output.

The partial differential values α ₀ (n) to α _L (n) are multiplied by the output signal (distortion component) v (n) from the extraction filter 17 in the multipliers 331 to 33L. The output signals of the multipliers 331 to 33L are further multiplied by the step coefficient μ in the multipliers 341 to 34L, thereby determining the coefficient update amount of the second term on the right side of Equation (9). The output signals of the multipliers 341 to 34L are input to the adders 351 to 35L. The output signals of the multipliers 341 to 34L are subtracted from the values of the output signals of the multipliers 341 to 34L held in the latches 361 to 36L by the adders 351 to 35L, respectively, and updated according to the equation (9). Determined filter coefficients are determined.

  As described above, the coefficient control unit 18 calculates the partial differential value of the evaluation function for the distortion component extracted by the extraction filter unit 17 using the received signal from the ADC 12. A coefficient update amount is determined by multiplying the partial differential value by a distortion component extracted by the extraction filter unit 17 and a step coefficient, and the filter coefficient is updated by the coefficient update amount. In this case, the multipath distortion can be effectively reduced by setting the step coefficient to an appropriate value (for example, 0.1).

  In the conventional CMA according to the equation (2), if the output signal amplitude of the variable coefficient filter is the same as the target amplitude, an error is not detected even if the phase is different, so that the filter coefficient converges slowly. On the other hand, in this embodiment, since the presence or absence of a distortion component in the pilot signal band is used as an error, the filter coefficients can be quickly converged, and equalization processing with excellent convergence characteristics can be performed.

(Second specific example of coefficient control unit)
Next, a second specific example of the coefficient control unit 18 will be described. As described above, the FM broadcast wave has a constant envelope if there is no multipath effect. Therefore, assuming that the equalization output signal y (n) from the variable coefficient filter 13 has a constant envelope, the evaluation function J is approximated as follows.

と書くことができ、これは式（１１）と等価である。ここで、ｌ＝０，１，・・・，Ｌである。 Consider the case where the variable coefficient filter 13 is a transversal filter. As shown in FIG. 5, the transversal filter includes a tapped delay device D, a multiplier x that multiplies the signals x _{0 to} x _L of each tap of the delay device D by weights (tap coefficients) w _{0 to} w _L , and It comprises an adder Σ that adds the signals after multiplication. In such a transversal filter, the relationship of x _l (n−m) = x ₀ (n−m−l) is generally established. That is, the signal value of a certain tap is equal to the signal value one time before the tap one stage before. From this relationship, equation (12) becomes

Which is equivalent to equation (11). Here, l = 0, 1,..., L.

FIG. 6 shows a coefficient control unit 18 when a transversal filter is used as the variable coefficient filter 13, and includes a β calculation unit 40 that calculates β _l (n) of Expression (13). The β calculator 40 receives the signal x ₀ of the first (0th) element of the digitized received signal (vector) from the ADC 12. The signal x ₀ is divided by the divider 411 by 1 + y (n) ² which is the sum of the square of the output signal y (n) of the variable coefficient filter 13 and the fixed value “1”. The output signal of the divider 411 is input to the multiplier 422, and the first element β ₀ (n) of the tap coefficient is calculated.

When a transversal filter is used for the variable coefficient filter 13 in this way, the filter required by the coefficient control unit 18 is only required to have one multiplier 422 for obtaining the first element β ₀ (n) of the tap coefficient. . The shift registers 401 to 40L obtain other elements β ₁ (n) to β _L (n) of the tap coefficients by sequentially shifting the output signal β ₀ (n) of the filter 422. The update unit 18 in FIG. 6 is expected to have a certain decrease in performance due to the introduction of approximation. However, since only one multiplier 422 in the coefficient control unit 18 is required and the amount of calculation is reduced, the processing time is high. And cost reduction.

(Third specific example of coefficient control unit)
In a third embodiment of the coefficient control unit 18 controls the filter coefficients of the variable coefficient filter 13 so as to minimize an evaluation function J ₃ defined by the following equation.

Here, J is an evaluation function for a distortion component extracted from the pilot signal band in the FM composite signal 104 described in the first and second specific examples of the coefficient control unit 18, and is, for example, the least mean square (Least). According to the Mean Square (LMS) algorithm, this is an evaluation function when an instantaneous average is used instead of an ensemble average. J _CMA is an evaluation function for an error in the conventional CMA (an error of the amplitude of the equalized output signal from the variable coefficient filter 13 with respect to the target amplitude). As shown in equation (14), in the third specific example of the coefficient control unit 18 while reducing the cost function J based on the first and second embodiment, an evaluation function J _CMA for errors in addition CMA As a result, the coefficient of the variable coefficient filter 13 can be controlled. When Expression (14) is expressed by a coefficient, the following expression is obtained.

Then, the coefficient of the variable coefficient filter 13 is updated according to the following equation.

  FIG. 7 is a block diagram illustrating the coefficient control unit 18 based on the third specific example. The coefficient control unit 18 in FIG. 7 uses both the β calculation unit 51 and the CMA calculation unit 52. 4 and 6 showing the first and second specific examples of the coefficient control unit 18, each block is described for each vector element, but in FIG. 7, the blocks are described collectively corresponding to the vector. ing. For example, the β calculation unit 51 is the same as the β calculation unit 40 in FIG. 6, and the output signal β is a vector. The multiplier 53 corresponds to the multipliers 431 to 43L in FIG. 6, and multiplies β (n) v (n) by the output signal β (n) of the β calculation unit 51 and the distortion component v (n) from the extraction filter 17. )I do. Instead of the β calculator 51, an α calculator shown in FIG. 4 may be used.

On the other hand, in the CMA calculator 52, the partial differential value y (n) of the evaluation function (hereinafter simply referred to as the CMA evaluation function) for the error of the target amplitude σ with respect to the equalized output signal y (n) from the variable coefficient filter 13 x (n) conj (y (n) ² −σ ² ) is calculated. Since the calculation of the CMA calculation unit 52 is known, a detailed description thereof will be omitted.

Next, the adder 54 adds the output signal from the multiplier 53 and the output signal from the CMA calculation unit 52, thereby generating a partial differential value of the evaluation function J ₃ represented by Expression (15). Further, the multiplier 55 multiplies the step coefficient μ. In accordance with the output signal of the multiplier 55, the filter coefficient of the variable coefficient filter 13 is updated by the adder 56 and the latch 57 as shown in Expression (16).

  As described above, according to the third specific example of the coefficient control unit 18, the filtering coefficient of the variable coefficient filter 13 is updated by the first update process based on the distortion component in the pilot signal band of the FM composite signal and the first update based on the CMA. 2 update processes are combined. Even in a multipath environment where the filter coefficient is not correctly updated in either one of the first update process and the second update process and the equalization operation is not normally performed, the filter is filtered in the other update process. There is a high probability that the coefficient will be updated normally. Accordingly, the convergence speed of the filter coefficient is improved by a kind of diversity effect.

ρはＦＭコンポジット信号を用いた評価関数Ｊの偏微分値に対する重み係数であり、ρ_CMAはＣＭＡを用いた評価関数Ｊ_CMAの偏微分値に対する重み係数である。これによって、ＦＭ放送波の伝搬状況に応じたより細かな制御が可能になり、等化速度をさらに速めることが可能である。 In equation (15), the partial differential value of the evaluation function J using the FM composite signal and the partial differential value of the evaluation function J _CMA of the _CMA are added by the adder 54. It is also possible to perform addition.

ρ is a weight coefficient for the partial differential value of the evaluation function J using the FM composite signal, and ρ _CMA is a weight coefficient for the partial differential value of the evaluation function J _CMA using the _CMA . As a result, finer control according to the propagation state of the FM broadcast wave is possible, and the equalization speed can be further increased.

  Next, another embodiment of the present invention will be described with reference to FIG. The FM receiver of FIG. 8 includes a plurality of (two in the example in the figure) reception branches, and reduces multipath distortion by space-time equalization. That is, FM broadcast waves are received by the antennas 10A and 10B, and output signals (FM reception signals) of the antennas 10A and 10B are input to the front ends 11A and 11B, respectively. The reception signals output from the front ends 11A and 11B are digitized by the ADCs 12A and 12B, respectively, and then input to the variable coefficient filters 13A and 13B. The variable coefficient filter 13 performs an equalization process for reducing multipath distortion included in each received signal by filtering the input received signal. The output signal from the variable coefficient filter 13 is synthesized by the adder 20 to generate an equalized output signal, and this equalized output signal is input to the FM demodulator 14. The FM composite signal generated by the FM demodulator 14 is input to the extraction filter 17, and the distortion component in the pilot signal band based on the multipath is extracted by the extraction filter 17.

  The coefficient control unit 18 receives the output signals of the ADCs 12A and 12B, the output signals of the variable coefficient filters 13A and 13B, and the signal of the distortion component in the pilot signal band output from the extraction filter 17, and has been described in the previous embodiment. Similarly, the filter coefficient is updated. Thus, the present invention can also be applied to FM receivers that reduce multipath distortion using space-time equalization.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the components without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

1 is a block diagram illustrating an FM receiver according to an embodiment of the present invention. The figure which shows the frequency arrangement | positioning of the composite signal in FM stereo broadcasting Diagram showing frequency characteristics of original composite signal and composite signal after being affected by multipath distortion The block diagram which shows the 1st specific example of the coefficient control part in FIG. Block diagram showing basic configuration of transversal filter The block diagram which shows the 2nd specific example of the coefficient control part in FIG. The block diagram which shows the 3rd specific example of the coefficient control part in FIG. FIG. 4 is a block diagram illustrating an FM receiver according to another embodiment of the present invention.

Explanation of symbols

10, 10A, 10B ... Antenna 11, 11A, 11B ... Front end 12, 12A, 12B ... A / D converter 13, 13A, 13B ... Variable constant filter 14 ... FM demodulator DESCRIPTION OF SYMBOLS 15 ... Arc tangent calculator 16 ... Differentiator 17 ... Extraction filter 18 ... Coefficient control part 19 ... Stereo demodulator 20 ... Adder 30 ... alpha calculation part (partial differentiation) Value calculator)
40 ... β calculation part (partial differential value calculation part)
51... Β calculator (partial differential value calculator)
52 ... CMA calculation part (partial differential value calculation part)

Claims (13)

A receiving unit that receives FM broadcast waves and outputs a received signal;
A first filter for filtering the received signal according to a variable filter coefficient;
An FM demodulator that demodulates the received signal filtered by the first filter to generate a composite signal including a main signal, a sub signal, and a pilot signal;
A second filter for extracting a distortion component generated in a target band other than the band in which the main signal and the sub signal are arranged from the composite signal;
And a coefficient control unit that controls the filter coefficient so as to minimize an evaluation function for the distortion component.   The FM receiver according to claim 1, wherein the target band is a band between the main signal and the sub signal.   The FM receiver according to claim 1, wherein the evaluation function is an ensemble average or an instantaneous average of power of the distortion component. The coefficient control unit
A calculation unit for calculating a partial differential value of the evaluation function;
A first multiplier for multiplying the partial differential value by the distortion component;
A second multiplier for multiplying the output signal of the first multiplier by a step coefficient;
The FM receiver according to claim 1, further comprising: an update unit that updates the filter coefficient according to an output signal of the second multiplier.   The FM receiver according to claim 4, wherein the calculation unit is configured to calculate the partial differential value for each element of the vector of the received signal.   The calculation unit calculates a first partial differential value corresponding to one element of the vector of the received signal, and delays the first partial differential value to delay a second partial differential value corresponding to another element of the vector. The FM receiver of claim 4 configured to approximate a value. A receiving unit that receives FM broadcast waves and outputs a received signal;
A first filter for filtering the received signal according to a variable filter coefficient;
An FM demodulator that demodulates the received signal filtered by the first filter to generate a composite signal including a main signal, a sub signal, and a pilot signal;
A second filter for extracting a distortion component generated in a target band other than the band in which the main signal and the sub signal are arranged from the composite signal;
A coefficient control unit that controls the filter coefficient so as to minimize the added value of the first evaluation function for the distortion component and the second evaluation function for the error with respect to the target amplitude of the received signal filtered by the first filter. And an FM receiver.   The FM receiver according to claim 7, wherein the target band is a band between the main signal and the sub signal.   The FM receiver according to claim 7, wherein the first evaluation function is an ensemble average or an instantaneous average of power of the distortion component. The coefficient control unit; a calculation unit for calculating a partial differential value of a first evaluation function for the distortion component;
A first multiplier for multiplying the partial differential value of the first evaluation function by the distortion component;
An adder for adding the output signal of the first multiplier and the partial differential value of the second evaluation function;
A second multiplier for multiplying the output signal of the first multiplier by a step coefficient;
The FM receiver according to claim 7, further comprising: an update unit that updates the filter coefficient according to an output signal of the second multiplier.   The FM receiver according to claim 7, wherein the calculation unit is configured to calculate the partial differential value for each element of the vector of the received signal.   The calculation unit calculates a first partial differential value corresponding to one element of the vector of the received signal, and delays the first partial differential value to delay a second partial differential value corresponding to another element of the vector. 8. The FM receiver of claim 7, configured to approximate a value.   The FM receiver according to claim 7, wherein the adder is configured to perform weighted addition of a partial differential value of a second evaluation function and an output signal of the first multiplier.

Images