Method for limiting radio noise, in particular in the FM band, by polynomial interpolation

Abstract

A method for limiting impulse noise in a radio signal, in particular in the FM band, that is multiplexed and intended to be received by an FM receiver that is for example located on-board a vehicle. The method includes detecting impulse noise by way of a calculation of a score IND, for a sequence of successive samples of the demodulated radio signal FM MPX, while taking into the account, on the one hand, a capacity to detect a drop in the modulus of the ratio I/Q, which is dependent in particular on data derived from a sensor level, and, on the other hand, a capacity to detect high-frequency noise, which in particular depends on data derived from a sensor modulation.

Claims

1. A method for limiting noise in a received radio signal, especially in an FM band, said received radio signal being multiplexed according to a quadrature amplitude modulation scheme, said modulated received radio signal comprising an in-phase component and a quadrature component, the received radio signal being demodulated in order to consist of a succession of samples, and said radio signal including a risk of impulse noise liable to degrade a returned radio signal, said method comprising: detecting a drop in a modulus of a ratio between the in-phase component and the quadrature component of the modulated received radio signal, depending on an average modulus over time of the modulated received radio signal; detecting high-frequency noise in the demodulated received radio signal, depending on a threshold, said threshold being determined dynamically depending on the modulation of the modulated received radio signal; calculating an impulse noise score, said impulse noise score being dependent on the drop on the modulus detection and on the high-frequency noise detection; depending on the impulse noise score calculated, determining a number of samples, forming a sequence of samples to be substituted, between a last preserved non-noisy sample and a restart non-noisy sample, in the demodulated received radio signal; creating a sequence of substitution samples, by polynomial interpolation between the last preserved non-noisy sample and the restart non-noisy sample; and returning a modified demodulated radio signal corresponding to the demodulated received radio signal in which the sequence of samples to be substituted has been replaced by the sequence of substitution samples.

2. The method as claimed in claim 1, wherein, the polynomial interpolation is based on a polynomial estimated depending on a plurality of non-noisy samples located before the sequence to be substituted in the demodulated received radio signal and on a plurality of non-noisy samples located after the sequence to be substituted in the demodulated received radio signal.

3. The method as claimed in claim 1, wherein the detecting detec.sub.drop.sub.n the drop in the modulus of the ratio between the in-phase component and the quadrature component of the received radio signal equals: $\mathrm{norm}\left(I_n,Q_n\right)\begin{array}{c}1\\ \\ 0\end{array}{\mathrm{Thr}}_{{\mathrm{drop}}_n}$ where Thr.sub.drop.sub.n=f(Sensor level) and f() is a linear function.

4. The method as claimed in claim 1, wherein the detecting detec.sub.noise.sub.n, depending on the dynamically determined threshold, the high-frequency noise in the demodulated received radio signal, equals: $\mathrm{HPF}\left({\mathrm{MPX}}_n\right)\begin{array}{c}1\\ \\ 0\end{array}{\mathrm{Thr}}_{{\mathrm{noise}}_n}$ where Thr.sub.noise.sub.n=f(Modulation sensor) and f() is a linear function.

5. The method as claimed in claim 1, wherein, a score IND is equal to:
.sub.n=0.sup.N1(detec.sub.drop.sub.n& detec.sub.noise.sub.n) where N is the number of samples of the demodulated received radio signal (FM MPX) of which account is taken, corresponding for example to the size of a dedicated buffer memory.

6. The method as claimed claim 1, wherein the polynomial interpolation is implemented between a preceding sequence of clean samples terminating with the last preserved sample and a subsequent sequence of clean samples starting with the restart sample.

7. A radio-receiving device able to receive and to demodulate a multiplexed radio signal, comprising a noise-limiting device comprising means for implementing the method for limiting noise in a received radio signal, as claimed in claim 1.

8. The radio receiving device as claimed on claim 7, wherein the received radio signal is in the FM band.

9. The method as claimed in claim 3, wherein the detecting detec.sub.drop.sub.n the drop in the modulus of the ratio between the in-phase component and the quadrature component of the received radio signal equals: $\mathrm{norm}\left(I_n,Q_n\right)\begin{array}{c}1\\ \\ 0\end{array}{\mathrm{Thr}}_{{\mathrm{drop}}_n}$ where Thr.sub.drop.sub.n=f(Sensor level) and f() is a linear function.

10. The method as claimed in claim 2, wherein the detecting detec.sub.noise.sub.n, depending on the dynamically determined threshold, the high-frequency noise in the demodulated received radio signal, equals: $\mathrm{HPF}\left({\mathrm{MPX}}_n\right)\begin{array}{c}1\\ \\ 0\end{array}{\mathrm{Thr}}_{{\mathrm{noise}}_n}$ where Thr.sub.noise.sub.n=f(Modulations sensor) and f() is a linear function.

11. The method as claimed in claim 3, wherein the detecting detec.sub.noise.sub.n, depending on the dynamically determined threshold, the high-frequency noise in the demodulated received radio signal, equals: $\mathrm{HPF}\left({\mathrm{MPX}}_n\right)\begin{array}{c}1\\ \\ 0\end{array}{\mathrm{Thr}}_{{\mathrm{noise}}_n}$ where Thr.sub.noise.sub.n=f(Modulation sensor) and f() is a linear function.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Aspects of the invention will be better understood upon reading the following description, given solely by way of example, and with reference to the appended drawings, in which:

(2) FIG. 1 shows a block diagram illustrating the operation of the method for limiting impulse noise in a radio signal, according to an aspect of the invention;

(3) FIGS. 2A-2C show graphs illustrating the quality of the processing obtained by implementing the method according to an aspect of the invention with respect to the prior art.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(4) It should be noted that the figures disclose an aspect of the invention in a detailed manner so as to enable the implementation thereof, said figures also being able to serve to better define an aspect of the invention, of course.

(5) An aspect of the invention is mainly presented with a view to an implementation in a radio receiver located on board a motor vehicle. However, other applications are also targeted by aspects of the present invention, in particular for the purpose of an implementation in contexts other than a motor vehicle.

(6) With reference to FIG. 1, the method for limiting multi-path or impulse noise in a received radio signal FM comprises calculating a score IND associated with noise present in the received radio signal FM. This score IND allows smart detection of noise.

(7) Thus, the FM received signal is modulated, as is known, using a multi-path modulation technique, the modulated FM received signal having an in-phase component I and a quadrature component Q.

(8) It is also known that the modulus of the ratio I/Q is able to reveal the probability of presence of impulse noise in the received radio signal FM. Thus, the method according to an aspect of the invention makes provision for a detection of the drop in the I/Q modulus, denoted I/Q in FIG. 1.

(9) Specifically, it is more probable to find impulse noise in the received radio signal FM when the I/Q modulus drops spectacularly.

(10) According to an aspect of the invention, the detection of the drop in the I/Q modulus consists in comparing the I/Q modulus at a given time to a threshold determined dynamically depending on the longer-term variation in the level of the radio signal, in other words depending on the average modulus over time of the modulated received radio signal, which is given by the sensor level SL, which indicates the radiofrequency (RF) level of the received radio signal FM.

(11) The method according to an aspect of the invention moreover makes provision for the detection of high-frequency noise in the demodulated radio signal FM MPX, via a sensor noise, which indicates the existence of high-frequency components in the demodulated radio signal FM MPX, while taking into account sensor modulation SM, which indicates the modulation level of the desired signal. Thus, the high-frequency-noise sensor SHF detects the occurrence of high-frequency peaks in the demodulated radio signal FM MPX, by comparing these peaks to a frequency threshold determined dynamically depending on the modulation of the desired signal, which is given by the sensor modulation SM. Thus, the higher the modulation index, the higher the threshold of detection of peaks.

(12) The threshold of detection of the peaks, which are characteristic of the impulse noise, thus varies dynamically. The dynamic determination of these thresholds allows the detection of false positives (peaks not corresponding to impulse noise) to be avoided, and thus distortion in the returned radio signal to be minimized. Thus, if the modulation of the signal increases, the threshold of detection of these peaks also increases.

(13) As is shown in FIG. 1, the information output by the high-frequency-noise sensor SHF and the information relating to the detection of any drop in the modulus of the ratio I/Q is then combined to calculate the score IND.

(14) To generate the information output by the high-frequency-noise sensor SHF, account is taken of the sensor modulation SM and, to allow the detection of a drop in the modulus of the ratio I/Q, account is taken of the sensor level SL. In this way, the method according to an aspect of the invention allows the presence of peaks in the received radio signal FM, which peaks are characteristic of the presence of impulse noise in said received audio signal FM, to be detected dynamically.

(15) To this end, said information output by the high-frequency-noise sensor SHF and the information relating to the detection of any drop in the modulus of the ratio I/Q is therefore compiled to calculate the score IND, on the basis of which the method according to an aspect of the invention determines the length of the signal portion to be replaced in the demodulated received radio signal FM MPX.

(16) Specifically, the demodulated radio signal FM MPX consists of successive samples. The score IND gives a probability of having impulse noise as a function of time. For example, in a sequence of N samples, from 0 to N samples will be considered to be noisy. Thus, the method according to an aspect of the invention makes provision to determine a number of samples to be replaced, between a last preserved non-noisy sample and a restart non-noisy sample.

(17) According to an aspect of the invention, a sequence of substitution samples is then created by polynomial interpolation IP between said last preserved non-noisy sample and said restart non-noisy sample. To further improve the quality of the polynomial interpolation IP, according to one embodiment, account is taken of a sequence of clean samples terminating with the last preserved sample and a sequence of clean samples starting with the restart sample.

(18) According to one embodiment, the polynomial interpolation IP is based on a polynomial the coefficients of which are estimated depending on a plurality of non-noisy samples located before the sequence of samples to be substituted in the demodulated received radio signal FM MPX and a plurality of non-noisy samples located after the sequence of samples to be substituted in the demodulated received radio signal FM MPX.

(19) It will be noted that the number of samples used to form said plurality of non-noisy samples, before and after the sequence of samples to be substituted, varies depending on the length of the sequence to be substituted, which corresponds to a radio signal portion to be replaced. Specifically, the longer the noisy received radio signal portion, the higher the number of clean samples, i.e. non-noisy samples, that will be required, before and after the sequence to be substituted, to achieve a satisfactory construction of the sequence of substitution samples.

(20) Moreover, the number of samples used to form said plurality of non-noisy samples, before and after the sequence of samples to be substituted, also varies depending on the frequency at which noise occurs in the received radio signal FM.

(21) Thus, in other words, the corrupted signal portion is replaced by a signal synthesized on the basis of a polynomial. The coefficients of this polynomial are estimated using the clean signals before and after the portion polluted by the noise.

(22) Given that the sequence of substitution samples has the same frequency component as its neighbors, the distortion created by this processing is minimal.

(23) The method according to an aspect of the invention in particular has the following advantages, namely: i) the modification of the threshold of detection of peaks depending on the modulation level of the desired signal allows the detection of false positives to be avoided; and ii) replacing the corrupted signal portion with a polynomial interpolation estimated on the basis of neighboring clean signals minimizes the distortion associated with the removal of the noise.

(24) FIG. 2 shows the results of simulations relating to the limitation of noise in a noisy received radio signal FM (FIG. 2A), according to prior-art methods (FIG. 2B) and according to an aspect of the invention (FIG. 2C).

(25) As these graphs show, the signal returned after processing using the method according to an aspect of the invention (FIG. 2C) is exempt both of impulse noise and discontinuities.

(26) In summary, an aspect of the present invention proposes a method for limiting impulse noise in a radio signal, especially in the FM band, that is multiplexed and intended to be received by an FM receiver that is for example located on-board a vehicle.

(27) The method according to an aspect of the invention comprises detecting and evaluating impulse noise by way of the calculation of a score IND, for a sequence of successive samples of the demodulated radio signal FM MPX, while taking into the account, on the one hand, a capacity to detect a drop in the modulus of the ratio I/Q, which is dependent in particular on data derived from a sensor level SL, and, on the other hand, a capacity to detect high-frequency noise, which in particular depends on data derived from a sensor modulation SM.

(28) According to one embodiment, the detection detec.sub.drop.sub.n of a drop in the modulus of the ratio I/Q equals:

(29) $\mathrm{norm}\left(I_n,Q_n\right)\begin{array}{c}1\\ \\ 0\end{array}{\mathrm{Thr}}_{{\mathrm{drop}}_n}$

(30) where Thr.sub.drop.sub.n=f(Sensor level) and f() is a linear function.

(31) According to one embodiment, the detection detec.sub.noise.sub.n of high-frequency noise equals:

(32) $\mathrm{HPF}\left({\mathrm{MPX}}_n\right)\begin{array}{c}1\\ \\ 0\end{array}{\mathrm{Thr}}_{{\mathrm{noise}}_n}$

(33) where Thr.sub.noise.sub.n=f(Modulation sensor) and f() is a linear function.

(34) Thus, according to one embodiment, in a sequence of N samples, for example corresponding to the size of a dedicated buffer memory, the score IND is given as a mark ranging from 0 to N1, where 0 indicates a very clean radio signal and N1 indicates a very noisy radio signal. In this context, the score IND may be obtained by virtue of the following equation:
IND=.sub.n=0.sup.N1(detec.sub.drop.sub.n & detec.sub.noise.sub.n),

(35) where N is the number of samples of the demodulated radio signal FM MPX of which account is taken, corresponding for example to the size of a dedicated buffer memory.

(36) Depending on this score IND, the polynomial interpolation proposed by an aspect of the present invention, to replace a polluted signal portion with a reconstructed signal portion, is based on the use of said score IND in order to cancel out a maximum of impulse noise, without generating distortion in the returned radio signal.

(37) It will moreover be noted that an aspect of the present invention also relates to a device for limiting impulse noise, which device is intended for a radio receiver, in particular for an FM radio receiver, implementing the method described above.

(38) Aspects of the present invention are in particular intended to be implemented in the automotive field, but may also be implemented in other fields.

(39) It is clarified furthermore that aspects of the present invention are not limited to the examples described above, and is open to many variants that are accessible to those skilled in the art.