WO2006114513A1

WO2006114513A1 - Method for transmitting a signal modulated with high amplitude dynamics, corresponding transmitter and receiver

Info

Publication number: WO2006114513A1
Application number: PCT/FR2006/000915
Authority: WO
Inventors: Christian Lereau; Salvatore Ragusa; Jacques Palicot
Original assignee: France Telecom
Priority date: 2005-04-28
Filing date: 2006-04-24
Publication date: 2006-11-02

Abstract

The invention relates to a method for transmitting a modulated radio signal (MRS) with high amplitude dynamics subjected to a power amplification before transmission. The method consists of at least subjecting, at the transmission (A), upstream from the power amplification, the signal (MRS) to a specific clipping (E) whereby shortening the distance from the backward point vis-à-vis the zone of saturation of amplification and, after transmission (B), at reception (C), submitting (D) the processed modulated radio signal (TMRS) to an inverse clipping (E-1) whereby restoring the modulated radio signal with unclipped amplitude dynamics (≃MRS). The invention is for use in radio frequency transmitters/receivers of mobile devices.

Description

METHOD FOR TRANSMITTING MODULATED SIGNAL WITH HIGH AMPLITUDE DYNAMICS, TRANSMITTER AND RECEIVER

CORRESPONDING

The invention relates to a method for transmitting a modulated signal with a high amplitude dynamic, a transmitter and a corresponding receiver.

At present, many problems, even technological challenges, are posed in the field of on-board equipment, with the aim of meeting a growing demand for high-speed transmission resources enabling the development of innovative services.

In particular, spectral congestion is subject to strong regulatory constraints, we are now witnessing the appearance of modulation processes whose spectral efficiency is high, processes such as EDGE, UMTS, OFDM or others.

Unfortunately, such modulation processes generate signals with envelope, or amplitude, which are not constant and which require a great linearity on the transmission and reception chains, in particular on the radiofrequency power amplifier. The latter, whose great linearity is required at the expense of efficiency, because the criteria of linearity and efficiency are substantially antagonistic, appears accordingly as a critical component, especially when such a component is integrated into a mobile equipment whose Electric power and available range are limited.

For signals with a substantially constant envelope generated for example by the GMSK modulation process, it is usual to operate the radio frequency power amplifier (RF power amplifier) in an operating zone close to saturation, so as to guarantee optimum performance at maximum output power.

However, in the context of a non-constant envelope modulation, the RF power amplifier must be able to reproduce the information contained in the amplitude and phase parameters of the signal. Consequently, the technique currently used consists in placing the aforementioned amplifier at a point of operation such that the peak power of the signal thus reached is less than the saturation power, to avoid clipping. By the abovementioned technique called the technique of the point of recoil, or back-off in English, this result is achieved, at the cost of a degradation of the efficiency and performance of the RF power amplifier, however. . ^• .

This deterioration of the yield is further accentuated by the process of controlling the electric power of emission, made necessary on many cellular networks, for example, for which, according to the conditions of interference and radioelectric propagation, it is imposed on the corresponding transmitter to limit the transmit power to lower levels.

In the case of a cellular network (GSM, UMTS, etc.), and according to its granularity, the statistics of the radio powers emitted as a percentage of the transmission time by the transmitter of a mobile telephone terminal, Gaussian type, significantly confirms that 90% of the time the RF power amplifier delivers a much lower power than the maximum allowable power, set by the standard.

As a corollary, the implementation of the above-mentioned technique of the setback point requires the obtaining of very low yields, at maximum power, transmitters equipping the aforementioned mobile telephone terminals. The control of the power emitted will accentuate the degradation of the efficiency because for the cellular networks ¹ , the power emitted is less than 90% of the service time at the maximum power specified, which leads to a consequent degradation of the autonomy of this type equipment.

Thus, in summary, it appears that the amplitude dynamics of the envelope signal of the modulated signal directly induces the degradation of the output of the power amplifier.

In order to improve the above-mentioned performance and autonomy, many studies have proposed: The reduction of the dynamics of envelope, these techniques making it possible to release the constraints related to the point of recoil, for joint digital and / or analog treatments, such as addition of signal, clipping or others.

The design of an optimized linearized amplifier chain by: "compensation of residual faults of the power amplifier, by digital processing techniques, pre-distortion, use of equivalent methods of amplification, using a saturated amplifier These include the Envelope Elimination and Restoration (EER) method, which includes the approach of not clipping the signal and operating the power amplifier. in linear area has the major disadvantage of increasing power amplifier consumption.

The approach of introducing clipping of the signal before amplification has the merit of optimizing the efficiency of the power amplifier, but requires additional filtering, in order to limit the rise of energy in the adjacent channels, while accepting some signal distortion.

The present invention aims to overcome the disadvantages and limitations of the aforementioned approaches of the prior art.

In particular, an object of the present invention is the implementation of a method for transmitting a modulated signal with a high amplitude dynamic, allowing, on the one hand, to improve the efficiency of the amplifier. power of the transmitters of such a signal, and, secondly, to obtain in reception a quality of service comparable to that which is obtained by the operating mode of such an amplifier in substantially linear mode, but with a higher efficiency. low. The method of transmitting a modulated signal with a high amplitude dynamic, this signal being subjected to a power amplification, prior to its transmission, which is the subject of the present invention, is remarkable in that it consists at least in subject, on transmission, upstream of this power amplification, this signal modulated to a specific clipping processing, so as to reduce the distance of the recoil point, with respect to the saturation zone, necessary to this amplification of power and, following the transmission of this processed modulated signal, subject, on reception, this processed modulated signal to a specific inverse clipping processing, so as to restore this modulated, uncapped signal.

The method for transmitting a modulated signal with a high amplitude dynamic, which is the subject of the present invention, applies to the transmission of all the non-constant envelope signal signals obtained by the modulation processes such as the OFDM multi-carrier modulation for Orthogonal Frequency Division Multiplex in English, used in 802.11a / g wireless LANs, with data rates of up to 54 Mbps in short-range, DAB digital broadcast modulation for Digital Audio Broadcasting in English, DVB-T terrestrial video modulation for Digital Video Broadcasting Terrestrial in English, DVD-H Mobile Video Modulation for Digital Video Broadcasting - Handset in English.

The invention furthermore relates to: a method for transmitting a modulated signal, with a strong amplitude dynamic, characterized in that it consists in subjecting this signal, upstream of the power amplification from the latter, to a specific clipping processing, so as to reduce the distance of the recoil point, vis-à-vis the saturation zone necessary for this amplification, to generate a processed modulated signal; a method of receiving a modulated signal, with a strong amplitude dynamic, subjected to the emission upstream of the power amplification of the latter to a specific clipping process, to generate a processed modulated signal, characterized by subjecting said processed modulated signal to an inverse specific clipping processing so as to render said modulated signal with a strong amplitude dynamic; a transmitter of a modulated signal with a high amplitude dynamic having at least one power amplifier connected to a transmitting antenna, characterized in that said radio frequency transmitter comprises at least one specific clipping stage of said modulated signal placed in upstream of said amplifier power, so as to reduce the distance from the recoil point to the saturation region of said power amplifier, and output a processed modulated signal; a receiver of a modulated signal with a strong amplitude dynamic and processed, on transmission, upstream of a power amplification by a specific clipping processing, so as to reduce the distance from the setback point, screws to the saturation zone, necessary for this power amplification, characterized in that this receiver comprises at least one inverse specific clipping stage, so as to restore said modulated signal with a strong amplitude dynamic, not clipped; a clipping module of a modulated signal, characterized in that it consists of a non-linear characteristic component whose sampled transfer function is represented by a positive, monotonic and increasing function of the amplitude of the samples of said signal modulated; and. an inverse clipping module of a modulated signal with a strong amplitude dynamic and processed, on transmission, upstream of a power amplification by a clipping processing represented by a monotonic and increasing positive function of the amplitude of the samples of the modulated signal, characterized in that said inverse clipping module is constituted by a non-linear characteristic component whose sampled transfer function is represented by the inverse function of said clipping function. It will be better understood on reading the following description and on observing the drawings in which: FIG. 1a is a purely illustrative flow chart of the essential steps of the method of transmitting a modulated signal with a strong dynamic object of the invention, in the context of a specific transmission phase and a reception phase; FIG. 1b represents, by way of illustration, the distribution law of the module of a symbol in OFDM modulation, used for example in 802.11a / g wireless local area networks, the aforementioned distribution law making it possible to justify the interest of the implementation of the method which is the subject of the present invention as illustrated in FIG. FIG. 2a represents, by way of illustration, an advantageous variant of implementation of the method that is the subject of the present invention in its transmission phase; FIG. 2b represents, by way of illustration, an advantageous variant of implementation of the method that is the subject of the present invention in its reception phase; FIG. 2c represents, by way of illustration, a specific choice diagram of the non-linearity generated by the clipping process on transmission respectively by the inverse clipping process on reception, with respect to the non-linearity intrinsic linearity of the amplification of power necessary for the transmission of a signal and making it possible to mask and then substantially eliminate this intrinsic nonlinearity; FIG. 3a represents, by way of illustration, a block diagram of a transmitter according to the object of the present invention, more particularly intended for the implementation of the method which is the subject of the invention, as illustrated in FIG. in its emission phase; FIG. 3b represents, by way of illustration, a nonlinear transfer characteristic or function applied to the transmission at the modulated signal in order to exert a clipping processing on the latter respectively a nonlinear transfer characteristic or function applied to the reception of the modulated signal processed to perform inverse clipping processing on the latter; FIG. 3 c represents, by way of illustration, a block diagram of a receiver according to the subject of the present invention, more particularly intended for the implementation of the method which is the subject of the invention, as illustrated in FIG. the, in its reception phase;

FIG. 4 a represents, in a non-limiting way, a three-dimensional diagram of the variation of the bit error rate during the transmission of a modulated signal according to the method that is the subject of the present invention, as a function of the clipping factor and the distance from the point of departure of the average input power with respect to the saturation power, such a diagram making it possible to highlight an optimum zone; FIG. 4b represents, in a non-limiting way, a timing diagram of a modulated signal to be transmitted, of a processed signal transmitted after clipping, thanks to the transmission phase, then of an uncapped signal restored after inverse clipping, thanks to at the reception phase, in accordance with the method that is the subject of the present invention; FIG. 4c represents, by way of illustration, a diagram enabling the gain gain to be demonstrated, without major distortion obtained, in a transmission chain of a modulated signal incorporating a transmitter and a receiver illustrated in FIG. 3a and 3; vs.

A more detailed description of the method of transmitting a modulated signal with a high MRS amplitude dynamic, in accordance with the subject of the present invention, will now be given in connection with FIG. 1a and the following figures.

The method which is the subject of the invention is implemented for the aforementioned signal, this signal being subjected to a power amplification prior to its transmission. With reference to FIG. 1a, the method which is the subject of the invention consists in subjecting the MRS modulated signal to a specific clipping processing, noted on transmission, in a step A and upstream of the power amplification. E, so as to allow the reduction of the distance of the operating back-up point, with respect to the saturation zone, necessary for the implementation of the power amplification. The notion of distance from the operating recoil point can advantageously be expressed as the difference expressed in dB between the average power of the signal obtained after clipping processing and the input power of the amplification causing the saturation of the clutch. amplification, as will be described later in the description. Following the amplification operation B and the transmission of the TMRS signal designating the obtained processed modulated signal, after specific clipping processing, the amplification and transmission operation being denoted B in FIG. the invention consists in subjecting, in a step C, the processed modulated signal TMRS to a specific inverse clipping processing, noted for this reason E ^"1 , so as to restore the modulated signal with a strong dynamic amplitude and substantially unclipped. The inverse specific clipping processing operation is performed in step D of FIG. Substantially, after the above-mentioned step D, the signal of MRS origin, noted for this reason = MRS, is obtained. In a more specific manner, with reference to FIG. 1a, it is indicated that the method which is the subject of the invention consists in applying, on transmission, the specific clipping processing E on the modulated signal MRS, either in a band of RBB base for Radio Base Band in English, either IF intermediate frequency or RF radio frequency. In a symmetrical manner, the method which is the subject of the invention consists in applying, on reception, the inverse specific clipping processing E ^"1 on the processed TMRS signal, either in baseband RBB or in intermediate frequency IF, or in RF radio frequency.

A justification of the implementation of the transmission method, object of the present invention, and the specific steps of specific clipping or inverse specific clipping respectively, applied to the transmission respectively to the reception, will now be given in connection with the FIG. 1b, in the nonlimiting case where the modulated signal is obtained by an OFDM multi-carrier type modulation for Orthognal Frequency Division Multiplex in English used in 802.11a / g wireless local area networks with data rates of up to 54 Mbps for short staves.

The aforesaid proof given in this situation is not. not limiting because this. Modulation type is also used in many digital broadcast networks such as DAB broadcast for Digital Audio Broadcasting in English, DVB-T terrestrial video for Digital Video Broadcasting-Terrestrial in English and DVB-H mobile video for Digital Video Broadcasting -Handset in English.

In principle, the OFDM modulation method consists of dividing a high speed serial data stream into N low-rate streams, respectively modulating and simultaneously N sub-carrier in narrow band, which have the distinction of being orthogonal to each other in order to optimize Spectral congestion The transformation of a frequency selective channel into N non-selective channels has the advantage of high robustness to frequency fading due to multiple paths and thus allows a simplified equalization on reception.

However, the major defect of such a modulation process is to generate a modulated signal whose time envelope has a great dynamic.

In order to quantify the aforementioned dynamics, it is customary to use a characteristic power ratio, the ratio PAPR for Peak to Average Power Ratio in English, ratio between the peak power and the average power of the signal.

In a transmission system using an OFDM type modulation process, a serial stream of binary information, after channel coding, time interleaving, is divided into packets to form N complex symbols, each denoted S _k , according to a modulation scheme. appropriate channel such as QPSK, 16 QAM, 64 QAM, or others.

The aforementioned complex symbols are, after parallelization, subjected to a fast inverse Fourier transform to form a complex symbol S _n in the time domain, each complex symbol satisfying the relation (1):

(1) S _n = X _n + JY _n = 1 / y [N * (Σs _k * _e ⁽ W * - ' ^{) (»} -υ ^{/ Λ} 0) _{5 for} i < _n <N

K = I

Assuming the statistically independent and uniformly distributed N complex symbols S _k , the respective statistical data representative of the real and imaginary parts of each symbol can be assimilated to Gaussian forms for a large number of subcarriers of zero mean and of variance σs ¹ according to the central limit theorem.

In addition, the aforementioned statistical data make it possible to deduce, with reference to FIG. 1b, that the distribution function of the module, that is to say the amplitude of each OFDM symbol S _n , this module satisfying the relation 2 :

follows a Rayleigh law, of the form:

( ³ ) P (P _n ) = P, ' In the previous relation P (p _n ) represents the probability of occurrence of a symbol of amplitude p _n and σs ² represents the variance of the aforementioned distribution law.

The distribution of the module of each OFDM symbol according to the Rayleigh law represented in the figure allows it to retain the following essential conclusions: the large amplitudes, as well as the small amplitudes, of the OFDM symbols have very low probabilities of appearance; the average value of the amplitudes has the highest probability of occurrence.

The first conclusion then confirms that it is possible to interrupt the signal up to a certain threshold level, in accordance with the transmission method, object of the present invention, and which should be defined, depending on the quality. desired service quality, for example the quality of service can be defined by the bit error rate designated BER for Bit Error Ratio in English.

By clipping the signal, it is therefore possible to reduce the distance of the average power from the power of saturation amplification and thus optimize the efficiency of power amplification.

However, the clipping operation is necessarily non-linear and consequently, this operation may advantageously be followed by a filtering operation which limits the rise in energy in the adjacent channels, as will be described later in the drawing. description.

The method according to the invention then consists, due to the specific clipping introduced on transmission, in implementing a specific inverse clipping technique on reception, this inverse specific clipping technique acting as a clipping technique. a process of decompression of the signal, on reception, in order to reproduce as faithfully as possible the original signal, in baseband for example.

The specific clipping and inverse specific clipping processes are then complementary and chosen because of their great simplicity as to their implementation, the clipping function being positive, monotonous and increasing. It is furthermore understood that the specific clipping or inverse specific clipping processes justified in particular for a base band modulated radio signal, with reference to FIG. 1b, can of course also be implemented either intermediate frequency IF, RF radio frequency within the scope of the subject of the present invention.

A preferred mode of implementation of the method for transmitting a modulated radio signal with a high amplitude dynamic, in accordance with the subject of the present invention, will now be described in connection with FIGS. 2a to 2c. With reference to FIG. 2a, it is indicated that the specific clipping processing executed in the sub-step A ₁ of the aforementioned figure is followed in a sub-step A ₂ by a filtering, such as Nyquist filtering, for example. so as to limit the inter-symbol interference and the rise of the secondary lobes in the adjacent channels, as mentioned previously in the description. For an OFDM type modulation as described above in the description in the context of an implementation of the method that is the subject of the invention with reference to the 802.11 a / g standard and for an OFDM type modulation comprising 52 subcarriers, the clipping processing E and the inverse clipping processing E ^"1 can be performed in baseband, in which case the above-mentioned filtering is a Nyquist type filtering of 10 MHz bandwidth with a rounding factor designated Roll-off factor in English of 0.35 It should be noted that many transmission systems already include such a filtering which allows to say that the innovative process does not add additional complexity.

Similarly, with reference to FIG. 2b, it is indicated that, on reception, the inverse specific clipping processing performed in the substep D ₂ of FIG. 2b is preceded by a filtering performed in the substep D ₁ . filtering type is however limited to noise filtering, simply limiting the noise frequency band, to maintain the signal-to-noise ratio of the modulated signal, restored, uncapped. With reference to FIGS. 2a and 2b, it is indicated that the filterings executed in sub-steps A ₂ , on transmission, respectively D ₁ on reception, are not symmetrical.

A specific mode of implementation of the method, object of the present invention, will now be described in connection with Figure 2c showing an input amplitude diagram IA output amplitude OA.

The transmission method object of the present invention is particularly suitable for industrial implementation, taking into account the imperfections and nonlinearities of the power amplification applied to the modulated signal, in particular by any commercially available consumer device.

Thus, with reference to FIG. 2c, for an emission power amplification having a specific gain / amplitude nonlinearity, this nonlinearity being represented by the dotted line curve denoted by PA for the power amplification, and illustrated by a non-linear curve of slope less than the line of slope 1 representing the perfect linearity, shown in phantom in the figure _. 2c, the specific clipping processing E is chosen to represent a nonlinear attenuation / amplitude, substantially greater than the non-linearity gain / amplitude of the above power amplification.

Thus, in FIG. 2c, the nonlinearity attenuation / amplitude of the specific clipping processing E has a distance greater than the slope line 1 of FIG. 2c, to that of the distance of the nonlinearity gain / amplitude of the PA power amplification.

Similarly, and in accordance with a remarkable aspect of the transmission method object of the present invention, the nonlinearity inverse attenuation / amplitude of the inverse specific clipping processing E ^" can then be chosen substantially symmetrical to that of the specific clipping processing E , with respect to the straight slope 1 in FIG. 2c.

With reference to FIG. 2c, for any modulated radio signal of amplitude I subjected to specific clipping E, after clipping, a clipped amplitude EI is obtained, then for any processed modulated signal of amplitude E1, subjected to inverse specific clipping processing E ^"1 is obtained, as shown in Figure 2c, a modulated signal restored amplitude E ^-1 EI, that is to say substantially equal to I.

It is understood, in particular, that the specific gain / amplitude nonlinearity of the power amplification located between the nonlinearity curve of the specific clipping E respectively of the specific inverse clipping E ^"1 is then masked, as represented 2c, the abscissa and ordinate axes are graduated in input amplitude IA respectively of output OA of the modulated signal subjected either to the specific clipping processing E, or to the specific processing of FIG. 'inverse clipping E ^{' 1} .

As regards the implementation of the specific clipping processing, it is indicated that when, in a non-limiting embodiment, the method which is the subject of the invention is applied to a baseband for example, the latter may be executed on a digital version of the modulated signal. In this case, the specific clipping processing is applied via a non-linear characteristic whose sampled transfer function is represented by a positive, monotonic and increasing function of the amplitude of the samples of the modulated radio signal. For example, a polynomial function with odd coefficients corresponding to a positive, monotonous and increasing function satisfies the relation (4):

By definition, the aforementioned polynomial function is non-linear and the choice of an odd-numbered polynomial function is justified by the advantage of limiting the level of second order intermodulation products, which are likely to be in the band. of useful signal and thus to degrade the signal-to-noise ratio.

Similarly, with respect to the inverse specific clipping processing, it is indicated that it can be applied under the same conditions via a characteristic whose sampled transfer function is represented by another polynomial function with odd coefficients of the amplitude of the samples of the processed signal.

This other polynomial function satisfies the relation (5):

(5) f (x) = Σlb _2M * x ^2M l

/ c = Q The relationship between the aforementioned polynomial functions is given by the symmetry with respect to the slope line 1 of Figure 2c.

A more detailed description of a transmitter and a radio frequency receiver of a modulated signal with a high amplitude dynamic, in accordance with the subject of the present invention, will now be given in connection with FIGS. 3a to 3c. In general, it is indicated that the emitter and the radio-frequency receiver of a modulated signal with a high amplitude dynamic, in accordance with the subject of the present invention, are implemented for a modulated signal. baseband in a non-limiting manner.

With reference to FIG. 3a, the transmitter that is the subject of the invention may comprise, as such, a digital part D, in which the baseband processing is performed, in particular the specific clipping processing E, and a analog part A, for ensuring, on the one hand, the transformation of the digital signal into baseband after clipping into an analog signal, and then a translation of the signal to the radio frequency carrier, before power amplification PA and, on the other hand, on the other hand, the radiation via a TA transmit antenna after RFF filtering.

As shown in FIG. 3a, the radio frequency transmitter, which is the subject of the present invention, comprises at least one power amplifier PA which is connected to a radio frequency stage conventionally comprising a frequency change stage on a phase-in-phase channel. I and a quadrature phase channel Q, the aforementioned analog channel A and each of the quadrature-phase phase-switched channels receiving, via a digital analog converter CNA, the processed modulated signal, which has been subjected to baseband, ie in the form of a digital signal, to the aforementioned specific clipping processing on each of the phase I and Q phase quadrature channels respectively. As shown in FIG. 3a, the radio frequency transmitter notably comprises, for this purpose, at least one specific clipping stage of the modulated signal placed upstream of the power amplifier.

In FIG. 3a, the specific clipping stage is considered to comprise a clipping module on the in-phase channel, module Ei and a clipping module on the quadrature-phase channel, module E _Q. This makes it possible to reduce the distance of the reversal point with respect to the saturation zone of the power amplifier PA, as mentioned previously in the description, and to deliver the processed modulated signal to the power amplifier PA. According to a remarkable aspect of the radio frequency transmitter object of the present invention, it is indicated that the specific clipping stage is constituted by a nonlinear characteristic component whose sampled transfer function is represented by a positive, monotonic function. and increasing, in this case, a polynomial function with odd coefficients of the amplitude of the samples of the modulated signal.

With reference to FIG. 3a, it is pointed out that the transmitter, which is the subject of the present invention, thus comprises the modules Ei on the phase I channel and the EQ module respectively on the _Q phase quadrature channel, the coefficient polynomial function. odd previously mentioned being none other than that given by the relation 4 previously mentioned in the description.

In FIG. 3b, the sampled transfer function illustrating the aforementioned odd-numbered polynomial function symmetric with respect to the origin 0 of the amplitude x of the modulated signal, subjected to the specific clipping processing, has also been represented. With respect to the aforementioned non-linear characteristic component whose transfer function is shown in FIG. 3b in solid lines, for the clipping processing E, it is indicated that the aforementioned component closes the modulated signal at a certain level, so that it is advisable to define the clipping factor CF for ÇJiping Factor in English, as the ratio of the amplitude of the processed modulated signal to The effective value of the modulated signal, that is to say before clipping processing, the aforementioned clipping factor CF being between 0.5 and 1.8.

In addition, the nonlinear transfer function of the aforementioned component makes it possible to minimize the distance from the recoil point to the saturation zone of the power amplifier PA, this distance being measured by a power difference between the saturation power of the PA power amplifier and the maximum clipping power between 2 and 3 dB. .

In addition, the clipping stage E comprises a filter, such as a Nyquist filter for example in cascade with the nonlinear characteristic component, the above-mentioned filter making it possible to limit the inter-symbol interference and / or the feedback of the signals. side lobes in adjacent channels. This filter already exists in many transmission systems which confirms that the process does not add a great complexity to the system object of the invention.

In FIG. 3a, each channel in phase I respectively in phase quadrature Q, thus comprises a Nyquist filter module denoted NF ₁ respectively

NF _Q.

With reference to FIG. 3c, it is pointed out that the receiver which is the subject of the present invention is specially adapted for the processing of a modulated signal with a high amplitude dynamic, this signal being processed at the upstream transmission of an amplifier of power by a specific clipping processing, as described in connection with Figures la, Ib, 2a, 2b and 3a, 3b, so as to reduce the distance of the recoil point vis-à-vis the saturation zone necessary for power amplification.

Thus, as will be observed in FIG. 3c, the receiver which is the subject of the invention conventionally comprises a reception antenna RA, an RFF filter, a low noise amplifier LNA and a frequency change stage at from a local oscillator RFO for translating the baseband signal to demodulate it and thus obtain a phase I channel and a Q phase quadrature channel. All the aforementioned elements constitute the analog part A of the object receiver of the present invention. After automatic gain control AGC and ADC digital analog conversion, the receiver object of the invention comprises a digital part D, comprising a filter for reducing the noise power in the channel and an inverse specific clipping stage, denoted E ^"1 , so as to restore the modulated signal with a strong amplitude dynamic, that is, that is, a non-clipped signal It is important to emphasize that this analog demodulation mentioned above in baseband (state of the art) can be performed by digital processing at an IF intermediate frequency.

It will be understood, of course, from the observation of FIG. 3c that the inverse specific clipping stage E ^"1 consists of a specific inverse clipping module, denoted Ef ¹ connected in cascade on the channel in phase I and in an inverse specific clipping module, denoted E _Q ^"1 connected in cascade on the Q phase quadrature channel.

As furthermore shown in FIG. 3c, the inverse specific clipping stage and finally each module Ef ¹ and E _Q ^"1 is constituted by a component of nonlinear characteristic whose sampled transfer function is represented by FIG. another polynomial function with odd coefficients of the amplitude of the samples of the processed modulated signal.

In FIG. 3b, the nonlinear characteristic of the component, that is to say the module Ef ¹ or E _Q ^"1 , is represented, the aforementioned polynomial function is given by the relation 5, previously given in the description.

This function is represented in dashed lines in FIG. 3b. The aforementioned non-linear characteristic for introducing the specific inverse clipping E ^"1 , is then symmetrical with respect to the straight line of slope equal to 1 with respect to the non-linear characteristic making it possible to apply the specific clipping E .

The transfer functions given by the aforementioned relations 4 and 5 are linked by the symmetry indicated previously in the description. Finally, as shown in FIG. 3c, the inverse clipping modules

Ef ¹ respectively E _Q ^"1 are preceded by an anti-noise filtering module, placed on the channel in phase I respectively on the quadrature Q phase channel and denoted for this reason ANFI and ANFQ, respectively. , we indicate that the above-mentioned noise filtering modules may be constituted by low-pass filters, for example.

Test simulation results of the implementation of the transmission method that is the subject of the present invention and in particular of a transmission chain incorporating an emitter and a receiver, as described above, in conjunction with FIGS. 3a to 3c. will now be given, hereinafter.

For the above simulation tests, it is indicated that these tests have been implemented relative to the 802.11 a / g standard in OFDM type modulation at 52 subcarriers. The specific clipping processing and the inverse specific clipping processing were performed in baseband on each of the channels, in phase quadrature phase respectively, as described previously in connection with FIGS. 3a to 3c, with a filtering of Nyquist, on transmission, of 10 MHz bandwidth and a roll-off factor of 0.35, as mentioned previously in the description.

Regarding the power amplifier, transmitter part, it was chosen with reference to a behavioral model, for which the saturation power is considered as a variable which simplifies the parameterization according to the desired PAPR. On the other hand, for this simulation it is specified that the simulation chain does not include an error correction algorithm, a function that will improve the results obtained.

The following simulation tests were carried out, taking into account: the quality of service in terms of BER bit error percentage, the spectral power rise in the adjacent channels, characterized by the rejection parameter, by the ACPR for Adjacent Çhannel Power Rejection in English.

The simulation tests were then performed for: a clipping factor: CF = 0.5. ; 0.7; 0.9; 1.2; 1.5; 1.8; a level of recoil of the average power at the input compared to the power of saturation at the input, this parameter representing, of course, the distance from the setback point to the saturation zone of the power amplifier, this parameter designated IBO for Input Back-Off, taking the values O; 1;

2; 3; 4; 5 in dB.

Firstly, the simulation results mentioned above have made it possible to highlight the influence of clipping and in particular of the aforementioned clipping factor and the distance of withdrawal from the point of withdrawal of the average power with respect to the saturation power of the power amplifier RP on the specific ratio

ACPR.

These tests have made it possible to illustrate that the rise of the secondary lobes is a function of the 2 parameters which are the clipping factor CF related to the clipped signal before amplification, and the withdrawal distance, that is to say the IBO variable mentioned above. .

For a given shrinkage value in dB, the degradation of the ACPR factor is related to the growth of the clipping factor CF. From the moment when the PAPR ratio of the clipped signal is lower than the retraction distance, that is to say of the IBO parameter, only the clipping factor CF affects the value of the ACPR. In particular, if it is not desired to have distortion on the part of the RF power amplifier, a linear zone operation of the RF power amplifier can be provided, but at the expense of a very clear degradation of effectiveness, more than

5%, at maximum average power. The aforementioned tests have furthermore shown that for a given ACPR value, -29 dB, it is possible to gain on the withdrawal distance, that is to say the IBO parameter, a value of 3 dB by clipping the signal.

The aforementioned simulation tests have also made it possible to evaluate the influence of the transmission method that is the subject of the present invention, in particular the specific clipping process, or the specific inverse clipping process, on the performances of a chain. transmitter constituted by a transmitter and a receiver, as illustrated in FIGS. 3a to 3c, previously described.

The performance of a transmission chain is evaluated from indicators, in particular an EVM indicator for Error Vector Magnitude in English where in addition the percentage or bit error rate, BER. The aforementioned simulation tests have confirmed the positive influence of the inverse specific clipping process applied to the reception.

In particular, it is possible to gain 10 dB on the bit error rate for a peak factor of more than 0.7, that is for a clipped signal level equal to 70% of the rms value of the signal. not clipped and for a distance of setback point vis-à-vis the saturation zone, that is to say an IBO parameter, of more than 2 dB.

On the other hand, under the same conditions, a gain of nearly 5 dB can be envisaged on the EVM factor, with a direct influence on the gain of the signal-to-noise ratio at the demodulation. FIG. 4a represents a three-dimensional diagram of the evolution of the bit error rate BER as a function of the clipping factor CR and of the retraction distance IBO. The observation of Figure 4a above confirms the existence of an optimum for an acceptable bit error rate.

In FIG. 4a, this compromise can be evaluated for a clipping factor CR between 0.85 and 0.93, typically CF = 0.9, which represents a good clipping (E = OJ) and a low IBO removal distance of 2 to 3. dB for a bit error rate BER of the order of 10. ^"4 to 10. ^{" 5} . In any case, a gain of 3 dB over the IBO removal distance may be considered for the same value of the BER bit error rate.

It is the same with regard to the EVM parameter, which allows in a three-dimensional diagram not shown in the drawings to highlight the presence of the aforementioned optimum, always for substantially the same values of clipping factor CF between 0.85 and 0.93 and distance from the IBO recoil point between 2 and 3 dB. In this situation, the optimum of the EVM parameter is

-18 to -20 dB. In this case, a gain of 3 dB on the distance parameter of the IBO retrieval point is possible for the same value of the EVM parameter.

The optimum zone for the aforementioned values of the clipping factor CF, or of the distance of the IBO retrieval point, as regards both the bit error rate BER and the EVM parameter, makes it possible to justify the implementation the method object of the present invention, and, of course, the transmitter and the receiver for the implementation of the latter. In particular, the implementation of the inverse specific clipping processing process makes it possible to improve the BER bit error rate and the EVM parameter following the application of the inverse specific clipping.

The improvement is important for the low clipping factors CF, ie for strong clipping, because in this region the amplifier is transparent and thus the inversion eliminates all the distortions introduced by the clipping.

When, on the contrary, we choose a lower clipping, that is to say a high clipping factor close to the value 1, then the power amplifier is no longer transparent and it also introduces distortions by saturation, which is then no longer compensated by the inverse specific clipping because the inverse specific clipping processing only eliminates distortions due to clipping applied to the transmit.

The timing diagram shown in FIG. 4b is made for the original modulated signal in dashed line, the clipped signal obtained after applying the emission-specific clipping processing represented continuously and the reconstituted signal, that is to say after applying the inverse clipping processing on reception, shown in phantom. With the observation of FIG. 4b, one can note the low amplitude distortion introduced by the method object of the present invention, insofar as the difference of amplitude between the returned signal and the original signal, does not exist. does not exceed 10% punctually.

Finally, the simulation tests mentioned above have made it possible to show that it is possible to reduce the distance of the recoil point with respect to the saturation of 3 dB. This reduction is reflected, of course positively on the performance of the power amplifier PA and of course on the overall consumption of the transmitting device, especially when it is a transmitter of a radio terminal mobile phone, for example.

With reference to FIG. 4c, from the curve 1 of the output of a power amplifier of a mobile terminal said 3G, for terminal of third generation, it is possible to establish that it is possible to gain 10% in performance for a decrease in IBO setback distance of 3 dB. This gain in yield seems particularly important, considering that the power amplifier PA consumption can represent more than 50% of the total consumption of the transmission chain. The autonomy of a mobile terminal equipped with such a transmitter and thus implementing the method that is the subject of the present invention can thus be greatly improved, with a substantially equal performance with respect to the timing diagram illustrated in FIG. 4b, establishing the virtual absence distortion.

In addition, the method of transmission respectively of transmission and / or reception of a signal, the transmitter and / or the receiver and the clipping modules or inverse clipping modules, objects of the present invention, can advantageously be implemented on a digital signal, in baseband for example.

In particular, in this case, they can be implemented via: a program product recorded on a storage medium for execution by a computer executing the specific clipping function A Figure la, Al Figure 2a, described previously in the description, so as to reduce the distance of the setback point vis-à-vis the saturation zone necessary for the amplification of the modulated signal and to generate a processed modulated signal. This program product can be implanted in the transmitter object of the invention shown in FIG. 3 a, to implement the modules E 1 and E _Q , as well as the filtering modules NF, on the in-phase or quadrature-phase channel, respectively. ; a program product recorded on a storage medium for execution by a computer executing the inverse clipping function D FIG. 1a, D2 FIG. 2b, previously described in the description, making it possible to subject a modulated signal with a strong amplitude dynamic but subject to the upstream transmission of the power amplification to a specific clipping processing, to a reverse clipping to restore the modulated signal with a high amplitude dynamic. This program product can be implanted in the receiver object of the invention shown in FIG. 3c, to implement the modules E ₁ ^"1 and E _Q ^{" 1} , as well as the ANFI and ANFQ filtering modules, on the in-phase or quadrature-phase channel, respectively.

Finally, the aforementioned program products can particularly advantageously be implemented in the form of a programmable component, forming the specific clipping module E or the inverse clipping module E ^'1 previously described in the description in connection with Figures 3a,

3b and 3c.

For this purpose, in a nonlimiting preferred implementation mode, the aforementioned programmable component can be implemented in the form of a DSP digital signal processor for Digital Signal Processor in English. The aforementioned component can then be programmed to perform one or both of the clipping or reverse clipping functions. . •

Claims

A method of transmitting a modulated signal with a high amplitude dynamic, said modulated signal being subjected to a power amplification, prior to its transmission, characterized in that it consists of at least:

- on transmission, subjecting said modulated signal, upstream of said power amplification, to a specific clipping processing, so as to reduce the distance of the recoil point, with respect to the saturation zone, necessary for this power amplification, said specific clipping processing, for an emission power amplification having a specific gain / amplitude non-linearity, having a non-linearity attenuation / amplitude substantially greater than the non-linearity gain / amplitude of said power amplification, to mask the non-linearity gain / amplitude of said power amplification; and, following the transmission of said processed modulated signal, subjecting, upon reception, said processed modulated signal to an inverse specific clipping processing, so as to render said modulated signal with a strong, uncapped amplitude dynamic.

2. Method according to claim 1, characterized in that it consists in applying, on transmission, said specific clipping processing on said modulated signal either in baseband, intermediate frequency or radiofrequency.

3. Method according to claim 2, characterized in that it consists in applying, on reception, said inverse specific clipping processing on said processed signal either in baseband, intermediate frequency or radiofrequency.

4. Method according to one of claims 1 to 3, characterized in that, on transmission, said specific clipping processing is followed by a specific filtering, so as to limit inter-symbol interference and recovery of side lobes in adjacent channels.

5. The method as claimed in claim 4, characterized in that, on reception, said inverse specific clipping processing is preceded by a filtering limiting the noise band in the channel, to maintain a suitable signal-to-noise ratio.

6. Method according to one of claims 1 to 5, characterized in that said specific clipping processing is applied via a nonlinear characteristic whose sampled transfer function is represented by a positive function, increasing and monotone of the amplitude of the samples of said modulated radio signal. ^•

7. Method according to claim 6, characterized in that said inverse specific clipping processing is applied via a non-linear characteristic whose sampled transfer function is represented by the inverse of said positive monotonic function and increasing clipping.

8. A method for transmitting a modulated signal, with a high amplitude dynamic, characterized in that it consists in subjecting this signal, upstream of the power amplification of the latter, to a processing of specific clipping, so as to reduce the distance of the recoil point, vis-à-vis the saturation zone necessary for this amplification, to generate a processed modulated signal.

9. A method of receiving a modulated signal, with a strong amplitude dynamic, subjected to the emission upstream of the power amplification of the latter to a specific clipping processing, to generate a processed modulated signal, characterized by subjecting said processed modulated signal to an inverse specific clipping processing so as to render said modulated signal with a strong amplitude dynamic.

10. Radio frequency transmitter of a modulated signal with a strong amplitude dynamic comprising at least one radiofrequency power amplifier connected to a transmitting antenna, characterized in that said radiofrequency transmitter comprises at least one specific clipping stage of said signal modulated upstream of said radiofrequency power amplifier, so as to reduce the distance from the recoil point to the saturation region of said power amplifier, and output a processed modulated signal.

An emitter according to claim 10, characterized in that said specific clipping step is constituted by a non-linear characteristic component whose sampled transfer function is represented by a positive, monotonic and increasing function of the amplitude of the samples. said modulated signal.

An emitter according to claim 11, characterized in that said non-linear characteristic component has a clipping factor in the baseband defined as the ratio of the amplitude of the processed modulated signal to the rms value of the modulated signal included. between 0.85 and 0.93 and a distance from the point of recoil, vis-à-vis the saturation zone of said power amplifier, between 2 and 3 dB.

13. Transmitter according to one of claims 11 or 12, characterized in that said clipping stage further comprises a specific filter in cascade with said nonlinear characteristic component and for limiting inter-symbol interference and recovery of side lobes in adjacent channels.

14. Radiofrequency receiver of a modulated signal with a strong amplitude dynamic and processed, on transmission, upstream of a power amplification by a specific clipping processing, so as to reduce the distance from the setback point , in relation to the saturation zone, necessary for this amplification of power, characterized in that this radio frequency receiver comprises at least one inverse specific clipping stage, so as to restore said modulated signal with a high dynamic range. amplitude, not clipped.

A receiver according to claim 14, characterized in that said inverse specific clipping stage is constituted by a non-linear characteristic component whose sampled transfer function is represented by the inverse function of said clipping function.

16. Clipping module of a modulated signal, characterized in that it consists of a nonlinear characteristic component whose sampled transfer function is represented by a positive, monotonic and increasing function of the amplitude of the samples of said modulated signal.

17. Module for inverse clipping of a modulated signal with a strong amplitude dynamic and processed, on transmission, upstream of a power amplification by a clipping processing represented by a monotonic and increasing positive function of the amplitude of the samples of the modulated signal, characterized in that said inverse clipping module is constituted by a nonlinear characteristic component whose sampled transfer function is represented by the inverse function of said clipping function.

18. Computer program product recorded on a storage medium for execution by a computer, characterized in that it makes it possible to subject, before amplification of power, a modulated signal with a strong amplitude dynamic to a processing of specific clipping, so as to reduce the distance of the recoil point vis-à-vis the saturation zone necessary for this amplification, to generate a processed modulated signal.

19. Computer program product recorded on a storage medium for execution by a computer, characterized in that it makes it possible to subject a modulated signal with a strong amplitude dynamic, subject to transmission upstream of the computer. power amplifier of the latter, to a specific clipping processing, to a specific inverse clipping processing, so as to restore said modulated signal with a high amplitude dynamics.

20. programmable component, characterized in that it comprises a programmable program product according to one of claims 18 and / or 19.

21. Programmable component according to claim 20, characterized in that said programmable component is a digital signal processor.