FR2795257A1 - Phase detector for phase locked loop includes calculator calculating input-output products in order to detect phase shift - Google Patents

Abstract

The system uses a product derived from input and corresponding output symbols, in order to detect the phase shift. The phase detector is intended for operation in a phase locked loop circuit. It comprises a circuit calculating the products of the type (qm.dm), where qm is the input symbol of the phase detector and dm is the conjugate of the corresponding output symbol (or the conjugate of the decision on this output symbol). The circuit delivers a voltage (wm) which is representative of a phase difference between two products (qm.dm) and (qm-x.dm-x), obtained for two distinct input symbols (qm and qm-x).

Description

 DESCRIPTION The invention relates to a transmission system comprising a transmitter for transmitting symbols, and a receiver comprising a phase-locked loop provided with a phase detector which is intended to receive an input symbol corresponding to an emitted symbol, and which comprises means for calculating a product of said input symbol by the conjugate of the corresponding transmitted symbol.

The invention also relates to a receiver for use in such a system, as well as a phase detection method, a frequency correction method and a computer program enabling a receiver to implement such methods.

The invention has particular applications in the field of broadcasting digital programs, cable or satellite for example, with return path allowing different users to transmit data to a head station.

Such a phase-locked loop is for example described in chapter 5.8 of the book "Digital Communication Receivers: Synchronization, Channel Estimation and Signal Processing" by Heinrich Meyr, Marc Moeneclaey and Stefan A. Fechtel, published in the "A Wiley-Interscience" collection. Publication "by John Wiley & Sons Inc.

The phase detector described in this work delivers a voltage which is equal to the imaginary part of the product of the input symbol of the phase detector by the conjugate of the corresponding transmitted symbol.

In particular, it has been found that such a phase-locked loop does not converge when used in a transmission system that uses a spread spectrum transmission technique.

The object of the invention is in particular to propose another phase detector which, in particular, allows the convergence of the phase-locked loop in a transmission system which uses a spread spectrum transmission technique.

For this, a transmission system and a receiver according to the invention and as described in the introductory paragraph are characterized in that said phase detector comprises means for delivering a voltage representative of a phase difference between two products obtained for two separate entry symbols.

Similarly, a phase detection method according to the invention is characterized in that it delivers a voltage representative of a phase difference between two products obtained for two distinct input symbols. And a frequency correction method according to the invention is characterized in that the phase detection step delivers a voltage representative of a phase difference between two products obtained for two distinct input symbols. Finally, a computer program according to the invention is characterized in that it allows a receiver to implement the methods according to the invention.

The invention will be better understood and other details will appear in the description which follows with reference to the accompanying drawings which are given by way of non-limiting examples and in which: FIG. 1 represents a phase-locked loop according to FIG. FIG. 2 represents an example of a transmission system according to the invention; FIG. 4 is a flowchart summarizing the various steps of an exemplary method; FIG. frequency correction device according to the invention.

In particular, the object of the invention is to propose a phase detector that allows in particular the convergence of a phase-locked loop in a transmission system that uses a spread spectrum transmission technique.

In the remainder of the description, only the notions of spread spectrum that are necessary for the understanding of the invention are recalled. For more details we can for example refer to paragraph 1.2 of the book by Samuel C. YANG entitled "CDMA RF system engineering" published by Artech House Inc. in 1998.

At the transmission the spread spectrum is performed for a given user by multiplying each symbol to be transmitted d ", by a code which is a sequence of L chips ck (k = 0, ..., L-1) which verify especially Ck.Ck = 1. At the reception, the received signal r (t) is sampled at the frequency chie 1 / Tc. In the following description the samples obtained for a user are denoted r ,, k. of.Tc the deviation in frequency, normalized with respect to the chip frequency, and relative to the emitter, the phase difference with which the sample r ,, k is received is written rp ,, k <I≥ </ I > 27c.Af.Tc. (mL <I> + k). </ I>

si l'on néglige le bruit. So

if we neglect the noise.

c'est-à-dire:

Les boucles à verrouillage de phase selon l'art antérieur utilisent un détecteur de phase qui délivre en boucle ouverte une tension vm telle que: Vm In 1(d. .@7'm ) où Im(z) est la partie imaginaire du nombre complexe z, p'R, est le symbole d'entrée de la boucle à verrouillage de phase, et dm est le conjugué du symbole émis correspondant au symbole p'r,, (ou le conjugué de l'estimation du symbole émis correspondant au symbole d'entrée p'R, lorsque le symbole émis n'est pas connu). The spectrum despreading operation consists of multiplying each sample r, m, k by the chip c, and adding the products obtained for a fixed value of m and for k varying from 0 to L-1. If we neglect noise and intersymbol interference that is comparable to a Gaussian noise, we get a received symbol pm that checks:

that is to say:

The phase-locked loops according to the prior art use a phase detector which delivers in an open loop a voltage vm such that: Vm In 1 (d · @ 7'm) where Im (z) is the imaginary part of the number complex z, p'R, is the input symbol of the phase-locked loop, and dm is the conjugate of the transmitted symbol corresponding to the symbol p'r ,, (or the conjugate of the estimate of the transmitted symbol corresponding to the entry symbol p'R, when the emitted symbol is not known).

Une telle boucle à verrouillage de phase ne peut pas converger à cause du terme 27r.df.Tc.m.L qui dépend de m. In a spread spectrum transmission system, pR, is given by equation (1). If p, n = p ', n, then equation (2) is written:

Such a phase locked loop can not converge because of the term 27r.df.Tc.mL which depends on m.

où qm est une estimation de la phase à corriger, et délivre un symbole q, qui vérifie

Le symbole q, est appliqué en entrée d'un détecteur de phase 2 selon l'invention. Le détecteur de phase 2 délivre une tension w, qui commande un oscillateur numérique 3 par l'intermédiaire d'un filtre 4. L'oscillateur numérique 4 délivre une exponentielle

qui est appliquée au multiplicateur 1. FIG. 1 shows an example of a phase-locked loop according to the invention. This phase-locked loop comprises a multiplier 1 for correcting the phase of a symbol p'm applied at the input of the phase-locked loop. This multiplier 1 multiplies the input symbol p 'by an exponential

where qm is an estimate of the phase to be corrected, and delivers a symbol q, which verifies

The symbol q is applied at the input of a phase detector 2 according to the invention. The phase detector 2 delivers a voltage w, which controls a digital oscillator 3 via a filter 4. The digital oscillator 4 delivers an exponential

which is applied to the multiplier 1.

According to the invention the voltage w is representative of a phase difference between two products (qm.dm ') and (qm_X.dm_X') obtained for two input symbols distinct from the phase detector, q, and q -x. For example, we choose: w = sin [arg (qm.dm *) - arg (qm-.r .d, n-, r @)] where sin indicates the function sine, and arg (z) is the argument of the complex number z. In open loop, qm = p'm. If p ', is given by equation (1) we obtain then: w. = sin [t27r.Af.Tc. (mL) <I> + </ I> n.Af.Tc. (L -1)} - {27r.Of.Tc. (m <I> - </ I> x) l <I> + </ I> 7r.Of.Tc. (L -1)}] that is to say:

On obtient une valeur maximale de (df .Ts)m,, pour x=1. Cette valeur maximale est égale à 50%. Sur la figure 2 on a représenté un exemple de système de transmission selon l'invention. Ce système de transmission comporte une pluralité d'émetteurs 10 qui émettent des données vers un récepteur 12 par l'intermédiaire d'un média de transmission 14. II peut s'agir par exemple d'une voie montante dans un système de diffusion de programmes par satellite ou par câble. Les terminaux de ce type de systèmes sont destinés au grand public. II est donc important de réduire au maximum leur prix de revient. Pour cela, il est notamment avantageux d'utiliser des oscillateurs locaux de faible prix pour générer les fréquences porteuses à utiliser pour les transmissions montantes. Les oscillateurs locaux de faible prix sont imprécis: typiquement on utilise des oscillateurs dont la précision varie entre 1 ppm (de l'anglais "part per million") et 10 ppm. L'erreur qui en résulte sur la fréquence porteuse générée est proportionnelle à la fréquence porteuse. Cette erreur est donc d'autant plus gênante que les fréquences utilisées sont élevées. A titre d'exemple, dans les systèmes de transmission par satellite qui utilisent la bande de fréquences 20Ghz- 30Ghz pour les transmissions montantes, l'écart en fréquence obtenu pour un oscillateur local ayant une précision de 1 ppm peut atteindre 30kHz (c'est- à-dire que pour une fréquence symbole 1/Ts de 100kHz, l'écart de fréquence normalisé par rapport à la fréquence symbole est de 30%). w ", <SEP> = <SEP> sin [27r.Af <I> .Tc (mL <SEP> - <SEP> (m <SEP> - <SEP> x) L)] <SEP> z- <SEP > 2n.Af.Tc.Lx </ I><SEP> if <SEP><I> - </ I>
<tb><I><<SEP> Of <SEP> .Tc.Lx <SEP><</I>
<tb> that is, <SEP> if
<tb><I> - <SEP> 1 <SEP><SEP> OfIs <SEP><<SEP> 1 </ I><SEP> where <SEP> 1 / Ts <SEP> is <SEP><SEP> frequency <SEP> symbol.
<tb> 2x <SEP> 2x The phase-locked loop according to the invention thus operates for a standardized frequency deviation with respect to the symbol frequency, the maximum value of which is

We obtain a maximum value of (df .Ts) m ,, for x = 1. This maximum value is equal to 50%. FIG. 2 shows an example of a transmission system according to the invention. This transmission system comprises a plurality of transmitters 10 which transmit data to a receiver 12 via a transmission medium 14. It may be for example an uplink in a program broadcasting system by satellite or cable. The terminals of this type of system are intended for the general public. It is therefore important to minimize their cost price. For this, it is particularly advantageous to use low-cost local oscillators to generate the carrier frequencies to be used for uplink transmissions. The local oscillators of low price are imprecise: typically one uses oscillators whose precision varies between 1 ppm (of English "part per million") and 10 ppm. The resulting error on the generated carrier frequency is proportional to the carrier frequency. This error is therefore all the more troublesome that the frequencies used are high. For example, in satellite transmission systems that use the 20Ghz-30Ghz frequency band for uplink transmissions, the frequency deviation obtained for a local oscillator having an accuracy of 1 ppm can reach 30 kHz (that is that is, for a symbol frequency 1 / Ts of 100 kHz, the standard frequency deviation with respect to the symbol frequency is 30%).

One solution for reducing this standardized frequency difference is to use a spread spectrum transmission technique. Indeed, the spread spectrum technique consists in adding L transitions in a symbol. We thus go from a symbol duration of Ts to a chip duration Tc equal to Ts / L. The standardized frequency difference is therefore divided by L.

At reception, the frequency difference relative to each transmitter must be corrected. For this purpose, a phase-locked loop is used as shown in FIG. 1. In order to reduce the convergence time of the phase-locked loop (which is all the more important as the standardized frequency difference is high) ) a frequency estimator and frequency correction means are used upstream of the phase-locked loop. Thus the phase locked loop only has to correct a residual error.

FIG. 3 shows an exemplary receiver 12 according to the invention. This receiver 12 comprises means 20 for transposing the received signal r (t) into a baseband, a sampler 22 for sampling the baseband signal at the chip 1 / Tc frequency, a filter 24 for filtering the sampled signal around zero, spectrum despreading means 26 which outputs p symbols, called received symbols, a frequency estimator 28 for estimating the frequency error Δf.Tc on the received symbols pR, and a frequency corrector 30 for correcting the frequency The corrected symbol p '"is then applied to a phase locked loop 40 of the invention to correct the residual phase error. The output of the phase locked loop is connected to a threshold decision device 50 which makes a decision dm on the received symbol. And the threshold decision device 50 is connected to conventional channel decoding means 60 which outputs the final data. The frequency estimator 28 is for example constituted by one of the estimators described in the article "Feedforward estimation for PSK: a tutorial review" by Michele Morelli and Umberto Mengali, published in the journal European Transactions on Communications, voume9, n 2, March - April 1998.

L'invention n'est pas limitée aux systèmes de diffusion par câble ou satellite qui ont été décrits ici à titre d'exemple. Elle est en particulier applicable à tout type de système de transmission qui utilise une technique de transmission par étalement de spectre. FIG. 4 summarizes in a flowchart the various steps of a frequency correction method according to the invention. This method comprises a frequency estimation step 100 for estimating a frequency error af.Tc on a received symbol p, n, a frequency correction step 110 for correcting this frequency error Δf.Tc (this step 110 consists of multiply the received symbol p by an exponential ej "-f"<I>,</I> and a phase-lock step 120 to correct the residual error (Af <I> - </ I> Af <I >) </ I> on the symbol

The invention is not limited to cable or satellite broadcast systems which have been described here by way of example. It is particularly applicable to any type of transmission system that uses a spread spectrum transmission technique.

Claims (10)

A transmission system comprising a transmitter (10) for transmitting symbols, and a receiver (12) comprising a phase locked loop provided with a phase detector (2) which is adapted to receive an input symbol ( qm) corresponding to an emitted symbol (dm), and which comprises means for calculating a product of said input symbol by the conjugate of the corresponding transmitted symbol (dm.), characterized in that said phase detector comprises means to provide a voltage (wm) representative of a phase difference between two products obtained for two distinct input symbols. 2. Transmission system according to claim 1, characterized in that it uses a spread spectrum transmission technique (26), and in that said receiver comprises means (26) spectrum despreading upstream of said detector of phase. 3. Transmission system according to claim 1, characterized in that it comprises, upstream of said phase detector, a frequency estimator (28) for estimating a frequency error relative to said transmitter, and a frequency corrector (30). to correct said frequency error. A receiver (12) comprising a phase locked loop having a phase detector (2) for receiving an input symbol (qm) corresponding to an emitted symbol (dm) and having means for calculating a product of said input symbol by the conjugate of the corresponding transmitted symbol, characterized in that said phase detector comprises means for delivering a voltage (w,) representative of a phase difference between two products obtained for two separate entry symbols. 5. Receiver according to claim 4, characterized in that it comprises, upstream of said phase-locked loop, a frequency estimator (28) for estimating a frequency error relative to said transmitter, and a frequency corrector (30). ) to correct said frequency error. A method of phase detection from a symbol (q,) corresponding to a transmitted symbol (dm), comprising a step of calculating a product of said input symbol by the conjugate of the corresponding transmitted symbol, characterized in it delivers a voltage (w,) representative of a phase difference between two products obtained for two distinct input symbols. A frequency correction method for correcting at a receiver a frequency error relating to a transmitter, said method comprising a phase-locking step (120) which notably comprises a phase detection step calculating from an input symbol (q,) corresponding to a transmitted symbol (d,), a product of said input symbol by the conjugate of the corresponding transmitted symbol, characterized in that said phase detection step outputs a voltage (w, ) representative of a phase difference between two products obtained for two distinct input symbols. 8. frequency correction method according to claim 8, characterized in that it comprises, upstream of said phase lock step, a frequency estimation step (100) for estimating a frequency error relative to said transmitter, and a frequency correction step (110) for correcting said frequency error. 9. Computer program enabling a receiver to implement a frequency correction method for correcting a frequency error relating to a transmitter, said method comprising a phase-locking step which notably comprises a phase detection step calculating from an input symbol corresponding to a transmitted symbol, a product of said input symbol by the conjugate of the corresponding transmitted symbol, characterized in that said phase detection step delivers a voltage representative of a phase difference between two products obtained for two distinct entry symbols. 10. Computer program according to claim 9, characterized in that said frequency correction method comprises, upstream of said phase-locking step, a frequency estimation step for estimating a frequency error relating to said transmitter, and a frequency correction step for correcting said frequency error.