The object of the present invention is a method for receiving spectrum spreading signals with frequency shift correction. It finds an application in digital transmissions.
With the invention, effects due to untimely frequency shifts may be corrected, regardless of the origin of these shifts. Most often, this will be a Doppler effect related to the moving speed of the receiver. But this might also be effects due to a frequency shift of the local oscillators. In the description which follows, it will be assumed that the effect to be corrected is a Doppler effect, without however limiting the scope of the invention to this case.
A great number of publications have been made on the correction of the Doppler effect. For example, U.S. Pat. No. 5,007,068 may be mentioned as well as the corresponding article of M. K. SIMON and D. DIVSALAR entitled “Doppler-Corrected Differential Detection of MPSK”, published in the journal, “IEEE Transactions on Communications”, Vol. 37, No. 2, February 1989, pages 99–109. These documents describe a technique wherein the Doppler shift is determined on one half of the symbol period. For this purpose, the receiver uses two circuits each with a half period delay, and a Doppler effect estimation circuit connected between the two delay circuits. The correction is then performed on the usual demodulated signal.
This technique may also be used for modulations of the multiple differential phase modulation type (abbreviated as MDPSK for “M-ary Differential Phase Shift Keying”), but it is not applicable to spectrum spreading transmissions where each symbol is multiplied by a pseudo-random sequence.
The article of F. D. NATALI, entitled “AFC Tracking Algorithms”, published in the journal, “IEEE Translations on Communications, vol. COM-32, No. 8, August 1984, pages 935–947, describes a technique in which preambles made up of known symbols are formed before transmitting the useful information. The working frequency is automatically controlled (“Automatic Frequency Control” or AFC) by a loop structure.
This technique is not adapted to the case when information data blocks separated by blanks are transmitted.
The following correction techniques may further be mentioned:
These techniques are generally expensive and complex and do not make the most out of the advantages of spectrum spreading, nor of the digital processing of the signals. The object of the present invention is precisely to overcome these drawbacks by providing a method which processes the signals in the base band (and not the signals in the radio portion) and which is well adapted to digital processing.
Document EP-A-0 822 668 describes a receiver for spectrum spreading signals wherein the Doppler effect correction is performed on the base band signal.
Specifically, the object of the invention is a method for receiving spectrum spreading signals with frequency shift correction, wherein:
Circuits F(I), F(Q) perform a correlation operation between the received signal and the pseudo-random sequence used at emission. This operation consists in storing a certain number of successive samples and in performing a weighted sum with the help of coefficients which are the coefficients for direct sequence spectrum spreading. These coefficients are equal to +1 and to −1, according to the sign of the chips forming the pseudo-random sequence.
The analog-digital converters ADC(I) and ADC(Q) operate at frequency neFc where Fc is the frequency of the chip (Fc=1/Tc), and ne is the number of samples taken in a chip period (Tc). To simplify the discussion, it will be assumed that one sample is taken per chip. The correlation signals CORR(I) and CORR(Q) contain one correlation peak per symbol period.
As for the DOT and CROSS signals delivered by the demodulation circuit DD, as a reminder, they are for the first, the sum IkIk−1+QkQk−1 and for the second, the difference QkIk−1−IkQk−1, where Ik and Qk designate the correlation samples of rank k for channels I and Q, the rank corresponding to a symbol period (Ik corresponds to CORR (Ik) and Qk to CORR (Qk)).
In order to estimate the time T (or 2T), according to the invention, the correlation signals are used as DOT and CROSS signals.
To carry out the correction, either the incident signals or the correlation signals are acted upon. There are therefore several possible alternatives which are illustrated in
In
The properties of the signals to be processed are used for determining the period T illustrated in
CORRdop(Ik)=P.cos(πkN/2T) (1)
CORRdop(Qk)=P.sin(πkN/2T) (2)
where P is an amplitude.
After differential demodulation, the following DOT and CROSS signals are obtained:
DOTdop(k)=CORRdop(Ik).CORRdop(Ik−1)+CORRdop(Qk).CORRdop(Qk−1) (3)
CROSSdop(k)=CORRdop(Qk).CORRdop(Ik−1)−CORRdop(Ik).CORRdop(Qk−1) (4)
By replacing in (3), (4) the quantities with their values given by (1) and (2) and taking into account the properties of trigonometric functions, it is found that:
DOTdop(k)=P2.cos(πN/2T) (5)
CROSSdop(k)=P2.sin(πN/2T) (6)
It is seen that both DOTdop and CROSSdop quantities are independent of the rank k of the preamble symbol.
By taking the ratio of these quantities, the tangent of angle πN/2T is formed from which the angle and the value of T may be extracted:
where tan−1(.) means “arc for which the tangent is equal to (.)”.
The Doppler effect estimation circuit ESTdop of
Cc=cos(πx/2T) (8)
Cs=sin(πx/2T) (9)
Such a signal is generated by a generator with two quadrature outputs.
This calculation may be changed by taking a sequence of weighted samples and calculating:
Still in a more general way, estimation of T is improved by proceeding with low pass filtering of signals DOTdop and CROSSdop, i.e., with A=f(CROSSdop(k)) and B=f(DOTdop(K)), where f represents the filtering function:
A generator receiving T, delivers components Cs and Cs as defined by (8) and (9).
Having described the means for obtaining both components Cc and Cs of the correction signal, a description will now be made on how the received signals are corrected accordingly. This correction processes the signals carrying the transmitted information and no longer the preamble.
Generally, the correlation signals CORR(I) and CORR(Q) of the phase and quadrature channels may be considered as real and imaginary components of a complex signal CORR(I)+jCORR(Q). The Doppler effect changes the phase of this signal (in other words, it rotates the vector which represents it) by a quantity ej(πx/2T). The obtained signal is the Doppler effect affected signal. Its components are CORRdop(I) and CORRdop(Q). This may be therefore written as:
CORRdop(I)+jCORRdop(Q)=[CORRdop(I)+jCORRdop(Q)]ej(πx/2T) (12)
Conversely, the components free from the Doppler effect may be expressed with respect to the components impaired by the Doppler effect as:
CORR(I)+jCORR(Q)=[CORRdop(I)+jCORRdop(Q)]e−j(πx/2T) (13)
By developing the right-hand member of this equation and by identifying the real and imaginary terms, it is found that:
CORR(1)=CORRdop(I).cos(πx/2T+CORRdop(Q).sin(πx/2T) (14)
CORR(Q)=CORRdop(Q).cos(πx/2T)−CORRdop(I).sin(πx/2T) (15)
The same relationships may be established with signals I and Q, i.e.:
[I+jQ]=[Idop+jQdop]e−j(πx/2T) (16)
i.e.:
I−Idop.cos(πx/2T)+Qdop.sin(πx/2T) (17)
Q=Qdop.cos(πx/2T)−Idop.sin(πx/2T) (18)
The correction circuit must therefore comprise multipliers for multiplying the signals to be corrected by both components of the correction signal, and adders in order to perform the sum of the obtained products.
In
In
The invention is not limited to the case when both channel I and channel Q are corrected. One or the other of these channels might as well be corrected. The correction circuit will then implement the portion of the means required for calculating expressions (18) or (19).
The invention, which has just been described, is applied to any type of pseudo-random sequence spectrum spreading signal receiver. In particular it may be applied to so-called Code Division Multiple Access (CDMA) transmissions. In these transmissions, several users share a same channel by means of different pseudo-random sequences. The receiver then comprises as many parallel channels as users. In a particular embodiment, such a CDMA receiver may comprise parallel suppression means for interference between users, with weighting means. Such a receiver is described in a French patent application filed by the present applicant on Mar. 24, 1998, under number 98 03586.
Finally
Curve 50 refers to a conventional method with one stage, without interference suppression. Curves 51 and 52 refer to the same method but with two different Doppler effects, the first with a relative shift of 10−6 at 2.45 GHz and the second with a relative shift of 10−5.
Curve 60 refers to an interference parallel suppression method with only one stage for parallel suppression of interferences and curves 61, 62 to the same method but with shifts of 10−6 and 10−5.
Curve 70 refers to a method with two stages for parallel suppression of interferences with associated curves 71, 72 for shifts at 10−6 and 10−5.
Finally, curve 80 marks the theoretical limit of the phase difference modulation technique (DQPSK).
Number | Date | Country | Kind |
---|---|---|---|
98 16316 | Dec 1998 | FR | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/FR99/03220 | 12/21/1999 | WO | 00 | 6/15/2001 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO00/39939 | 7/6/2000 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
4481640 | Chow et al. | Nov 1984 | A |
4706286 | Sturza | Nov 1987 | A |
5260969 | Kato et al. | Nov 1993 | A |
5271034 | Abaunza | Dec 1993 | A |
5623485 | Bi | Apr 1997 | A |
5799034 | Walley et al. | Aug 1998 | A |
5856997 | Mochizuki et al. | Jan 1999 | A |
Number | Date | Country |
---|---|---|
0 822 668 | Feb 1998 | EP |
WO 9622561 | Jul 1996 | WO |