FIELD OF THE INVENTION

The present invention relates to signal descrambling.

BACKGROUND OF THE INVENTION Known signal descrambling methods known use a code sequence. Each original code of the sequence is a complex number corresponding to a vector in a complex plan which is aligned neither to an imaginary axis nor to a real axis of the complex plan. The code sequence used to descramble the signal corresponds to the one used to scramble the signal. Typically, the descrambling code sequence is a complex-conjugate of the sequence used by a scrambler. An example of such a method may be found in 3GPP (3 ^rd Generation Partnership Project) Technical Specification 25.213 version 3.1.1 (3GTS 25 213 V3.1.1) "Spreading and modulation (FDD)", for example, available at ftp://3gpp.org and in particular, ftp://ftp.3gpp.org/Specs/December_99/25_series.

Generally, known descrambling methods involve a great number of operations.

SUMMARY OF THE INVENTION

Accordingly, it is an object of the invention to provide a descrambling method implementing a reduced number of operations. The invention provides a descrambling method comprising a step of multiplying two items to obtain the descrambled signal, one item being a bit of information from a scrambled signal, the other item being a code from a descrambling code sequence, each item being a complex number corresponding to a vector in a complex plan which is aligned neither to an imaginary axis nor to a real axis of the complex plan, wherein, prior to the multiplying step, the method comprises a step of modifying one of the items by performing a mathematical rotation to align the vector corresponding to the modified item either to the imaginary axis or to the real axis.

In the above method, the vector corresponding to the modified item is aligned either to the imaginary axis or to the real axis of the complex plan. This means that either the imaginary part or the real part of the item is null. Therefore, either the modified

descrambling code sequence or the modified scrambled signal contains a great number of "zero" values for reducing the number of operations necessary to descramble the signal. The above method is therefore faster and requires less computing resources.

The feature wherein the modifying step further comprises setting the module of the vector corresponding to the modified item to one using a predetermined mathematical proportional transformation, and the feature wherein the modifying step comprises replacing the original value of one of the items with one complex number chosen in the group {l ; - l ;j ; - j} depending on the original value of the item and according to a predetermined rule, where j is the imaginary unit, further reduce the number of operations needed to descramble the signal.

The invention also relates to a signal descrambler for a receiver having a channel estimator to correct a descrambled signal, the descrambler being adapted to perform the above descrambling method.

The invention also relates to telecommunication equipment comprising the above descrambler.

These and other aspects of the invention will be apparent from the following description, drawings and claims.

BRIEF DESCRIPTION OF THE DRAWINGS Fig.l is a schematic diagram of a scrambling/descrambling system using a descrambler;

Figs.2 and 3 are diagrams of two code constellations used in the descrambler of Fig.l, and

Fig.4 is a flowchart of a descrambling method.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Fig.l shows a CDMA (Code Division Multiple Access) telecommunication system 2 like the one used for UMTS (Universal Mobile Telecommunication System). For simplicity, only one transmitter 4 and one receiver 6 are represented. The transmitter 4 includes a conventional spreader and scrambler unit 10. Unit

10 has an input 12 which receives complex information symbols and an output 14 which outputs the corresponding spread and scrambled symbols to other conventional units (not shown for simplicity), before they are transmitted through the air over an antenna 16.

Unit 10 includes a spreading and scrambling code sequence generator 20. Generator 20 outputs a code sequence S that is used to scramble and spread the input information symbols. For example, the code sequence is a so-called pseudo-noise (PN) bit sequence. More precisely, the codes of sequence S are chosen in a QPSK (Quadrature Phase Shift Keying) constellation in which each possible code is a complex number corresponding to a vector which is aligned neither to the real axis nor to the imaginary axis in the complex plan.

By way of illustration Fig.2 represents the QPSK constellation used by transmitter 4 in a complex plan. Referring to Fig.2, the real axis of the complex plan is designated "R" and the imaginary axis of the complex plan is designated "I". Four vectors Vi, V ₂ , V ₃ and V ₄ represent the four possible positions for a vector corresponding to a code of sequence S. Each vector Vi, V ₂ , V ₃ and V ₄ is perpendicular to two other vectors of the constellation. Angle α between Vi and the real axis is equal to π/4. Here, the module of each vector Vj is equal to -Jl . Thus, each code of sequence S is chosen from the complex number group {1 + j ; 1 — j ; - 1 + j ; - 1 - j}.

In CDMA telecommunication systems, each elementary information bit which is output by unit 10 is called a chip. Several chips (usually 2 ⁿ ) correspond to one information symbol. The chips are complex numbers having a real part and an imaginary part.

For example, a detailed description of unit 10 can be found in 3GPP technical specification 25.213.

Receiver 6 is user equipment such as a mobile phone.

Receiver 6 has an antenna 30 to receive radio signal transmitted by transmitter 4. Antenna 30 is connected by conventional units, which are not represented in Fig.1 for simplicity, to an input 34 of a descrambler and despreader integrated in one unit 36.

Input 34 receives the spread and scrambled signals transmitted by transmitter 4 through a transmission channel 32.

Unit 36 has an output 38 which is connected to an input of a conventional channel corrector 40. Output 38 is used to transmit the descrambled and despread signal to the channel corrector 40.

Channel corrector 40 is adapted to correct the descrambled and despread signal. Conventionally, channel corrector corrects the signal by using an estimation of the

distortion introduced by the transmission channel provided by a channel estimator 41. Channel distortions shift the phase and/or amplitude of the transmitted signal. As an example, the channel estimator uses predetermined symbols in the received signal to establish an estimation of the distortion due to the channel and then the channel corrector corrects with such estimation the other symbols in the same signal. Methods to correct channel distortion are well known. For example, one may refer to "Digital Communication Receivers: Synchronization, Channel Estimation & Signal Processing"; Meyr, Moeneclaey, Fechtel; Wiley Editions.

Unit 36 includes a despreading and descrambling code sequence generator 42, a complex multiplier 44 and a complex integrator 45.

Generator 42 outputs an original code sequence S* to be used to descramble and despread the received signal. There exists a relationship between sequence S and sequence S*. Sequence S* is the complex-conjugate sequence of sequence S. As a result, the codes PN* of sequence S* are chosen from the QPSK constellation of Fig.2. Multiplier 44 multiplies the spread and scrambled received signals by a modified code sequence S' to output a modified descrambled signal to integrator 45. More precisely, multiplier 44 carries out the following operation:

outputChipi outputChipi. PN' i - inputChipQ.PN' Q \ outputChipQ outputChipi, PN' Q - inputChipQ.PN' A ( ¹ )

where:

- PlSP ₁ is the real part of a code of sequence S',

- PN' _Q is the imaginary part of a code of sequence S',

- inputChipi is the real part of a received chip, - inputChip _Q is the imaginary part of a received chip,

- outputChip _! is the real part of the chip which is output by multiplier 44, and

- outputChip _Q is the imaginary part of the chip which is output by multiplier 44.

Integrator 45 integrates or adds the signal provided by multiplier 44 over one period, the period being equal to the duration of one symbol. The integrated signal is output to channel corrector 40.

Unit 36 includes a modifier 46 to generate the modified code sequence S' from the original code sequence S*. Modifier 46 performs a mathematical rotation and a proportional transformation of each original code PN* of sequence S* to obtain a modified code PN' of sequence S'. More precisely, modifier 46 implements the following operation:

where:

- PN' is a code of sequence S', - PN* is a code of sequence S*,

- e is the exponential function,

- j is the imaginary unit,

- |V| is the module of the vector corresponding to code PN*, and

- α is the angle of the constellation used by generator 42 as defined on Fig.2. In the present example, α is equal to π/4 and |V| is equal to-χ/2 .

The exponential function e in relation (2) corresponds to a mathematical rotation by angle α of the vector corresponding to code PN*. As a result of such a rotation, the vector V'i corresponding to code PN' is aligned either to the imaginary axis "R" or to the real axis "I" as illustrated on Fig.3.

The multiplication of code PN* by 1/|V| performs a proportional transformation to set the module of vector V'i to one.

Consequently, modifier 46 transforms a code sequence where each code is chosen in the constellation of Fig.2 in a modified code sequence where each code PN' is chosen in the constellation of Fig.3. In the constellation of Fig.3, the possible positions of a vector corresponding to a code PN' are represented by vectors V'i, V ₂ , V ₃ and V ₄ which correspond to complex numbers 1 ; j ; - 1 ; - j, respectively.

The operation of system 2 will now be described with reference to Fig.4. First, in step 60, each information symbol to be transmitted to receiver 6 is spread and scrambled. More precisely, in step 60, generator 20 generates sequence S that is used to spread and scramble the input symbol stream.

The resulting chips are transmitted over transmission channel 32.

In step 62, the signal is received by receiver 6, and transmitted to input 34 of unit 36.

Then, in step 64, generator 42 generates the original descrambling and despreading code sequence S*.

In step 66, each code PN* of sequence S* is modified by modifier 46 to align the corresponding vector either to the real axis or to the imaginary axis using a rotation by the predetermined angle α and a proportional transformation. More precisely, to perform such mathematical rotation and transformation, modifier 46 may perform as follows: - if the real part and the imaginary part of code PN* are positive, then it replaces this code with the complex number + 1 , or else

- if the real part of code PN* is positive and its imaginary part is negative, then modifier 46 replaces the code PN* with the complex number - j, or else

- if the real part of the code PN* is negative and its imaginary part is positive, then modifier 46 replaces code PN* with the complex number + j, or else

- if the real part of code PN* is negative and its imaginary part is also negative, it replaces code PN* with the complex number - 1.

By doing so, modifier 46 generates sequence S' without carrying out any operation such as multiplications or additions. Consequently, such an implementation of relation (2) is simpler than using multiplications.

Subsequently, in step 68, multiplier 44 processes the received signal by multiplying the received signal by sequence S' instead of sequence S*.

In step 70, integrator 45 integrates the signal received from multiplier 44 and outputs a descrambled and despread signal.

Thereafter, the descrambled and despread signal is transmitted to channel corrector 40 and, in step 72, the signal is corrected by channel corrector 40.

It should be noted that the signal generated by unit 36 is different from a descrambled and despread signal that would be obtained if the received signal was directly descrambled and despreaded with sequence S*. In fact, each symbol of the signal generated by unit 36 has a shifted phase and shifted amplitude in comparison to the phase and amplitude of a symbol that would be obtained by descrambling and despreading directly the received signal by sequence S*. This phase and amplitude shift is the same for

every symbol generated by unit 36. Consequently, such a phase and amplitude shift of the descrambled and despread signal is considered by corrector 40 to be a channel distortion introduced by the channel and corrected as such. Therefore, the signal corrected by corrector 40 is the same as the one that would be obtained if the received signal was directly despread and descrambled by sequence S*. However, the total number of operations required to descramble and despread the received signal is reduced using unit 36.

In fact, in step 68, since either the real part or the imaginary part of each code PN' is null, this means that two of the multiplications in the right term of relation (1) need not be carried out.

Furthermore, since the module of PN' is equal to 1, this means that it is only necessary to change the sign of inputChipi and/or inputChipQ in relation (1) and no multiplications need be performed. As a result, it should be understood that the use of sequence S' instead of sequence S* replaces four multiplications with only two sign transformations. Thus, the number of operations required to descramble and despread the received signal using sequence S' is smaller than the number of operations required to descramble and despread the same received signal by sequence S*.

Many additional embodiments are possible. For example, the above system and method can be adapted to telecommunication systems in which the transmitted signal is only scrambled and descrambled without being spread and despread. More generally, the above system and method apply to every system in which the signal is scrambled using a scrambling code sequence in which each code is chosen in the constellation of Fig.2 having a non-null angle α . In fact, angle α may have any value chosen in ]0; — [. The

above method also applies to a descrambler using a BPSK (Binary Phase Shift Keying)

TC constellation having an angle α comprised in ]0; — [. A BPSK constellation includes only

two vectors that are in each other's opposite directions.

System 2 and the descrambling method have been described in the case of radio transmission. However, the descrambling method may also be applied to any type of wireless or wired scrambled transmission. For example, it applies to optical transmission.

In another embodiment, the proportional transformation is not implemented and only the mathematical rotation of sequence S* is used. In this embodiment, the number of operations to multiply the received signal by the code sequence is divided by two.

Relation (1) may be written as follows: outputChip = inputChip. PN' (3) where:

- outputChip = outputChipi + j. outputChipQ,

- inputChip = inputChip _t + j. inputChipQ, and - PN' = PN'i +j. PN' _Q

Using relations (2) and (3) there may be written:

outputChip = (4) where PN* is a complex number corresponding to a code of sequence S*.

Thus relation (1) is equivalent to relation (4). The main embodiment described here implements relation (1). In another embodiment, relation (4) is implemented instead of relation (1). The embodiment implementing relation (4) can be carried out by placing modifier 46 on the path of the received signal before multiplier 44. For example, the input of modifier 46 is connected to input 34 and the output of modifier 46 is connected to an input of multiplier 44. Generator 42 is directly connected to the other input of multiplier 44. In this embodiment, the number of operations to descramble the received signal is reduced because either inputChipi or inputChip _Q is null and the module of the vector corresponding to inputChip is equal to 1. This embodiment implementing relation (4) may be combined with any of the previous embodiments.