JP2687783B2 - Spread spectrum demodulator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a direct sequence spread spectrum demodulator, and more particularly to a spread spectrum demodulator using a rake system.

[0002]

2. Description of the Related Art The rake method is a method for realizing path diversity in which the energy of delay waves generated in a selective fading communication channel is realized by utilizing the characteristics of a spread spectrum signal of the direct spread method. Maximum ratio combining of delayed waves is performed. A conventional spread spectrum demodulation device using the rake method is disclosed in, for example, Mitsuo Yokoyama: "Spread Spectrum Communication System", Science & Technology Publishing Company (1988).
And Proakis: “DIGITAL COMMUN
ICATIONS (2nd Edition) ", Mc
See Graw-HiLL (1989).

A conventional technique will be described with reference to the drawings.
FIG. 3 is a block diagram showing the configuration of a conventional spread spectrum demodulation device using the rake method. In the figure, 200 is an oscillator that oscillates a local carrier wave, 201 is a phase shifter that shifts the phase of the local carrier wave by π / 2 radians, 202 and 203 are multipliers, 204 and 205 are low-pass filters, and 206,
Reference numeral 207 is a sampler, and 208 is a correlator. Reference numeral 100 is a weighting coefficient generation means, 209 is a delay element, 210
Is a multiplier, 211 is a conjugate circuit that outputs the complex conjugate number of the input signal, 212 is a value obtained by correcting the value of the input signal according to the magnitude relationship between the squared value of the absolute value of the input signal and a given threshold value. Is a correction circuit that outputs 213 is a correlator 208
From the correlation signal output from the correction circuit 212 and the corrected weighting coefficient output from the correction circuit 212 to generate a combined correlation signal. Reference numeral 214 indicates a value of the combined correlation signal output from the signal combining circuit 213. It is a determiner that determines the value of demodulated data.

Next, the operation will be described. Here, the symbols used in the following description are defined as follows. The symbol period of the data is T _d , P used for spread spectrum
The repetition period of the N signal is M (M is an integer of 2 or more) chips, and the chip period is T _c = T _d / M. m (m = 1,
The value of the Mth PN signal is u _m ε {−1,1}, and the value of the transmission data at time nT _d (n is an integer) is a _n ε.
{-1,1} is and shall generate a spread spectrum signal of the direct spread system by multiplying a PN signal {u _m} in the data sequence {a _n} is the transmission side. That is, the value of the spread spectrum signal at time nT _d + mT _c is a _n
· A u _m, the transmitted signal is binary phase shift keying by the spread spectrum signal (hereinafter, the phase shift keying abbreviated as PSK) is assumed to be modulated.

In FIG. 3, an oscillator 200 oscillates a local carrier wave. The phase shifter 201 shifts the phase of the local carrier wave output from the oscillator 200 by π / 2 (radian). The received signal is multiplied by the local carrier output from the oscillator 200 by the multiplier 202 and then the low pass filter 20.
The high frequency component is removed by 4 and converted into a baseband signal having a component in phase with the local carrier (hereinafter referred to as an inphase baseband signal). At the same time, the received signal is the multiplier 2
03 is also multiplied by the π / 2 (radian) phase-shifted local carrier wave output from the phase shifter 201, and then the low-pass filter 205 removes high-frequency components, and a baseband signal of a component orthogonal to the local carrier wave (hereinafter , Quadrature baseband signal). The in-phase baseband signal and the quadrature baseband signal form a complex baseband signal. That is, the in-phase baseband signal is the real number component of the complex baseband signal, and the quadrature baseband signal is the imaginary number component of the complex baseband signal.

The in-phase baseband signal output from the low pass filter 204 is sampled by the sampler 206 every chip period T _c . Similarly, the low pass filter 2
The quadrature baseband signal output from 05 is sampler 2
07 is sampled every chip period T _c . The sampled in-phase baseband signal and the sampled quadrature baseband signal form a sampled complex baseband signal. Here, the values of the sampled in-phase baseband signal, the sampled quadrature baseband signal, and the sampled complex baseband signal y at time kT _c (k is an integer) are y _pk , y _qk, and When y _k is set, the relationship represented by Expression 1 is established.

[0007]

(Equation 1)

[0008] the sampled complex baseband signal is performed mutual correlation operation of the PN sequence is input to a correlator 208 {u _m}, the correlation signal z is output. Here, the time (n
_{M + i) T c = nT} d + iT c (i = 0, ..., M-1)
When the value of the correlation signal in is z _{nM + i} , the relation expressed by the equation 2 is established.

[0009]

(Equation 2)

Under selective fading, the received signal is a combination of various signal waves having different arrival times. Hereinafter, the signal wave with the shortest arrival time is referred to as the preceding wave,
Other signal waves are called delayed waves. In general, in the spread spectrum communication direct diffusion system, PN signal {u _m} are sequences such as autocorrelation coefficients is impulsive (e.g. M-sequence, etc.) is selected. Such self-correlation properties of PN signal {u _m}, in the correlation signal output from the correlator 208, the preceding wave and the energy of each delay wave is separated. That is, it is a value indicating the energy of the correlation signal |
z _{nM + i} | ² has a remarkable maximum at a time corresponding to the arrival time of the preceding wave or the delayed wave. Hereinafter, it is assumed that the maximum of the energy of the correlation signal by the preceding wave corresponding to the n-th transmission data a _n occurs at time nMT _c . Similarly, the maximum of the energy of the correlation signal by the delayed wave having the longest arrival time corresponding to the n-th transmission data a _n is the time (nM + L) T _c (L
Is an integer of 1 or more and less than M).

In the weighting coefficient generating means 100, the correlation signal z output from the correlator 208 is input to the delay element 209 whose delay time is equal to the symbol period T _d and becomes a delayed correlation signal. That is, time (nM + i) T _c = nT _d
The value of the delayed correlation signal at + iT _c is z _{(n-1) M + i} . The delayed correlation signal output from the delay element 209 is multiplied by the demodulated data output from the determiner 214 by the multiplier 210, and becomes the unmodulated correlation signal c. As described later, the value of the demodulated data output from the determiner 214 at time nT _d is α _n−1 ε {−1,1} corresponding to the (n−1) th transmission data a _n−1. is there. Therefore, the time (nM +
i) When the value of the non-modulated correlation signal at T _c = nT _d + iT _c is _{cnM + i} , the relationship expressed by the equation 3 is established.

[0012]

(Equation 3)

As described above, since z _{(n-1) M + i} is the correlation signal of the received signal corresponding to the (n-1) th transmission data a _n-1, it depends on the transmission data a _n-1 . Data is being modulated. The unmodulated correlation signal c _{nM + i} is the correlation signal z
_The data modulation component contained in the correlation signal is removed by multiplying _{(n-1) M + i} by the demodulated data α _n-1 .

The unmodulated correlation signal c output from the multiplier 210 is input to the conjugate circuit 211, and the conjugate circuit 211 outputs a weighting coefficient which is the complex conjugate number of the unmodulated correlation signal. That is, time (nM + i) T _c = nT
The value of the weighting coefficient output from the conjugate circuit 211 at _d + iT _c is expressed by Equation 4.

[0015]

(Equation 4)

When the rake method is used for an antipodal signal such as a two-phase PSK modulated signal, the complex conjugate number of the correlation signal delayed by one symbol period as described above is multiplied by the demodulated data one symbol before. The use as a coefficient is described in the above-mentioned document "DIGITAL C".
OMMUNICATIONS ”.

The weighting coefficient output from the conjugate circuit 211 is input to the correction circuit 212. The correction circuit 212 corrects the weighting coefficient according to the magnitude relationship between the square of the absolute value of the input weighting coefficient and the given threshold value, and outputs the corrected weighting coefficient w. Here, when the value of the modified weighting coefficient at time (nM + i) T _c = nT _d + iT _c is w _{nM + i} and the threshold value is h,
w _{nM + i} is determined as follows.

[0018]

(Equation 5)

However, since α _n-1 ε {-1,1}, the relation expressed by the equation 6 is clearly established.

[0020]

(Equation 6)

That is, the correction circuit 212 operates so that the corrected weighting coefficient w at the time when the energy | z _{(n-1) M + i} | ² of the delayed correlation signal becomes equal to or less than the threshold value h becomes zero. . Therefore, by appropriately selecting the threshold value h,
The corrected weighting coefficient w at the time when it can be estimated that there is no remarkable maximum of the energy of the delayed correlation signal corresponding to the preceding wave or the delayed wave, that is, only the noise component is present in the delayed correlation signal is set to 0. Therefore, the reliability of the combined correlation signal d output from the correlation signal combining circuit 213 described later is improved.

The modified weighting coefficient w output from the correction circuit 212 is input to the correlation signal synthesis circuit 213 together with the correlation signal output from the correlator 208. The correlation signal synthesizing circuit 213 weights the input correlation signal z with the modified weighting coefficient w and synthesizes it to generate a synthesized correlation signal d, which is output for each symbol period T _d . Here, if the value of the composite correlation signal at the time nMT _c = nT _d is d _n−1 , the relationship expressed by Expression 7 is established.

[0023]

(Equation 7)

By the above signal processing, signal combination (path diversity) by the rake method is realized.

The synthesized correlation signal d output from the correlation signal synthesis circuit 213 is input to the determination circuit 214. The determination circuit 214 determines the demodulated data according to the value of the input combined correlation signal d. Depending on the value of the combined correlation signal d _{n-1 at} the time nMT _c = nT _d , the demodulation data α _n-1 ε {-1,1} corresponding to the (n-1) th transmission data a _n-1 .
Is determined as follows.

[0026]

(Equation 8)

Finally, from the judgment circuit 214, time nMT _c
= NT _d , the demodulated data α _n-1 is output.

[0028]

The conventional spread spectrum demodulation device using the rake system is configured as described above, and the correlation signal delayed by one symbol period is multiplied by the demodulation data one symbol before. The complex conjugate number is used as the weighting coefficient. Therefore, when the carrier-to-noise power ratio of the received signal is low, the signal-to-noise power ratio of the correlation signal that is the output of the correlator is also low, and the error due to the noise included in the weighting coefficient becomes large, resulting in an accurate error. Since the weighted correlation signal synthesis becomes impossible, there is a problem that the error rate characteristic of the demodulated data deteriorates.

The present invention has been made in order to solve the above problems. Even when the carrier-to-noise power ratio of a received signal is low, a weighting coefficient with a small error due to noise can be obtained. An object of the present invention is to obtain a spread spectrum demodulation device using a rake method that enables correlation signal synthesis with various weights and prevents deterioration of error rate characteristics of demodulated data.

[0030]

In order to achieve the above object, a spread spectrum demodulation device according to the present invention uses a signal which is spread spectrum by a direct spread method by a PN signal as a received signal, and the received signal Baseband signal generation means for generating a more complex baseband signal, sampling means for sampling the complex baseband signal, and correlation between the complex baseband signal sampled by the sampling means and the PN signal Correlation signal generating means for generating a signal,
Weighting coefficient generating means for generating a weighting coefficient by performing moving average processing on the unmodulated correlation signal obtained by removing the data modulation component contained in the correlation signal from the correlation signal, and the correlation generated by the correlation coefficient generating means. Correlation signal synthesizing means for generating a synthesized correlation signal by weighting the signals by the weighting coefficient and synthesizing the signals, and data determining means for deciding the value of the demodulated data by the value of the synthetic correlation signal are provided. .

The spread spectrum restoration according to claim 2
The adjusting device uses a pseudo noise (hereinafter referred to as PN) signal.
The received signal is the signal that has been spread spectrum by the direct spread method.
To generate a complex baseband signal from the received signal.
And a baseband signal of the complex baseband signal.
Sampling means for sampling and sampling by the sampling means
Phase between the generated complex baseband signal and the PN signal
Correlation signal that generates a relation number (hereinafter referred to as correlation signal)
Generating means and data included in the correlation signal from the correlation signal.
Weighting coefficient from the unmodulated correlation signal with the modulation component removed
And a correlation coefficient generating means for generating
The correlation signal generated by the means is
Correlation signal that generates a composite correlation signal that is weighted and combined.
Demodulation data according to the signal synthesis means and the value of the synthesized correlation signal
Data determining means for determining the value of
The coefficient generation means converts the unmodulated correlation signal into a forgetting coefficient.
Characterized by having means for performing weighted average processing by
Things.

[0032]

In the spread spectrum demodulator of the present invention according to claim 1 configured as described above, the weighting coefficient generating means removes the data modulation component contained in the correlation signal from the unmodulated correlation signal. By performing the moving average process, even when the carrier-to-noise power ratio of the received signal is low, it is possible to obtain a weighting coefficient with a small error due to noise, thereby enabling accurate weighted correlation signal synthesis.

Further, in the spread spectrum demodulating apparatus of the present invention according to claim 2 configured as described above, the unweighted correlation signal is obtained by removing the data modulation component contained in the correlation signal from the correlation signal by the weighting coefficient generating means. By performing weighted averaging using the forgetting factor, it is possible to obtain a weighting factor with a small error due to noise even when the carrier-to-noise power ratio of the received signal is low. Synthesis is possible.

[0034]

[Embodiment 1] Embodiment 1 of the present invention will be described below with reference to the drawings. First Embodiment FIG. 1 is a block diagram showing the configuration of a spread spectrum demodulation device using a rake method according to a first embodiment of the present invention.
In the figure, 300 is a weighting coefficient generation means, and 3
A moving average processing unit 01 performs a moving average process on the unmodulated correlation signal c. It should be noted that the same or corresponding parts as those in FIG.

Next, the operation will be described. Here, the symbols and the like used in the following description are defined as follows, as in the conventional example. The symbol period of data is T _d , the repetition period of the PN signal used for spread spectrum is M (M is an integer of 2 or more) chips, and the chip period is T _c = T _d / M. Further, the value of the m-th (m = 1, ..., M) PN signal is u _m ε {−1,1}, and the value of the transmission data at time nT _d (n is an integer) is a _n ε {−1, 1} and is, on the transmission side is assumed to generate a spread spectrum signal of the direct spread system by multiplying a PN signal {u _m} in the data sequence {a _n}. That is, it is assumed that the value of the spread spectrum signal at time nT _d + mT _c is a _n · u _m , and the transmission signal is 2-phase PSK modulated by this spread spectrum signal.

In FIG. 1, the received signal is converted into a complex baseband signal y sampled for each chip period T _c via the baseband signal generating means and then the sampling means as in the conventional example. The sampled in-phase baseband signal that constitutes the sampled complex baseband signal y is y _p , the sampled quadrature baseband signal is y _q, and
When the values at the time kT _c (k is an integer) are y _k , y _pk , and y _qk , respectively, the relationship expressed by the equation 1 is established.

[0037]

(Equation 9)

The sampled complex baseband signal y is
Is input to the correlator 208 cross-correlation calculation between the PN sequence {u _m} is performed, the correlation signal z is output. Here, the time _{(nM + i) T c =} nT d + iT c (i = 0, ..., M
If the value of the correlation signal in -1) is z _{nM + i} , the relationship expressed by Equation 2 holds.

[0039]

(Equation 10)

Under selective fading, the received signal is a combination of various signal waves having different arrival times. Hereinafter, the signal wave with the shortest arrival time is referred to as the preceding wave,
Other signal waves are called delayed waves. In general, in the spread spectrum communication direct diffusion system, PN signal {u _m} are sequences such as autocorrelation coefficients is impulsive (e.g. M-sequence, etc.) is selected. Such self-correlation properties of PN signal {u _m}, in the correlation signal output from the correlator 208, the preceding wave and the energy of each delay wave is separated. That is, it is a value indicating the energy of the correlation signal |
z _{nM + i} | ² has a remarkable maximum at a time corresponding to the arrival time of the preceding wave or the delayed wave. Hereinafter, it is assumed that the maximum of the energy of the correlation signal by the preceding wave corresponding to the n-th transmission data a _n occurs at time nMT _c . Similarly, the maximum of the energy of the correlation signal by the delayed wave having the longest arrival time corresponding to the n-th transmission data a _n is the time (nM + L) T _c (L
Is an integer of 1 or more and less than M).

Now, the correlation signal z output from the correlator 208 is input to the correlation signal synthesizing circuit 213 together with the weighting coefficient w generated by the weighting coefficient generating means 300 having a new configuration with the correlation signal z as an input. . The correlation signal synthesizing circuit 213 generates a synthesized correlation signal d by weighting the inputted correlation signal z with the weighting coefficient w and synthesizing the weighted weight.

In the weighting coefficient generating means 300 having the above-mentioned new structure, the moving average processing using the unmodulated correlation signal c obtained by removing the data modulation component contained in the correlation signal from the correlation signal z output from the correlator 208 is input. In section 301, moving average processing is performed to obtain moving average correlation signal v. The moving average correlation signal v is supplied to the correction circuit 2 via the conjugate circuit 211.
It is input to 12, and is corrected in accordance with the magnitude relationship between the square of the absolute value and the given threshold value h, and is output as the corrected weighting coefficient w.

The details will be described below. The correlation signal z output from the correlator 208 is input to the delay element 209 whose delay time is equal to the symbol period Td and becomes a delayed correlation signal.
That is, the value of the delayed correlation signal at the time _{(nM + i) T c =} nT d + iT c is _{z (n-1) M +} i. Delay element 20
The delayed correlation signal output from 9 is output by the multiplier 210.
The demodulated data output from the determiner 214 is multiplied to form the unmodulated correlation signal c. As will be described later, the determiner 214
The value of the demodulated data output at time nTd is (n-
1) α _n-1 ε {− corresponding to the transmission data a _n-1
1, 1}. Therefore, time (nM + i) T _c = nT
_If the value of the non-modulated correlation signal at _d + iT _c is c _{nM + i} , the relationship expressed by Equation 3 holds.

[0044]

[Equation 11]

As described above, z _{(n-1) M + i} is (n-
It is a correlation signal of the reception signal corresponding to the 1) -th transmission data a _n-1 and is data-modulated by the transmission data a _n-1 . The unmodulated correlation signal c _{nM + i} is the correlation signal z
_The data modulation component contained in the correlation signal is removed by multiplying _{(n-1) M + i} by the demodulated data α _n-1 .

The unmodulated correlation signal c output from the multiplier 210 is input to the moving average processing unit 301, and from the moving average processing unit 301, N times the symbol interval of the unmodulated correlation signal (N is an integer of 2 or more). ) A moving average correlation signal v, which is a signal obtained by multiplying the moving average value by N, is output. That is, time (nM +
i) When the value of the moving average correlation signal at T _c = nT _d + iT _c is v _{nM + i} , the relationship expressed by Expression 9 holds.

[0047]

(Equation 12)

By this moving average processing, the moving average correlation signal v has a higher signal-to-noise power ratio than the unmodulated correlation signal c.

The moving average correlation signal v output from the moving average processing unit 301 is input to the conjugate circuit 211, and the weighting coefficient which is the complex conjugate number of the moving average correlation signal v is output. That is, the time _{(nM + i) T c =} nT d + i
The value of the weighting coefficient output from the conjugate circuit 211 at T _c is v ^* _{nM + i} (* means complex conjugate).

The weighting coefficient output from the conjugate circuit 211 is input to the correction circuit 212. The correction circuit 212 corrects the weighting coefficient according to the magnitude relationship between the square of the absolute value of the input weighting coefficient and the given threshold value,
The modified weighting factor w is output. Here, when the value of the modified weighting coefficient at time (nM + i) T _c = nT _d + iT _c is w _{nM + i} and the threshold value is h, w
_{nM + i} is determined as in Expression 10.

[0051]

(Equation 13)

The modified weighting coefficient w output from the correction circuit 212 is input to the correlation signal synthesis circuit 213 together with the correlation signal z output from the correlator 208.
The correlation signal synthesizing circuit 213 weights the input correlation signal z with the modified weighting coefficient w and synthesizes it to generate a synthesized correlation signal d, which is output for each symbol period T _d . Here, if the value of the combined correlation signal at time nMT _c = nT _d is d _n−1 , the relationship expressed by Expression 11 is established.

[0053]

[Equation 14]

By the above signal processing, signal combination (path diversity) by the rake method is realized.

The combined correlation signal d output from the correlation signal combining circuit 213 is input to the determination circuit 214. The determination circuit 214 determines the demodulated data according to the value of the input combined correlation signal d. Depending on the value of the combined correlation signal d _{n-1 at} the time nMT _c = nT _d , the demodulation data α _n-1 ε {-1,1} corresponding to the (n-1) th transmission data a _n-1 .
Is determined as follows as in the conventional example.

[0056]

(Equation 15)

Finally, the determination circuit 214 outputs the time nMT.
_The demodulated data α _n-1 is output at _c = nT _d .

In the above embodiment, the case where the two-phase PSK modulation system is used as the modulation system has been illustrated, but the present invention is not limited to this, and other multi-phase PSK modulation systems, for example, four-phase PSK system.
The SK modulation method or the like may be used. Further, other modulation methods such as MSK modulation method and GMSK modulation method may be used. In the above embodiment, the case where the sampling interval of the sampler is equal to the chip period is illustrated, but the sampling interval may be 1 / K of the chip period (K is an integer of 1 or more). It may be ½ or ¼.

Embodiment 2 FIG. Next, FIG. 2 is a block diagram showing the configuration of a spread spectrum demodulation device using the rake method according to a second embodiment of the present invention. In the figure, 400 is a weighting coefficient generation means, 401 is a weighted average processing unit for performing a weighted average processing of the unmodulated correlation signal c, 402 is an adder, 403 is a delay element, 4
Reference numeral 04 is a multiplier. It should be noted that the same or corresponding parts as those in FIG.

Next, the operation will be described. Here, the symbols and the like used in the following description are defined as follows, as in the conventional example. The data symbol period is T _d , the repetition period of the PN signal used for spread spectrum is M (M is an integer of 2 or more) chips, and the chip period is T _c = T _d / M.
The value of the m-th (m = 1, ..., M) PN signal is u _m.
∈ {-1,1}, (where n is an integer) times nT _d is the value of the transmission data in a a _n ∈ {-1,1}, the data sequence on the transmission side {a _n} PN signal {u _m} in A direct spread spectrum spread signal is generated by multiplying by. That is, it is assumed that the value of the spread spectrum signal at time nT _d + mT _c is a _n · u _m , and the transmission signal is 2-phase PSK modulated by this spread spectrum signal.

In FIG. 2, the received signal is converted into a complex baseband signal y sampled for each chip period Tc through the baseband signal generating means and then the sampling means as in the conventional example. The sampled in-phase baseband signal that constitutes the sampled complex baseband signal y is y _p , the sampled quadrature baseband signal is y _q, and
When the values at the time kT _c (k is an integer) are y _k , y _pk , and y _qk , respectively, the relationship expressed by the equation 1 is established.

[0062]

(Equation 16)

The sampled complex baseband signal y is
Is input to the correlator 208 correlation calculation between the PN sequence {u _m} is carried out, the correlation signal is output. Here, the time _{(nM + i) T c =} nT d + iT c (i = 0, ..., M-
When the value of the correlation signal in 1) is z _{nM + i} , the relationship expressed by Equation 2 is established.

[0064]

[Equation 17]

Under the selective fading, the received signal is a combination of various signal waves having different arrival times. Hereinafter, the signal wave with the shortest arrival time is referred to as the preceding wave,
Other signal waves are called delayed waves. In general, in the spread spectrum communication direct diffusion system, PN signal {u _m} are sequences such as autocorrelation coefficients is impulsive (e.g. M-sequence, etc.) is selected. Such self-correlation properties of PN signal {u _m}, in the correlation signal output from the correlator 208, the preceding wave and the energy of each delay wave is separated. That is, it is a value indicating the energy of the correlation signal |
z _{nM + i} | ² has a remarkable maximum at a time corresponding to the arrival time of the preceding wave or the delayed wave. Hereinafter, it is assumed that the maximum of the energy of the correlation signal by the preceding wave corresponding to the n-th transmission data a _n occurs at time nMT _c . Similarly, the maximum of the energy of the correlation signal by the delayed wave having the longest arrival time corresponding to the n-th transmission data a _n is the time (nM + L) T _c (L
Is an integer of 1 or more and less than M).

Now, the correlation signal z output from the correlator 208 is input to the correlation signal synthesizing circuit 213 together with the weighting coefficient w generated by the weighting coefficient generating means 400 having a new configuration with the correlation signal z as an input. . The correlation signal synthesizing circuit 213 generates a synthesized correlation signal d by weighting the inputted correlation signal z with the weighting coefficient w and synthesizing the weighted weight.

In the weighting coefficient generating means 400 having the above-mentioned new structure, the weighted average processing is performed by inputting the non-modulated correlation signal c obtained by removing the data modulation component contained in the correlation signal from the correlation signal z output from the correlator 208. In the unit 401, weighted average processing using the forgetting factor λ is performed to obtain a weighted average correlation signal x. The weighted average correlation signal x is the conjugate circuit 211.
Is input to the correction circuit 212 via the, the correction is performed according to the magnitude relationship between the square of the absolute value and the given threshold value h, and the corrected weighting coefficient w is output.

The details will be described below. The correlation signal z output from the correlator 208 is input to the delay element 209 whose delay time is equal to the symbol period T _d and becomes a delayed correlation signal.
That is, the value of the delayed correlation signal at the time _{(nM + i) T c =} nT d + iT c is _{z (n-1) M +} i. Delay element 20
The delayed correlation signal output from 9 is output by the multiplier 210.
The demodulated data output from the determiner 214 is multiplied to form the unmodulated correlation signal c. As will be described later, the determiner 214
The value of the demodulated data output at time nT _d is (n−
1) α _n-1 ε {− corresponding to the transmission data a _n-1
1, 1}. Therefore, time (nM + i) T _c = nT
_If the value of the non-modulated correlation signal at _d + iT _c is c _{nM + i} , the relationship expressed by Equation 3 holds.

[0069]

(Equation 18)

As described above, z _{(n-1) M + i} is (n
It is a correlation signal of the reception signal corresponding to the -1) th transmission data a _n-1 , and is data-modulated by the transmission data a _n-1 . The unmodulated correlation signal c _{nM + i} is the correlation signal z
_The data modulation component contained in the correlation signal is removed by multiplying _{(n-1) M + i} by the demodulated data α _n-1 .

The unmodulated correlation signal c output from the multiplier 210 is input to the adder 402, and the multiplier 404 described later is used.
The delayed weighted average correlation signal multiplied by the forgetting factor output from is added to form the weighted average correlation signal x. This weighted average correlation signal x becomes a delayed weighted average correlation signal via the delay element 403 whose delay time is equal to the symbol period T _d , and is input to the multiplier 404 to be multiplied by the forgetting factor.
However, the value λ of the forgetting factor is assumed to be represented by Expression 12.

[0072]

[Equation 19]

The multiplier 404 outputs the delayed weighted average correlation signal multiplied by the forgetting factor as described above. Time (nM + i) when the value of the weighted average correlation signal at T _c = nT _d + iT _c and x _{nM + i} relation of Formula 13 is established.

[0074]

(Equation 20)

This equation is a recurrence equation for the weighted average correlation signal x _{nM + i} . Equation 14 is obtained from this.

[0076]

(Equation 21)

That is, the weighted average correlation signal x is a weighted average of the unmodulated correlation signal c with the forgetting factor λ. This weighted average processing improves the signal-to-noise power ratio of the weighted average correlation signal x as compared with the unmodulated correlation signal c. The value of the weighted average correlation signal x does not diverge by setting the value of the forgetting factor λ to 0 or more and less than 1 as in the above Expression 12. Compared with the circuit for taking the moving average in the first embodiment, the circuit for taking the weighted average has an advantage that the circuit configuration is simple.

The weighted average correlation signal x output from the adder 402 is input to the conjugate circuit 211, and the weighting coefficient which is the complex conjugate number of the weighted average correlation signal x is output. That is, the value of the weighting factor that is output from the conjugate circuit 211 at time _{(nM + i) T c =} nT d + iT c becomes x ^* _{nM +} i ^(* denotes the complex conjugate).

The weighting coefficient output from the conjugate circuit 211 is input to the correction circuit 212. The correction circuit 212 corrects the weighting coefficient according to the magnitude relationship between the square of the absolute value of the input weighting coefficient and the given threshold value, and outputs the corrected weighting coefficient w. Here, when the value of the modified weighting coefficient at time (nM + i) T _c = nT _d + iT _c is w _{nM + i} and the threshold value is h,
w _{nM + i} is determined as in Expression 15.

[0080]

(Equation 22)

The modified weighting coefficient w output from the correction circuit 212 is input to the correlation signal synthesis circuit 213 together with the correlation signal z output from the correlator 208.
The correlation signal synthesizing circuit 213 generates a synthesized correlation signal d by weighting and synthesizing the input correlation signal z with the modified weighting coefficient w, and outputs it at every symbol period Td. Here, if the value of the composite correlation signal at time nMT _c = nT _d is d _n−1 , the relationship expressed by Expression 16 is established.

[0082]

(Equation 23)

By the above signal processing, signal combination (path diversity) by the rake method is realized.

The combined correlation signal d output from the correlation signal combination circuit 213 is input to the determination circuit 214. The determination circuit 214 determines the demodulated data according to the value of the input combined correlation signal d. Depending on the value of the combined correlation signal d _{n-1 at} the time nMT _c = nT _d , the demodulation data α _n-1 ε {-1,1} corresponding to the (n-1) th transmission data a _n-1 .
Is determined as follows, as in the conventional example.

[0085]

(Equation 24)

Finally, the determination circuit 214 outputs the time nMT.
_The demodulated data α _n-1 is output at _c = nT _d .

In the above embodiment, the case where the two-phase PSK modulation method is used as the modulation method is illustrated, but the present invention is not limited to this, and other multi-phase PSK modulation methods, for example, four-phase PSK.
The SK modulation method or the like may be used. Further, other modulation methods such as MSK modulation method and GMSK modulation method may be used. In the above embodiment, the case where the sampling interval of the sampler is equal to the chip period is illustrated, but the sampling interval may be 1 / K of the chip period (K is an integer of 1 or more). It may be ½ or ¼.

[0088]

As described above, according to the invention of claim 1, the weighting coefficient generating means removes the data modulation component contained in the correlation signal from the correlation signal to perform the moving average processing on the unmodulated correlation signal. Even if the carrier-to-noise power ratio of the received signal is low, it is possible to obtain a weighting coefficient with a small error due to noise, which enables accurate weighted correlation signal synthesis and the error rate of the demodulated data. It is possible to obtain a spread spectrum demodulation device using a rake method that can prevent deterioration of characteristics.

According to the second aspect of the present invention, the weighted coefficient generating means performs the weighted average processing using the forgetting coefficient on the non-modulated correlation signal from which the data modulation component included in the correlation signal is removed from the correlation signal. By doing so, even when the carrier-to-noise power ratio of the received signal is low, it is possible to obtain a weighting coefficient with a small error due to noise, which enables accurate weighted correlation signal synthesis and error rate characteristics of demodulated data. It is possible to obtain a spread spectrum demodulation device using the rake system that can prevent the deterioration of

[Brief description of the drawings]

FIG. 1 is a configuration block diagram of a spread spectrum demodulation device using a rake system according to a first embodiment of the present invention.

FIG. 2 is a configuration block diagram of a spread spectrum demodulation device using a rake system according to a second embodiment of the present invention.

FIG. 3 is a configuration block diagram showing a spread spectrum demodulation device using a conventional rake method.

[Explanation of symbols]

 200 oscillator 201 phase shifter 202, 203 multiplier 204, 205 low-pass filter 206, 207 sampler 208 correlator 209 delay element 210 multiplier 211 conjugate circuit 212 correction circuit 213 correlation signal synthesis circuit 214 determiner 300 weighting coefficient generation means 301 Moving Average Processing Unit 400 Weighting Coefficient Generating Unit 401 Weighted Average Processing Unit 402 Adder 403 Delay Element 404 Multiplier

Claims (2)

(57) [Claims] 1. A baseband signal generating means for generating a complex baseband signal from a received signal, which is a signal spread in spectrum by a direct spread method with a pseudo noise (hereinafter referred to as PN) signal, and the complex signal. Sampling means for sampling the baseband signal, and correlation signal generation for generating a cross-correlation coefficient (hereinafter referred to as a correlation signal) between the complex baseband signal sampled by the sampling means and the PN signal Means for generating a weighting coefficient by performing a moving average process on an unmodulated correlation signal obtained by removing a data modulation component included in the correlation signal from the correlation signal, and generating by the correlation coefficient generation means Correlation signal synthesizing means for generating a synthesized correlation signal by synthesizing the weighted correlation signal by the weighting coefficient, and a value of the synthesized correlation signal Spread spectrum demodulation apparatus characterized by comprising: a data determination means for determining the value of the demodulated data, the Ri. 2. A pseudo noise (hereinafter referred to as PN) signal
Receives signals that have been spread spectrum using a more direct spread method.
Signal to generate a complex baseband signal from the received signal.
Baseband signal generating means, sampling means for sampling the complex baseband signal, and complex baseband signal sampled by the sampling means
And the above-mentioned PN signal cross-correlation coefficient (hereinafter referred to as a correlation signal
And a data modulation component included in the correlation signal from the correlation signal.
A weight that generates a weighting coefficient from the removed unmodulated correlation signal
And a correlation coefficient generated by the correlation coefficient generation means.
Synthetic correlation signal weighted by weighting coefficient and synthesized
And a correlation signal synthesizing means for generating a value of the demodulated data according to the value of the synthesized correlation signal.
A spread spectrum demodulation device including data determination means
Therefore, the weighting coefficient generation means applies to the unmodulated correlation signal.
It is equipped with a means for weighted averaging with the forgetting factor.
Characteristic spread spectrum demodulator.