JPH02230845A - Psk signal demodulation system - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a coherent detection demodulation method for PSK modulated signals.

(Prior Art) Demodulation methods for PSK modulated signals are broadly classified into synchronous detection methods and delayed detection methods. Among these, the synchronous detection method has a slightly more complex demodulator configuration than the differential detection method, but has excellent demodulation bit error rate characteristics. Therefore, P.S.K.
In recent years, a synchronous detection method has been widely used as a demodulation method for modulated signals.

In the synchronous detection method, it is necessary to recover a reference carrier wave that is phase-synchronized with the received signal carrier wave on the receiving side.
The L (Phase Locked Loop) method and the like have been proposed.

(Problems to be Solved by the Invention) In these conventional systems, when the line condition is poor, for example, when the received signal is weak, that is, when the carrier wave power to noise power ratio (C/N) of the received signal is low, the received signal It takes a long time to establish phase synchronization between the carrier wave and the regenerated carrier wave.
Furthermore, there is a problem in that synchronization is likely to be lost due to variations in reception frequency or C/N, and once synchronization is lost, it takes a long time to establish synchronization again.

As mentioned above, in the conventional PSK signal synchronous detection method, there are limits to the C/N of the received signal that allows carrier wave recovery, the size of the steady frequency deviation of the received signal, or the size of the frequency fluctuation. Therefore, if the line condition is poor, carrier wave regeneration cannot be performed with high precision, and the application of communication systems using PSK signals is also limited.

The present invention was made to solve the above-mentioned drawbacks of the prior art, and it is possible to demodulate with high precision even when the C/N of the received signal is very low and there is a large frequency deviation in the received signal. The purpose is to provide a PSK signal demodulation method.

(Means for Solving the Problems) The feature of the present invention is to perform quasi-synchronous detection on a phase-modulated PSK received signal, and then perform digital signal processing on the PSK signal, which is sampled at a predetermined period and converted into a digital value. In the PSK signal demodulation method, the digitized PSK signal is branched and the branched P
One of the SK signals is converted from a received signal in the time domain to a signal in the frequency domain by FFT to obtain its power spectrum, and the carrier frequency of the received signal is estimated from the obtained electric scum vector, and the estimated carrier frequency is After correcting the frequency of the other branched PSK signal in the time domain using After performing clock regeneration on one side and removing the signal modulation component of the other branched signal to obtain an unmodulated signal, the signal is converted into the frequency domain using a sliding DFT method that performs adaptive processing to frequency fluctuations of the input signal. The purpose of this method is to demodulate the PSK signal after correcting the frequency deviation and phase of the clock-regenerated received signal with the noise component reduced using the unmodulated signal converted to the frequency domain. (Principle of the Invention) The present invention performs quasi-coherent detection of a PSK signal once on the receiving side using a reference carrier wave having a frequency close to the carrier frequency of the received signal, and the received signal and the reference signal obtained by the quasi-coherent detection. A PSK modulated signal with a carrier frequency difference of
It is converted into digital data by A/D conversion. The power vector of the time-domain received signal converted into a digital value is determined by the FFT method, and the carrier frequency of the received PSK signal is estimated in the frequency domain. The received signal in the time domain is corrected by a bandpass filter having the estimated frequency as the center frequency or by using the estimated frequency.
Noise is reduced by a low-pass filter. The received signal in the time domain with reduced noise is multiplied by, for example, M to become an unmodulated signal in order to remove modulation components. here,
M represents the number of phases of the PSK signal, for example, 2-phase PSK (
BPSK) signal, M=2. The unmodulated signal in the time domain is then processed using the sliding DFT method (rS-
(referred to as DFTJ), the power spectrum is determined by
Carrier frequency and phase are estimated. Then, if the quasi-synchronous detection signal in the time domain is corrected using these estimated frequencies and phase values, the PSK signal will be synchronously detected.

Here, the S-DFT method does not convert an input consisting of a certain N points into a frequency domain signal block by block, unlike the normal DFT or FFT method. R. Rabiner et al.”Th
theory and application of D
digitalSignal Processing,
As described in Prentice-Hall 1975), DFT operations are performed sequentially on sample values in each input time domain.

In general, in the DFT or FFT method, the longer the input observation time, the more accurate the frequency spectrum can be obtained, but since the number of points increases, the calculation time also becomes longer.

On the other hand, if the S-DFT method is used, conversion to the frequency domain can be performed with high accuracy without delay.

Therefore, in the present invention, S
- By performing carrier wave recovery in the frequency domain using the DFT method, it is possible to demodulate a PSK signal with high precision even when the reception C/N is low and the frequency deviation is large.

(Example) The present invention will be explained in detail below using the drawings.

FIG. 1 is a block diagram of a PSK signal demodulation method according to the present invention. Note that in the following, the input PSK modulation signal is an M-phase PSK signal.

In the figure, 100 is a manual terminal for the received PSK signal, 1 is a band pass filter (hereinafter referred to as rBPFJ), 2.2
' is a multiplier, 3 is a π/2 phase shifter, 4 is a fixed frequency oscillator, 5.5' is a low pass filter (hereinafter referred to as rLPFJ), 6 is an A/D (analog/digital) converter, 7
is a digital signal processing type demodulator which is a feature of the present invention, 1
11 is a demodulated data output terminal, and 101 to 110 indicate respective signals.

Here, the input terminal 100 has an M-phase PSK signal expressed by the following equation.
A signal is input.

Sk(t) = JAcos ((IJ c't+θe+
θo)--(t) Equation (1) does not take into account noise added on the transmission path. A indicates the amplitude level, ω
.゛ is the carrier angular frequency, θ. indicates the initial phase, and θ is the information phase according to the k-th information bit, for example,
O or π for 2-phase PSK (M=2), 4-phase PSK
(M=4), take O, π/2, π, 3π/2. Here, the angular frequency ω of the carrier wave of the received signal is expressed by equation (1).

゛ is the carrier wave frequency ω on the transmitting side depending on the stability of the transmitting carrier wave and the stability of the frequency converter, fixed oscillator, etc. used in the line. has a different value. That is, the receiving side will have a frequency deviation of (ω.' - ωC). Therefore, the bandwidth of the bandpass filter 1 must be determined in consideration of the maximum frequency deviation occurring in the transmission path. The signal from which out-of-band signals and noise have been removed by the bandpass filter 1 is subjected to quasi-synchronous detection using a reference carrier obtained from a fixed oscillator 4. Fixed oscillator 4
Reference carrier 104 (R.) generated at , and π/2
The reference carrier wave 103 (R.) output from the phase shifter 3 is expressed as follows.

Re(t) :17TcOs ((xJ c ”t
10I)---(2)R*(t) =Jsin(ωe
"t+θ,)...(3) ω in equations (2) and (3).

′゛, θ1 indicate the angular frequency and initial phase value of the fixed oscillator. Here, ω.゛′ is the transmitted angular frequency ω. It has a value close to ω in the following explanation. =ω. ″.

The received signal 101 of equation (1) is expressed by equation (2). By multiplying by the reference carrier waves 103 and 104 shown in (3), signals 105 and 106 as shown in the following equation are obtained.

el(t)=S, l(t)・Rc(t)= Acos
{(ωc'- ωe)t+θ5+(θ.-θ1)}+
Acos {(ω.′+ω.)t÷θ5+θ. +θ1}θ
5+θ. +θ1}...(4) ex(t) =Sv(t') ・R-(t):-As
tn{((aJc'- Cdc)j10+(θ.10+)) + Asin{(ωe” ωc)t+θ5+
θ. +01}...(5) The signals shown by equations (4) and (5) are passed through a low-pass filter (
The harmonic components are removed by LPF) 5.5' to obtain signals 107 and 108 as shown in the following equations.

e+'(t) = Acos (Δωt 105 ÷ Δθ)
−−−(6) ex'(t) = −Asin(Δωt+
θ8÷Δθ) − − (7) Here, Δω, Δθ
respectively indicate the frequency deviation amount and phase difference (ω.゛−ω.) and (θ.−ω.) between the received signal and the reference carrier wave.

The signals represented by equations (6) and (7) are converted into digital signals 109 by the A/D converter 6 as shown in the following equation. It becomes 110.

X1 ” e+'(t.) = Acos (ΔωtI+θ ÷ Δθ) --- (8) y
I= e2'(tt) =-A sin(Δωtl÷θk÷Δθ) − − (
9) However, 1+ represents the i-th sample time.

These digital data signals are sequentially input to a digital signal processing type demodulation circuit 7, which is a feature of the present invention.

Here, these signals are signals that have passed through a band pass filter (BPF) l, and if it is assumed that the frequency deviation of the received signal is large, the bandwidth of the BPFI must be widened as described above. In this case, these signals x. ,y
．． contains noise corresponding to the bandwidth of the BPFI, and the larger the received frequency deviation, the worse the C/N of the BPFI output signal becomes, making it difficult to recover the carrier wave. Therefore, in the present invention, in order to demodulate the received signal when the line condition is poor, a method is used to reduce the noise level by processing the signal converted into the frequency domain using digital signal processing. There is.

FIG. 2 is a concrete block diagram of the digital signal processing type demodulation circuit 7 according to the present invention. Input terminal 1 in Figure 2
09, 110 is the formula (8). A signal X**Vt expressed by (9) is input. These signals are FF
The signal is converted into a frequency domain signal by a T circuit 8. F
As the output of the FT circuit 8, real value and imaginary value amplitude spectra in the frequency domain are obtained. The frequency estimator 9 calculates the power spectrum of the input signal from these amplitude spectra. In the frequency domain, the white noise component has a constant power level, whereas the modulated signal component has a large power level only within the band in which the signal exists. Therefore,
A modulated signal that has a correlation in frequency and a noise component that is uncorrelated are:
In the frequency domain, it can be easily identified despite the low received C/N. In this way, the center frequency of the received signal can be estimated relatively accurately using the FFT technique even when the line condition is poor. Depending on the frequency estimated by the frequency estimator 9, the signals Xi, y. is 10. in the time domain. At step 10', the frequency is corrected to produce a signal with a center frequency of O. In this way, the almost baseband signal has a passband width comparable to the modulation bandwidth.
Noise components are removed by PFII. By the above operations, the C/N of the received signal is (LPFII bandwidth/BP
FI bandwidth), and carrier wave regeneration, which will be described later, becomes easier. The frequency estimation using the FFT method performed here is as follows:
This is performed only at the time of initial synchronization of continuous mode signals, or when out-of-synchronization occurs after synchronization has been established.

The LPFII output signal is multiplied by M in the multiplier 12 and is expressed by the formula (
10) as an unmodulated signal S2.

s*(t+) "A"6-j('Δωtt+MΔθ'
---(10) This unmodulated signal represents the received carrier wave component itself. The output of the multiplier 12 actually includes a noise component in the signal expressed by equation (lO). The unmodulated signal is then passed through a sliding DFT circuit (
(hereinafter referred to as rS-DFTJ). S
-DFT 13 is for converting a signal inputted sequentially for each time sample into the frequency domain. In other words, frequency components that change over time can be determined for each time sample. S-DFT
As the output of 13, up to N frequency components are obtained for N time domain signals from the lth sample to the Nth sample, as in the normal DFT. Thereafter, for each input sample, frequency components corresponding to the N time signals input up to that point are output.

FIG. 3 is a concrete block diagram of the S-DFT13 circuit used in the present invention. here,. The signal is a complex signal. In FIG. 3, 30, to 30, are delay circuits (Z-N
), N is the number of delay time samples, 31. ~31n is an adder, 32. ~32. represent multipliers, respectively. In addition, input terminal 120. ~120,,1 are given weighting coefficients determined by the frequency/phase estimator 14 in FIG.

In this way, for the i-th input at a certain time sample, N
It is possible to sequentially obtain up to 3 frequency values.

Figure 5 shows the input signal (S-DF) of the frequency/phase estimator l4.
FIG. 4 is a frequency component diagram of the output of Tl3. Yo, YI,・
...y. ．． , is the output signal of S-DFTl3,
As shown in FIG. 5, the unmodulated signal has high-level frequency components in the frequency domain. Therefore, in the maximum value level determining section 41 of FIG. 4, the input signals Yo, Y. ,...
, Y1, determine the frequency component of the highest level,
The frequency/phase calculation unit 42 performs interpolation calculation (
estimate) and output. Next, the correction signal generating section 43 generates a correction signal for correcting the received signal using the obtained carrier wave frequency and phase. On the other hand, the frequency/phase calculation section 42
From the carrier frequency estimated at f. After calculating m frequency components around , the weighting coefficient determination unit 44 determines and outputs weighting coefficients W(0) to W(m).

If the received carrier frequency shifts to f 1+1, it is sufficient to give new weighting coefficients W(0) to W(m) to calculate m frequency components centered around the frequency f lli+. It turns out. That is, as shown in FIG. 4, from the frequency estimation results of the frequency/phase estimator 14, the weighting coefficients W(0) to W(m) are set so that only m frequency components are calculated around the received carrier frequency. Even if the frequency of the unmodulated signal deviates, frequency components less than N centered around that frequency can always be found.

In this way, in the present invention, by performing calculations only on frequency components near the frequency of interest, it is possible to achieve a configuration with a reduced circuit scale. Further, even if there is a sudden change in the carrier frequency, it is possible to always follow the change in mΔf. Furthermore, in the S-DFT method for frequency estimation, by taking a sufficiently long observation time of the input signal, we aim to improve the carrier power to noise power ratio in the frequency domain. It is also possible to reduce the circuit scale of the S-DFTl3 by using this. In this way, unlike the conventional PLL carrier wave regeneration system, the frequency pull-in range is not limited, and the pull-in range can be made very wide for slow frequency fluctuations.

When determining the frequency components, the frequency spectrum of the unmodulated signal obtained here has a large power as a line spectrum at the frequency MΔω. On the other hand, since the noise component exists on average at all frequencies within the band, it can be said that a method performed in the frequency domain is superior in estimating the frequency and phase of an unmodulated signal. In S-DFT13, as with normal DFT and FFT methods, the longer the observation time of the input time signal, that is, the greater the number of points, the higher the resolution of the frequency spectrum obtained, so the accuracy is higher. Frequency estimation becomes possible. Since the input unmodulated signal is a line spectrum, high frequency resolution means that the frequency step size is small, so the power of the noise component is reduced, and a relatively high level line spectrum can be obtained. Therefore, the line condition is poor,
Even when the received C/N is low, the unmodulated signal that is the received carrier wave component can be easily identified in the frequency domain.

However, the scale of the circuit increases in proportion to the number of points N,
The normal DFT and FFT methods also take a long time to calculate. However, here we only need to know the frequency with the largest line spectrum within the band, and the S-DF
Unlike normal DFT and FFT methods, T13 does not need to perform calculations on all N components as described above.
If the frequency fluctuation is continuous over time, the frequency/phase estimator l
Using the results of frequency estimation according to 4.4, it is sufficient to perform calculations to obtain only frequency components near the carrier frequency, thereby reducing the circuit scale.

Estimating the carrier frequency in the frequency domain requires low received C/N.
In addition to being able to perform the estimation with high accuracy, it also has the advantage of being easy to estimate even if there is a large frequency deviation in the carrier frequency. This is different from the conventional multiplication/branching method and PL
This is a feature that cannot be obtained with the carrier wave regeneration method using the L method. However, the estimated value obtained in the frequency domain still only has an accuracy of the resolution Δf of the S-DFT13, and the S
-Errors due to DFT13 also exist. Therefore, even if the frequency and phase are corrected using the estimation results, a very small frequency error remains. Therefore, in the present invention, a more accurate frequency and phase can be obtained by using an interpolation formula for frequency domain data obtained with a small number of samples, and a PLL with a simple configuration that has good followability in spite of a narrow frequency range. This enables highly accurate correction.

The frequency deviation and phase value estimated by the frequency/phase estimator 14 are input to the frequency/phase corrector 16.

Here, the frequency and phase corrector 16 corrects the frequency deviation and phase of the signal whose clock component has been recovered by the clock recovery circuit 15, and carrier wave recovery has been performed here.

The carrier-regenerated signal is determined as information bit data by a determiner 17, and demodulated data is outputted to an output terminal 111.

(Effects of the Invention) As described in detail above, the present invention adaptively obtains weighting coefficients for a time domain PSK signal using the S-DFT 13 and the frequency/phase estimator 14, converts it into a frequency domain signal, and converts the PSK signal into a frequency domain signal. Because of demodulation, the amount of deviation of the received carrier frequency is large, and even in low C/N conditions, accuracy can be improved in a short time.
Can be demodulated.

Further, in the present invention, by using the S-DFT technique, it is possible to demodulate sequentially inputted time-continuous signals without delay time.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a PSK signal demodulation system according to the present invention, FIG. 2 is a block diagram of a digital signal processing type demodulation circuit according to the present invention, FIG. 3 is a schematic diagram of an S-DFT used in the present invention, and FIG. 4 is a specific block diagram of the frequency/phase estimator according to the present invention, and FIG. 5 shows the input signal (S-DF
FIG. 4 is a frequency component diagram of the output of Tl3. 1 ● 2, 4 ・ 5, 6 ● 7 ・ 8 ・ lO・ 12・ l4● 15・ 16・ 30. ... Band pass filter (BPF), 2' ... Multiplier, 3 ... π/2 phase shifter, ... Fixed frequency oscillator, 5' ... Low pass filter (LPF), ... A /D
(analog/digital) converter,...digital signal processing type demodulator,...FFT circuit, 9...frequency estimator,...multiplier, 1l...LPF.・Multiplier, l3・
...S-DFT, ...Frequency/phase estimator, ...Crossic regenerator, ...Frequency, phase corrector, l7...Judgment device, ~30n
...Delay circuit (Z1''), 31.-31.... Adder, 32.-32,... Multiplier, 4l... Maximum level determination unit, 42... Frequency/phase Calculation unit, 43... Correction signal generation unit, 44... Weighting coefficient generation unit, 100... Input terminal for received PSK signal, 101-1
10...Signal, I'll...Demodulated data output terminal, l20. ~}2
0. ...input terminal

Claims (2)

[Claims] (1) In a PSK signal demodulation method that performs quasi-synchronous detection on a phase-modulated PSK received signal, samples it at a predetermined period, and demodulates the PSK signal converted into a digital value by digital signal processing. One of the branched PSK signals is converted from a time domain signal to a frequency domain signal using FFT to obtain its power spectrum, and the carrier frequency of the received signal is estimated from the obtained power spectrum. After correcting the frequency of the other branched PSK signal in the time domain using the estimated carrier frequency, noise added on the transmission path by a bandpass filter having the estimated carrier frequency as the center is corrected. After reducing the components, perform clock recovery on one of the branched signals, remove the modulation component of the other branched signal in the time domain to make it an unmodulated signal, and then convert it to a frequency domain signal. A PSK signal demodulation method, characterized in that the PSK signal is demodulated after correcting the frequency deviation and phase of the clock-regenerated received signal with the noise component reduced using the unmodulated signal. (2) When converting the unmodulated signal in the time domain to the frequency domain, a sliding DFT method is used that performs highly accurate domain conversion by performing adaptive processing on input frequency fluctuations or deviations. The PSK signal demodulation method according to claim 1.