JP3405507B2 - Demodulation apparatus and method for code transmission - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a demodulation system for OFDM transmission in which modulation / demodulation used in transmission / reception of digital signals, in particular, a digital information signal modulated by QAM or the like is transmitted using a plurality of carriers.

[0002]

2. Description of the Related Art OFDM transmission is a digital modulation method of frequency division multiplexing that transmits digital information using a plurality of orthogonal carrier waves. It is strong against multipath and does not easily interfere with other transmission systems. It has features such as difficulty in receiving and relatively high frequency use efficiency, and in recent years, it has been attracting attention as a modulation method suitable for mobile digital audio broadcasting and digital television broadcasting. Multiple carriers can be generated using an IFFT circuit that performs inverse Fourier transform on the transmission side, and at reception the carrier can be separated by an FFT path that performs Fourier transform. With the progress of the IFFT circuit and FFT circuit packaging technology, OFDM transmission is becoming a reality.

FIG. 2 shows a part of a modulation circuit in the conventional OFDM transmission, and FIG. 3 shows a part of a demodulation circuit in the OFDM transmission. The information data to be transmitted is encoded by the encoding circuit 10 such as PSK, QAM, etc., and the encoded digital data is added with a reference reference signal applied to a specific carrier by an adder 11 to obtain an IFFT operation unit. After assigning frequencies to 12 real parts and imaginary parts, IFFT
The calculation is performed, the I signal and the Q signal, which are the calculation results, are sent to the quadrature modulator 13, and after modulation, the OFDM wave is output.

On the demodulation side, the received and quadrature demodulated I and Q signals are input to the real part and the imaginary part of the FFT calculation unit 1, the FFT calculation is performed, and the calculation result is input to the reference signal calibrator 14. Then, correction is performed based on the change in the amplitude and phase of the standard reference signal applied during transmission,
The result is sent to the decoding circuit 9 to perform decoding such as PSK and QAM, and output decoded data. At this time, the applied standard reference signal is transmitted at a constant interval for each carrier, and the receiving side obtains a correction value from the change component of the amplitude and phase of the signal,
Perform signal correction. Until the next standard reference signal is transmitted, the correction value is corrected, and the correction value is updated by the next standard reference signal.

[0005]

It is said that OFDM transmission is generally resistant to multipath. This is because it is possible to avoid interference between signals by providing a signal period called a guard interval for each transmission signal. However, when QAM is used for code modulation, waveform equalization (signal correction) is essential for demodulation.
Further, when mobile reception is assumed in a multipath environment, the received signal fluctuates greatly due to movement, and therefore more accurate signal correction is required than in a static state.

The fluctuations of the received signal are as follows:
Further, the larger the difference in the received electric field strength due to multipath, the larger the variation amount in the unit time. In addition, QA
The more multi-valued M-modulation, the more severe the demodulation. In order to cope with such fluctuations, a reference reference signal is inserted into each transmission signal in each carrier and transmitted, and a change in the amplitude and phase of the signal is extracted on the receiving side, and based on that information, the information signal Although correction is performed, since the standard reference signal is transmitted at regular intervals, the correction value is not updated until the next standard reference signal is transmitted. Therefore, due to movement or changes in the surrounding environment, it is not possible to deal with a received signal that fluctuates within a minute time, and demodulation is impossible. If the number of standard reference signals is increased, the possibility of correction can be increased in proportion to the number of standard reference signals, but the amount of transmitted information decreases accordingly. Therefore, considering the realization of mobile reception in OFDM transmission, the higher the multilevel modulation system,
The important issue is how to absorb and correct the fluctuation of the received signal.

[0007]

SUMMARY OF THE INVENTION In order to solve the above problems, the present invention extracts a change in the amplitude and phase of a reference reference signal which is received and then transmitted, and a correction value based on the change component. The reference signal calibrator 2 which determines the signal and performs the signal correction and outputs the primary correction signal, and the amplitude and phase difference information read from the memory 6 for the primary correction signal, is obtained in the past along the time axis. Amplitude / phase correction circuit 5 that sequentially corrects the signal using the correction value obtained by predicting the correction value for the current signal based on the change in the correction value and outputs the secondary correction signal 5 And the amplitude / phase difference detection circuit 7 that detects the amplitude / phase difference between the secondary correction signal output as described above and the original transmission signal on the signal point arrangement decoded by the decoding circuit 9 and outputs the amplitude / phase difference information. Self consisting of memory 6 for storing amplitude / phase difference information And Yaribureta 4, composed of a decoding circuit 9 to output the decoded data by decoding the second order correction signal. In the amplitude / phase correction circuit 5, the memory 6 is reset at the first symbol immediately after the reference signal is transmitted, and as a result, the primary correction signal, which is the output of the reference signal calibration, is passed through as it is.

Based on the idea of self-calibration, the following is performed in order to increase the reliability of signal correction. First, since the amplitude and phase change is continuous in the correction value that is updated sequentially, it is possible to predict the tendency of the change and obtain the correction value, and use slope prediction or regression model prediction as the means. . Secondly, by self-calibration, the correction values are updated for each signal for all carriers and calibration is performed.At this time, the correction values obtained in the past and the correction values of the adjacent carriers are averaged and obtained. Calibration is performed by using the value obtained as a correction value.

Operation According to the above method, the received signal largely fluctuates in a short time due to movement or change in characteristics, so that the environment changes before the next reference reference signal is sent from the reference reference signal. However, the correction cannot catch up and accurate decoding is impossible, but by performing self-calibration, correction is performed while updating the correction value for each carrier for each signal, so dynamic transmission due to movement etc. Follow the fluctuations of the received signal in the path characteristics.

In the self-calibration, by averaging the time axis and the frequency axis, it is possible to derive an accurate correction value with little variation, and when a change in variation, such as acceleration / deceleration during movement, is performed by prediction processing. Since it is possible to derive the correction value by catching the change in the above, the reliability of the signal correction becomes high by using the both together to obtain the correction value.

In the self-calibration, among the correction values updated for each signal, the correction value obtained by the low-reliability signal is deleted to correct the amplitude and phase, thereby obtaining an accurate correction value with little variation. Can lead
The reliability of signal correction is high.

[0012]

DETAILED DESCRIPTION OF THE INVENTION

Example (corresponding to claims 1 and 5) The specifications of the entire OFDM device will be described. The transmitting side is the same as the conventional example. Transmission information is transmitted using a carrier of 256 waves. Use 2x oversampling to facilitate designing filters in analog circuits.
The IFFT operation of the points is performed to generate the OFDM wave. Figure 2
Shows a part of the OFDM modulator. 25 for modulation of each carrier
It is assumed that 6QAM is used, and the input information data is encoded by the encoding circuit 10, and 8-bit information is applied to one carrier, that is, 4 bits are applied to each of the real part and the imaginary part. In addition to the transmission information data, reference data for calibration and synchronization data are inserted in one symbol. The frequency allocation of transmission information to the IFFT calculation unit 12 is as follows when numbers are assigned in order from the lowest frequency in the IFFT window.

F0 to f127 Information transmission signals to be transmitted are given. f128 to f383 Set carrier level to 0 and do not generate signal. f384 to f511 The information transmission signal to be transmitted is given. The reference reference signal shall be inserted together with the coded data by the adder 11 into one carrier specific to one symbol,
The insertion carrier is transferred for each symbol. In this example, 25
Since it uses a carrier of 6 waves, once every 256 symbols,
The reference reference signal for each carrier will be transmitted. The frequency allocation is performed as described above, and the quadrature modulator 13 generates an OFDM wave from the I signal and the Q signal, which are time series signals output by the IFFT operation.

An embodiment of the OFDM demodulator of the present invention will be described below with reference to the drawings. Figure 1 shows the OFD of the present invention
It is a block diagram of one Example of the M demodulation device. FIG. 1 shows a configuration in which a self-calibrator 4 is inserted between the reference signal calibrator 14 and the decoding circuit 9 of FIG. First, the signal flow in each circuit of the OFDM demodulator of the present invention will be described. As shown in FIG. 1, the OFDM wave received and quadrature demodulated is supplied to the FFT calculation unit 1, and after the FFT calculation,
The calculation result is supplied to the reference signal calibrator 2. In the reference signal calibrator 2, for the carrier to which the reference reference signal is transmitted, the change component of the amplitude and phase of the reference reference signal is calculated, and the correction value that can correct the change is calculated. The correction value is updated by performing the correction by using, and transmitting a new standard reference signal.

The primary correction signal 3 corrected by the reference signal calibrator 2 is the amplitude / phase correction circuit 5 and the memory.
6 and the amplitude / phase difference detection circuit 7 are supplied to the self-calibrator 4. Where the primary correction signal
3 is supplied to the amplitude / phase correction circuit 5, and reset signals from the reference signal calibrator 2 described later are supplied to the memory 6, respectively. In the self-calibrator 4, the amplitude / phase correction circuit 5 fetches the information of the amplitude / phase difference of the sequentially received information signals from the memory 6, and based on the value,
The amplitude and phase of the primary correction signal 3 are corrected, and the resulting secondary correction signal 8 is output to the amplitude and phase difference detection circuit.
7 and decoding circuit 9.

In the amplitude / phase difference detection circuit 7, the secondary correction signal
The amplitude and phase difference between 8 and the decoded signal is detected, and the amplitude and phase difference information is sent to the memory 6. The memory 6 holds the information, and resets the memory 6 in response to a request (reset signal) from the reference signal calibrator 2 for the carrier to which the reference reference signal is transmitted. Second-order correction signal corrected by self-calibrator 4
The signal 8 is input to the decoding circuit 9 and subjected to QAM decoding, then output as decoded data and returned to the amplitude / phase difference detection circuit 7 of the self-calibrator 4.

Next, the operation from the reference signal calibrator 2 to the decoding circuit 9 will be described below. A new signal point frame is generated by obtaining and correcting the correction value using the standard reference signal. In the case of 256QAM, it has a frame of 256 signal points, and if the received signal exceeds this frame, it will be regarded as another signal and an error will occur. After the next symbol, this signal point frame becomes the reference until a new reference reference signal is transmitted and the correction value is updated. Self-calibration is a correction between the reference reference signal and the next reference reference signal that is transmitted using the amplitude and phase difference of the signals that are received sequentially. The signal is accurately corrected to any point on the 256QAM constellation.

This received signal is used as a reference signal in the next symbol. That is, by using the amplitude and phase change components in the received signal of the next symbol, correction of the received signal in the next symbol, that is, a new signal point frame is generated.
This state is shown in FIG. Signal point frame with standard reference signal 15
Is constant until the next standard reference signal is transmitted,
The received signal 16 changes due to movement or the like over time.
Signal point frame using the received signal so as to follow the fluctuation
17 changes. This signal point frame has changed due to multipath and movement, and is sequentially updated for each symbol. The signal point frame tracking by the self-calibration is reset by the next transmission of the standard reference signal, and the iterative process is performed with the signal point frame newly generated by the standard signal calibration as a reference.

Embodiment (corresponding to claim 3) In the self-calibration, since the correction value used for generating the signal point frame uses the change component of the received signal in the previous symbol, it is always based on the past variation. Generate a signal point frame at. Therefore, if there is a sudden change in the fluctuation, such as acceleration / deceleration during movement, or if the amount of fluctuation itself is large, such as when the moving speed is high or the difference in the received electric field strength is large, the change speed is supported. If this is not possible, the received signal will jump over the signal point frame and errors will occur easily. Therefore, high-order prediction is performed using the slope component to determine the correction value. In the method, when generating the signal point frame, not only the correction value obtained by the previous symbol but also the past correction value is used to find the slope component.When the number of correction values used is 2 or more, Higher order differences are recursively calculated, and correction values are calculated in consideration of their slopes.

FIG. 6 shows the situation. When n points of the correction values y1, y2, y3,… yn obtained in the past are given, the approximate expression passing through those points is calculated by using Newton's interpolation polynomial.
(n-1) The following polynomial can be expressed as a curve 19 in the figure, and the correction value Pk for the current symbol is given by the following equation (Equation 1).

[0021]

[Equation 1]

By using this equation, it is possible to predict the change in the correction value in consideration of the inclination. For example,
If the past correction values, y1 and y2, are used, Pk =
The signal point frame is generated based on the calculated Pk given by the simple formula of 2y1-y2. If prediction is not used, the correction value P used to generate the signal point frame in the current symbol is the correction value y1 obtained in the previous symbol, and the signal point frame is generated from that value. .

Embodiment (corresponding to claim 4) In the self-calibration, since the correction value used for generating the signal point frame uses the change component of the received signal in the previous symbol, it is always based on the past change amount. Generate a signal point frame at. Therefore, if there is a sudden change in the fluctuation, such as acceleration / deceleration during movement, or if the amount of fluctuation itself is large, such as when the moving speed is high or the difference in the received electric field strength is large, the change speed is supported. If this is not possible, the received signal will jump over the signal point frame and errors will occur easily.

Therefore, higher-order prediction using a regression model is performed to determine the correction value. The method is not only the correction value obtained by the previous symbol, but also the past correction value 2 when generating the signal point frame.
The regression model by the least-squares method is calculated using the points above,
The correction value for the current symbol is calculated from the model. The regression model to be obtained can be represented by a straight line or a curve of quadratic or higher. In this example, prediction using a linear regression model will be described. This is shown in FIG. When n points of correction values y1, y2, y3, ... yn obtained in the past are given, the minimum
The linear regression model is represented by a straight line 20 in the figure by the square method, and the correction value Pc at the current symbol position on the straight line is given by the following equation (Equation 2).

[0025]

[Equation 2]

By using this equation, it is possible to predict the change in the correction value by the linear regression model, and since the least squares method is used, it is effective in a series including an error such as white noise, and was found. A signal point frame is generated based on Pc.

Embodiment (corresponding to claims 2 and 6) In the self-calibration, the correction value is updated for each carrier for each symbol, and a new signal point frame is generated in the next reception signal based on this correction value. However, at this time, if the correction is performed using only the correction value obtained from the previously received signal, there are many variations, and there is a possibility that the reverse correction may be performed in a place where the S / N is bad. Therefore, the correction value is determined by averaging in the time axis direction and the frequency axis direction. In the time axis direction, correction values updated for each symbol for each carrier are held for the number of averages, and a new signal point frame is generated from the average of those correction values. After the next symbol, averaging is performed while replacing the oldest correction value with the newly obtained correction value. In the frequency direction, a new signal point frame is generated by performing averaging using the correction values of adjacent carriers centering on the carrier to be determined.

FIG. 5 shows the average state. Each circle shows the correction value by self-calibration, time axis,
In the case of n carriers when averaging is not performed on the frequency axis, only the correction value 18 is used to generate the signal point frame.
When averaging the average number m on the time axis and the average number t on the frequency axis to determine the correction value for n carriers, the correction value is determined by the average of the black circle m * t correction values shown in the figure. , Generate a signal point frame.

By performing this processing, it is possible to derive a highly reliable correction value with little variation. Also,
Regarding the average number, if the average total number m * t is fixed and the average number m on the time axis is increased and the average number t on the frequency axis is decreased, the past correction values are used a lot, which weakens the fluctuation speed. However, when an error occurs, the average number t on the frequency axis
Since it is small, it is easy to recover by transmitting the standard reference signal, and if the average number m on the time axis is decreased and the average number t on the frequency axis is increased, the influence of the past correction value is small and it becomes stronger against the fluctuation speed. When an error occurs, it takes time to recover because the average number t on the frequency axis is large.

Embodiment (corresponding to claims 7, 8 and 9) FIG. 7 is a block diagram of another embodiment of the OFDM demodulator of the present invention. FIG. 7 shows the memory 6 and the amplitude / phase difference detection circuit 7 of FIG.
The signal deleting circuit 21 is inserted between and.
The configuration of a self-calibrator (self-calibration means) 4 different from the configuration of FIG. 1 will be described below.

In the self-calibrator 4, the amplitude / phase correction circuit 5 fetches the information of the amplitude / phase difference of the sequentially received information signals from the memory 6, and based on the value, the first
The amplitude and phase of the secondary correction signal 3 are corrected, and the secondary correction signal 8 output as a result is output to the amplitude and phase difference detection circuit 7
And send it to the decoding circuit 9.

In the amplitude / phase difference detection circuit 7, the secondary correction signal
8 detects the amplitude phase difference between the decoded signal and the low reliability signal deletion circuit 21 deletes the amplitude phase difference information derived from the low reliability signal from the amplitude phase difference information, and stores Send to 6. The memory 6 retains the information and resets the memory 6 for the carrier to which the reference signal is transmitted in response to a request from the reference signal calibrator 2.

The secondary correction signal 8 corrected by the self-calibrator 4 is supplied to the decoding circuit 9 and QA
After M decoding, it is output as decoded data and returned to the amplitude / phase difference detection circuit 7 and the low reliability signal deletion circuit 21 of the self-calibrator 4.

In the self-calibration, the correction value is updated for each carrier for each symbol, and a new signal point frame is generated based on this correction value in the next reception signal. At this time, it is obtained by the previous reception signal. In order to increase the reliability, the correction values of adjacent carriers centering on the carrier to be determined are averaged, and the correction values are calculated based on the unreliable signal among these correction values. If there is a signal, the value is deleted and the correction value is determined.

For signals with low reliability, in self-calibration, signals having various amplitudes and phases are used in order to utilize information data for signal correction. For this reason,
When a signal having a small level is sent, it is more disadvantageous in terms of noise than a signal having a large amplitude level when normalizing the amplitude / phase difference detection. Therefore, when the decoding circuit 9 considers that there are four signals 22 having the smallest level as shown in FIG. 5 in the 256QAM signal point constellation, the amplitude calculated based on the signals,
Delete the phase difference component.

[0036]

As described above, according to the present invention, by performing the self-calibration in addition to the reference signal calibration, the correction value is updated for each transmission signal, and the correct signal can be sequentially corrected. In the calibration, the average on the frequency axis and the time axis,
More accurate signal correction can be performed by performing high-order prediction using a slope or regression model.

Therefore, since it is possible to perform the correction by following the transmission line characteristic which changes in a short time, it is possible to reduce the code error and improve the reliability in communication. Further, as a result, mobile reception can be made possible in a high-multilevel modulation method, so that when the amount of transmission information is constant, the symbol time can be lengthened and the transmission frequency band can be narrowed.

Further, in self-calibration, by deleting the amplitude / phase difference information introduced by the low reliability signal, more accurate signal correction can be performed.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an embodiment of an OFDM wave demodulation device according to the present invention.

FIG. 2 is a configuration diagram of a conventional OFDM wave modulator.

FIG. 3 is a block diagram of a conventional OFDM wave demodulation device.

FIG. 4 is a conceptual diagram of self-calibration according to the present invention.

FIG. 5 is a diagram showing a correction value averaging method in self-calibration according to the present invention.

FIG. 6 is a diagram showing a correction value prediction method in self-calibration according to the present invention.

FIG. 7 is a configuration diagram of another embodiment of an OFDM wave demodulator according to the present invention.

FIG. 8 is a diagram showing a 256QAM signal point constellation diagram.

[Explanation of symbols]

1 FFT calculator 2 Reference signal calibrator (reference signal calibration means) 3 Primary correction signal 4 Self calibrator (self-calibration means) 5 Amplitude and phase correction circuit 6 Memory 7 Amplitude and phase difference detection circuit 8 Secondary correction signal 9 Decoding circuit (decoding means) 10 Encoding circuit 11 Adder 12 IFFT operation unit 13 Quadrature modulator 14 Reference signal calibrator 15 Signal point frame generated by reference signal 16 Received signal 17 Signal point generated by self-calibration Frame 18 Correction value in n carriers 19 Curve approximated by Newton's interpolation polynomial 20 Linear regression model 21 Unreliable signal elimination circuit 22 Signal points with minimum amplitude level in 256QAM signal constellation m Average in time direction t Average in frequency direction Number Pc, Pk correction value

Continuation of front page (56) References JP-A-8-274749 (JP, A) JP-A-8-265293 (JP, A) JP-A-10-290208 (JP, A) JP-A-7-273741 (JP , A) JP 7-321766 (JP, A) JP 5-219021 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04J 11/00

Claims (9)

(57) [Claims] 1. In an OFDM system in which a standard reference signal is added to a specific carrier for transmission, a signal generated and transmitted on a transmission side is received, and the received signal is corrected based on the standard reference signal. Reference signal calibrating means for obtaining a primary correction signal, and amplitude and phase difference between sequentially received information signals are detected to obtain amplitude / phase difference information, the result is stored in a memory, and the signal is read from the memory. For each time, based on the amplitude phase difference information, the correction value obtained in the past on the time axis is used to predict the correction value for the current signal, and the correction value obtained by this prediction is used to perform sequential signal correction. A demodulation device for code transmission, comprising: a self-calibration means for performing a secondary correction signal to be executed; and a decoding means for decoding the secondary correction signal to obtain decoded data. 2. In an OFDM system in which a standard reference signal is added to a specific carrier for transmission, the signal generated and transmitted on the transmission side is received, and the received signal is signal-corrected based on the reference reference signal. Reference signal calibrating means for obtaining a primary correction signal, and amplitude and phase difference between sequentially received information signals are detected to obtain amplitude / phase difference information, the result is stored in a memory, and the signal is read from the memory. Self-calibration means for sequentially performing signal correction using a correction value obtained by averaging in the time axis direction and the frequency axis direction based on the amplitude / phase difference information to obtain a secondary correction signal for each time; A demodulator in code transmission, comprising a decoding means for decoding a secondary correction signal to obtain decoded data. 3. The demodulator for code transmission according to claim 1, wherein when determining a correction value, a slope component is calculated based on a correction value calculated in the past on a time axis, and the correction value is calculated in consideration of the slope. A demodulator in code transmission, which is characterized in that 4. The demodulator for code transmission according to claim 1, wherein, when determining a correction value, a regression model by a least squares method is obtained based on a correction value obtained in the past in terms of time, and the regression model is obtained. A demodulation device in code transmission, characterized in that a correction value is derived by using. 5. In an OFDM system in which a standard reference signal is added to a specific carrier for transmission, a signal generated and transmitted on a transmission side is received, and the received signal is signal-corrected based on the standard reference signal. Performing a primary correction signal, detecting the amplitude and phase difference of the sequentially received information signals, obtaining the amplitude and phase difference information, accumulating the results, reading and receiving each signal,
A correction value obtained in the past on the time axis based on the amplitude / phase difference information is used to predict a correction value for the current signal, and the correction value obtained by the prediction is used to sequentially perform signal correction. A demodulation method in code transmission, characterized in that a secondary correction signal is obtained, the secondary correction signal is decoded, and decoded data is obtained. 6. In an OFDM system in which a standard reference signal is added to a specific carrier for transmission, a signal generated and transmitted on a transmission side is received, and the received signal is corrected based on the reference reference signal. Performing a primary correction signal, detecting the amplitude and phase difference of the sequentially received information signals, obtaining the amplitude and phase difference information, accumulating the results, reading and receiving each signal,
Based on the amplitude / phase difference information, the correction values obtained by averaging in the time axis direction and the frequency axis direction are used to sequentially perform a signal correction to obtain a secondary correction signal, and the secondary correction signal is decoded, A demodulation method in code transmission, characterized in that decoded data is obtained. 7. In an OFDM system in which a standard reference signal is added to a specific carrier for transmission, the received signal is used as a reference in order to receive the signal generated and transmitted at the transmission side and demodulate the transmission information. A standard signal calibration means for performing a signal correction based on a reference signal to obtain a first-order correction signal, an amplitude and phase difference between sequentially received information signals are detected to obtain amplitude / phase difference information, and a low reliability signal is deleted. Then, the result is stored, read as needed, and for each received signal, a self-calibration means for sequentially performing signal correction using the correction value obtained based on the amplitude and phase difference information to obtain a secondary correction signal. A demodulation device for code transmission, comprising: a decoding means for decoding the secondary correction signal output from the self-calibration means to obtain decoded data. 8. The demodulator for code transmission according to claim 7, wherein the self-calibration means has a small amplitude level in a signal point arrangement in multilevel modulation among the detected amplitude / phase difference information. A demodulator in code transmission, wherein the amplitude / phase difference information calculated from a signal is deleted. 9. In an OFDM system in which a standard reference signal is added to a specific carrier for transmission, the signal generated and transmitted on the transmission side is received, and the received signal is signal-corrected based on the reference reference signal. The primary correction signal is obtained, the amplitude and phase difference of the sequentially received information signals are detected, the amplitude and phase difference information is obtained, the low reliability signal is deleted, the result is accumulated, and each signal is read and received. In addition, using the correction value obtained in the past on the time axis based on the amplitude phase difference information, the correction value in the current signal is predicted, and the signal correction is sequentially performed using the correction value obtained by the prediction. A demodulation method in code transmission, wherein a secondary correction signal is obtained, the secondary correction signal is decoded, and decoded data is obtained.