JP2012221513A - Encoding modulation method and demodulation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve high error correction capability by enabling error correction utilizing redundancy and furthermore by properly removing chromatic noise in an encoding modulation system used in an optical disc and digital communications, and in a demodulation system.SOLUTION: An encoding modulation method comprises: an error correction encoding step which performs error correction encoding of digital information and outputs an error correction code word; and a modulation step which demodulates the error correction code word and outputs a modulation code word. The number of symbols of the error correction code word included in one modulation code word is the same as that of one error correction code word outputted in the error correction encoding step.

Description

  The present invention relates to a modulation method and a demodulation method technology for encoding and modulating digital information used in a recording medium such as an optical disk and a digital communication channel such as digital broadcasting and wireless LAN.

  For the transmission of digital information, a coded modulation method combining an error correction code and modulation is used. For example, when digital information is recorded and reproduced using an optical disc such as a BD as a medium, a Reed-Solomon code is used as an error correction code, and an RLL (Run Length Limited) code is used as a modulation code (for example, Non-Patent Document 1). ).

  FIG. 2 is a diagram showing an outline of a BD coded modulation scheme. In BD error correction coding, digital information is subjected to a predetermined interleaving process in units of bytes, arranged in 216 bytes, and a Reed-Solomon code of (248, 216, 33) to which 32 bytes of parity are added is used. Yes. Modulation encoding is performed in a predetermined order, with a plurality of codewords arranged as a set. RLL17 code is used as the modulation code, 2 bits are 3 channel bit recording pattern, 4 bits are 6 channel bit recording pattern, 6 bits are 9 channel bit recording pattern, 8 bits are bit pattern after error correction encoding. It is converted into a 12-channel bit recording pattern. With the error correction coding as described above, it is possible to correct a large error that occurs due to fingerprints or scratches attached while using the BD. Also, by modulation encoding, the frequency characteristics of the binary recording pattern signal are concentrated in the range of the transfer characteristics of the optical disk determined by the laser wavelength and the numerical aperture of the lens.

  FIG. 7 is a diagram showing a configuration of a portion related to a processing system for recording digital information and a processing system for reproducing digital information in an optical disc apparatus that records and reproduces digital information to and from an optical disc such as a BD. The recording digital information input for recording on the optical disc 701 is converted into a recording pattern by the error correction encoding 702 using Reed-Solomon code and the modulation encoding 703 using RLL17 code as described above, and then the optical pickup 705 converts the recording digital information into the optical disc. The output of the laser irradiated to the track on 701 is driven by a laser drive circuit 704, and a recording mark corresponding to the recording pattern is formed on the track.

  On the other hand, the recorded digital information is reproduced by utilizing the fact that the amount of reflected light with respect to the laser irradiated to the track from the optical pickup 705 changes corresponding to the recording mark formed on the track. The optical pickup 705 detects a reproduction signal according to the amount of reflected light, and the HPF and LPF 706 reduce noise components other than the frequency band of the RLL17 code included in the reproduction signal. At a reference linear velocity where the frequency of one channel bit is 66 MHz, the cutoff frequency of HPF is 10 kHz, and the cutoff frequency of LPF is 33 MHz, which is the Nyquist frequency of the one channel bit frequency. The output signals of the HPF and LPF 706 are converted into a digital signal sampled at an interval of 1 channel bit in the synchronization processing circuit 707. A PRML (Partial Response Maximum Likelihood) circuit 707 binarizes the digital signal. PRML is a technique that combines partial response (PR) and maximum likelihood decoding (ML), and is a signal processing that selects the most probable signal sequence from the waveform of a digital signal on the assumption that known intersymbol interference occurs. It is a method. Specifically, the synchronized digital signal is subjected to partial response equalization so as to have a predetermined frequency characteristic using an FIR filter or the like, and then the most probable state transition sequence is selected using Viterbi decoding or the like. Thus, it is converted into a corresponding binary signal. The demodulating circuit 709 demodulates the binary signal into a bit string before modulation and coding in accordance with the RLL17 code. The error correction decoding circuit 710 performs error correction processing according to the Reed-Solomon code after rearranging the demodulated bit string in a predetermined procedure. Through the above processing, the original digital information recorded can be reproduced.

  In digital communications such as digital broadcasting and wireless LAN, a Reed-Solomon code or LDPC (Low Density Parity Check) code is used as an error correction code, and an OFDM (Orthogonal Frequency Division Multiplexing) method is used as a modulation method (for example, non-decoding). Patent Document 2).

  FIG. 4 is a diagram showing an outline of a coded modulation system for digital communication. In the error correction coding of digital communication, as in the case of error correction coding of an optical disc, digital information uses a Reed-Solomon code or an LDPC code that is subjected to a predetermined interleaving process and a parity is added in a predetermined unit. . In the DVB-T2 standard, which is one of the digital broadcasting standards, an LDPC code having a code length of 64 k bits or 16 k bits and a variable coding rate from a minimum of 1/2 to a maximum of 5/6 is used. This codeword is divided into a plurality of pieces, and subcarrier modulation is performed. Subcarrier modulation methods include QPSK, 16-QAM, 64-QAM, 256-QAM, etc., which are selected and used. Further, using the inverse discrete Fourier transform, modulation is performed by an OFDM method in which a plurality of subcarrier modulation results are respectively allocated and multiplexed to orthogonal frequencies within a predetermined frequency band, and the obtained signal is transmitted from the antenna. Sent. In the DVB-T2 standard, the size of the inverse discrete Fourier transform is selected from 1 k samples to 32 k samples. With the coded modulation system as described above, an error caused by multipath propagation is less likely to occur, and the error can be strongly corrected, thereby realizing high-speed data transmission.

  FIG. 8 is a diagram illustrating a configuration of a transmission device and a reception device of a digital communication system using the OFDM scheme. As described above, digital information to be transmitted is first subjected to error correction coding 801 using an LDPC code or the like. This codeword is divided into a plurality of subcarriers, and subcarrier modulation 802 is performed for each divided subcarrier. For subcarrier modulation 802, QPSK for transmitting 2 bits / symbol per subcarrier, 16-QAM for transmitting 4 bits / symbol, or the like is used. A plurality of subcarrier signals are arranged and multiplexed at a predetermined frequency interval to generate a multicarrier signal, and the multicarrier signal expressed on the frequency axis is converted into an OFDM signal on the time axis by inverse discrete Fourier transform 803. . Further, a redundant signal called a guard interval (GI) is inserted by the GI addition circuit 804 and transmitted from the transmission antenna 805 in order to avoid the influence of interference of multipath delayed waves generated by the communication path.

  In the receiver, a signal transmitted from the reception antenna 806 is received, and synchronization processing 807 such as symbol timing synchronization and carrier frequency synchronization is performed, and the guard interval inserted by the GI removal circuit 808 is removed. The received OFDM signal from which the guard interval has been removed is converted into a received multicarrier signal expressed on the frequency axis by discrete Fourier transform 809. The subcarrier demodulation circuit 810 divides the received multicarrier signal into subcarrier signals, and performs demodulation in accordance with a modulation scheme such as PSK, 16-QAM, and 64-QAM for each subcarrier. The error correction decoding circuit 811 arranges the demodulated bit strings in a predetermined order and performs error correction processing corresponding to error correction encoding such as LDPC code. Through the above processing, digital information can be transmitted and received.

Blu-ray Disc Association “White Paper Blu-ray Disc Format 1. A Physical Format Specification for BD-RE” DVB (Digital Video Broadcasting Project) BlueBook “Frame structure channel coding and modulation for a second generation digital terrestrial digital terrestrial terrestrial digital terrestrial

  In the encoding modulation method used in the optical disk and digital communication as described above, the error correction encoding increases the error correction codeword in order to increase the error correction capability, The purpose of modulation coding is to match the frequency characteristics of the signal to be transmitted with the transmission characteristics of the transmission path of the optical disc or digital communication. In order not to be biased, the code length of the modulation codeword processed at one time in the modulation coding is shortened so that the bias can be controlled to be small. Therefore, the code length of the modulation codeword is considerably shorter than the code length of the error correction codeword. For this reason, in modulation and coding, only the frequency characteristic of a signal to be transmitted is converted, and an error generated due to noise on the transmission path is low in correction capability.

  FIG. 3 is a diagram showing a state of a recording signal and a reproduction signal when digital information is recorded and reproduced using an optical disc as a transmission path. A binary recording pattern signal modulated and encoded by the RLL17 code is recorded on the track of the optical disc. The binary recording pattern includes a low frequency component of about several hundred kHz in a long section. On the other hand, not only white noise but also various colored noises are superimposed on the reproduction signal detected from the track of the optical disk. The recording surface of the BD is multi-layered up to four layers, and colored noise of 100 kHz or more is generated in the reproduction signal due to interference of reflected light from the recording surface other than the recording surface accessed at that time. There is. Since the frequency band of the colored noise is close to the frequency component of the binary recording pattern, the colored noise component cannot be removed by the conventional HPF and LPF 706, PRML circuit 708, and demodulating circuit 709, and an error occurs in the demodulated bit string. It will leave a lot. In addition, as a method for whitening the colored noise in order to reduce the influence of the colored noise, there is an NPML (Noise Predictive Maximum Likelihood) method. The number of taps of the FIR filter that whitens the colored noise as a constituent element is the same. Because of its short length, low-frequency colored noise cannot be sufficiently reduced, and furthermore, it copes with colored noise that is partially generated due to low response of tap coefficient control of the FIR filter that determines whitening characteristics. I can't do that either.

  Optical disks are also increasing in linear density, and intersymbol interference and SNR degradation are factors that generate many errors, and all these errors and errors caused by scratches and fingerprints are corrected by error correction decoding 710. This makes it difficult to reproduce digital information.

  FIG. 5 is a diagram illustrating states of a transmission signal and a reception signal of digital communication using the OFDM method. In digital communication, similar to an optical disc, various colored noises are superimposed on a received signal. Since the frequency of colored noise is different from the frequency orthogonal to each other where subcarriers are multiplexed, not only the frequency band of the main lobe of the frequency spectrum but also the wide frequency band up to the side lobe is obtained by the discrete Fourier transform. This affects the detection result and increases errors. Even in the case of colored noise having a period longer than the time length of the OFDM signal generated by one inverse discrete Fourier transform, the detection result is affected over a wide frequency band.

  Although the LDPC code has a very high error correction capability, in order to completely correct the increased error, the transmission speed of digital information can be increased by increasing the number of repeated decoding or lowering the coding rate. There is a problem that the remarkably decreases.

  In order to solve the above-mentioned problem, an encoding modulation method of the present invention is an encoding modulation method of digital information, and an error correction encoding step of outputting an error correction codeword by performing error correction encoding on the digital information; A modulation step of modulating the error correction codeword and outputting a modulation codeword, wherein the number of symbols of the error correction codeword included in one modulation codeword is output in the error correction encoding step It is the same as the number of symbols of one error correction codeword.

  Further, the longest period of the modulation signal in which the modulation codewords are arranged in a predetermined order may be equal to or shorter than the length of one modulation codeword.

  In addition, the modulation codeword may be selected in the modulation step so that a direct current component in a predetermined section of the modulation signal in which the modulation codewords are arranged in a predetermined order becomes zero.

  Further, in the transmission path for outputting the modulation signal in which the modulation codewords are arranged in a predetermined order, colored noise having an arbitrary frequency is superimposed on the modulation signal, and the length of one modulation codeword is You may make it longer than the period of colored noise.

  The demodulation method of the present invention is a demodulation method for demodulating a modulated signal that has been modulated, a demodulation step for demodulating the modulation signal, an expected value generation step for generating an expected value signal from the demodulation result, and the modulation A difference detection step for detecting a difference between the signal and the expected value signal, a spectrum calculation step for calculating a spectrum of the difference signal detected in the difference detection step, and an error signal for calculating an error signal from a predetermined section component of the spectrum A calculating step; and a removing step of removing the error signal from the modulated signal, wherein the demodulating step re-demodulates the modulated signal from which the error signal has been removed by the removing step. To do.

  In the error signal calculation step, the processing may be repeatedly performed while updating a predetermined section for calculating the error signal.

  In the spectrum calculation step, the spectrum may be calculated by discrete Fourier transform.

  In addition, the time interval for performing the discrete Fourier transform may be controlled to a length according to the frequency band in which the error signal is calculated in the error signal calculation step.

  The modulation signal is error-correction encoded before modulation, and further includes a decoding step for decoding the output of the demodulation step corresponding to the error correction encoding, and the expected value generation step includes The expected value signal may be generated from the decoding result of the decoding step.

  The maximum length of the time interval for obtaining the spectrum in the spectrum calculating step may be an interval including the number of symbols of one error correction codeword by the error correction encoding.

  The error correction coding is an error correction code that is decoded by iterative decoding. In the decoding step, iterative decoding is performed, and the modulation signal from which the error signal has been removed is demodulated together with one decoding. May be.

  An encoding modulation apparatus according to the present invention is an encoding modulation apparatus for digital information, wherein the digital information is error correction encoded and error correction encoding means for outputting an error correction code word, and the error correction code word is modulated. A modulation codeword that outputs a modulation codeword, and the number of symbols of the error correction codeword included in one modulation codeword is equal to one error correction codeword output from the error correction coding means The number of symbols is the same.

  Further, the longest period of the modulation signal in which the modulation codewords are arranged in a predetermined order may be equal to or shorter than the length of one modulation codeword.

  Further, the modulation codeword may be selected by the modulation means so that a direct current component in a predetermined section of the modulation signal in which the modulation codewords are arranged in a predetermined order becomes zero.

  Further, in the transmission path for outputting the modulation signal in which the modulation codewords are arranged in a predetermined order, colored noise having an arbitrary frequency is superimposed on the modulation signal, and the length of one modulation codeword is You may make it longer than the period of colored noise.

  The demodulating device of the present invention is a demodulating device that demodulates a modulated signal that is modulated, a demodulating unit that demodulates the modulated signal, an expected value generating unit that generates an expected value signal from the demodulation result, and the modulation A difference detecting means for detecting a difference between the signal and the expected value signal, a spectrum calculating means for calculating a spectrum of the difference signal detected by the difference detecting means, and an error signal for calculating an error signal from a predetermined section component of the spectrum Comprising calculating means and removing means for removing the error signal from the modulated signal, the demodulating means performs demodulation again on the modulated signal from which the error signal has been removed by the removing means. To do.

  Further, the error signal calculating means may perform the processing repeatedly while updating a predetermined section for calculating the error signal.

  In the spectrum calculation means, the spectrum may be calculated by a discrete Fourier transform.

  In addition, the time interval for performing the discrete Fourier transform may be controlled to a length according to the frequency band in which the error signal is calculated in the error signal calculation step.

  The modulation signal is error correction encoded before modulation, and further includes decoding means for decoding the output of the demodulation means corresponding to the error correction encoding, and the expected value generation means The expected value signal may be generated from the decoding result of the decoding means.

  Further, the maximum length of the time interval for obtaining the spectrum in the spectrum calculating means may be an interval including the number of symbols of one error correction codeword by the error correction encoding.

  The error correction coding is an error correction code that is decoded by iterative decoding. The decoding means performs iterative decoding, and performs demodulation on the modulated signal from which the error signal has been removed together with one decoding. May be.

  According to the present invention, by making the code length of the modulation coding equal to the code length of the error correction coding, it becomes possible to perform error correction utilizing redundancy in the demodulation corresponding to the modulation coding, and the error correction code In addition, high error correction capability can be obtained.

  In addition, by repeating the process of accurately extracting and removing the colored noise based on the frequency spectrum obtained from the time interval in which the colored noise exists, the influence of the colored noise that causes a large error is reduced, A high error correction capability can be obtained together with the error correction code.

It is a figure which shows the encoding modulation system by Embodiment 1. It is a figure which shows the encoding modulation system of the conventional optical disk. It is a figure which shows the state on which the colored noise was superimposed on the reproduction signal of the conventional optical disk. It is a figure which shows the encoding modulation system of the conventional digital communication. It is a figure which shows the state by which colored noise was superimposed on the received signal of the conventional digital communication. It is a figure which shows the state by which colored noise was superimposed on the received signal by Embodiment 1. It is a figure which shows the structure of the conventional optical disk apparatus. It is a figure which shows the structure of the conventional digital communication system. It is a figure which shows the flow of the colored noise removal process by Embodiment 2. FIG. FIG. 5 is a diagram illustrating a configuration of an optical disc device according to a second embodiment. (A), (b) and (c) is a figure which shows the spectrum calculation result by Embodiment 2. FIG. It is a figure which shows the update process of the colored noise extraction frequency band by Embodiment 2. It is a figure which shows the discrete Fourier-transform area by Embodiment 2. It is a figure which shows the structure of the digital radio | wireless communications system by Embodiment 3.

  Hereinafter, embodiments of an encoding modulation method and a demodulation method according to the present invention will be described.

(Embodiment 1)
FIG. 1 is a diagram showing a coded modulation scheme in Embodiment 1 of the present invention.

  In digital communication channels such as digital broadcasting and wireless LAN, digital information to be recorded or transmitted is arranged with a predetermined interleave process in units of bytes or bits. The arranged predetermined digital information is divided for every M bytes, subjected to error correction coding using an LDPC code, and N-byte parity is added. Therefore, the code length of one error correction codeword generated by one error correction encoding process is M + N bytes.

  One error correction codeword is divided into K pieces, each assigned to K subcarriers, and subjected to subcarrier modulation. QAM is used for subcarrier modulation. K subcarrier signals are arranged and multiplexed at predetermined frequency intervals within a predetermined frequency band, and one multicarrier signal expressed in the frequency domain is generated.

  In order to convert a multicarrier signal expressed in the frequency domain into an OFDM signal in the time domain, inverse discrete Fourier transform processing is performed. One OFDM signal generated by one inverse discrete Fourier transform process is one modulation codeword, and its length is the code length of the modulation codeword.

  As described above, the M-byte digital information is converted into one error correction codeword and further converted into one OFDM signal, that is, a modulation codeword. The code length of the error correction codeword and the codelength of the modulation codeword are It has the same section.

  As many OFDM signals as necessary for transmitting all digital information are sequentially generated and arranged in order on the time axis. Between each OFDM signal, for example, a redundant signal having a predetermined length called a guard interval is inserted in order to avoid the influence of interference between adjacent OFDM signals due to multipath delay occurring in the digital communication path. Output signal.

  FIG. 6 is a diagram illustrating a state of a transmission signal that is coded and modulated as described above and transmitted to the transmission line, and a reception signal that is received through the transmission line. As described above, one modulation code word corresponds to an OFDM signal in one section, and the OFDM signals for the transmission of all digital information are arranged in order to become a transmission signal.

  While being transmitted on the transmission path, various colored noises are superimposed on the transmission signal to become a reception signal. Since the frequency of the colored noise is different from the orthogonal frequency at which the subcarrier signals are multiplexed, the discrete noise is applied not only to the main lobe frequency band of the colored noise spectrum but also to a wide frequency band up to the side lobe. This affects the detection result of the received signal by conversion and increases errors. Even in the case of colored noise having a period longer than the code length of the modulation codeword, the detection result is affected over a wide frequency band. For example, if the code length of the modulation codeword is short with respect to the period of the colored noise as in the conventional digital communication shown in FIG. 5 described above, the frequency of the colored noise cannot be accurately specified, and the frequency bandwidth of the spectrum Will spread. Also, in the method of extracting the colored noise component included in the received signal using the error correction decoding result as in the second and third embodiments described later, many errors remain in the error correction decoding result. In such a situation, it is difficult to accurately extract the colored noise component, so that the number of repetitions necessary to reduce the colored noise component increases.

  However, as shown in FIG. 6, the code length of the modulation codeword is set to be equal to the code length of the error correction codeword that is long enough to cope with various error factors. The noise frequency spectrum can be accurately identified and separated without spreading. By combining the code length of the error correction codeword and the code length of the modulation codeword, an ideal received signal is predicted using the result of decoding the error correction code, and the frequency spectrum of the difference signal from the actual received signal is used. Colored noise can be accurately separated. For example, as shown in Embodiments 2 and 3 to be described later, the error correction capability for the colored noise component can be further enhanced by a combination of iterative decoding of the LDPC code and the removal processing of the colored noise component. In addition, the redundancy obtained by increasing the code length of the modulation codeword can be determined by, for example, providing a non-arranged band in the frequency interval at which the subcarrier signal is multiplexed or using a convolutional code between the subcarrier signals. By using this for providing the correlation, it is possible to further improve the effect of reducing errors caused by the influence of colored noise.

  As described above, appropriate error correction capability at each step so that errors due to the effects of colored noise can be reduced by modulation and demodulation, and errors due to the effects of other white noise and burst errors due to missing received signals can be corrected by error correction codes. Thus, it is possible to prevent a decrease in transmission speed of digital communication.

  In the first embodiment described above, the example corresponding to the modulation / demodulation of the OFDM method used in digital communication is shown, but the present invention is not limited to this. The same effect can be obtained even in the case of modulation / demodulation by the RLL code used in a recording medium such as an optical disk. The length of the bit unit for converting the bit pattern after error correction coding into the recording pattern is the code length of the modulation codeword, and the same value continues by taking this as long as the code length of the error correction codeword. In addition to the limitation on the run length, redundancy can be increased. This redundancy can be utilized when demodulating a reproduction signal from an optical disk into a bit pattern of an error correction codeword by a PRML circuit and a demodulation circuit. For colored noise that cannot be removed by the conventional HPF and LPF, the colored noise component is separated and removed by detecting the frequency component that does not exist within the range of one modulation codeword. Errors that have occurred can be reduced before error correction decoding. As shown in Embodiments 2 and 3 to be described later, an ideal reproduction signal is predicted using the result of decoding the error correction code by combining the code length of the error correction codeword and the code length of the modulation codeword. Thus, it is possible to accurately separate the colored noise from the frequency spectrum of the difference signal from the actual reproduction signal.

  As a result, the effects of colored noise and intersymbol interference are reduced by the PRML circuit and the demodulation circuit, and each error is corrected appropriately so that scratches and fingerprint errors, which are major problems when using optical disks, are dealt with by error correction decoding. It has the ability to read digital information stably.

(Embodiment 2)
FIG. 10 is a diagram showing a configuration of a part related to a processing system for recording digital information on the optical disc 1001 and a processing system for reproducing digital information from the optical disc 1001 in the optical disc apparatus according to Embodiment 2 of the present invention.

  The recording digital information input for recording on the optical disc 1001 is error correction encoded by the LDPC encoding circuit 1002 and recorded by the modulation method shown in the first embodiment or the modulation encoding circuit 1003 using the conventional RLL17 code. After being modulated into a pattern, the output of the laser applied to the track on the optical disc 1001 is driven by the laser drive circuit 1004 by the optical pickup 705, and a recording mark corresponding to the recording pattern is formed on the track of the optical disc 1001. Information is recorded.

  On the other hand, the recorded digital information is reproduced by utilizing the fact that the amount of reflected light with respect to the laser irradiated onto the track from the optical pickup 1005 changes corresponding to the recording mark formed on the track. The optical pickup 1005 detects a reproduction signal corresponding to the amount of reflected light, and the HPF and LPF 1006 reduce noise components in low and high frequencies other than the frequency band of the modulation code included in the reproduction signal. Output signals of the HPF and LPF 1006 are converted into a digital signal sampled at an interval of 1 channel bit in the synchronization processing circuit 1007. The noise removal circuit 1008 outputs the digital signal to the PRML circuit 1009 after subtracting and removing the colored noise signal detected by the colored noise detection circuit 1017 from the digital signal. Further, the digital signal after removing the colored noise signal is buffered. The PRML circuit 1009 binarizes the digital signal. The PRML circuit 1009 includes an adaptive equalization filter circuit that PR equalizes the digital signal by an FIR filter whose tap coefficients are adaptively controlled so that the frequency characteristic of the digital signal becomes a target frequency characteristic, and PR equalization. The digital signal is composed of a Viterbi decoding circuit which converts the most probable state transition sequence into a corresponding binary signal. The demodulating circuit 1010 demodulates the binary signal into a bit string before modulation encoding according to the modulation code. The LDPC decoding circuit 1011 performs error correction on the demodulated bit string by iterative decoding processing and outputs reproduced digital information.

  Hereinafter, the colored noise detection circuit 1017 that performs the process of reducing the colored noise in the optical disc apparatus will be described in detail.

  The colored noise detection circuit 1017 includes a remodulation coding circuit 1012, an expected value generation circuit 1013, an error detection circuit 1014, a spectrum calculation circuit 1015, and a noise extraction circuit 1016.

  Each time the iterative decoding process is performed in the LDPC decoding circuit 1011, the decoding result is extracted, the re-modulation encoding circuit 1012 performs modulation encoding similar to the recording process, and outputs an expected value recording pattern. The expected value generation circuit 1013 generates an expected value digital signal having frequency characteristics targeted by the PRML circuit 1009 based on the expected value recording pattern. The error detection circuit 1014 buffers the digital signal PR-equalized by the PRML circuit 1019 in advance, and calculates an error between the generated expected value digital signal and the PR-equalized digital signal. As a method for generating an expected value digital signal, a method for generating an expected value recording pattern by performing discrete Fourier transform, performing multiplication with a target frequency characteristic of PR equalization in the frequency domain, and then performing inverse discrete Fourier transform There is. When the target frequency characteristic of PR equalization is, for example, PR (1, 2, 2, 2, 1) characteristic, the ratio of tap coefficients to the bit string of the expected value recording pattern is 1: 2: 2: It can also be generated by passing through a 2: 1 FIR filter.

  The spectrum calculation circuit 1015 calculates the frequency spectrum of the error signal calculated by the error detection circuit 1014, and calculates the frequency spectrum of the error signal before equalization based on the equalization characteristics of the FIR filter of the PRML circuit 1009. . FIG. 11 is a diagram showing processing by the spectrum calculation circuit 1015. FIG. 11A shows the amplitude characteristics of the frequency spectrum obtained by subjecting the error signal calculated by the error detection circuit 1014 to discrete Fourier transform. It can be seen that a particularly high colored noise component remains in the vicinity of 100 kHz to 300 kHz without being removed. FIG. 11B shows the gain characteristic of the equalization characteristic of the FIR filter of the PRML circuit 1009. The equalization characteristic is calculated from the tap coefficient of the FIR filter. The frequency of one channel bit is 66 MHz, and the modulation code is RLL17. Since the shortest run length is 2 channel bits, the gain increases so as to equalize to the target frequency characteristic up to a frequency of about 16.5 MHz corresponding to a 4 channel bit period in which a recording mark and a space of 2 channel bits are continuous. I understand that FIG. 11C shows the amplitude characteristics of the frequency spectrum of the error signal before equalization. This is calculated by complex-dividing the frequency spectrum of the error signal shown in FIG. 11 (a) by the equalization characteristic spectrum (gain characteristic and phase characteristic) shown in FIG. 11 (b). In the frequency band of 16.5 MHz or higher, the gain of the equalization characteristic spectrum in FIG. 11B is zero, and thus cannot be obtained accurately. However, in the frequency band of 16.5 MHz or lower, equalization is performed by the PRML circuit 1009. The frequency spectrum of the noise component contained in the previous digital signal is represented.

  The noise extraction circuit 1016 leaves only a predetermined frequency band for extracting the error component from the frequency spectrum of the error signal before equalization calculated by the spectrum calculation circuit 1015, and sets the values of the spectrums in the other frequency bands to zero. Later, inverse discrete Fourier transform is performed to extract a colored noise signal to be removed. The extracted colored noise signal is removed from the digital signal by the noise removal circuit 1008 as described above.

  By repeating the above-described colored noise removal processing in the same manner while updating a predetermined frequency band for extracting a colored noise signal in the noise extraction circuit 1016 for each iteration of the iterative decoding processing in the LDPC decoding circuit 1011. The colored noise components are sequentially reduced, and accordingly, the equalization characteristic in the PRML circuit 1009 is controlled to a more appropriate characteristic, the binarization error in the Viterbi decoding circuit included in the PRML circuit 1009 and the demodulation in the demodulation circuit 1010 Errors are reduced.

  FIG. 12 shows a process of updating a predetermined frequency band for extracting a colored noise signal in the noise extraction circuit 1016. The frequency spectrum in FIG. 13 is a low-frequency part of the frequency spectrum of the error signal before equalization calculated by the spectrum calculation circuit 1015, and the colored noise removal process has not yet been performed. The frequency band for extracting the colored noise is moved in order from the low frequency side to the high frequency side. In the original state, as seen in the frequency spectrum of FIG. 13, there is high colored noise in the vicinity of 100 kHz to 300 kHz, but this colored noise is sufficiently removed when the first and second processing is performed, It cannot be seen in the frequency spectrum at that time. In the case of this colored noise, a large error reduction effect can be obtained by the first and second removal processing in the demodulation results of the PRML circuit 1009 and the demodulation circuit 1010. Further, the frequency band from which the colored noise is extracted may always be a predetermined frequency bandwidth, or may be changed to a fine frequency bandwidth particularly for a frequency band where a large amount of colored noise is likely to occur. .

  FIG. 13 shows a discrete Fourier transform interval for the frequency calculation circuit 1015 to calculate the frequency spectrum of the error signal. If the length of the section where the discrete Fourier transform is performed is always constant, it is possible to remove the colored noise that is always generated without any problem. Cannot be extracted accurately. A component having a low frequency has a smaller number of waves included in the time interval TL in which discrete Fourier transform is performed, so that the amplitude and phase of the component can be accurately detected. However, a component having a high frequency is included in the time interval TL. This is because the number of detected waves increases, the detected amplitude and phase become average values, and a noise component that partially changes cannot be accurately detected. Therefore, when the frequency band of the colored noise extracted by the noise extraction circuit 1016 is lower than the time interval TL in which the error signal is detected by the error detection circuit 1014, the discrete Fourier analysis performed by the spectrum calculation circuit 1015 as in the Fourier transform interval L is performed. A discrete Fourier transform performed by the Fourier transform section M, the Fourier transform section H, and the sequential spectrum calculation circuit 1015 in accordance with the fact that the length of the transform section is long and the frequency band of the colored noise extracted by the noise extraction circuit 1016 is updated high. If the time interval TL is shortened and the time interval TL is divided into a plurality of times to detect the colored noise, the partially generated colored noise can be accurately removed.

  The procedure for removing colored noise by iterative processing will be organized with reference to FIG. First, demodulation processing of a modulation code is performed by the PRML circuit 1009 and the demodulation circuit 1010 in a state where colored noise is not removed from the digital signal. An error correction code decoding process is performed on the demodulated bit string. As a result, if no error remains, the digital information reproduction process is completed. If it remains, an error signal is detected based on the reproduction result at this stage. The spectrum calculation circuit 1015 calculates the frequency spectrum of the error signal. A colored noise signal is extracted from the spectral component of the frequency band section N. A signal from which the colored noise signal is removed from the first digital signal is calculated, and this is converted into a digital signal, the second demodulation process is performed, and the error correction code decoding process is performed. As a result, if no error remains, the digital information reproduction process is completed. If the error remains, the same process is repeated while updating the section N. Here, the original bit string for obtaining the expected value recording pattern is in a state including an error, and the detected error signal is erroneous in the portion where the error remains. If this error signal is directly removed from the digital signal, the digital signal has a signal waveform closer to the error side. However, the error does not continue for a long time, but occurs at most about one channel bit in several tens of channel bits, and since the influence on the frequency spectrum is small, in the frequency domain as in the present embodiment, By limiting the section N in which the colored noise is extracted and repeating the processing, the colored noise can be accurately removed without being affected by the remaining error.

  In the second embodiment, the error correction code is an LDPC code. However, the present invention is not limited to this. It may be a turbo code or a Reed-Solomon code.

(Embodiment 3)
FIG. 14 is a diagram illustrating a configuration of a transmission device and a reception device of a digital wireless communication system using the OFDM scheme.

  The digital information to be transmitted is first subjected to error correction encoding by the LDPC encoding circuit 1401. In subcarrier modulation circuit 1402, the codeword is divided into a plurality of subcarriers, and subcarrier modulation is performed for each of the divided subcarriers. Subcarrier modulation, QPSK that transmits 2 bits / symbol per subcarrier, 16-QAM that transmits 4 bits / symbol, and the like are used. A plurality of subcarrier signals are arranged and multiplexed at a predetermined frequency interval to generate a multicarrier signal, and the inverse discrete Fourier transform circuit 1403 converts the multicarrier signal represented on the frequency axis into an OFDM signal on the time axis. . Further, a guard interval signal is inserted by the GI addition circuit 1404 and transmitted from the transmission antenna 1405 in order to avoid the influence of multipath delay wave interference generated by the communication path.

  In the receiver, a signal transmitted from the receiving antenna 1406 is received, and synchronization processing such as symbol timing synchronization and carrier frequency synchronization is performed by the synchronization processing circuit 1407. The GI removal circuit 1408 removes the guard interval inserted in the received signal. The noise removal circuit 1409 subtracts the colored noise signal detected by the colored noise detection circuit 1418 from the received OFDM signal from which the guard interval is removed, and then outputs the received OFDM signal to the discrete Fourier transform circuit 1410. Further, the received OFDM signal after removing the colored noise signal is buffered. Discrete Fourier transform circuit 1410 converts the received OFDM signal on the time axis into a received multicarrier signal expressed on the frequency axis. The subcarrier demodulation circuit 1411 divides the received multicarrier signal into subcarrier signals, and performs demodulation processing according to a modulation scheme such as PSK, 16-QAM, and 64-QAM for each subcarrier. The LDPC decoding circuit 1412 performs error correction by iterative decoding processing on the demodulated bit string, and outputs the received digital information.

    Hereinafter, the colored noise detection circuit 1418 that performs processing for reducing the colored noise in the receiving apparatus of the digital wireless communication system will be described in detail.

  The colored noise detection circuit 1418 includes a re-subcarrier modulation circuit 1413, a re-inverse discrete Fourier transform circuit 1414, an error detection circuit 1415, a spectrum calculation circuit 1416, and a noise extraction circuit 1417.

  Each time the iterative decoding process is performed in the LDPC decoding circuit 1412, the decoding result is extracted, and the modulation process similar to the transmission process is performed by the re-subcarrier modulation circuit 1413 and the re-inverse discrete Fourier transform circuit 1414. Generate a signal. The error detection circuit 1415 calculates an error between the generated expected value reception OFDM signal and the reception OFDM signal buffered by the noise removal circuit 1409.

  The spectrum calculation circuit 1416 calculates the frequency spectrum of the error signal calculated by the error detection circuit 1415. The noise extraction circuit 1417 leaves only a predetermined frequency band from which an error component is extracted from the frequency spectrum of the error signal calculated by the spectrum calculation circuit 1416, sets the spectrum values of other frequency bands to zero, and then performs inverse discrete processing. A colored noise signal to be removed is extracted by performing Fourier transform. As described above, the extracted colored noise signal is removed from the received OFDM signal by the noise removal circuit 1409.

  By repeating the above-described processing for removing colored noise in the same manner while updating a predetermined frequency band for extracting a colored noise signal in the noise extraction circuit 1417 for each iteration of the decoding processing in the LDPC decoding circuit 1412. The colored noise components are sequentially reduced, and accordingly, the demodulation errors in the discrete Fourier transform circuit 1410 and the subcarrier demodulation circuit 1411 are reduced.

  In the third embodiment described above, the error correction code is an LDPC code, but the present invention is not limited to this. It may be a turbo code or a Reed-Solomon code.

  In Embodiment 3 described above, the modulation scheme is the OFDM modulation scheme, but the present invention is not limited to this.

  In the third embodiment, the digital wireless communication system is used. However, the present invention is not limited to this. The same effect can be obtained even in a wired communication system.

  The components of the optical disc device of the second embodiment or the digital wireless communication system of the third embodiment can be realized as an LSI (Large Scale Integration) that is an integrated circuit. The components included in the optical disk device and the digital wireless communication system may be individually made into one chip, or may be made into one chip so as to include a part or all of them.

  Here, the integrated circuit is referred to as an LSI, but may be referred to as an IC (Integrated Circuit), LSI, super LSI, or ultra LSI depending on the degree of integration.

Further, the integrated circuit of the present embodiment is not limited to an LSI, and may be realized by a dedicated circuit or a general-purpose processor. Also, an FPGA (Field Programmable Gate Array) that can be programmed after manufacturing the LSI, or a reconfigurable processor that can reconfigure the connection and setting of circuit cells inside the LSI may be used.
Furthermore, if integrated circuit technology comes out to replace LSI's as a result of the advancement of semiconductor technology or a derivative other technology, it is naturally also possible to carry out function block integration using this technology. There is a possibility of adaptation of biotechnology.

  INDUSTRIAL APPLICABILITY The present invention is useful in increasing the capacity of an optical disk by increasing the number of layers and increasing the transfer rate of digital communication, and can be used in a method for reproducing a large capacity optical disk, an optical disk device, a digital communication method, a digital communication system, and an integrated circuit.

701, 1001 Optical disk 702 Error correction coding circuit 703, 1003 Modulation coding circuit 704, 1004 Laser drive circuit 705, 1005 Optical pickup 706, 1006 HPF and LPF
707, 1007 Synchronization processing circuit 708, 1009 PRML circuit 709, 1010 Demodulation circuit 710 Error correction decoding circuit 801 Error correction coding circuit 802, 1402 Subcarrier modulation circuit 803, 1403 Inverse discrete Fourier transform circuit 804, 1404 GI addition circuit 805, 1405 Transmitting antenna 806, 1406 Receiving antenna 807, 1407 Synchronization processing circuit 808, 1408 GI removal circuit 809, 1410 Discrete Fourier transform circuit 810, 1411 Subcarrier demodulation circuit 811 Error correction decoding circuit 1002 LDPC encoding circuit 1008 Noise removal circuit 1011 LDPC Decoding circuit 1012 Remodulation coding circuit 1013 Expected value generation circuit 1014 Error detection circuit 1015 Spectrum calculation circuit 1016 Noise extraction circuit 1017 Color noise detection circuit 1401 LDPC encoder 1409 noise removing circuit 1412 LDPC decoding circuit 1413 re subcarrier modulation circuit 1414 re-inverse discrete Fourier transform circuit 1415 error detection circuit 1416 spectrum calculation circuit 1417 noise extraction circuit 1418 colored noise detection circuit

Claims (22)

A method for encoding and modulating digital information, comprising:
An error correction encoding step for error correction encoding the digital information and outputting an error correction codeword;
A modulation step of modulating the error correction codeword and outputting a modulation codeword,
The number of symbols of the error correction codeword included in one modulation codeword is the same as the number of symbols of one error correction codeword output in the error correction coding step Modulation method.   The encoding modulation method according to claim 1, wherein the longest period of a modulation signal in which the modulation codewords are arranged in a predetermined order is equal to or shorter than the length of one modulation codeword.   The modulation codeword is selected in the modulation step so that a direct current component in a predetermined section of a modulation signal in which the modulation codewords are arranged in a predetermined order becomes zero. Coded modulation method.   In the transmission path for outputting the modulation signal in which the modulation codewords are arranged in a predetermined order, colored noise having an arbitrary frequency is superimposed on the modulation signal, and the length of one modulation codeword is the colored noise. The encoded modulation method according to claim 1, wherein the period is longer than the period of the code. A demodulation method for demodulating a modulated signal being modulated,
A demodulation step of demodulating the modulated signal;
An expected value generating step for generating an expected value signal from the demodulation result;
A difference detecting step for detecting a difference between the modulated signal and the expected value signal;
A spectrum calculating step for calculating a spectrum of the difference signal detected in the difference detecting step;
An error signal calculating step of calculating an error signal from a predetermined section component of the spectrum;
A demodulation step of removing the error signal from the modulation signal, wherein the demodulation step re-demodulates the modulation signal from which the error signal has been removed by the removal step.   The demodulation method according to claim 5, wherein the error signal calculation step performs an iterative process while updating a predetermined section for calculating an error signal.   6. The demodulation method according to claim 5, wherein in the spectrum calculation step, the spectrum is calculated by a discrete Fourier transform.   8. The demodulation method according to claim 7, wherein a time interval for performing discrete Fourier transform is controlled to a length according to a frequency band in which the error signal is calculated in the error signal calculation step. The modulation signal is error-correction-encoded before modulation, and further includes a decoding step for decoding the output of the demodulation step corresponding to the error-correction encoding,
The demodulation method according to claim 5, wherein the expected value generation step generates an expected value signal from a decoding result of the decoding step.   The demodulation method according to claim 9, wherein the maximum length of a time interval for obtaining a spectrum in the spectrum calculating step is an interval including the number of symbols of one error correction codeword by the error correction encoding.   The error correction coding is an error correction code that is decoded by iterative decoding. In the decoding step, iterative decoding is performed, and demodulation is performed on the modulated signal from which the error signal has been removed together with one decoding. The demodulation method according to claim 9, wherein: An apparatus for encoding and modulating digital information,
Error correction encoding means for error correction encoding the digital information and outputting an error correction codeword;
A modulation means for modulating the error correction codeword and outputting a modulation codeword;
The number of symbols of the error correction codeword included in one modulation codeword is the same as the number of symbols of one error correction codeword output in the error correction coding means Modulation device.   The coded modulation apparatus according to claim 12, wherein the longest period of a modulation signal in which the modulation codewords are arranged in a predetermined order is equal to or shorter than the length of one modulation codeword.   The modulation codeword is selected by the modulation means so that a direct current component in a predetermined section of a modulation signal in which the modulation codewords are arranged in a predetermined order becomes zero. Coded modulation device.   In the transmission path for outputting the modulation signal in which the modulation codewords are arranged in a predetermined order, colored noise having an arbitrary frequency is superimposed on the modulation signal, and the length of one modulation codeword is the colored noise. The encoded modulation apparatus according to claim 12, wherein the encoding modulation apparatus is longer than the period. A demodulator that demodulates a modulated signal being modulated,
Demodulation means for demodulating the modulated signal;
Expected value generating means for generating an expected value signal from the demodulation result;
Difference detection means for detecting a difference between the modulated signal and the expected value signal;
Spectrum calculation means for calculating the spectrum of the difference signal detected by the difference detection means;
Error signal calculating means for calculating an error signal from a predetermined section component of the spectrum;
The demodulating device comprises a removing unit that removes the error signal from the modulated signal, and the demodulating unit performs demodulation again on the modulated signal from which the error signal has been removed by the removing unit.   The demodulator according to claim 16, wherein the error signal calculation means performs iterative processing while updating a predetermined section for calculating an error signal.   The demodulator according to claim 16, wherein in the spectrum calculation means, the spectrum is calculated by discrete Fourier transform.   19. The demodulator according to claim 18, wherein a time interval for performing discrete Fourier transform is controlled to a length according to a frequency band in which the error signal is calculated in the error signal calculation step. The modulation signal is error correction encoded before modulation, and further includes decoding means for decoding the output of the demodulation means corresponding to the error correction encoding,
17. The demodulator according to claim 16, wherein the expected value generation means generates an expected value signal from the decoding result of the decoding means.   21. The demodulator according to claim 20, wherein the maximum length of a time interval for obtaining a spectrum in the spectrum calculating means is a interval including the number of symbols of one error correction codeword by the error correction encoding.   The error correction encoding is an error correction code that is decoded by iterative decoding. The decoding means performs iterative decoding, and performs demodulation on the modulated signal from which the error signal has been removed together with one decoding. 21. The demodulator according to claim 20, characterized in that

Images