The present invention relates to an apparatus for reading and/or writing information from/on an information storage medium.
To increase the storage capacity of an optical disk medium, which is one of various types of information storage media, it is effective to increase the numerical aperture of the objective lens and shorten the wavelength of the laser beam. This is because the size of a condensed laser beam spot is inversely proportional to the numerical aperture of an objective lens and directly proportional to the wavelength of the laser beam. That is why if the numerical aperture of the objective lens is increased and if the wavelength of the laser beam is shortened, the size of the condensed laser beam spot can be decreased and read/write operations can be done with marks and spaces of reduced lengths. Such a method of writing information with the mark/space widths varied is called a pulse width modulation (PWM) recording.
Hereinafter, conventional methods for reading/writing information from/on an optical disk medium will be described with reference to the accompanying drawings.
Such a method of writing using a beam spot works like a sort of low pass filtering because the condensed spot has a finite size. That is to say, the higher the frequency of the recording signal (i.e., the smaller the marks or spaces to be record), the more difficult it is to get the write operation done accurately. That is why in a conventional optical disk medium, to increase the recording density, the write data is converted into a run-length limited code, thereby making the lengths of the marks and spaces to be record greater than one bit and trying to get the write operation done using a write signal with as low frequency as possible. Such a code adapted to write properties is called a “recording code”.
In converting the recording code shown in portion (b) of
By controlling the light intensity of the write beam spot with the write-compensated write signal and moving such a write beam spot, data is written on the track. Considering a single location on the track of the recording layer, that single location is irradiated with pulsed beams a number of times. The temperature at that location rises every time the location is irradiated with the pulsed beam. And if the sum of the temperature increases exceeds the threshold value of the phase change material, the reflectance changes at that location. Otherwise, the reflectance does not change. That is to say, the reflectance at that location has one of two values depending on whether the sum of the temperature increases has exceeded the threshold value or not.
The pattern shown in portion (d) of
As described above, the read signal has had its high frequency components attenuated. That is why after its high frequency components have been amplified with a cosine equalizer, for example, its signal levels are determined to be zero or one, i.e., digitized. The digitization is realized by setting a predetermined slice level or by performing most likelihood decoding by PRML, for example. A PLL circuit for comparing the phases of a VCO with respect to the edges of the digitized signal generates a read clock signal. And by sampling the digitized signal with the read clock signal generated, the read digitized data shown in portion (c) of
The read digitized data is converted inversely to that of the write operation, i.e., the recording code is converted into binary data, thereby obtaining the read binary data shown in portion (d) of
Meanwhile, a method for increasing the recording density differently from the approach adopted in the conventional optical disk medium described above, i.e., the PWM method in which zeros and ones of digital data are recorded as marks and spaces, has also been proposed. An analog recording method has also been suggested. For example, according to Patent Document No. 1, a master for an optical disk medium is made by cutting track grooves as the modulated recording data, and the track groove shapes of the optical disk medium, transferred from the master, are recognized, thereby reading and writing data.
Another example, different from the simple PWM recording, is disclosed in Patent Document No. 2, which teaches a method of writing a signal by orthogonal frequency division multiplexing with the PWM and pulse amplitude modulation (PAM) methods combined.
These recording methods, however, have the following problems.
Specifically, according to the conventional methods for optical disk media, the technique of increasing the storage capacity, represented by the recording density in the PWM recording, by decreasing the spot size of a laser beam is about to reach the limit. As for BDs (Blu-ray discs), for example, the laser beam should have a wavelength of 405 nm and the objective lens should have a numerical aperture of 0.85. In order to further shorten the laser wavelength to increase the storage capacity, an ultraviolet laser beam will be needed but nobody knows when such a laser beam is available for use in actual products. Also, if the numerical aperture were set greater than 0.85, then a lens with such an NA would not only be hard to make perfectly but also have to be attached with much stricter accuracy. Furthermore, if the numerical aperture exceeded one, then a near-field recording should be performed using an immersion lens, not an ordinary lens. Each of these requirements is very hard to realize. That is why the technique of increasing the recording density by decreasing the laser beam spot has almost reached its limit.
Also, the efficiency of converting binary data to be written into a run-length limited code is 2:3 according to the RLL (1, 7) code for use in BDs and 8:16 according to the 8/16 code for use in DVDs. These conversion efficiencies are not so good, which is one of the reasons for the insufficient increase in recording density. The data shown in portions (a) and (b) of
Furthermore, the read digitized data may have errors. It is known that if the data shown in portion (c) of
Next, according to the method of Patent Document No. 1, either the width or the radial location of a track groove is changed with a signal, obtained by modulating the amplitude of write data, thereby recording write data on a master. That is to say, the master of an optical disk medium is made by cutting the track grooves with the modulated signal and the optical disk medium is formed with a stamper made of the master. However, such cutting should be done too accurately to be performed by an optical disk drive. For example, to eliminate external vibrations, the drive itself should be insulated dynamically from the outside world with a servo bench, for example. Besides, as the tracking needs to be subjected to open control, an accurate head feeding mechanism is required. What is more, the degree of planarity of the optical disk medium used and the accuracy of mounting it on the motor should also be sufficiently high. For these reasons, an optical disk drive that can perform read/write operations cannot do such cutting.
Furthermore, the radial shift of a track groove is very small in a normal optical disk system because the shift affects the tracking control. That is why the signal amplitude cannot be so large as to realize a high SNR, and not so much information can be stored. In the wobbled groove of a DVD-RAM, the track groove is also shifted radially as in the method of Patent Document No. 1. In DVD-RAMS, the grooves are shifted radially with respect to the centerline of the tracks. According to chapter 19.5 of the ECMA-33 standard, the magnitude of a wobble signal representing this groove shift is defined to be 5% to 10% of that of a tracking differential signal. This signal amplitude is very small.
Also, if the width of a track groove were changed, then the tracking signal itself would be modulated and the tracking control should be affected. That is why the track groove width cannot be changed significantly. As a result, the signal amplitude cannot be so large as to realize a high SNR, and not so much information can be stored.
According to the method of Patent Document No. 2, a single time unit length T is further divided into n1×2 sections as shown in
For example, suppose data for four sub-channels (i.e., 16 bits) needs to be written in a single symbol by 16 QAM modulation. The waveform of 16 QAM needs to have 16 levels in the amplitude direction. Also, supposing there are four sub-channels, the amplitude could be increased fourfold at most, recording will be performed in 64 gray scales (=16 gray scales×4), and n1=64. Therefore, the time unit length T is divided into 66 (=64+2) sections, where the last two bits added correspond to the data with the height a at both ends of the time unit length. Since one division length corresponds to one bit of the PWM recording code, one time unit length T will be 66 bits×74.5 nm=4917 nm supposing that recording can be done as accurately as on a BD. This length is much greater than the beam spot size of approximately 582 nm in a BD system. One time unit length T is preferably approximately several times as large as the beam spot size. If one time unit length T needs to be twice as large as the beam spot size, the one time unit length T should be approximately equal to 582 nm×2/4917 nm≈1/4.2. Thus, one division length (T/(n1×2)) should be 74.5 nm/4.2=17.7 nm, at which recording cannot be done. If the number of gray scales is decreased to a fourth, then the division length will be approximately equal to one bit length on the recording code of a BD. However, the number of sub-channels will be one and only four bits can be written per time unit length T, which is no different from the recording density of a BD.
Consequently, according to the writing method of Patent Document No. 2, the recording density realized will be at most equal to, or even lower than, that achieved by the PWM method.
In order to overcome the problems described above, an object of the present invention is to increase the recording density of an information storage medium.
An apparatus according to the present invention includes a write compensation section, which generates a write signal to write information on an information storage medium, and a writing section for irradiating the information storage medium with a pulsed beam based on the write signal generated by the write compensation section. The information storage medium includes a recording layer, which has an optical constant that changes continuously with the total quantity of light received. The writing section radiates and condenses multiple pulsed beams on the recording layer at an interval that is shorter than the diameter of the pulsed beams on the recording layer. The write compensation section generates the write signal such that the each sum of variations in the optical constant at each irradiated spot of the pulsed beam on the recording layer through the end of a write operation forms a predetermined variation pattern.
In one preferred embodiment, the optical constant is a refractive index.
In this particular preferred embodiment, the refractive index varies in response to a two-photon absorption reaction of the material of the recording layer, and the probability of the two-photon absorption reaction is proportional to the square of the intensity of the pulsed beam.
In an alternative preferred embodiment, the refractive index varies as the molecules of the recording layer change their directions perpendicularly to the plane of polarization of the pulsed beam, and the probability of change of the directions of the molecules is proportional to the square of the intensity of the pulsed beam.
In another alternative preferred embodiment, the refractive index varies in response to a one-photon absorption reaction of the material of the recording layer, and the probability of the one-photon absorption reaction is proportional to the intensity of the pulsed beam.
In another alternative preferred embodiment, the refractive index varies as the molecules of the recording layer change their directions perpendicularly to the plane of polarization of the pulsed beam, and the probability of change of the directions of the molecules is proportional to the intensity of the pulsed beam.
In a specific preferred embodiment, the material is diarylethene.
In another specific preferred embodiment, the recording layer includes photoaddressable polymers (PAPs).
In still another preferred embodiment, the apparatus further includes a modulating section for generating a signal representing the information by combining a plurality of sub-channel signals with each other. The information is written on the basis of a symbol with a predetermined length. A difference in frequency between carrier signals of the sub-channel signals is an integral multiple of the product of the spatial frequency of the symbol and a relative velocity of a beam spot with respect to the information storage medium.
In this particular preferred embodiment, the sub-channel signals have been subjected to phase modulation and the number of phase divisions is defined for each said sub-channel.
In an alternative preferred embodiment, the sub-channel signals have been subjected to orthogonal amplitude modulation and a signal point is set for each said sub-channel.
In another alternative preferred embodiment, the write compensation section generates the write signal such that there is a non-recorded area with a predetermined length between the symbols.
In a specific preferred embodiment, the write compensation section generates the write signal such that the non-recorded area is located between areas irradiated with a pulsed beam with prescribed power.
In yet another preferred embodiment, the information storage medium includes a plurality of recording layers including the recording layer.
Another apparatus according to the present invention is designed to generate a write signal to write information on an optical disk medium in the cast of use in an optical disk drive for writing data on the optical disk medium. The optical disk drive includes a writing section for irradiating the optical disk medium with a pulsed beam based on the write signal. The optical disk medium includes a recording layer, which has an optical constant that changes continuously with the total quantity of light received. The writing section radiates and condenses multiple pulsed beams on the recording layer at an interval that is shorter than the diameter of the pulsed beams on the recording layer. And the apparatus generates the write signal such that the each sum of variations in the optical constant at each irradiated spot of the pulsed beam on the recording layer through the end of a write operation forms a predetermined variation pattern.
A writing method according to the present invention includes the steps of: generating a modulation signal to write encoded information on an information storage medium; and irradiating the information storage medium with a pulsed beam based on the write signal generated by a write compensation section. The information storage medium includes a recording layer, which has an optical constant that changes continuously with the total quantity of light received. The step of irradiating includes radiating and condensing multiple pulsed beams on the recording layer at an interval that is shorter than the diameter of the pulsed beams on the recording layer. The step of generating the write signal includes generating the write signal such that the each sum of variations in the optical constant at each irradiated spot of the pulsed beam on the recording layer through the end of a write operation forms a predetermined variation pattern.
A program according to the present invention is set up to make an apparatus for writing data on an information storage medium perform recording processing. The recording processing includes the steps of: generating a modulation signal to write information on the information storage medium; and irradiating the information storage medium with a pulsed beam based on the write signal generated by a write compensation section. The information storage medium includes a recording layer, which has an optical constant that changes continuously with the total quantity of light received. The step of irradiating includes radiating and condensing multiple pulsed beams on the recording layer at an interval that is shorter than the diameter of the pulsed beams on the recording layer. The step of generating the write signal includes generating the write signal such that the each sum of variations in the optical constant at each irradiated spot of the pulsed beam on the recording layer through the end of the recording processing forms a predetermined variation pattern.
Still another apparatus according to the present invention is designed to read information from an information storage medium. The information storage medium includes a recording layer, which has an optical constant that changes continuously with the total quantity of light received. The information has been stored on the recording layer as a combination of a plurality of sub-channel signals. The apparatus includes: a reading section, which irradiates the recording layer with a beam to generate a read signal based on the light that has been reflected from the information storage medium; and a demodulating section, which multiplies the read signal and each of carrier signals together to generate each of sub-channel signals, associated with the respective carrier signal, and detect the information based on each sub-channel signal.
In one preferred embodiment, the information has been stored on the recording layer on the basis of a symbol with a predetermined length. There is a non-recorded area that stores no information between the symbols of the recording layer. And the demodulating section detects the top of the symbol by reference to a signal indicating the non-recorded area in the read signal.
In this particular preferred embodiment, the demodulating section generates a clock signal based on the signal indicating the non-recorded area.
In an alternative preferred embodiment, the non-recorded area is interposed between areas with a predetermined optical constant pattern.
An information storage medium according to the present invention includes a base material and a recording layer for storing information. The information is stored in the recording layer on the basis of a symbol with a predetermined length. And there is a non-recorded area that stores no information between the symbols of the recording layer.
In one preferred embodiment, the recording layer has an optical constant that changes continuously with the total quantity of light received, and the non-recorded area is interposed between areas with a predetermined optical constant pattern.
Paying attention to one particular location on an information storage medium, the information storage medium of the present invention includes a recording layer, which has an optical constant that changes continuously with the integrated quantity of light received at that location. When the recording layer is irradiated with multiple pulsed beams at an interval that is shorter than the diameter of the pulsed beams, a write compensation section generates a write signal such that the sum of variations in the optical constant at each irradiated spot of the pulsed beam on the recording layer through the end of a write operation has a predetermined variation pattern. By using such an information storage medium that has a continuously changing optical constant, information can be written using a multiplexed signal (e.g., an orthogonal frequency division multiplexed signal) and the recording density of the information storage medium can be increased. Also, by generating the write signal such that the sum of variations in the optical constant has a predetermined variation pattern, information can be written on such an information storage medium that has a continuously changing optical constant.
According to the present invention, a non-run-length-limited code can be used. That is why the error rate never deteriorates due to propagation of errors of a recording code and the errors can also be corrected. As a result, the recording density can be further increased.
In a preferred embodiment of the present invention, a difference in frequency between carrier signals of respective sub-channel signals is an integral multiple of the product of the inverse number of one symbol length and the linear velocity of the optical disk medium. Thus, the respective sub-channel signals have an orthogonal relation with respect to each other. By superposing those sub-channel signals with each other, a write signal is generated. And following the pattern of such a write signal, the optical constant of the recording layer changes and a write operation is carried out on a symbol-by-symbol basis. That is to say, this write operation is performed without threshold values, i.e., as analog recording.
In another preferred embodiment of the present invention, information is written on an information storage medium by superposing a number of mutually orthogonal sub-channel signals on a symbol-by-symbol basis. Also, a read signal is generated based on the light that has been reflected from such an information storage medium, and is multiplied by each of carrier signals, thereby generating each of sub-channel signal. Then, the stored information is detected using those sub-channel signals. By reading and writing information by a multiplexing method such as orthogonal frequency division multiplexing in this manner, the recording density of the information storage medium can be increased.
In still another preferred embodiment of the present invention, when irradiated with a beam, the molecules of the material of the recording layer change their directions perpendicularly (or parallel) to the plane of polarization of the beam, thereby writing information there. When the molecules of the material of the recording layer change their directions, an optical rotatory phenomenon arises, thus changing the refractive index. And when the refractive index changes, the reflectance also changes. Also, the directions of the molecules of the material of the recording layer are proportional to the square of the beam intensity. Furthermore, as there are two optical rotatory directions, two different pieces of information can be written per sub-channel.
In yet another preferred embodiment of the present invention, the optical constant (such as refractive index) changes due to the two-photon absorption reaction of the material of the recording layer. Therefore, the optical constant is proportional to the square of the intensity of the pulsed beam. That is why even if the recording power has varied for some reason, the influence on the optical constant is proportional to the root of the power variation. That is to say, the influence of the power variation on the optical constant can be reduced. As a result, the recording power margin can be increased and more stabilized recording is realized.
Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.
Portion (a) of
The recording layer 67 is made of a photon mode recording material with almost no threshold value. Also, the recording layer 67 has an optical constant that changes continuously with the total quantity of light received. For example, the variation in the optical constant of the recording material may be substantially a function (e.g., a liner function or a quadratic function) of the intensity of a write beam. A write operation in which the variation in the optical constant of the recording material is a linear function of the intensity of the write beam is called “one-photon absorption recording”. On the other hand, a write operation in which the variation in the optical constant of the recording material is a quadratic function of the intensity of the write beam is called “two-photon absorption recording”.
The optical constant may be a refractive index, for example. The refractive index changes due to the two-photon or one-photon absorption of the material of the recording layer 67. The probability of the two-photon absorption reaction is proportional to the square of the intensity of the pulsed beam. The probability of the one-photon absorption reaction is proportional to the intensity of the pulsed beam. A recording material with such a property may be diarylethene, for example, and the recording layer 67 includes such a material.
Also, in a preferred embodiment, the refractive index varies as the molecules of the recording layer change their directions perpendicularly to the plane of polarization of the pulsed beam. The probability of change of the directions of the molecules is proportional to either the square of the intensity of the pulsed beam or the intensity of the pulsed beam itself. A recording material with such a property may be a photoaddressable polymer (PAP), for example, and the recording layer 67 includes such a material.
Examples of materials for the recording layer 67 include fulgide, diarylethene, and PAP, which enable both one-photon absorption recording and two-photon absorption recording alike. These materials produce recording by changing their refractive index, which is one of optical constants. Also, it depends on the recording conditions (such as the wavelength of the beam) whether the recording material is proportional to the squared intensity of the pulsed beam or the intensity of the pulsed beam itself. In the following description, a recording material for two-photon recording such as diarylethene is supposed to be used. However, the principle, functions and effects of the present invention remain the same even if one-photon absorption recording is carried out.
The area B 65 on the optical disk medium 61 shows one of the recording layers 67 of the optical disk medium 61 and an enlarged view of the area B shows the details of the area B. As shown in the enlarged view of the area B, the recording layer 67 includes either concentric or spiral tracks 68, each of which is further divided into symbols 62 with a predetermined length. In each of those symbols 62, data is stored as a refractive index variation pattern as shown in portion (b) of
Hereinafter, an optical disk drive 100 for writing data on the optical disk medium 61 of this preferred embodiment and the writing method thereof will be described with reference to
The optical disk drive 100 includes: a modulating section 101 for generating a multiplexed signal, representing information to be written on the optical disk medium 61, by combining a number of sub-channel signals together; a write compensation section 95 for generating a write signal to write information on the optical disk medium 61; and an optical head section 103 for irradiating the recording layer 67 of the optical disk medium 61 with a converged pulsed beam based on the write signal generated by the write compensation section 95. In writing information on the disk medium 61, the optical head section 103 functions as a recording section. While moving the beam spot of the pulsed beam, the optical head section 103 irradiates the recording layer 67 with a plurality of pulsed beams at intervals that are shorter than the diameter of the pulsed beams on the recording layer 67. The write compensation section 95 generates the write signal such that the sum of variations in the optical constant at each irradiated spot of the pulsed beam on the recording layer 67 through the end of the write operation has a predetermined variation pattern. As used herein, the diameter of the pulsed beams on the recording layer 67 represents the extent of a pulsed beam spot area with a predetermined energy density. For example, the diameter of a pulsed beam may refer to the extent of an area that has generated power I/e2 times as high as that at the center of the pulsed beam (where e is a natural logarithm).
Taking a particular location on the recording layer 67, the particular location is irradiated with at least a portion of each of multiple pulsed beams. The write compensating section 95 generates the write signal such that the sum of the variations in the optical constant at the particular location, irradiated with each of the multiple pulsed beams, has a desired value.
The optical disk drive 100 may be either a recorder or a recorder/player. Alternatively, the optical disk drive 100 may also be a player that does not perform a write operation but does perform a read operation (to be described later). Some of those components of the optical disk drive 100 may be fabricated as a semiconductor integrated circuit. For example, the write compensation section 95 may be fabricated as a semiconductor integrated circuit. Those components of the optical disk drive 100 will be described in detail later.
Hereinafter, a writing method according to a preferred embodiment of the present invention will be described with reference to
The binary data shown in portion (a) of
The data to be written is stored on the basis of a symbol with a predetermined length. For example, the binary data shown in portion (a) of
The binary data for a single symbol is distributed to a plurality of sub-channels and modulated on a sub-channel basis. The data included in each pair of parentheses in portion (a) of
Each sub-channel signal is subjected to phase modulation. In the example shown in portion (b) of
By combining these QAM modulated sub-channel signals, the orthogonal frequency division multiplexed signal shown in portion (c) of
Since the two beam impulses have the same intensity, the resultant refractive index distribution should be flat between the beam spots of these two beam impulses. Comparing
As the method of write compensation, a write signal that will result in the desired read signal may be figured out asymptotically by hill climbing or any other simulation method. In this case, the desired read signal is similar to the orthogonal frequency division multiplexed signal to be written. The simulation algorithm may have the steps of:
(Step 1) set an initial write signal appropriately, generate write beam impulses by sampling the initial write signal at predetermined intervals, perform a write operation using the impulses, and obtain a refractive index pattern produced;
(Step 2) consider that the write operation is the two-photon absorption recording in which the probability is proportional to the square of the write beam intensity and that the data is written non-linearly by being irradiated with write beam spots a number of times, and obtain a read signal as a convolution of the intensity pattern of a read beam spot and the refractive index pattern;
(Step 3) compare the read signal obtained in step (2) to the desired read signal, and change the write signal such that the read signal obtained approaches to the desired read signal. More specifically, compare the read signal obtained in step (2) to the desired read signal and increase the intensity of the write beam impulse where the read signal obtained has a lower level or decrease the intensity of the write beam impulse where the read signal obtained has a higher level. And at the same time, calculate the integrated value of the absolute values of the differences between the read signal obtained and the desired read signal and use this value as an estimate of the write signal;
(Step 4) perform a write operation as in step (1) using the write signal changed; and
(Step 5) go back to step (2).
These processing steps are repeatedly carried out until the estimate of step (3) becomes equal to or smaller than a predetermined value. As a result, the orthogonal frequency division multiplexed signal shown in portion (c) of
Portion (e) of
A method of seeking a solution asymptotically such as the hill climbing method needs s lot of computations. That is why a method of finding a solution more analytically would require fewer computations and figure out the write signal more quickly. Hereinafter, a write compensation method according to a preferred embodiment of the present invention will be described.
The write compensation according to this preferred embodiment of the present invention is characterized by obtaining the intensity pattern of a write beam within a symbol such that a desired orthogonal frequency division modulated wave can be produced during a read operation. The intensity pattern of the write beam can be calculated in two steps. Specifically, in the first step, an optical constant pattern that will result in the desired orthogonal frequency division modulated wave is obtained. Then, in the next step, the intensity pattern of the write beam to record this optical constant pattern is obtained.
First, an optical constant pattern that will result in the desired read signal is obtained. More specifically, an optical constant pattern for one symbol, in which an orthogonal frequency division modulated wave is produced as a read wave, is obtained. In this example, the optical constant is refractive index, a Gaussian beam is used as a read beam, and a read signal is supposed to be obtained by scanning the refractive index pattern with a continuously radiated read beam. The read signal is obtained by performing a convolution operation on the refractive index pattern using the read beam. Thus, its inverse convolution operation may be performed.
The middle portion of
1−(8×N0̂2×N12̂2/(N12̂2+N0̂2)̂2+4×N0̂2×N12̂2−(N0̂2−N12̂2)̂2×cos(4×π×N12×D/λ)))
Since N1, N2, N0, λ and D are known and constant, the reflectance Rf[L] and the normalized refractive index n[L] have one-to-one correspondence. That is why the pattern shown in this middle portion will be referred to herein as a “reflectance pattern”.
The lower portion of
Taking these results into consideration, the relation between the read beam, the reflectance distribution and the read signal for a single symbol can be represented by the matrix equation shown in
Next, the intensity pattern of the write beam that forms this optical constant pattern, i.e., the refractive index pattern, will be obtained.
Consider how the refractive index at a particular location varies with the write beam. The normalized intensity of the write beam that irradiates the particular location for the kth time is supposed to be b[k] and the normalized refractive index after the write beam has been radiated for the kth time is supposed to be n[k]. As used herein, the “normalized write beam intensity” is supposed to be a beam intensity that causes the two-photon absorption recording reaction 100% when b[k]=1.0 as in
n[k]=n[k−1]+(1−n[k−1])×b[k]̂2
This equation can be modified into:
(1−n[k])=(1−n[k−1])×(1−b[k]̂2)
This is a kind of a geometrical series. If N[k]=1−n[k], then
N[k]=N[k−1](1−b[k]̂2)
Therefore, N[k] can be represented by b[0] through b[k]. That is to say,
N[k]=N[0](1−b[0]̂2)×(1−b[1]̂2)× . . . ×(1−b[k−1]̂2)×(1−b[k]̂2)
N[0] shows 1−n[0]. Consequently, if the initial refractive index and the write beam intensity at a particular point are known, the refractive index after the write beam has been radiated for an arbitrary number of times can be calculated.
In the same way, if the intensity of the write beam is known, the refractive index pattern recorded within a symbol can also be calculated analytically.
The parameters are re-defined by adding location parameters to the parameters of the beam distribution described above. Suppose the normalized intensity of the write beam radiated toward a sample point L for the kth time is b[k][L], the normalized refractive index after the sample point L has been irradiated with the write beam for the kth time is n[k][L], and N[k][L]=1-n[k][L]. Also, the normalized beam power distribution of the write beam at the sample point L is g[k][L] (of which the peak beam power is supposed to be 1.0). A[k] shows the peak power of the kth beam. The beam power distribution of the kth write beam is represented by A[k]×g[k][L].
Also, the beam power distribution of one pulse of the write beam within a single symbol can be represented by the matrix equation shown in
Using these re-defined parameters, the refractive index distribution after the kth write beam has been radiated is obtained based on the refractive index distribution after the (k−1)th write beam has been radiated and the beam power distribution of the kth write beam. In
The refractive index distribution within a single symbol after the kth write beam has been radiated is represented by the matrix equation shown in
As described above, the target read signal vector, obtained by substituting the target read waveform into plysg shown in
By formulating the write compensation computations into a mathematical equation and solving that equation as described above, the write beam intensity pattern can be calculated quickly so as to obtain the desired orthogonal frequency division modulated waveform. As a result, a write-compensated write signal can be obtained.
The optical disk drive 100 shown in
The modulating section 101 includes a turbo code modulator 91 and an orthogonal frequency division modulator 93. The write compensation section 95 includes a refractive index pattern calculator 96 and a write pulse intensity calculator 98. The optical head section 103 includes a pulse laser driver 110, a pulse laser diode 112, a collimator lens 113, a beam splitter 114, a quarter wave plate 115, an objective lens 116, an objective lens actuator 117, a condenser lens 121, a group of photosensors 122 and a servo circuit 126. The optical disk medium 118 including the recording layer 119 is rotated by the spindle motor 120.
In reading information from the optical disk medium, the optical head section 103 serves as a read section, which irradiates the recording layer with a beam and generates a read signal based on the light that has been reflected from the optical disk medium. The optical disk drive 100 further includes a demodulating section 104, which multiplies the read signal, generated by the optical head section 103, by a plurality of carrier signals, thereby generating a plurality of sub-channel signals associated with the carrier signals. Also, the demodulating section 104 detects the information stored on the information storage medium based on those sub-channel signals. The read operation will be described later.
Hereinafter, the write compensation operation performed by the optical disk drive 100 will be described.
The data to be written (i.e., the write data) is input from an information source coding block (not shown) to the turbo code modulator 91 (see
The ⅚ recursive systematic convolutional code modulator 222 receives the five-bit write data and outputs a single parity bit of a six-bit recursive systematic convolutional code.
The random interleaver 223 rearranges the arrangement order of input data/its associated parity bit pairs at random on a pair-by-pair basis. The interleave length corresponds to m symbols and is ordinarily less than one track round length. In this example, one codeword is supposed to be made up of six bits and one interleave length is supposed to be made up of 255×129 words (=197,370 bits). This interleave unit will be referred to herein as “one turbo codeword”.
The random interleaver 223 supplies the randomly rearranged data to another ⅚ recursive systematic convolutional code modulator 224. In this case, the ⅚ recursive systematic convolutional code modulators 222 and 224 have the same circuit. That is why the ⅚ recursive systematic convolutional code modulator 224 also outputs a single parity bit. The parity bits supplied from the two ⅚ recursive systematic convolutional code modulators are input to the selector 225, which selectively outputs the two parity bits alternately. That is to say, this turbo code is a puncture code.
The alternately output single parity bit and the five-bit data, supplied from the random interleaver 223, are output as a turbo code modulated signal 92 including six-bit parallel data to the orthogonal frequency division modulator 93.
Since this turbo code modulator 91 receives five-bit write data and outputs a six-bit turbo code, the coding rate is ⅚=0.83.
The orthogonal frequency division modulator 93 modulates the input turbo code modulated signal 92, including the six-bit parallel data, after having distributed the data to the respective sub-channels, and outputs an orthogonal frequency modulated signal 94 in which the modulated data has been added to the respective sub-channels.
The frequency f1 through f9 sub-channel encoders 232 through 240 receive the six-bit parallel data and a write clock signal from the sub-channel data distributor 230, and subject the six-bit parallel data to the 64 QAM modulation with carriers having respective frequencies (e.g., frequency f1 for the frequency f1 sub-channel encoder 232), thereby outputting 64 QAM modulated waveform signals synchronously with the write clock signal. The frequency fw of the write clock signal and the carrier frequencies fsc (=f1 through f9) satisfy the following equation (f1:n=1, f9:n=9):
fsc=fw/(n×2m)
where n=1, 2, 3, . . . and 9 and n and m are natural numbers.
The frequency f1 through f9 sub-channel encoders 232 to 240 output the 64 QAM modulated waveform signals to an adder 231, which outputs the sum as the orthogonal frequency division modulated signal 94 responsive to every pulse of the write clock signal.
The write compensation section 95 is made up of the refractive index pattern calculator 96 and the write pulse intensity calculator 98. The refractive index pattern calculator 96 receives the orthogonal frequency division modulated signal 94 and outputs a refractive index signal 97. The write pulse intensity calculator 98 receives the refractive index signal 97 and outputs a write signal 99.
The input orthogonal frequency division modulated signal 94 is once stored in a symbol modulated data memory 242 by way of a selector 241. The symbol modulated data memory 242 has double buffers. That is to say, while the orthogonal frequency division modulated signal 94 is being written on one symbol modulated data memory 242, the other symbol modulated data memory 242 is calculating the refractive index pattern. The orthogonal frequency division modulated signal 94 that has been output from the symbol modulated data memory 242 is supplied to a floating point sum of products calculator 244 by way of a selector 243.
The floating point sum of products calculator 244 performs floating point multiplications and additions and stores and outputs the results of the calculations so as to calculate the matrix products of the output values of the Gaussian beam inverse matrix value table 245 and the orthogonal frequency division modulated signal 94. Data reading for the purpose of calculating the matrix products is controlled using the addresses output by a storage address/calculated address generator 247 to the symbol modulated data memory 242 and to the Gaussian beam inverse matrix value table 245. The sum-of-products calculations are controlled in accordance with a sum-of-products calculation instruction issued by an arithmetic controller 246 to the floating point sum of products calculator 244. When the matrix product has been calculated, the symbol modulated data memories 242 and the selector 243 are switched to calculate the matrix product of the orthogonal frequency division modulated signal stored in the other symbol modulated data memory 242.
The refractive index signal 97 that has been output from the refractive index pattern calculator 96 is input to the write pulse intensity calculator 98. Receiving the refractive index signal 97 and its associated orthogonal frequency division modulated signal 94, the write pulse intensity calculator 98 solves the simultaneous equations shown in
The write pulse intensity calculator 98 shown in
The refractive index signal that has been output from the refractive index data memory 252 is input to an fk calculator 258 by way of a selector 256. The asymptotic write signal that has been output from the write signal data memory 255 is also input to the fk calculator 258 by way of a selector 257 and a ΔA[k] adder 25A. The ΔA[k] adder 25A adds zero or a predetermined constant to the asymptotic write signal in a prescribed sequence. That is to say, the write signal data memory outputs asymptotic write signals once (zero added) plus 256 times (A[k]) corresponding to one equation shown in
The zero-added value that has been output from the fk calculator 258 is stored in a register 259 and written in an fk memory 25B. A subtractor 25J subtracts the fk calculated values of the following asymptotic write signals, to which ΔA[k] has been added, from the value stored in the register 259 and writes the remainders in an αk memory 25C. 256 values corresponding to one equation shown in
When these operations are performed for one symbol, 256 fk values and 65,536 αk values for one symbol are output to an LUV decomposer 25D that solves the following simultaneous equations including the fk and αk values:
−fk=αk×xk
These equations will be 256 simultaneous equations. The resultant xk is input to an adder 25E, which adds it to the asymptotic write signal stored in the write signal data memory 255. Then, the resultant sum is written on the write signal data memory 255 again. As a result of this write operation, the value of the asymptotic write signal is updated to approach to the write-compensated write signal.
By repeatedly performing this calculation loop, the fk value goes closer and closer to zero, which means that the asymptotic write signal is approaching to the write compensated value. When a convergence decision circuit 25H senses that the fk value is equal to or smaller than a predetermined value, this calculation loop ends, the write-compensated write signal 99 and a write compensation fixed signal 25F are output, and the calculations are started all over again by switching to the other refractive index memory 255 and the other write signal data memory 255.
The write signal 99 that has been output from the write compensation section 95 and the write clock signal 135 are supplied to a pulse laser driver 110, which generates a pulse laser drive signal 111 based on the write signal 99 supplied. The pulse laser drive signal 111 is synchronous with the write clock signal 135. The pulse laser diode 112 transforms the pulse laser drive signal 111 into a laser beam. Since the two-photon absorption recording requires a laser beam with very high peak power, the pulse width should be much smaller than one period of the write clock signal and may be about 1 ns, for example.
The laser beam emitted from the pulse laser diode 112 is transformed into a parallel beam by the collimator lens 113, passed through the beam splitter 114 and the quarter wave plate 115, and then incident on the objective lens 116.
The objective lens 116 is controlled by the objective lens actuator 117 so as to focus the laser beam on the recording layer 119. The laser beam that has been focused on the recording layer 119 changes the composition of the recording layer 119, thereby varying the refractive index and writing data there.
In the foregoing description, every data is supposed to be floating point data. However, as long as the error can be estimated, calculations may also be done on fixed point data as well.
Next, a reading method according to the present invention will be described.
Portion (a) of
Portion (b) of
By multiplying the read signal shown in portion (b) of
Portion (d) of
If this binary data is generated as a turbo code or a low-density parity code, for example, robust error correction is realized and the recording density can be further increased substantially.
Hereinafter, the reading and writing methods of this preferred embodiment of the present invention will be further described.
When the reading method shown in
Optionally, a write signal, which has been subjected to write compensation with the powers of the last write beam impulse 143 of the former symbol and the first write beam impulse 144 of the latter symbol maximized from the beginning, may be generated to perform a write operation.
As described above, an area of a recording layer between two symbols includes a non-recorded area where no information is stored. The non-recorded area is interposed between two areas with a predetermined optical constant pattern. The demodulating section 104 locates the top of a symbol by reference to a signal component of a read signal representing the non-recorded area. Also, the demodulating section 104 generates a clock signal based on the signal representing the non-recorded area.
Hereinafter, it will be described in further detail how the optical disk drive 100 (see
The demodulating section 104 includes a read circuit 128, an orthogonal frequency division demodulator 131, and a turbo code demodulator 133. During reading, the pulse laser driver 110 controls the pulse laser diode 112 such that the laser diode 112 has continuous and constant optical power. However, its power is much smaller than the laser power for writing. The laser beam emitted from the pulse laser diode 112 is transformed into a parallel beam by the collimator lens 113, transmitted through the beam splitter 114 and the quarter-wave plate 115, and then incident on the objective lens 116.
The objective lens 116 is controlled by the objective lens actuator 117 so as to focus the laser beam on the recording layer 119. Unlike writing, the laser beam that has been focused on the recording layer 119 is too weak to change the composition of the recording layer 119. The laser beam that has been reflected from the recording layer 119 is transmitted through the objective lens 116 and the quarter-wave plate 115, reflected by the beam splitter 114, and then condensed by the condenser lens 121 onto the group of photosensors 122. The group of photosensors 122 output not only a focus sensor error signal 123 and a tracking sensor error signal 124, but also an analog read signal 125 as a sum of these sensor signals. The focus sensor error signal 123 and the tracking sensor error signal 124 are input to the servo circuit 126, which outputs the actuator drive signal 127 to the objective lens actuator 117, thereby controlling the objective lens actuator such that the laser beam is focused on the recording layer and that the laser beam spot is located right on the target track.
The analog read signal 125 is input to the read circuit 128.
The digital read signal 129 that has been output from the read circuit 128 is supplied to the orthogonal frequency division demodulator 131.
The incoming digital read signal 129, along with the top-of-symbol detection signal 130 and the sample clock signal 134 (which is an operating clock signal for the orthogonal frequency division demodulator 131), is input to frequency f1 through f9 sub-channel decoders 271 through 279, each of which performs a Hilbert transform on a carrier with an associated one of the frequencies f1 through f9 by reference to the top-of-symbol detection signal 130, thereby carrying out the 64 QAM demodulation and outputting two signals with I and Q values. The sample clock frequency fp and the carrier frequency fsc (=one of f1 through f9) satisfy the following Equation (in which n=1 for f1 and n=9 for f9):
fsc=fp/(n×2m)
where is n is a natural number and n=1, 2, 3 . . . , or 9 and m is also a natural number.
The 64 QAM demodulated data I and Q values are output from the frequency f1 through f9 sub-channel decoders 271 through 279 to a decoded value selector 27A, which selectively and sequentially supplies the outputs of the frequency f1 through f9 sub-channel decoders 271 through 279. This output signal will be an orthogonal frequency division demodulated signal 132.
The orthogonal frequency division demodulated signal 132 is output from the orthogonal frequency division demodulator 131 to the turbo code demodulator 133.
The incoming orthogonal frequency division demodulated signal 132 is supplied to decoder #2281, a channel value calculator 283, a variance value calculator 282, and a random deinterleaver 286.
The decoder #2281 performs maximum a posteriori probability (MAP) decoding on the incoming series of orthogonal frequency division demodulated signal 132 corresponding to one turbo codeword, thereby outputting a posteriori probability on a data-by-data basis. In this case, the series of orthogonal frequency division demodulated signal 132 is input as a series coming from the ⅚ recursive systematic convolutional code modulator 224. That is why the parities are input in the order in which the data has been interleaved. In the following description, the order of this series will be identified by n and the data of this series by dn. In the same way, as for the series of data to which parities have been added from the series coming from the ⅚ recursive systematic convolutional code modulator 222, the order of the series will be identified by k and the data of the series by dk.
The variance value calculator 282 calculates a Euclidean distance from the incoming orthogonal frequency division demodulated signal 132 to its closest signal point, figures out a variance value, and outputs it to the decoder #2281, the channel value calculator 283, decoder #1287 and a channel value calculator 288. When decoding is performed for the first time, zero is input from the selector 28B as a priori probability for the decoder #2281. That is to say, a priori probability becomes zero, which is not used for MAP decoding.
The channel value calculator 283 calculates, for each dn, a probability that the orthogonal frequency division demodulated signal 132 is the data dn. In this case, the orthogonal frequency division demodulated signal 132 is supposed to have a Gaussian distribution and is calculated based on the variance value supplied from the variance value calculator 282 and the Euclidean distance from the orthogonal frequency division demodulated signal 132 to each signal point.
The output ̂ (dn) of the decoder #2281 and the output Lch(dn) of the channel value calculator 283 are subjected to the following calculation to derive an external value Le(dn):
̂(dn)−La(dn)−Lch(dn)=Le(dn)
When this calculation is done for the first time, a priori probability is zero. That is why La(dn)=0.
The external value calculated is deinterleaved by a random deinterleaver 285 and converted into the series of La(dk), which is used as a priori probability for the decoder #1287 and is supplied to a priori probability input terminal of the decoder #1.
The series of deinterleaved orthogonal frequency division demodulated signal 132 that have been output from the random deinterleaver 286 at the same time are supplied to the decoder #1287 and the channel value calculator 288.
The decoder #1287 performs maximum a posteriori probability (MAP) decoding on the series of deinterleaved orthogonal frequency division demodulated signal 132, corresponding to one turbo codeword and supplied from the random deinterleaver 286, and a priori probability La(dk) supplied from the random deinterleaver 285, thereby outputting a posterior probability ̂(dk). In this case, the series of deinterleaved orthogonal frequency division demodulated signal 132 is input as a series coming from the ⅚ recursive systematic convolutional code modulator 222. That is why the parities are input in the order of the data.
The channel value calculator 288 calculates, for each dk, a probability that the orthogonal frequency division demodulated signal 132 is the data dk. In this case, the orthogonal frequency division demodulated signal 132 is supposed to have a Gaussian distribution and is calculated based on the variance value supplied from the variance value calculator 282 and the Euclidean distance from the deinterleaved orthogonal frequency division demodulated signal 132 to each signal point.
The output ̂(dk) of the decoder #1287, the output Lch(dk) of the channel value calculator 288 and a priori probability La(dk) supplied from the random deinterleaver 285 are subjected to the following calculation to derive an external value Le(dk):
̂(dk)−La(dk)−Lch(dk)=Le(dk)
The external value calculated is interleaved again by a random interleaver 28A so as to be converted into the series of La(dn), which is used as a priori probability for the decoder #2281 and is input to a priori probability input terminal of the decoder #2281 by way of a selector 28B.
The decoder #2281 calculates a posteriori probability ̂ (dn) again based on the incoming a priori probability La(dk) and the intermediate data (γ value) of previous MAP decoding, which is stored in the decoder #2281. As in the previous time, the decoder #2281 makes the following calculation to derive an external value Le(dk) and outputs it to the random deinterleaver 285 again.
̂(dn)−La(dn)−Lch(dn)=Le(dn)
It should be noted that La(dn) is the data that has been input to the decoder #2281 and the channel value Lch(dn) is the same data as the previous time.
When the difference between La(dk) and its previous value becomes equal to or smaller than a predetermined value by performing decoding recursively a prescribed number of times, a hard decision circuit 28C outputs dk that has the highest a posteriori probability in the respective words received.
Next, it will be described in detail how the decoders #1 and #2 operate.
(Step 11) The incoming orthogonal frequency division demodulated signal 132 is supplied to a Euclidean distance calculator 301, which calculates and outputs a Euclidean distance from the input orthogonal frequency division demodulated signal 132 to a reference read signal (i.e., a read signal with no errors), which has been generated by a reference read signal generator 302 in a predetermined order for every state produced as a result of the transition. In a ⅚ recursive systematic convolutional code, there are 32 states. Thus, 32×32=1,024 different Euclidean distances (i.e., all outputs shown in
The reference read signal generator 302 sequentially outputs signals rightward from F00 of the combination of the target state “00” and original state “00” shown in
This turbo code is punctured alternately. The punctured portions are different parity bits of multiple codewords. That is why as for the punctured portion, the average of the probability when the parity bit is “0” and the probability when the parity bit is “1” is output as the probability described above. The upper table shown in
(Step 12) The Euclidean distance thus obtained, supplied from the Euclidean distance calculator 301, is converted into a probability value based on the variance value supplied from the variance value calculator 282 and the Euclidean distance and then output by regarding the orthogonal frequency division demodulated signal 132 as having a Gaussian distribution.
(Step 13) The orthogonal frequency division demodulated signal that has been converted into a probability value gets stored in a γ (m, m′) memory 305 by way of a selector 304. The γ (m, m′) memory 305 retains the probability of transition from respective states for one turbo codeword into all states.
(Step 14) The orthogonal frequency division demodulated signal that has been converted into a probability value is also input to an adder 308 by way of a selector 306.
The adder 308 adds together the probability value and the output of a register 307 and outputs the sum. The probability value that has been added is input to, and stored in, the register 307 by way of a selector 30D. Thus, the output value of the register 307 becomes equal to that of the previous adder. The register 307 is reset every time the probability value of the first orthogonal frequency division demodulated signal of a turbo codeword to a symbol has been input and the target states are changed. That is why the sum of the probability values that have been input for the respective target states is output. The sum of the probability values for the respective target states is output from the adder 308, passed through selectors 309 and 30B, and then stored in an αS(m) memory 30A and an α(m) memory 30C, respectively. The αS(m) memory 30A stores the sum of the probabilities of transition from respective states of the orthogonal frequency division demodulated signal for one symbol into a single state. The α(m) memory 30C stores the sum of the probabilities of transition from respective states of the orthogonal frequency division demodulated signal for one symbol into a single state.
If the probabilities of the second or posterior orthogonal frequency division demodulated signal of a turbo codeword to a symbol are input, the values that have been stored in the order of transition in the register 307 by way of the αS(m) memory 30A and selector 30D are reset every time the target states are changed. Then, these values are used as initial values and added again by the adder, thereby updating the sum of the respective target state probability values.
(Step 15) When these processing steps are performed for one turbo codeword, the sum of the transition probabilities from the top of the turbo codeword through the state of each dn are stored in the α(m) memory. On the other hand, all transition probabilities within a single turbo codeword are stored in the γ(m, m′) memory 305.
(Step 16) The data stored in the γ(m, m′) memory 305 is read out reversely to the order of input of the orthogonal frequency division demodulated signals 132 and sequentially on a transition-by-transition basis. Specifically, in
(Step 17) The outputs of the γ(m, m′) memory 305 are supplied to an adder 30F by way of the selector 304. As in the αS(m) memory, the symbol of the orthogonal frequency division demodulated signal input to the adder 30F for the first time is added by resetting the value of the register 30E to zero every time the original states are changed. As for the second symbol to be added and symbols that follow it, the sum for each original state stored in the βS(m) memory is loaded into the register 30E and used as the initial value for addition every time the original states are changed.
(Step 18) When these processing steps are performed for one turbo codeword, the sum of the transition probabilities from the top of the turbo codeword through the state of each dn are stored in the β(m) memory.
(Step 19) For all combinations of transitions m→m′ for the zeroth symbol d0=0, the sum of the products α(m)×γ (m, m′)×β(m) is calculated. The sum of the probabilities for d0=0 is input from the α(m) memory 30C to the multiplier 30L by way of the selector 30B. The transition probabilities for d0=0 are input from the γ(m, m′) memory 305 to the multiplier 30L. And the sum of the probabilities for d0=0 is input from the β(m) memory 30J to the multiplier 30L. In response, the multiplier 30L multiplies together these input values. And the product and the value stored in a register 30N are added together by an adder 30M. The value of the register 30N is retained until the input to the multiplier 30L is changed from dn=0 and is reset when dn=1. That is why the output value of the adder 30M becomes the sum of the values that have been added while dn=0.
(Step 20) The processing step 19 is performed on each symbol from dn=0 through dn=31 (i.e., for one turbo codeword).
Next, it will be described what if the orthogonal frequency division demodulated signal is input to the decoder #1 or what if a turbo codeword is input to the decoder #2 from the second round on. Hereinafter, it will be described how the decoder #2 operates. The decoders #1 and #2 have the same configuration but are different in the order of data included in the incoming orthogonal frequency division demodulated signal 132, for example, i.e., whether the data has been interleaved (dn) or not (dk).
(Step 21) A priori value La(dn) is input to the multiplier 30O. At the same time, a transition probability for dn is input from the γ(m, m′) memory 305 to the multiplier 30O by way of the selector 304. There are 32×32=1,024 transition probabilities (i.e., all shown in
(Step 22) The product calculated by the multiplier 30O gets stored in the γ(m, m′) memory 305 by way of the selector 304. In the γ(m, m′) memory 305, the probability of transition from each of the states for one turbo codeword into every state is updated. After that, the same operation as that of step 14 is carried out.
In the foregoing description, every data is supposed to be floating point data. However, as long as the error can be estimated, calculations may also be done on fixed point data as well.
Next, the recording density achieved by the writing method of the present invention and that of the PWM recording, which is a conventional writing method, will be compared to each other with reference to
The uppermost graph of portion (a) of
In portion (b) of
As described above, 274 bits can be stored for 256 T according to the writing method of the present invention, whereas only 170 bits can be stored according to the conventional PWM recording method. Consequently, the recording density achieved by the writing method of the present invention can be approximately 1.6 times as high as that achieved by the conventional method.
In the two-photon absorption recording, the variation in the optical constant of the recording material is a quadratic function of the intensity of a write beam, and therefore, the recording pattern can be controlled three-dimensionally. That is why if this method is applied to three-dimensional recording on an optical disk medium with multiple recording layers, for example, the recording density can be further increased.
It should be noted that the operation performed by the optical disk drive 100 described above could be realized by means of software at least partially. For example, the optical disk drive 100 may include a memory device to store a program for performing the operations of the respective components described above and a central processing unit (CPU) that reads and executes the program. The optical disk drive 100 performs the operation described above following the program. The program may be either stored in advance in the memory device or downloaded and installed, for example.
The present invention is particularly effectively applicable for use in the field of reading and writing information from/on information storage media.
Number | Date | Country | Kind |
---|---|---|---|
2004-305456 | Oct 2004 | JP | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/JP2005/019310 | 10/20/2005 | WO | 00 | 4/18/2007 |