JP3892970B2 - Viterbi detection device, optical information reproducing device, and optical information recording / reproducing device - Google Patents

Description

ここで、
【数１０】
ｂ_ij＝ａ_k，０≦ｋ≦Ｎ−１
とすると、
【数１１】
ｉ＝ｋ mod N₁、ｊ＝ｋ mod N₂
である。
【００９２】
そして、ステップＳ１における既知信号Ｓrとしては、上記Ｍ配列｛ｂ_ij｝に対応する既知信号Ｓrが用いられる。
【００９３】
次に、当該既知信号Ｓrを用いた光メモリシステムＳの光学系の伝達関数の設定について説明する。
【００９４】
今、伝達関数を設定すべき光学系において情報の記録再生を行った場合に、液晶空間光変調器１内の各画素とこれらに夫々対応するＣＣＤ素子６内の各画素との間で、相互に上下左右に長さＬの符号間干渉が起こるものと仮定する。
【００９５】
この符号間干渉を表す光学系伝達関数（すなわち、ステップＳ９において設定すべき光学系伝達関数ｈij）を有する光メモリシステムＳの光学系に上記Ｍ配列｛ｂij｝に対応する既知信号Ｓrを入力した時、雑音が全くない場合に出力される符号（すなわち、理想出力）ｓijは、以下の畳み込みで示される。
【００９６】
【数１２】

ところで、実際にＣＣＤ出力信号Ｓpとして出力される符号ｇ_ijには、上記符号ｓ_ijに対して光学系自体が有している固有の雑音ｎ_ijが重畳されたものであり、
【数１３】
ｇ_ij＝ｓ_ij＋ｎ_ij
となる。
【００９７】
ここで、既知信号ＳrとしてのＭ配列｛ｂ_ij｝とＣＣＤ出力信号Ｓpとしての符号ｇ_ijとの間の相互相関は、以下の式により求められる。
【００９８】
【数１４】

このとき、式（２）の右辺第２項はＭ配列｛ｂ_ij｝と雑音ｎ_ijとの相関であるから零とおけるので、式（２）は式（１）を用いて、
【数１５】

と書ける。そして、これはＭ配列｛ｂ_ij｝の自己相関と伝達関数ｈ_ijの畳み込みである。
【００９９】
今、仮に、上述のＭ系列｛ａ_k｝におけるａ_iの「０」を「−１」、「１」を「＋１」に対応させた系列｛ｍ_i｝を用いたＭ配列｛ｂ_ij｝の自己相関は、
【数１６】

であるから、この式と式（３）より、
【数１７】
Φ_bb（０，０）＝１
Φ_bb（ｍ，ｎ）＝−１／Ｎ₁Ｎ₂ （但し、ｍ≠０，ｎ≠０）
であるから、
【数１８】

となる。ここで、Ｎ₁Ｎ₂が十分大きいとすると、近似的に
【数１９】
Φ_bg（ｋ，ｌ）≒ｈ_kl
となって、最終的に光学系伝達関数ｈ_ijの各要素を算出することができる（ステップＳ９）。
【０１００】
次に、ステップＳ１０で実行される液晶空間光変調器１の各画素と夫々に対応するＣＣＤ素子６の各画素との空間的な位置関係のずれ方向の検出について図４及び図５を用いて詳述する。
【０１０１】
始めに、ステップＳ１０において検出される位置のずれ方向の定義について、図４を用いて説明する。なお、図４は、ＣＣＤ素子６の受光面に液晶空間光変調器１の液晶面を投影した図であり、ＣＣＤ素子６の受光面は実線で示す各画素に区分されており、液晶空間光変調器１の液晶面は破線で示す各画素に区分されている。
【０１０２】
本実施形態における位置ずれは、図４中符号δx及びδyで示すように、液晶空間光変調器１の一の画素（図４中、左下斜め斜線で示す画素）と、これに対応するＣＣＤ素子６の一の画素（図４中、右下斜め斜線で示す画素）との相対的な位置のずれであり、以下の説明では、右方向及び下方向を夫々正として取り扱う。
【０１０３】
次に、ステップＳ１０における位置ずれ方向の検出方法について、図５に示すフローチャートを用いて説明する。
【０１０４】
図５に示すように、ステップＳ１０の位置ずれ方向検出処理においては、始めに、ステップＳ９までの処理で設定された光学系伝達関数の各要素のうち、最も大きい値の要素を検出し、当該検出した要素の光学系伝達関数行列中における位置を［Ｘ，Ｙ］とする（ステップＳ２０）。
【０１０５】
次に、設定された光学系伝達関数の各要素のうち、ステップＳ２０で設定された最大要素に対して一つ右隣の要素の値を［Ｘ＋１，Ｙ］とし、最大要素に対して一つ左隣の要素の値を［Ｘ−１，Ｙ］とし、最大要素に対して一つ上隣の要素の値を［Ｘ，Ｙ−１］とし、最大要素に対して一つ下隣の要素の値を［Ｘ，Ｙ＋１］とする（ステップＳ２１）。
【０１０６】
そして、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ２２）、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］であるときは（ステップＳ２２；ｙｅｓ）、最大要素の上隣の要素より下隣の要素の方が大きく、且つ左隣の要素よりも右隣の要素の方が大きいので、この場合は位置ずれ方向を右下と判定して（ステップＳ２３）次のステップＳ１１に移行する。
【０１０７】
一方、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］でないときは（ステップＳ２２；ｎｏ）、次に、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ２４）、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］であるときは（ステップＳ２４；ｙｅｓ）、最大要素の上隣の要素より下隣の要素の方が大きく、且つ左隣の要素と右隣の要素とが等しい値であるので、この場合は位置ずれ方向を下と判定して（ステップＳ２５）次のステップＳ１１に移行する。
【０１０８】
一方、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］でないときは（ステップＳ２４；ｎｏ）、次に、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ２６）、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］であるときは（ステップＳ２６；ｙｅｓ）、最大要素の上隣の要素より下隣の要素の方が大きく、且つ右隣の要素よりも左隣の要素の方が大きいので、この場合は位置ずれ方向を左下と判定して（ステップＳ２７）次のステップＳ１１に移行する。
【０１０９】
一方、［Ｘ，Ｙ−１］＜［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］でないときは（ステップＳ２６；ｎｏ）、次に、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ２８）、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］であるときは（ステップＳ２８；ｙｅｓ）、最大要素の上隣の要素と下隣の要素とが等しい値であり、且つ左隣の要素よりも右隣の要素の方が大きいので、この場合は位置ずれ方向を右と判定して（ステップＳ２９）次のステップＳ１１に移行する。
【０１１０】
一方、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］でないときは（ステップＳ２８；ｎｏ）、次に、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ３０）、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］であるときは（ステップＳ３０；ｙｅｓ）、最大要素の上隣の要素と下隣の要素とが等しい値であり、且つ左隣の要素と右隣の要素とが等しい値であるので、この場合は位置ずれがないと判定して（ステップＳ３１）次のステップＳ１１に移行する。
【０１１１】
一方、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］でないときは（ステップＳ３０；ｎｏ）、次に、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ３２）、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］であるときは（ステップＳ３２；ｙｅｓ）、最大要素の上隣の要素と下隣の要素とが等しい値であり、且つ右隣の要素よりも左隣の要素の方が大きいので、この場合は位置ずれ方向を左と判定して（ステップＳ３３）次のステップＳ１１に移行する。
【０１１２】
一方、［Ｘ，Ｙ−１］＝［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］でないときは（ステップＳ３２；ｎｏ）、次に、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ３４）、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］であるときは（ステップＳ３４；ｙｅｓ）、最大要素の上隣の要素が下隣の要素よりも大きく、且つ左隣の要素よりも右隣の要素の方が大きいので、この場合は位置ずれ方向を右上と判定して（ステップＳ３５）次のステップＳ１１に移行する。
【０１１３】
一方、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＜［Ｘ＋１，Ｙ］でないときは（ステップＳ３４；ｎｏ）、次に、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ３６）、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］であるときは（ステップＳ３６；ｙｅｓ）、最大要素の上隣の要素が下隣の要素よりも大きく、且つ左隣の要素と右隣の要素とが等しい値であるので、この場合は位置ずれ方向を上と判定して（ステップＳ３７）次のステップＳ１１に移行する。
【０１１４】
一方、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＝［Ｘ＋１，Ｙ］でないときは（ステップＳ３６；ｎｏ）、次に、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］か否かが判定され（ステップＳ３８）、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］であるときは（ステップＳ３８；ｙｅｓ）、最大要素の上隣の要素が下隣の要素よりも大きく、且つ左隣の要素が右隣の要素よりも大きい値であるので、この場合は位置ずれ方向を左上と判定して（ステップＳ３９）次のステップＳ１１に移行する。
【０１１５】
一方、［Ｘ，Ｙ−１］＞［Ｘ，Ｙ＋１］且つ［Ｘ−１，Ｙ］＞［Ｘ＋１，Ｙ］でないときは（ステップＳ３８；ｎｏ）、今までのステップＳ２０乃至Ｓ３８の処理に何らかの誤動作があったとして再度ステップＳ２０に戻って今までの処理を繰り返す。
【０１１６】
なお、図５に示したような最大要素の隣にある要素の大小を比較することで位置ずれ方向が検出できるのは、ステップＳ９で設定される伝達関数が、理論上、液晶空間光変調器１とＣＣＤ素子６との間の上記位置ずれと再生信号光Ｌpの広がり関数とにより一義的に決められるものであることによる。
【０１１７】
ここで、再生信号光Ｌpの広がり関数について説明する。
【０１１８】
再生信号光ＬpがＣＣＤ素子６に照射されるとき、当該再生信号光Ｌp自体が有する二次元強度分布性に起因して、液晶空間光変調器１の一の画素に対応する再生信号光Ｌpが、ＣＣＤ素子６上の対応する一の画素以外に当該一の画素に隣接する画素にも照射される。そして、このときに、再生信号光Ｌpに含まれる情報の当該隣接する画素への漏れが生じるが、この漏れの程度を示すのが上記広がり関数である。
【０１１９】
このとき、当該広がり関数に含まれる各要素の値は、再生信号光Ｌpの二次元強度分布性により、最大値を有する要素を中心として上下左右対称な大きさの値を夫々に有している。従って、ステップＳ９で設定された伝達関数においても、もし位置ずれがなければ、各要素の値が最大値を有する要素を中心として上下左右対称な値となる。これにより、ステップＳ９で設定された光学系伝達関数においてその最大要素を中心とした場合にその上下左右の要素の値に大小の差があれば、それは、液晶空間光変調器１とＣＣＤ素子６との位置ずれによるものであると判断できるのであり、その大小の分布状況を検出すれば、必然的に当該位置ずれの方向も検出できるのである。
【０１２０】
次に、本実施形態に係るビタビ検出について図６乃至図９を用いて説明する。
【０１２１】
先ず、簡単のために１次元の場合を考える。
【０１２２】
今、符号間干渉（ＩＳＩ：Inter Symbol Interferance）の長さをＬ＝２とすると、前後からの符号間干渉があるので伝達関数は、
【数２０】
ｈ_i＝［ｈ_-1，ｈ₀，ｈ₁］
となる。
【０１２３】
今、入力信号をｆ_iとすると、雑音がない場合の理想出力ｓ_iは入力信号と伝達関数の畳み込みで与えられ、
【数２１】

である。
【０１２４】
次に、図６に示すように１次元信号系列を２ビットづつに分けてビタビ検出を行う場合を考える。なお、図６において、実線で囲まれた部分が時刻ｋ−１での状態、破線で囲まれた部分が現時刻ｋでの状態である。
【０１２５】
図６において、状態Ｓの数は２²＝４であり、具体的には、
【数２２】

の状態をとる。ここで、時刻ｋ−１と状態と時刻ｋの状態とはオーパーラップしているので、両時刻での状態、すなわち状態遷移が決まれば検出値は一意に決まる（図６の円で囲まれた値）。
【０１２６】
次に、状態遷移図を図７に示す。‘０’及び‘１’は各状態遷移に対応する検出出力である。
【０１２７】
更に、図７に示す状態遷移図を時系列に展開したトレリス線図を図８に示す。
【０１２８】
ここで、取り得る状態遷移の数は図７に示すように８種類に制限される。そして、各状態遷移に対応する理想出力ｓ_kは表１に示す如くになるが、この理想出力ｓ_kは伝達関数が解っているので事前に計算しておくことが可能である。
【０１２９】
【表１】

これらを用いると、ビタビ検出とは測定値系列ｒ＝（ｒ₁，…，ｒ_K）から条件付き確率
【数２３】

を最大にするデータ系列ｄ＝（ｄ，…，ｄ_K）を見いだすことと等価となる。
【０１３０】
そこで、雑音が無相関であるとすると、結合確率は次のように表される。
【０１３１】
【数２４】

今、符号間干渉長が２であるから、ｄ_k-Lからｄ_kまでのＬ＋１のデータが状態遷移を決定する。
【０１３２】
この点について、これまで説明した例では、符号間干渉長Ｌは２であるから３個のデータが状態遷移を決定し、時刻ｋのデータに対してｋ−１とｋ＋１からの符号間千渉がある。
【０１３３】
また、データ系列はｆ＝（ｆ₁，…，ｆ_K）である。従って、
【数２５】

となる。
【０１３４】
ここで、ｓ_kは時刻ｋでの状態遷移の理想出力（表１参照）である。
【０１３５】
そして、ビタビ検出では、この確率を最大とするパスを選択する。
【０１３６】
すなわち、両辺の負の対数をとったものを対数尤度（パスメトリック）Ｌとすると、
【数２６】

となり積は加算になる。
【０１３７】
今、雑音を平均０の標準偏差σのガウス雑音と仮定すると、理想出力がｓkの時に測定値がｒkとなる確率密度関数は、
【数２７】

となり、この両辺の負の対数をとると、
【数２８】

となる。この右辺第１項と第３項は定数項で全てのブランチ（図８のトレリス線図の矢印）に共通であるから除くと、
【数２９】

がブランチメトリックＢkとなる。
【０１３８】
一方、パスメトリックＬ_kは、ブランチメトリックＢ_kの加算で与えられ
【数３０】

となる。
【０１３９】
以上説明した内容を図９を用いて纏めると、ビタビ検出のアルゴリズムは以下の通りとなる。すなわち、
時刻ｋの測定値ｒ_kを用いて各状態遷移に対してブランチメトリックＢ_k
【数３１】

を計算する。
【０１４０】
時刻ｋにおいて状態Ｓ_n（ｎ＝１，…，４）のそれぞれに至る２本のパスに対して
パスメトリックＬk
【数３２】

を計算し、パスメトリックＬkが小さいパスを各状態に至る生き残りパスとして残す（図９の矢印参照）。
【０１４１】
（３）トレリス線図においてマージ（合流）が生じてパスが確定した部分（図９の太線）の検出値（‘０’又は‘１’）を出力する。
【０１４２】
次に、実施形態のステップＳ１１における判定帰還方向の検出及び状態ブロックの形状の決定の実際について、図１０を用いて説明する。
【０１４３】
ここで、本実施形態における状態ブロックの形状について説明する。
【０１４４】
先ず、ＣＣＤ素子６からのＣＣＤ出力信号Ｓｐをビタビ検出する場合の状態の定義について図１０（ａ）を用いて説明する。なお、図１０（ａ）において「０」又は「１」は液晶空間光変調器１に印加された駆動信号Ｓcsを示している。
【０１４５】
例えば、光学系伝達関数が３行３列であると仮定すると、ＣＣＤ素子６上の一の画素からの上記ＣＣＤ出力信号Ｓｐの値は、当該画素（図１０（ａ）において円で囲まれた画素）に対応する液晶空間光変調器１の画素と当該液晶空間光変調器１上でその周辺にある８個の画素とに印加される駆動信号Ｓcsに依存して決定される。ここで、当該駆動信号Ｓcsの縦３ビットからなる値をビタビ検出におけるシンボルアルファベットと定義すると、そのシンボルアルファベットの数は０乃至７の計８個である。
【０１４６】
今、符号間干渉の長さを「２」とした場合、ビタビ検出における状態は、図１０（ａ）に実線で示すように、当該シンボルアルファベットを２個纏めて定義される。ここで、実線で囲まれた部分が前の状態であり、破線で囲まれた部分が現時点での状態である。このとき、状態ブロックの形状とは、図１０（ａ）中の実線又は破線の図形の形状をいう。図１０（ａ）の場合は、３行２列の長方形となる。また、状態の数は、
【数３３】
Ｍ^L＝８²＝６４
で、以下に示す各値をとる。
【０１４７】
【数３４】

一方、取り得る状態遷移は、
【数３５】
Ｓm＝（ｉ，Ｊ）→Ｓn＝（Ｊ，ｋ）
の場合のみで、トレリス線図上で各状態に対して計８本の入出力がある。また、各状態遷移に対して復号シンボルアルファベットは一意に決まり、上記の例では、Ｊが復号シンボルアルファベットとなる。
【０１４８】
次に、本実施形態の場合のように判定帰還を用いたビタビ検出処理における状態について図１０（ｂ）を用いて説明する。
【０１４９】
上述したように、判定帰還法においては、例えば下方向に帰還を行う場合、直上の行が正しく検出されたと仮定して当該直上の行のデータの影響を差し引いて当該ビタビ検出を行う画素（図１０（ｂ）で円で囲まれた画素）のＣＣＤ出力値が推定されるので、３行３列の場合、その状態は図１０（ｂ）に実線で示すようになる。
【０１５０】
従って、図１０（ｂ）に示す場合に、シンボルアルファベットの数Ｍは「０」乃至「３」の四つであり、状態の数は、判定帰還法を用いない場合に比して大幅に減少し、
【数３６】

となる。そして、各状態について具体的には以下に示すような状態をとる。
【０１５１】
【数３７】

また、トレリス線図上で各状態に至るパスの数も夫々に４本と少なくなり、計算量を大幅に削減することができる。
【０１５２】
ここで、本実施形態のように当該位置ずれが存在し、且つ、上述したように再生信号光Ｌｐについて広がり関数が適用される場合の状態の定義及び状態ブロックの形状について、図１１乃至図１３を用いて説明する。
【０１５３】
液晶空間光変調器１の各画素とＣＣＤ素子６の各画素との間に図１１に示すような位置ずれ（図４参照）のみがあった場合の光学的伝達関数Ｑは、
【数３８】

となる。
【０１５４】
次に、上記位置ずれと再生信号光Ｌpにおける上記広がり関数が混在している場合の光学系伝達関数（すなわち、上記ステップＳ５で設定される光学系伝達関数）及びその規模を縮小した縮小伝達関数について説明する。
【０１５５】
今、位置ずれの光学的伝達関数として上記式（４）を用い、広がり関数として以下に示す３行３列の伝達関数Ｐを用いるとすると、
【数３９】

位置ずれと広がり関数が混在しているときの光学的伝達関数は、以下に示す４行４列の光学的伝達関数Ｈとなる。
【０１５６】
【数４０】

そして、各要素は、以下の通りとなる。
【０１５７】
【数４１】

一方、４行４列の光学的伝達関数Ｈを判定帰還ビタビ検出する場合は、図１２に示すように状態を定義する必要があるので、状態数は５１２（＝２⁹）状態となり、各状態に対するパスの数は各８本となって、実際上ビタビ検出器１０で計算できない規模の計算量となる。
【０１５８】
そこで、本実施形態では、上述のように、光学的伝達関数の規模を縮小した縮小伝達関数を用いてビタビ検出を実行する。
【０１５９】
すなわち、今、広がり関数として
【数４２】

を用い、位置ずれとして図１１又は図４においてδｘ＝δｙ＝０．１とした場合は、光学的伝達関数Ｈは、
【数４３】

となる。ここで、各要素の内最大値を有する要素ｈ［０，０］を中心として大小関係を見てみると、第４行及び第４列並びに要素ｈ［−１，−１］については他の要素に比して極端に小さいので、これらを「０」と置き換えた縮小伝達関数、すなわち、元の光学的伝達関数において２行前、２列前並びに右下からの符号間干渉を無視した縮小伝達関数Ｈ’を求めると、
【数４４】

となる。
【０１６０】
この場合に、縮小伝達関数Ｈ’を用いた理想出力と元の光学的伝達関数Ｈを用いた理想出力とを比較すると、その誤差は、上記の例の場合で２．６％であることがシミュレーションにより確認されており、これは、二次元情報の再生においては、許容し得る誤差範囲内の誤差である。
【０１６１】
そして、３行３列の上記縮小伝達関数Ｈ’を用いる場合の状態の定義を図１３に示す。なお、図１３（ａ）はシンボルアルファベットを示す図であり、図１３（ｂ）は状態を定義を示す図である。
【０１６２】
この図１３に示す場合、シンボルアルファベットと状態は相互に同一で、（ｂ₀，ｂ₁，ｂ₂）の３ビットで定まる８状態であり、また、許容される状態遷移は各状態に対して４種類となって、ビタビ検出器１０で計算可能な規模となる。
【０１６３】
ここで、図１３（ｂ）において実線で囲まれた前状態から破線で囲まれた現状態への状態遷移が確定すれば、図１３（ｂ）中円で囲まれたビットが出力信号Ｓoutとして出力される。このときの許容される状態遷移と検出出力について、表２に纏めている。
【０１６４】
【表２】

次に、特に図１３に示した状態の定義を前提として、上記ステップＳ１１及びＳ１２においてＣＰＵ９において実行される判定帰還方向の検出及び状態ブロック形状の設定について、図１４を用いて説明する。
【０１６５】
上記ステップＳ１１及びＳ１２においては、図１４に示すような参照テーブルが予めＣＰＵ９内の図示しないメモリに記憶されており、ステップＳ１０において検出された位置ずれ方向に基づいて当該参照テーブルが参照され、制御信号Ｓ_cdに含ませるべき判定帰還方向の検出及び状態ブロック形状の設定が行われる。
【０１６６】
ここで、図１４に示す各位置ずれの場合について説明すると、先ず、ＣＣＤ素子６上の対応する画素に対して、再生信号光Ｌpが右下にずれた場合（上記式（６）で示される縮小伝達関数に対応する場合）には、式（６）において、第３行の値が第１行の値より大きい。これは、上の行からの符号間干渉が下の行からの符号間干渉よりも大きいことを示しており（これは、光学系の理想出力が入力信号Ｓinと縮小伝達関数の畳み込みであることに起因している。）、この場合は判定帰還方向は上から下に向かうのが有効であり、更に、上述のように右下からの符号間干渉は無視し得る。従って、状態ブロック形状は、図１４の最上欄に示される形状となる。
【０１６７】
次に、再生信号光Ｌpが左下にずれた場合は、対応する縮小伝達関数は、式（６）の縮小伝達関数の左右が反転したものとなる。この場合は判定帰還方向は上から下に向かうのが有効であり、更に、左下からの符号間干渉は無視し得る。従って、状態ブロック形状は、図１４の上から２欄目に示される形状となる。
【０１６８】
次に、再生信号光Ｌpが右上にずれた場合は、対応する縮小伝達関数は、式（６）の縮小伝達関数の上下が反転したものとなる。この場合は判定帰還方向は下から上に向かうのが有効であり、更に、右上からの符号間干渉は無視し得る。従って、状態ブロック形状は、図１４の上から３欄目に示される形状となる。
【０１６９】
更に、再生信号光Ｌpが左上にずれた場合は、対応する縮小伝達関数は、式（６）の縮小伝達関数の上下左右が反転したものとなる。この場合は判定帰還方向は下から上に向かうのが有効であり、更に、左上からの符号間干渉は無視し得る。従って、状態ブロック形状は、図１４の最下欄に示される形状となる。
【０１７０】
これらと同様に、再生信号光Ｌpが左右に位置ずれがなく、下方向のみにずれている場合は、状態ブロック形状は図１４の最上欄に示される形状となり、上方向のみにずれている場合は、状態ブロック形状は図１４の上から３欄目に示される形状となる。
【０１７１】
更に、再生信号光Ｌpが上下に位置ずれがなく、右方向のみにずれている場合は、状態ブロック形状は図１４の最上欄に示される形状となり、左方向のみにずれている場合は、状態ブロック形状は図１４の上から２欄目に示される形状となる。
【０１７２】
最後に、ステップＳ１５において実行される判定帰還ビタビ検出処理について、説明する。
【０１７３】
上述したように、ステップＳ１５においてビタビ検出器１０で実行される判定帰還ビタビ検出処理は、制御信号ＳcdとしてＣＰＵ９から入力されている位置ずれ方向に対応した縮小伝達関数及び判定帰還方向並びに状態ブロック形状に基づいて実行される。
【０１７４】
すなわち、図１４の円で囲まれたビットを判定帰還ビタビ検出する場合、このビットに対応するＣＣＤ出力信号Ｓpの出力値ｇ［ｉ，ｊ］から、帰還される行の既に推定済みの３ビット（図１４において矩形実線で囲まれている）からの符号間干渉を減算した出力値ｇ'［ｉ，ｊ］を算出する。
【０１７５】
一方、理想出力値ｓ'［ｉ，ｊ］を、帰還される行からの符号間干渉を考慮に入れないで、許容されるすべての状態遷移（図１４最上欄のブロック形状の場合は表２に示される）に対して計算しておく（すなわち、伝達関数が解っているので、ビタビ検出を開始する前に予め計算しておくことが可能である。）。
【０１７６】
そして、帰還される行からの符号間干渉を滅算した出力値ｇ'［ｉ，ｊ］と理想出力値ｓ'［ｉ，ｊ］を用いて全ての許容される状態遷移（トレリス線図のブランチ）に対してブランチメトリックＢ_kを
【数４５】
Ｂ_k＝（ｇ'［ｉ，ｊ］−ｓ'［ｉ，ｊ］）²
として算出する。
【０１７７】
次に、各状態に至る４本の入力パスに対して夫々にパスメトリックを計算し、各パスのうち、パスメトリックが最小のパスを生き残りパスとして残置する。
【０１７８】
その後、トレリス線図上でパスにマージが生じてパスが確定したら、確定したパス内の各状態遷移に対応した検出出力（図１４最上欄のブロック形状の場合は表２の値）を検出データビットとし、これを出力信号Ｓoutとして出力する。
【０１７９】
【実施例】
次に、本発明の効果を確認するために、シミュレーションによりビット誤り率を算出した結果を図１５及び図１６に示す。
【０１８０】
なお、図１５及び図１６において、「ＢｂＢ」は、再生方法としてビタビ検出を用いないビットバイビット方式を用いた場合を示しており、「DF-VA-16」は、再生方法として１６状態の判定帰還ビタビ検出処理（図１０（ｂ））を用いた場合を示しており、「DF_-VA-8」は、再生方法として８状態の判定帰還ビタビ検出処理（図１３（ｂ））を用いた場合を示している。
【０１８１】
また、位置ずれ方向としては、（ｒ，ｄ）が右下である場合を示しており、（ｌ，ｕ）が左上である場合を示している。
【０１８２】
更に、判定帰還の方向は上から下へ、ビタビ検出方向は左から右としている。
【０１８３】
また、広がり関数としては、式（５）の広がり関数を用い、入力信号Ｓinのレベルは、ビット‘１’に対応する白（Ｓw）が２００、ビット‘０’に対応する黒（Ｓb）が５０としている。更に、雑音は平均値が０のガウス雑音であり、４８０×６４０画素の白黒ランダムパターンの液晶空間光変調器１に対して４行４列の光学的伝達関数により符号間干渉を負荷した後に雑音を加えて、０から２５６の８ビットで量子化している。
【０１８４】
更にまた、Ｓ／Ｎ（信号／雑音）比は、
【数４６】

で定義する。ここで、σは、雑音の標準偏差である。
【０１８５】
更に、図１５（ａ）に示すシミュレーション結果は位置ずれが（δx，δy）＝（０．１，０．１）である場合について示しており、図１５（ｂ）に示すシミュレーション結果は位置ずれが（δx，δy）＝（０．２，０．２）である場合について示しており、図１６に示すシミュレーション結果は位置ずれが（δx，δy）＝（０．３，０．３）である場合について示している。
【０１８６】
図１５及び図１６から解るように、判定帰還法を用いたビタビ検出とこれを用いないビットバイビットによる再生では、明らかに判定帰還法を用いた方がビット誤り率が低くなっている。また、この効果は、Ｓ／Ｎ比が高いほど顕著である。
【０１８７】
更に、１６状態の判定帰還ビタビ検出処理を用いた場合と、８状態の判定帰還ビタビ検出処理を用いた場合とを比較すると、その差は殆どないことが解る。従って、３行３列の縮小伝達関数を用いて状態数を１６状態から８状態に削減したことによる性能劣化はほとんどなく、この点で、８状態の判定帰還法を用いる効果は大きい。
【０１８８】
更にまた、図１５及び図１６において、（ｒ，ｄ）と記したグラフは位置ずれ方向と判定帰還方向及び状態ブロック形状とが対応している場合であり、（ｌ，ｕ）と記したグラフは位置ずれ方向と判定帰還方向及び状態ブロック形状とが対応せず判定方向が逆である場合を示しているが、これらを比較すると、位置ずれが大きくなるにしたがって、（ｌ，ｕ）と記したグラフの方がビット誤り率の劣化が大きくなる。
【０１８９】
一方、位置ずれの方向に応じて判定帰還の方向と状態ブロック形状とを変えることにより、（ｒ，ｄ）と記したグラフまでビット誤り率を改善することができる。
【０１９０】
以上説明したように、実施形態の光メモリシステムＳの動作によれば、液晶空間光変調器１の画素と対応するＣＣＤ素子６の画素の間における相対的な位置ずれ方向に基づいて設定された判定帰還方向及び状態ブロック形状を用いて判定帰還ビタビ検出処理を施すので、当該位置ずれ方向と判定帰還方向及びビタビ検出方向とが異なることに起因する再生精度の劣化を防止することができる。
【０１９１】
また、位置ずれ方向が、既知信号Ｓｒに基づいて液晶空間光変調器１を駆動することにより光メモリ３に記録された二次元情報に対して参照光Ｌ_rを照射して得られた再生信号光Ｌpを受光して得られるＣＣＤ出力信号Ｓpと当該既知信号Ｓrとに基づいて算出された光学的伝達関数に基づいて検出されているので、正確に当該位置ずれ方向を検出して判定帰還方向及び状態ブロック形状を的確に設定することができる。
【０１９２】
従って、入力信号Ｓinに対応する二次元情報を光メモリ３上に記録することができると共に、正確に入力信号Ｓinを再生することができる。
【０１９３】
なお、上述の実施形態においては、既知信号Ｓｒとして既知のＭ配列に対応する既知信号Ｓｒを用いたが、これ以外に、例えば、液晶空間光変調器１の一画素だけを「白」とし、その周囲の画素を「黒」とするいわゆるインパルスパターンを用いて光学系伝達関数を設定してもよいし、また、既知の一般的なランダムパターンを用いて光学系伝達関数を設定してもよい。
【０１９４】
【発明の効果】
以上説明したように、請求項１に記載の発明によれば、変調用画素と対応する受光用画素との間における相対的位置のずれ方向に基づいて設定された判定帰還方向及び状態ブロック形状を用いて判定帰還ビタビ検出処理を施すので、当該ずれ方向と判定帰還方向及びビタビ検出方向とが異なることに起因する再生精度の劣化を防止することができる。
【０１９５】
従って、記録媒体に記録されている二次元情報に基づいて、より正確に入力信号を再生することができる。
【０１９６】
請求項２に記載の発明によれば、請求項１に記載の発明の効果に加えて、上記ずれ方向が、既知入力信号に基づいて空間光変調手段を駆動することにより記録媒体に記録された二次元情報に対して再生参照光を照射して得られた反射光を受光して得られる出力信号と当該既知入力信号とに基づいて算出された伝達関数に基づいて検出されているので、正確に当該ずれ方向を検出して判定帰還方向及び状態ブロック形状を的確に設定することができる。
【０１９７】
請求項３に記載の発明によれば、請求項１又は２に記載の発明の効果に加えて、正確に入力信号を再生することができる。
【０１９８】
請求項４に記載の発明によれば、請求項１又は２に記載の発明の効果に加えて、入力信号に対応する二次元情報を記録することができると共に、当該入力信号の再生時において正確に入力信号を再生することができる。
【図面の簡単な説明】
【図１】実施形態の光メモリシステムの概要構成を示すブロック図である。
【図２】実施形態の情報記録再生処理を示すフローチャートであり、（ａ）は情報記録処理を示すフローチャートであり、（ｂ）は情報再生処理を示すフローチャートである。
【図３】Ｍ配列発生器の構成を示すブロック図である。
【図４】液晶空間光変調器とＣＣＤ素子の位置ずれを示す説明図である。
【図５】位置ずれ方向の検出動作を示すフローチャートである。
【図６】符号間干渉長が２のときの１次元ビタビ検出を説明する図である。
【図７】符号間干渉長が２のときのビタビ検出の状態遷移図である。
【図８】符号間干渉長が２のときのトレリス線図である。
【図９】生き残りパスとマージを説明する図である。
【図１０】状態の定義を示す図であり、（ａ）はビタビ検出における状態の定義を示す図であり、（ｂ）は判定帰還ビタビ検出における状態の定義を示す図である。
【図１１】位置ずれがある時の伝達関数を示す図である。
【図１２】４行４列の伝達関数を用いた判定帰還ビタビ検出における状態の定義を示す図である。
【図１３】位置ずれと広がり関数とが混在した場合の判定帰還ビタビ検出を示す図であり、（ａ）はシンボルアルファベットを示す図であり、（ｂ）は状態の定義を示す図である。
【図１４】位置ずれ方向と判定帰還方向及び状態ブロック形状の関係を示す図である。
【図１５】実施形態の効果を示す図であり、（ａ）は位置ずれが（δx，δy）＝（０．１，０．１）の場合を示す図であり、（ｂ）は位置ずれが（δx，δy）＝（０．２，０．２）の場合を示す図である。
【図１６】位置ずれが（δx，δy）＝（０．３，０．３）の場合の実施形態の効果を示す図である。
【符号の説明】
１…液晶空間光変調器
２、５…レンズ
３…光メモリ
４…Ｙθステージ
６…ＣＣＤ素子
７、１８…ミラー
８…アナモフィック光学系
９…ＣＰＵ
１０…ビタビ検出器
１１…メモリ
１２…既知信号発生器
１３…スイッチ
１４、１５…シャッタ
１６…アルゴンレーザ
１７…ハーフミラー
２２…遅延回路
２３…排他的論理和回路
Ｓ…光メモリシステム
Ｓcl1、Ｓcl2、Ｓci、Ｓcw、Ｓcd…制御信号
Ｓz…ずれ方向信号
Ｓp…ＣＣＤ出力信号
Ｓout…出力信号
Ｓcs…駆動信号
Ｓin…入力信号
Ｓr…既知信号
Ｌa、Ｌb…アルゴンレーザ光
Ｌs…信号光
Ｌp…再生信号光
Ｌr…参照光[0001]
BACKGROUND OF THE INVENTION
The present invention belongs to the technical field of an optical memory system for recording and reproducing information two-dimensionally using so-called holography (coherent interference phenomenon between two lights), and more specifically, reproduction in the optical memory system. It belongs to the technical field of processing.
[0002]
[Prior art]
In recent years, optical memory systems that record and reproduce information two-dimensionally using holography are attracting attention as high-density recording and reproducing devices.
[0003]
  In this optical memory system, when an input signal is recorded, a transmissive spatial light modulation element composed of a plurality of pixels (specifically, for example, a liquid crystal having a plurality of pixels).Light modulatorEtc.), the recording signal light is spatially modulated based on the input signal to be recorded, and the spatially modulated recording signal light and the recording reference light emitted separately are made to interfere with each other to produce two-dimensional diffraction Light is generated and an image formed by the diffracted light is recorded on a recording medium as two-dimensional information corresponding to the input signal.
[0004]
At this time, as the recording medium, for example, a volume hologram recording element formed by laminating a plurality of lithium niobate layers is used.
[0005]
On the other hand, when the input signal is reproduced by reading the two-dimensional information from the recording medium on which the two-dimensional information is recorded, reproduction reference light (the same incidence as the recording reference light) is applied to the two-dimensional information. The reflected light generated by irradiating the recording medium with an angle is received by a light receiving element having a plurality of pixels to generate an output signal, and using the generated output signal Play the original input signal.
[0006]
Here, in the optical memory system, various noises may be mixed in the recording process and the reproducing process (in particular, noises are often generated due to the inhomogeneity of the recording medium). In some cases, the output signal is deformed due to the influence of the above, and the original input signal cannot be accurately reproduced.
[0007]
Therefore, conventionally, as a technique for accurately reproducing an input signal while reducing the influence of the noise, performing Viterbi detection processing on the output signal output from the light receiving element has been studied.
[0008]
At this time, when the Viterbi detection process is applied to the optical memory system, the relative positional relationship between the corresponding pixels is accurate for each pixel in the spatial light modulator and each pixel in the light receiving element. It is important to agree with This is because when a so-called positional shift occurs due to a shift in the relative positional relationship, a transfer function (transfer function in the Viterbi detection process) of the entire optical memory system set based on the input signal and the output signal is obtained. This is due to the change.
[0009]
On the other hand, using the fact that information recorded in an optical memory system is two-dimensional information, a decision feedback Viterbi detection process has been developed in which a so-called two-dimensional decision feedback method (Decision Feedback) is applied to the Viterbi detection process. Yes.
[0010]
Here, the outline of the decision feedback method will be described. For example, the light receiving element is often configured by arranging a large number of pixels in a rectangle, and the rectangular light receiving element receives light in the uppermost row. When the input signal is reproduced based on the output signal, the amount of light received in the uppermost row (virtual row) of the uppermost row is known to be zero. When the input signal is reproduced based on the output signal, it is reproduced as it is and no decision feedback is given.
[0011]
Next, when the input signal is reproduced based on the output signal from the second row from the top, the uppermost row is accurately based on the output signal (from the uppermost row of the uppermost row). Assuming that the input signal has been reproduced (without being affected), it is reproduced by subtracting the influence from the input signal in the uppermost row.
[0012]
Furthermore, when the input signal is reproduced based on the output signal from the third row from the top, it is assumed that the input signal is accurately reproduced based on the output signal in the second row from the top. Subtracting the influence from the input signal of the second row from the second row.
[0013]
At this time, the direction of the Viterbi detection process is, for example, laterally from left to right in the light receiving element, and in the above example, the direction of determination feedback is downward.
[0014]
In this way, assuming that the playback process has been performed accurately in the line above the current line, the current line is subtracted (ie, the judgment feedback is performed in the downward direction) by subtracting the effect of the line above. If the Viterbi detection process is performed based on the output signal from the output signal, the Viterbi detection process is eventually executed using the transition of the information change in the vertical direction. Therefore, the bit error rate is increased compared to the normal Viterbi detection process. The reproduction process can be accurately performed by lowering.
[0015]
Furthermore, the Viterbi detection process using the determination feedback method also has an advantage that the number of states in the Viterbi detection process can be reduced as compared with the Viterbi detection process not using the determination feedback method.
[0016]
Here, in the conventional determination feedback Viterbi detection process, the direction of the determination feedback is always constant.
[0017]
Further, in the conventional determination feedback Viterbi detection process, the shape of the state block used for the Viterbi detection process is always constant.
[0018]
[Problems to be solved by the invention]
However, in the conventional optical memory system described above, the relative positional relationship between each pixel in the spatial light modulation element and each corresponding pixel in the light receiving element is actually shifted in various directions. Thus, there is a problem in that the determination feedback Viterbi detection process may not be performed correctly.
[0019]
That is, in an actual optical memory system, it is physically necessary to match the relative positional relationship between each pixel in the spatial light modulation element and each corresponding pixel in the light receiving element so accurately that Viterbi detection processing can be performed effectively. Is difficult.
[0020]
Accordingly, there is a case where the direction of the positional deviation between the corresponding pixels and the direction of the determination feedback (always constant) do not match. In this case, the determination feedback Viterbi detection process cannot be performed accurately and effectively. There are cases.
[0021]
Furthermore, since the shape of the state block is constant as described above, the inaccuracy of the reproduction process is further increased when the above-described misalignment occurs in various directions.
[0022]
Therefore, the present invention has been made in view of the above-described problems, and the problem is that relative displacement between each pixel in the spatial light modulation element and each corresponding pixel in the light-receiving element. A Viterbi detector capable of accurately and effectively performing a decision feedback Viterbi detection process to improve the reproduction accuracy of an output signal, and an information reproducing apparatus and an information recording / reproducing apparatus using the Viterbi detector. It is to provide.
[0023]
[Means for Solving the Problems]
In order to solve the above-mentioned problem, the invention according to claim 1 is characterized in that the recording signal light is converted into a spatial light based on an input signal by a spatial light modulation means such as a liquid crystal spatial light modulator constituted by a plurality of modulation pixels. By modulating and irradiating a recording medium such as an optical memory, the reflected reference light from the recording medium irradiated with the two-dimensional information recorded on the recording medium is received and corresponds to the reflected light. A light receiving means such as a CCD element for generating an output signal, and a determination feedback Viterbi detection process for the output signal from the light receiving means constituted by a plurality of light receiving pixels respectively corresponding to the plurality of modulation pixels. A Viterbi detection device that detects the input signal by applying a pre-detected relative positional shift between the one modulation pixel and the light receiving pixel corresponding to the one modulation pixel. And setting means such as a CPU for setting the determination feedback direction and the state block shape in the determination feedback Viterbi detection process based on the direction, and using the set determination feedback direction and the state block shape for the output signal And a detection means such as a Viterbi detector that performs a determination feedback Viterbi detection process and detects the input signal.
[0024]
According to the operation of the first aspect of the present invention, the setting means determines based on the relative displacement direction between one modulation pixel and the light receiving pixel corresponding to the one modulation pixel. The judgment feedback direction and the state block shape in the feedback Viterbi detection process are set.
[0025]
Then, the detection means performs a determination feedback Viterbi detection process on the output signal using the set determination feedback direction and state block shape, and detects an input signal.
[0026]
Therefore, since the determination feedback Viterbi detection process is performed using the determination feedback direction and the state block shape set based on the relative position shift direction between the modulation pixel and the corresponding light receiving pixel, It is possible to prevent the reproduction accuracy from being deteriorated due to the difference between the determination feedback direction and the Viterbi detection direction.
[0027]
In order to solve the above problem, the invention according to claim 2 is the Viterbi detection device according to claim 1, wherein the shift direction is based on a known input signal that is the preset input signal. The output signal and the known input obtained by receiving the reflected light obtained by irradiating the reproduction reference light to the two-dimensional information recorded on the recording medium by driving a spatial light modulator. Of the optical system including at least the spatial light modulation means, the recording medium, a reproduction reference light irradiation means such as an argon laser for irradiating the reproduction reference light, and the light reception means. It is configured to be detected in advance based on the transfer function.
[0028]
According to the operation of the invention described in claim 2, in addition to the operation of the invention described in claim 1, the deviation direction is recorded on the recording medium by driving the spatial light modulation means based on the known input signal. A spatial light modulation means, which is a transfer function calculated based on an output signal obtained by receiving reflected light obtained by irradiating the reproduced reference light on the two-dimensional information and the known input signal Since the detection is based on the transfer function of the optical system including at least the recording medium, the reproduction reference light irradiating means, and the light receiving means, the deviation direction is accurately detected and the determination feedback direction and the state block shape are accurately set. be able to.
[0029]
In order to solve the above-mentioned problem, the invention described in claim 3 is the Viterbi detector according to claim 1, a recording medium such as an optical memory in which the two-dimensional information is recorded, and the reproduction reference light. Irradiation means and the light receiving means.
[0030]
According to the operation of the invention described in claim 3, in addition to the operation of the invention described in claim 1 or 2, the recording medium records two-dimensional information.
[0031]
The reproduction reference light irradiation means irradiates the recording medium with reproduction reference light during reproduction.
[0032]
Further, the light receiving means generates an output signal based on the reflected light of the reproduction reference light.
[0033]
Therefore, since the determination feedback Viterbi detection process is performed using the determination feedback direction and the state block shape set based on the relative position shift direction between the modulation pixel and the corresponding light receiving pixel, It is possible to prevent the reproduction accuracy from being deteriorated due to the difference between the determination feedback direction and the Viterbi detection direction, and to accurately reproduce the input signal.
[0034]
  In order to solve the above-mentioned problem, the invention according to claim 4 is a Viterbi detection device according to claim 1 or 2,The spatial light modulation means;Recording signal light emitting means such as an argon laser that emits the recording signal light, and an argon laser that emits recording reference light to a position on the recording medium where the two-dimensional information is to be recorded at the time of recording the two-dimensional information Recording reference light emitting means such as a recording medium, reproduction reference light irradiating means such as an argon laser for irradiating the recording medium with the reproduction reference light when reproducing the input signal, and the light receiving means. Prepare.
[0035]
According to the operation of the invention described in claim 4, in addition to the operation of the invention described in claim 1 or 2, the recording signal light emitting means emits recording signal light.
[0036]
The recording reference light emitting means emits the recording reference light to the position on the recording medium where the two-dimensional information is to be recorded when recording the two-dimensional information.
[0037]
Next, the recording medium records the two-dimensional information.
[0038]
On the other hand, the reproduction reference light irradiating means irradiates the recording medium with reproduction reference light when reproducing the input signal.
[0039]
Further, the light receiving means generates an output signal based on the reflected light of the reproduction reference light.
[0040]
Therefore, it is possible to record two-dimensional information corresponding to the input signal, and at the time of reproducing the input signal, the two-dimensional information is set based on the relative positional deviation direction between the modulation pixel and the corresponding light receiving pixel. Since the judgment feedback Viterbi detection process is performed using the judgment feedback direction and the state block shape, it is possible to prevent deterioration in reproduction accuracy due to the difference between the deviation direction and the judgment feedback direction and the Viterbi detection direction, and to accurately input signals. Can be played.
[0041]
DETAILED DESCRIPTION OF THE INVENTION
Next, preferred embodiments of the present invention will be described with reference to the drawings. The embodiment described below is an embodiment when the present invention is applied to an optical memory system that records and reproduces information two-dimensionally by holography using argon laser light.
[0042]
First, the overall configuration of the optical memory system according to the present embodiment will be described with reference to FIG.
[0043]
As shown in FIG. 1, the optical memory system S of the embodiment includes a liquid crystal spatial light modulator 1 as a spatial light modulation means, lenses 2 and 5, an optical memory 3 as a recording medium, a Yθ stage 4, and a light receiving means. A CCD (Charge Coupled Device) element 6, a mirror 7, an anamorphic optical system 8, a CPU 9 as setting means, a Viterbi detector 10 as detection means, a memory 11, a known signal generator 12, The switch 13, the argon laser 16, a half mirror 17 and a shutter 14 as recording signal light emitting means, and a mirror 18 and a shutter 15 as recording reference light emitting means and reproduction reference light emitting means are configured.
[0044]
Next, an outline operation of the optical memory system S will be described.
[0045]
First, the operation during information recording will be described.
[0046]
  At the time of recording information, the switch 13 is switched to the input signal Sin side corresponding to the information to be recorded based on the control signal Scw from the CPU 9, and the input signal Sin is used as the drive signal Scs in the liquid crystal space.lightIt is output to the modulator 1.
[0047]
On the other hand, in parallel with this, the shutter 14 controls the control signal scl from the CPU 9.₁Based on the above, the entire surface of the liquid crystal spatial light modulator 1 is irradiated with argon laser light La as recording signal light emitted from the argon laser 16 and reflected by the half mirror 17.
[0048]
Then, the liquid crystal spatial light modulator 1 driven by the drive signal Scs performs a spatial light modulation process based on the input signal Sin on the argon laser light La. The spatial light modulation process performed at this time is executed two-dimensionally by controlling transmission or non-transmission of the argon laser light La based on the drive signal Scs for each pixel in the liquid crystal surface of the liquid crystal spatial light modulator 1. Is done.
[0049]
Next, the argon laser beam La subjected to the spatial light modulation processing is condensed by the lens 2 and irradiated to the information recording surface of the optical memory 3 as the signal light Ls.
[0050]
Here, the optical memory 3 can move or rotate the optical memory 3 in the Y direction (vertical direction in FIG. 1) and the θ direction (rotation direction in a plane perpendicular to the central axis of the optical memory 3 parallel to the Y direction). For example, it is formed by laminating thinned lithium niobate crystals. Then, the two-dimensional information corresponding to the drive signal Scs (that is, the input signal Sin) is recorded by irradiating the signal light Ls that is two-dimensionally spatially modulated by the liquid crystal spatial light modulator 1.
[0051]
  At this time, in parallel with the irradiation of the signal light Ls to the optical memory 3, the shutter 15 is emitted from the argon laser 16 and transmitted through the half mirror 17 and reflected by the mirror 18 based on the control signal Scl2 from the CPU 9. Argon laser light Lb is emitted as recording reference light, and the argon laser light Lb is modulated by the anamorphic optical system 8 (that is, an optical system having a different magnification depending on each direction in a plane perpendicular to the optical axis of the argon laser light Lb). Mirror after being7Are simultaneously irradiated to the irradiation position of the signal light Ls on the optical memory 3 as the reference light Lr. As a result, the signal light Ls and the reference light Lr interfere with each other at the irradiation position of the signal light Ls on the optical memory 3, and the information contained in the drive signal Scs (that is, the input signal Sin) is detected by the hologram based on this interference. It is recorded on the optical memory 3 as dimension information.
[0052]
Next, the operation at the time of reproducing the two-dimensional information recorded in the optical memory 3 will be described.
[0053]
When reproducing the two-dimensional information, first, the shutter 15 is controlled by the control signal Scl from the CPU 9.₂The argon laser light Lb as the reproduction reference light emitted from the argon laser 16 and transmitted through the half mirror 17 and then reflected by the mirror 18 is emitted, and the argon laser light Lb is anamorphic as in the information recording. A position where the two-dimensional information to be reproduced on the optical memory 3 is recorded is irradiated as the reference light Lr via the system 8 and the mirror 7.
[0054]
Then, the reference light Lr is modulated and reflected by the two-dimensional information on the optical memory 3, and the reflected light is converted into the reproduction signal light Lp which is parallel light by the lens 5, and then is reflected on the light receiving surface of the CCD element 6. Incident light is converted into a CCD output signal Sp which is an electrical signal by each pixel on the CCD element 6 and output.
[0055]
Here, in principle, the spatial position of each pixel in the CCD element 6 matches the spatial position of each pixel included in the liquid crystal spatial light modulator 1 for each corresponding pixel. Although desirable, it is difficult to make it coincide exactly as described later.
[0056]
Then, the CCD output signal Sp output from the CCD element 6 is input to the Viterbi detector 10, and a decision feedback Viterbi detection process is performed on the CCD output signal Sp, which serves as an estimation signal for estimating the input signal Sin. The output signal Sout is output, and the input signal Sin is reproduced.
[0057]
In the optical memory system S shown in FIG. 1, for example, in an initial state such as when the optical memory 3 is replaced with a new one, a known signal generator 12 described later is based on the control signal Sci from the CPU 9. The known signal Sr corresponding to the M array is output to the liquid crystal spatial light modulator 1 as the drive signal Scs via the switch 13 (previously switched to the known signal generator 12 side based on the control signal Scw from the CPU 9). As a result, the above-described information recording operation (information recording operation on the optical memory 3 after replacement) is executed.
[0058]
On the other hand, at the time of reproduction, the above-described information reproduction operation is performed on the recorded known signal. As a result, the CCD output signal Sp output from the CCD element 6 is analyzed by the CPU 9 by the processing described later, and after this, Transfer function (hereinafter referred to as an optical system including the liquid crystal spatial light modulator 1, lenses 2 and 5, optical memory 3, Yθ stage 4, CCD element 6, mirror 7 and anamorphic optical system 8) , Referred to as an optical system transfer function).
[0059]
Next, based on this optical system transfer function, the positional relationship between the liquid crystal spatial light modulator 1 and the CCD element 3 is shifted (as described above, it is desirable that the positional relationship is not shifted in principle. In the optical memory system S, the positional relationship is shifted due to the fact that the pixels of the liquid crystal spatial light modulator 1 and the CCD element 3 are extremely small and it is difficult to accurately align the mounting position of the optical memory 3. ) Is detected by the CPU 9, and the determination feedback direction and the state block shape in the determination feedback Viterbi detection process in the Viterbi detector 10 are set based on this misalignment direction, and the determined determination feedback direction A control signal Scd indicating the shape of the state block is output to the Viterbi detector 10.
[0060]
At this time, the judgment feedback direction, the shape of the state block, and the optical system transfer function are set to one-to-one with respect to the optical system of the optical memory system S after the replacement of the optical memory 3, and thereafter the optical memory system is replaced by the replacement of the optical memory 3 described above. The S optical system is unchanged until the next initialization.
[0061]
Further, in the present embodiment, in the determination feedback Viterbi detection process of the Viterbi detector 10, not the optical system transfer function itself, but the set optical system in order to reduce the calculation amount in the determination feedback Viterbi detection process. Viterbi detection is performed using a reduced transfer function with a reduced transfer function scale (a reduced transfer function that varies depending on the direction of displacement). At this time, information indicating the reduced transfer function corresponding to each displacement direction is also included in the control signal Scd and output to the Viterbi detector 10.
[0062]
When the actual input signal Sin is reproduced, the Viterbi detector 10 outputs the CCD output signal Sp based on the determination feedback direction, the state block shape, and the reduced transfer function included in the input control signal Scd. The determination feedback Viterbi detection process is performed, and an output signal Sout is output.
[0063]
Next, information recording processing and information reproduction processing according to the embodiment will be described with reference to the flowchart shown in FIG. Note that the information recording process and the information reproducing process described below are for a case where a plurality of (for example, three) optical memories 3 are used to record and reproduce mutually different information.
[0064]
In addition, the information reproduction process described below includes an operation at the time of initialization of the optical memory system S by replacing the optical memory 3 described above.
[0065]
Further, FIG. 2A is a flowchart showing information recording processing to the optical memory 3, and FIG. 2B is a flowchart showing information reproducing processing from the optical memory 3.
[0066]
First, the information recording operation for the optical memory 3 will be described with reference to the flowchart shown in FIG.
[0067]
In the information recording process of the embodiment, as shown in FIG. 2A, first, a known signal Sr corresponding to the known M array described above is first generated by the known signal generator 12, and then driven through the switch 13. The signal Scs is input to the liquid crystal spatial light modulator 1 (step S1). The contents of the known signal Sr corresponding to this known M array will be described later in detail together with the setting of an optical system transfer function described later.
[0068]
Then, the liquid crystal spatial light modulator 1 is driven based on the input drive signal Scs to generate the signal light Ls corresponding to the drive signal Scs, which is irradiated to the optical memory 3 to correspond to the known signal Sr. The dimension information is recorded on the optical memory 3 by the hologram described above (step S2).
[0069]
Next, the input signal Sin to be originally recorded in the optical memory 3 is input to the liquid crystal light modulation element 1 as the drive signal Scs (step S3), whereby the argon laser light La is spatially modulated to generate the signal light Ls. By irradiating the optical memory 3 together with the reference light Lr, two-dimensional information corresponding to the input signal Sin is recorded on the optical memory 3 (step S4).
[0070]
Next, it is determined whether or not the recording of the input signal Sin to be recorded has been completed (step S5). When the recording has been completed (step S5; yes), the input signal is subsequently continuously input to the other optical memory 3. It is determined whether or not the optical memory 3 has been exchanged to record Sin (step S6). If it has not been exchanged (step S6; no), the process is terminated, and if it has been exchanged (step S6; yes) Returning to step S1, the recording operation described so far is repeated for the replaced optical memory 3.
[0071]
On the other hand, if it is determined in step S5 that the recording of the input signal Sin has not ended (step S5; no), the process returns to step S3 to continue the recording process.
[0072]
By the recording process described above, the recording of the two-dimensional information corresponding to the known signal Sr and the recording of the two-dimensional information corresponding to the input signal Sin are performed on the plurality of optical memories 3.
[0073]
Next, the reproduction process of the recorded two-dimensional information will be described with reference to FIG.
[0074]
At the time of reproduction processing, when the optical memory 3 in which the two-dimensional information to be reproduced is recorded is placed on the Yθ stage 4, next, the argon information is compared with the two-dimensional information on the placed optical memory 3. The reference light Lr is irradiated using the laser 16, the half mirror 17, the mirror 18, the shutter 15, the anamorphic optical system 8, etc. (step S7), and the reproduction signal light Lp, which is the reflected light, is received by the CCD element 6 and known. Two-dimensional information corresponding to the signal Sr is reproduced as a CCD output signal Sp (step S8), and the CPU 9 sets an optical system transfer function of the optical memory system S using the CCD output signal Sp and the known M array. (Step S9). Then, the CPU 9 uses the set optical system transfer function to generate the reduced transfer function by a method described later.
[0075]
The setting method of this optical system transfer function will be described later in detail.
[0076]
Next, in the same way, the CPU 9 detects the direction of positional deviation between each pixel of the liquid crystal spatial light modulator 1 and each corresponding pixel of the CCD element 6 based on the generated optical system transfer function (step S10). The shift direction signal Sz is temporarily stored in a memory 11 such as a RAM (Random Access Memory). A method for detecting this displacement will also be described in detail later.
[0077]
Here, when the two-dimensional information on the optical memory 3 is reproduced, the liquid crystal spatial light modulator 1 and the signal light Ls are not used, but the liquid crystal spatial light modulation is performed in the Viterbi detection process at the time of reproduction in step S10. In the optical memory system S, it is necessary to detect a positional shift between each pixel of the device 1 and each pixel of the corresponding CCD element 6 by using the liquid crystal spatial light modulator 1 when recording the input signal Sin. This is because the input signal Sin is dimensionally recorded on the optical memory 3, and thus the recorded two-dimensional information includes position information of each pixel of the liquid crystal spatial light modulator 1. .
[0078]
That is, a two-dimensional image is formed by turning each pixel on / off at the time of recording, and this two-dimensional image is recorded as it is as two-dimensional information in the optical memory 3, so that the two-dimensional information includes liquid crystal spatial light modulation. The two-dimensional information is recorded by the position information of each pixel of the element 1 and its on / off. Therefore, the two-dimensional information includes the position information of each pixel of the liquid crystal spatial light modulator 1.
[0079]
Therefore, the two-dimensional image formed on the CCD element 6 as the reproduction signal light Lp when reproducing the two-dimensional information is the same as the two-dimensional image formed by the liquid crystal spatial light modulator 1 at the time of recording. Therefore, if the Viterbi detection process is not performed in consideration of the positional deviation between the position of each pixel of the liquid crystal spatial light modulator 1 and the corresponding pixel on the CCD element 6 that receives light, an accurate output signal Sout is not obtained. Will not be obtained.
[0080]
Next, by reading out the detected misalignment from the memory 11, using this, the determination feedback direction in the determination feedback Viterbi detection process in the Viterbi detector 10 is detected (step S11), and the state block shape is set. (Step S12), and these are output to the Viterbi detector 10 as a control signal Scd together with information indicating the reduced transfer function.
[0081]
Here, in the present embodiment, in the Viterbi detector 10, Viterbi detection is performed from the left to the right in the transfer function, so the determination feedback direction is either the above-described downward direction or the downward-upward direction.
[0082]
Next, as a reproduction process of the two-dimensional information corresponding to the input signal Sin, first, the reference light Lr is irradiated to the two-dimensional information corresponding to the input signal Sin (step S13), and the reflected light is used as the reproduction signal light Lp as the CCD. The element 6 receives the light and reproduces the two-dimensional information as a CCD output signal Sp (step S14). Based on the CCD output signal Sp, the Viterbi detector 10 performs a determination feedback Viterbi detection process (step S15). An output signal Sout corresponding to the signal Sin is generated. In the determination feedback Viterbi detection process at this time, the determination feedback direction, the state block shape, and the reduced transfer function that are input from the CPU 9 as the control signal Scd are used.
[0083]
Then, whether or not the reproduction processing for all the two-dimensional information recorded in the currently mounted optical memory 3 has been completed, whether or not the CCD output signal Sp is continuously input by the Viterbi detector 10. If all the two-dimensional information has not been reproduced (step S16; no), the process returns to step S13 and the reproduction process is continued.
[0084]
On the other hand, when the reproduction process has been completed for all the two-dimensional information in the determination in step S16 (step S16; yes), it is next determined whether or not the optical memory 3 has been replaced with another one (step S16). S17) When exchanged (step S17; yes), the process returns to step S7 again to perform the processing at the time of optical system initialization (steps S7 to S12), and then input to the newly placed optical memory 3 Two-dimensional information reproduction processing corresponding to the signal Sin is performed (steps S13 to S17).
[0085]
If it is determined in step S17 that the optical memory 3 has not been replaced (step S17; no), the process is terminated as it is.
[0086]
Next, the known signal Sr corresponding to the known M array executed in steps S1, S2 and S7 to S9 and the setting of the optical system transfer function using the same will be described in detail.
[0087]
The M array described above is an n-th order M-sequence {a_k}, The M series {a_k} For each element in a₀, A₁, A₂... and when it reaches the end of the row or column, it moves to the opposite end and repeats the same thing and arranges it. Here, the n-th order M array is a period 2 generated by an n-stage shift register with feedback.ⁿIn the −1 (0, 1) periodic sequence, the feedback tap is defined by a generator polynomial. At this time, the generator polynomial must be a primitive polynomial.
[0088]
More specifically, for example, a third-order primitive polynomial
[Expression 1]
G (x) = x^Three+ X + 1
As shown in FIG. 3, the M array generator using can be composed of three delay circuits 20 to 22 and an exclusive OR circuit 23.
[0089]
This M array is an apparently random sequence, and is widely used as a pseudo-noise sequence for simulation or a synchronization sequence.
[0090]
More specifically, for example, an n-th order M array {a_k} Period
[Expression 2]
N = 2ⁿ-1
Product of prime numbers
[Equation 3]
N = N₁× N₂
And, and
[Expression 4]
N₁= 2^v-1
N₂= N / N₁
n = v × e
e ≧ 2
GCD (N₁, N₂) = 1 (where “GCD (N₁, N₂) "Is N₁And N₂Indicates the greatest common divisor of. )
If you can write
[Equation 5]
N = 2ⁿ-1
In the case of
[Formula 6]
N₁= 2^{n / 2}-1
N₂= 2^{n / 2}+1
Then,
[Expression 7]
N = N₁× N₂
Is satisfied. So, for example,
[Equation 8]
N = 2^Four-1
N₁= 3, N₂= 5
In the case of M series {a_k} To the M array {b_ij} Can be obtained.
[0091]
[Equation 9]

here,
[Expression 10]
b_ij= A_k, 0 ≦ k ≦ N−1
Then,
## EQU11 ##
i = k mod N₁, J = k mod N₂
It is.
[0092]
As the known signal Sr in step S1, the M array {b_ij}, A known signal Sr corresponding to.
[0093]
Next, setting of the transfer function of the optical system of the optical memory system S using the known signal Sr will be described.
[0094]
Now, when information is recorded / reproduced in an optical system in which a transfer function is to be set, each pixel in the liquid crystal spatial light modulator 1 and each pixel in the CCD element 6 corresponding to each of the pixels are mutually connected. Assume that intersymbol interference of length L occurs vertically and horizontally.
[0095]
  An optical system transfer function representing the intersymbol interference (that is, the optical system transfer function hij to be set in step S9))When the known signal Sr corresponding to the M array {bij} is input to the optical system of the optical memory system S having the above, the code (that is, the ideal output) sij that is output when there is no noise is shown by the following convolution: It is.
[0096]
[Expression 12]

By the way, the code g actually output as the CCD output signal Sp._ijThe above code s_ijInherent noise n of the optical system itself_ijIs superimposed,
[Formula 13]
g_ij= S_ij+ N_ij
It becomes.
[0097]
Here, the M array {b_ij} And the symbol g as the CCD output signal Sp_ijIs obtained by the following equation.
[0098]
[Expression 14]

At this time, the second term on the right side of Expression (2) is M array {b_ij} And noise n_ijSince the correlation is zero, equation (2) can be expressed using equation (1)
[Expression 15]

Can be written. And this is an M array {b_ij} Autocorrelation and transfer function h_ijIs a convolution.
[0099]
Suppose now that the M series {a_k} In a_iOf “0” in “-1” and “1” in “+1” corresponding to {m_i} Using M array {b_ij} Autocorrelation is
[Expression 16]

Therefore, from this equation and equation (3),
[Expression 17]
Φ_bb(0,0) = 1
Φ_bb(M, n) = − 1 / N₁N₂  (However, m ≠ 0, n ≠ 0)
Because
[Expression 18]

It becomes. Where N₁N₂Is sufficiently large, approximately
[Equation 19]
Φ_bg(K, l) ≒ h_kl
Finally, the optical system transfer function h_ijCan be calculated (step S9).
[0100]
Next, detection of the shift direction of the spatial positional relationship between each pixel of the liquid crystal spatial light modulator 1 and each corresponding pixel of the CCD element 6 executed in step S10 will be described with reference to FIGS. Detailed description.
[0101]
First, the definition of the position shift direction detected in step S10 will be described with reference to FIG. FIG. 4 is a diagram in which the liquid crystal surface of the liquid crystal spatial light modulator 1 is projected onto the light receiving surface of the CCD element 6. The light receiving surface of the CCD element 6 is divided into each pixel indicated by a solid line. The liquid crystal surface of the modulator 1 is divided into pixels indicated by broken lines.
[0102]
In the present embodiment, as shown by reference numerals δx and δy in FIG. 4, one pixel of the liquid crystal spatial light modulator 1 (a pixel indicated by a diagonal line at the lower left in FIG. 4) and a CCD element corresponding thereto. 6 is a relative position shift with respect to one pixel (a pixel indicated by a diagonal line at the lower right in FIG. 4). In the following description, the right direction and the lower direction are treated as positive.
[0103]
Next, the method for detecting the misalignment direction in step S10 will be described with reference to the flowchart shown in FIG.
[0104]
As shown in FIG. 5, in the positional deviation direction detection process in step S10, first, the element having the largest value among the elements of the optical system transfer function set in the process up to step S9 is detected. The position of the detected element in the optical system transfer function matrix is set to [X, Y] (step S20).
[0105]
Next, among the elements of the set optical system transfer function, the value of the element right next to the maximum element set in step S20 is set to [X + 1, Y], one for the maximum element. The value of the element on the left is [X-1, Y], the value of the element immediately above the maximum element is [X, Y-1], and the element immediately below the maximum element Is set to [X, Y + 1] (step S21).
[0106]
Then, it is determined whether [X, Y-1] <[X, Y + 1] and [X-1, Y] <[X + 1, Y] (step S22), and [X, Y-1] <[X , Y + 1] and [X−1, Y] <[X + 1, Y] (step S22; yes), the element adjacent to the lowermost element is larger than the element adjacent to the uppermost element and the element adjacent to the left. Since the element on the right side is larger than the right side, in this case, the position shift direction is determined as lower right (step S23), and the process proceeds to the next step S11.
[0107]
On the other hand, if [X, Y-1] <[X, Y + 1] and [X-1, Y] <[X + 1, Y] are not satisfied (step S22; no), then [X, Y-1] < It is determined whether [X, Y + 1] and [X-1, Y] = [X + 1, Y] (step S24), and [X, Y-1] <[X, Y + 1] and [X-1, Y]. ] = [X + 1, Y] (step S24; yes), the element next to the uppermost element is larger than the element adjacent to the uppermost element, and the element adjacent to the left is equal to the element adjacent to the right. Therefore, in this case, it is determined that the direction of displacement is downward (step S25), and the process proceeds to the next step S11.
[0108]
On the other hand, when [X, Y-1] <[X, Y + 1] and [X-1, Y] = [X + 1, Y] are not satisfied (step S24; no), then [X, Y-1] < It is determined whether [X, Y + 1] and [X-1, Y]> [X + 1, Y] (step S26), and [X, Y-1] <[X, Y + 1] and [X-1, Y]. ]> [X + 1, Y] (step S26; yes), the element adjacent to the lowermost element is larger than the element adjacent to the uppermost element, and the element adjacent to the left is larger than the element adjacent to the right element. Therefore, in this case, the position shift direction is determined to be lower left (step S27), and the process proceeds to the next step S11.
[0109]
On the other hand, if [X, Y-1] <[X, Y + 1] and [X-1, Y]> [X + 1, Y] are not satisfied (step S26; no), then [X, Y-1] = It is determined whether [X, Y + 1] and [X-1, Y] <[X + 1, Y] (step S28), and [X, Y-1] = [X, Y + 1] and [X-1, Y]. ] <[X + 1, Y] (step S28; yes), the element adjacent to the uppermost element and the element adjacent to the lower element are equal in value, and the element adjacent to the right is more than the element adjacent to the left. In this case, the positional deviation direction is determined to be right (step S29), and the process proceeds to the next step S11.
[0110]
On the other hand, when [X, Y-1] = [X, Y + 1] and [X-1, Y] <[X + 1, Y] are not satisfied (step S28; no), then [X, Y-1] = It is determined whether [X, Y + 1] and [X-1, Y] = [X + 1, Y] (step S30), and [X, Y-1] = [X, Y + 1] and [X-1, Y]. ] = [X + 1, Y] (step S30; yes), the upper adjacent element and the lower adjacent element have the same value, and the left adjacent element and the right adjacent element are equal. In this case, it is determined that there is no displacement (step S31), and the process proceeds to the next step S11.
[0111]
On the other hand, when [X, Y-1] = [X, Y + 1] and [X-1, Y] = [X + 1, Y] are not satisfied (step S30; no), then [X, Y-1] = It is determined whether [X, Y + 1] and [X-1, Y]> [X + 1, Y] (step S32), and [X, Y-1] = [X, Y + 1] and [X-1, Y]. ]> [X + 1, Y] (step S32; yes), the element adjacent to the uppermost element and the element adjacent to the lower element have the same value, and the element adjacent to the left than the element adjacent to the right In this case, the misalignment direction is determined to be left (step S33), and the process proceeds to the next step S11.
[0112]
On the other hand, when [X, Y-1] = [X, Y + 1] and [X-1, Y]> [X + 1, Y] are not satisfied (step S32; no), then [X, Y-1]> It is determined whether [X, Y + 1] and [X-1, Y] <[X + 1, Y] (step S34), [X, Y-1]> [X, Y + 1] and [X-1, Y]. ] <[X + 1, Y] (step S34; yes), the element adjacent to the uppermost element is larger than the element adjacent to the lower element, and the element adjacent to the right is larger than the element adjacent to the left. In this case, the misalignment direction is determined to be upper right (step S35), and the process proceeds to the next step S11.
[0113]
On the other hand, when [X, Y-1]> [X, Y + 1] and [X-1, Y] <[X + 1, Y] are not satisfied (step S34; no), then [X, Y-1]> It is determined whether [X, Y + 1] and [X-1, Y] = [X + 1, Y] (step S36), and [X, Y-1]> [X, Y + 1] and [X-1, Y]. ] = [X + 1, Y] (step S36; yes), the element on the upper side of the largest element is larger than the element on the lower side, and the element on the left and the element on the right are equal. Therefore, in this case, it is determined that the positional deviation direction is upward (step S37), and the process proceeds to the next step S11.
[0114]
On the other hand, if [X, Y-1]> [X, Y + 1] and [X-1, Y] = [X + 1, Y] are not satisfied (step S36; no), then [X, Y-1]> It is determined whether or not [X, Y + 1] and [X-1, Y]> [X + 1, Y] (step S38), and [X, Y-1]> [X, Y + 1] and [X-1, Y]. ]> [X + 1, Y] (step S38; yes), the element adjacent to the uppermost element is larger than the element adjacent to the lower element, and the element adjacent to the left is larger than the element adjacent to the right. Therefore, in this case, the position shift direction is determined to be upper left (step S39), and the process proceeds to the next step S11.
[0115]
On the other hand, when [X, Y−1]> [X, Y + 1] and [X−1, Y]> [X + 1, Y] are not satisfied (step S38; no), the processing of steps S20 to S38 so far is If there is a malfunction, the process returns to step S20 again to repeat the processing so far.
[0116]
Note that the displacement direction can be detected by comparing the size of the elements adjacent to the largest element as shown in FIG. 5 because the transfer function set in step S9 is theoretically a liquid crystal spatial light modulator. This is because it is uniquely determined by the positional deviation between 1 and the CCD element 6 and the spread function of the reproduction signal light Lp.
[0117]
Here, the spread function of the reproduction signal light Lp will be described.
[0118]
When the reproduction signal light Lp is irradiated on the CCD element 6, the reproduction signal light Lp corresponding to one pixel of the liquid crystal spatial light modulator 1 is caused by the two-dimensional intensity distribution of the reproduction signal light Lp itself. In addition to the corresponding one pixel on the CCD element 6, the pixel adjacent to the one pixel is also irradiated. At this time, information contained in the reproduction signal light Lp leaks to the adjacent pixels. The spread function indicates the degree of the leakage.
[0119]
At this time, the value of each element included in the spread function has a value that is symmetrical in the vertical and horizontal directions around the element having the maximum value due to the two-dimensional intensity distribution of the reproduction signal light Lp. . Therefore, even in the transfer function set in step S9, if there is no positional deviation, the value of each element becomes a value that is symmetrical vertically and horizontally about the element having the maximum value. Thus, if the optical element transfer function set in step S9 is centered on the maximum element, if there is a large or small difference between the values of the upper, lower, left, and right elements, that is the difference between the liquid crystal spatial light modulator 1 and the CCD element 6. It can be determined that this is due to the positional deviation, and if the distribution status is detected, the direction of the positional deviation can be inevitably detected.
[0120]
Next, Viterbi detection according to the present embodiment will be described with reference to FIGS.
[0121]
First, consider the one-dimensional case for simplicity.
[0122]
Now, assuming that the length of intersymbol interference (ISI: Inter Symbol Interferance) is L = 2, there is intersymbol interference from the front and back, so the transfer function is
[Expression 20]
h_i= [H_-1, H₀, H₁]
It becomes.
[0123]
Now the input signal is f_iThen, ideal output s when there is no noise_iIs given by the convolution of the input signal and transfer function,
[Expression 21]

It is.
[0124]
Next, consider a case where Viterbi detection is performed by dividing a one-dimensional signal sequence into two bits as shown in FIG. In FIG. 6, the portion surrounded by a solid line is the state at time k−1, and the portion surrounded by a broken line is the state at current time k.
[0125]
In FIG. 6, the number of states S is 2.²= 4, specifically,
[Expression 22]

Take the state. Here, since the time k−1, the state, and the state at the time k overlap each other, the detection value is uniquely determined when the state at both times, that is, the state transition is determined (circled in FIG. 6). value).
[0126]
Next, a state transition diagram is shown in FIG. '0' and '1' are detection outputs corresponding to each state transition.
[0127]
Further, FIG. 8 shows a trellis diagram obtained by developing the state transition diagram shown in FIG. 7 in time series.
[0128]
Here, the number of possible state transitions is limited to eight types as shown in FIG. And the ideal output s corresponding to each state transition_kAs shown in Table 1, this ideal output s_kCan be calculated in advance since the transfer function is known.
[0129]
[Table 1]

Using these, Viterbi detection is a measurement value series r = (r₁, ..., r_K) Conditional probability
[Expression 23]

D = (d,..., D_K) Is equivalent to finding.
[0130]
Therefore, assuming that the noise is uncorrelated, the coupling probability is expressed as follows.
[0131]
[Expression 24]

Now, since the intersymbol interference length is 2, d_kLTo d_kUp to L + 1 data determines the state transition.
[0132]
In this example, since the intersymbol interference length L is 2 in the example described so far, three pieces of data determine the state transition, and the intersymbol interference from k−1 and k + 1 with respect to the data at time k. There is.
[0133]
The data series is f = (f₁, ..., f_K). Therefore,
[Expression 25]

It becomes.
[0134]
Where s_kIs the ideal output of state transition at time k (see Table 1).
[0135]
In Viterbi detection, a path that maximizes this probability is selected.
[0136]
That is, if the logarithmic likelihood (path metric) L is taken as the negative logarithm of both sides,
[Equation 26]

Next product is addition.
[0137]
  Assuming that the noise is Gaussian noise with a standard deviation σ having an average of 0, the probability density function with a measured value rk when the ideal output is sk is,
[Expression 27]

  And taking the negative logarithm of both sides,
[Expression 28]

  It becomes. The first and third terms on the right side are constant terms and all branches (ReliBecause it is common to the arrows in the diagram)
[Expression 29]

  Becomes the branch metric Bk.
[0138]
On the other hand, path metric L_kIs branch metric B_kGiven by the addition of
[30]

It becomes.
[0139]
The contents described above can be summarized with reference to FIG. 9, and the Viterbi detection algorithm is as follows. That is,
Measured value r at time k_kBranch metric B for each state transition using_k
[31]

Calculate
[0140]
State S at time k_nFor two paths to each (n = 1, ..., 4)
Path metric Lk
[Expression 32]

And a path with a small path metric Lk is left as a surviving path to each state (see arrow in FIG. 9).
[0141]
(3) In the trellis diagram, a detection value ('0' or '1') of a portion (thick line in FIG. 9) where a merge (merging) has occurred and a path has been determined is output.
[0142]
Next, actual detection of the determination feedback direction and determination of the shape of the state block in step S11 of the embodiment will be described with reference to FIG.
[0143]
Here, the shape of the state block in the present embodiment will be described.
[0144]
First, the definition of the state when the CCD output signal Sp from the CCD element 6 is viterbi-detected will be described with reference to FIG. In FIG. 10A, “0” or “1” indicates the drive signal Scs applied to the liquid crystal spatial light modulator 1.
[0145]
For example, assuming that the optical system transfer function is 3 rows and 3 columns, the value of the CCD output signal Sp from one pixel on the CCD element 6 is circled in the pixel (FIG. 10A). Is determined depending on the drive signal Scs applied to the pixel of the liquid crystal spatial light modulator 1 corresponding to the pixel) and the eight pixels on the periphery of the liquid crystal spatial light modulator 1. Here, if a value consisting of three vertical bits of the drive signal Scs is defined as a symbol alphabet in Viterbi detection, the number of symbol alphabets is 0 to 7 in total.
[0146]
Now, assuming that the length of intersymbol interference is “2”, the state in Viterbi detection is defined by combining two symbol alphabets as shown by a solid line in FIG. Here, the portion surrounded by the solid line is the previous state, and the portion surrounded by the broken line is the current state. At this time, the shape of the state block means the shape of a solid line or a broken line in FIG. In the case of FIG. 10A, it is a rectangle with 3 rows and 2 columns. The number of states is
[Expression 33]
M^L= 8²= 64
Then, the following values are taken.
[0147]
[Expression 34]

On the other hand, the possible state transitions are
[Expression 35]
Sm = (i, J) → Sn = (J, k)
In this case, there are a total of 8 inputs and outputs for each state on the trellis diagram. Also, the decoded symbol alphabet is uniquely determined for each state transition. In the above example, J is the decoded symbol alphabet.
[0148]
Next, the state in the Viterbi detection process using determination feedback as in the case of the present embodiment will be described with reference to FIG.
[0149]
As described above, in the decision feedback method, for example, when feedback is performed in the downward direction, the pixel that performs the Viterbi detection by subtracting the influence of the data in the immediately above row on the assumption that the immediately above row is correctly detected (see FIG. Since the CCD output value of the pixel (circled in 10 (b)) is estimated, in the case of 3 rows and 3 columns, the state is as shown by a solid line in FIG. 10 (b).
[0150]
  Therefore, in the case shown in FIG.AThe number of bets M is four from “0” to “3”, and the number of states is greatly reduced as compared to the case where the judgment feedback method is not used.
[Expression 36]

  It becomes. Each state specifically takes the following states.
[0151]
[Expression 37]

In addition, the number of paths reaching each state on the trellis diagram is reduced to four, and the amount of calculation can be greatly reduced.
[0152]
Here, with respect to the definition of the state and the shape of the state block when the positional deviation exists as in the present embodiment and the spread function is applied to the reproduction signal light Lp as described above, FIGS. Will be described.
[0153]
The optical transfer function Q when there is only a displacement (see FIG. 4) as shown in FIG. 11 between each pixel of the liquid crystal spatial light modulator 1 and each pixel of the CCD element 6 is
[Formula 38]

It becomes.
[0154]
Next, an optical system transfer function (that is, the optical system transfer function set in step S5) when the positional deviation and the spread function in the reproduction signal light Lp are mixed, and a reduced transfer function whose scale is reduced. Will be described.
[0155]
Now, assuming that the above equation (4) is used as the optical transfer function of misalignment and the transfer function P of 3 rows and 3 columns shown below is used as the spread function,
[39]

The optical transfer function when the positional deviation and the spread function are mixed is an optical transfer function H of 4 rows and 4 columns shown below.
[0156]
[Formula 40]

Each element is as follows.
[0157]
[Expression 41]

On the other hand, in the case of detecting the feedback feedback Viterbi of the 4 × 4 optical transfer function H, it is necessary to define the states as shown in FIG. 12, so the number of states is 512 (= 2⁹) States, and the number of paths for each state is eight, which is a calculation amount that cannot be calculated by the Viterbi detector 10 in practice.
[0158]
Therefore, in this embodiment, as described above, Viterbi detection is performed using a reduced transfer function obtained by reducing the scale of the optical transfer function.
[0159]
That is, as a spread function now
[Expression 42]

11 and FIG. 4 as δx = δy = 0.1, the optical transfer function H is
[Expression 43]

It becomes. Here, looking at the magnitude relationship around the element h [0, 0] having the maximum value among the elements, the fourth row and the fourth column and the element h [-1, -1] Since it is extremely small compared to the elements, a reduced transfer function in which these are replaced with “0”, that is, a reduction neglecting the intersymbol interference from the second row, the second column, and the lower right in the original optical transfer function. When the transfer function H ′ is obtained,
(44)

It becomes.
[0160]
In this case, when the ideal output using the reduced transfer function H ′ is compared with the ideal output using the original optical transfer function H, the error is 2.6% in the case of the above example. This is confirmed by simulation, and this is an error within an allowable error range in the reproduction of two-dimensional information.
[0161]
FIG. 13 shows the definition of the state when the reduced transfer function H ′ of 3 rows and 3 columns is used. FIG. 13A is a diagram showing symbol alphabets, and FIG. 13B is a diagram showing definitions of states.
[0162]
In the case shown in FIG. 13, the symbol alphabet and state are the same as each other, and (b₀, B₁, B₂) And 8 states determined by 3 bits, and the allowable state transitions are 4 types for each state, which is a scale that can be calculated by the Viterbi detector 10.
[0163]
Here, if the state transition from the previous state surrounded by the solid line to the current state surrounded by the broken line in FIG. 13B is confirmed, the bit surrounded by the circle in FIG. 13B becomes the output signal Sout. Is output. Table 2 summarizes the allowable state transition and detection output at this time.
[0164]
[Table 2]

Next, assuming the definition of the state shown in FIG. 13 in particular, the detection feedback direction detection and the state block shape setting executed by the CPU 9 in steps S11 and S12 will be described with reference to FIG.
[0165]
In steps S11 and S12, a reference table as shown in FIG. 14 is stored in advance in a memory (not shown) in the CPU 9, and the reference table is referred to based on the positional deviation direction detected in step S10. Signal S_cDetection of the decision feedback direction to be included in d and setting of the state block shape are performed.
[0166]
Here, the case of each positional shift shown in FIG. 14 will be described. First, when the reproduction signal light Lp is shifted to the lower right with respect to the corresponding pixel on the CCD element 6 (shown by the above formula (6)). In the case of corresponding to the reduced transfer function), the value of the third row is larger than the value of the first row in Equation (6). This indicates that the intersymbol interference from the upper row is larger than the intersymbol interference from the lower row (this means that the ideal output of the optical system is a convolution of the input signal Sin and the reduced transfer function) In this case, it is effective that the decision feedback direction is from the top to the bottom, and the intersymbol interference from the lower right can be ignored as described above. Accordingly, the state block shape is the shape shown in the uppermost column of FIG.
[0167]
Next, when the reproduction signal light Lp is shifted to the lower left, the corresponding reduced transfer function is an inverted version of the reduced transfer function of Equation (6). In this case, it is effective that the decision feedback direction is from top to bottom, and intersymbol interference from the lower left can be ignored. Accordingly, the state block shape is the shape shown in the second column from the top in FIG.
[0168]
Next, when the reproduction signal light Lp is shifted to the upper right, the corresponding reduced transfer function is obtained by inverting the reduced transfer function of Expression (6) upside down. In this case, it is effective that the decision feedback direction is from bottom to top, and intersymbol interference from the upper right can be ignored. Therefore, the state block shape is the shape shown in the third column from the top in FIG.
[0169]
Further, when the reproduction signal light Lp is shifted to the upper left, the corresponding reduced transfer function is an inverted version of the reduced transfer function of Expression (6). In this case, it is effective that the decision feedback direction is from bottom to top, and intersymbol interference from the upper left can be ignored. Accordingly, the state block shape is the shape shown in the bottom column of FIG.
[0170]
Similarly to the above, when the reproduction signal light Lp is not displaced in the right and left direction and is shifted only in the downward direction, the state block shape is the shape shown in the uppermost column of FIG. 14 and is shifted only in the upward direction. The state block shape is the shape shown in the third column from the top in FIG.
[0171]
Further, when the reproduction signal light Lp is not displaced in the vertical direction and is displaced only in the right direction, the state block shape becomes the shape shown in the uppermost column of FIG. The block shape is the shape shown in the second column from the top of FIG.
[0172]
Finally, the determination feedback Viterbi detection process executed in step S15 will be described.
[0173]
As described above, the determination feedback Viterbi detection process executed by the Viterbi detector 10 in step S15 is performed by the reduced transfer function, the determination feedback direction, and the state block shape corresponding to the misalignment direction input from the CPU 9 as the control signal Scd. It is executed based on.
[0174]
That is, when the decision feedback Viterbi detection is performed for the bit surrounded by the circle in FIG. 14, from the output value g [i, j] of the CCD output signal Sp corresponding to this bit, the already estimated 3 bits of the row to be fed back An output value g ′ [i, j] is calculated by subtracting the intersymbol interference from (enclosed by a solid rectangular line in FIG. 14).
[0175]
On the other hand, the ideal output value s ′ [i, j] is all allowed state transitions (Table 2 in the case of the block shape in the uppermost column in FIG. 14) without taking into account the intersymbol interference from the fed back row. (Ie, since the transfer function is known, it is possible to calculate in advance before starting the Viterbi detection).
[0176]
Then, using the output value g ′ [i, j] and the ideal output value s ′ [i, j] obtained by subtracting the intersymbol interference from the fed back row, all allowed state transitions (in the trellis diagram) Branch metric B_kThe
[Equation 45]
B_k= (G ′ [i, j] −s ′ [i, j])²
Calculate as
[0177]
Next, a path metric is calculated for each of the four input paths leading to each state, and a path with the smallest path metric among the paths is left as a surviving path.
[0178]
Thereafter, when the paths are merged on the trellis diagram and the paths are determined, the detection output corresponding to each state transition in the determined path (the value in Table 2 in the case of the block shape in the uppermost column in FIG. 14) is detected data. Bits are output as output signals Sout.
[0179]
【Example】
Next, in order to confirm the effect of the present invention, the result of calculating the bit error rate by simulation is shown in FIG. 15 and FIG.
[0180]
In FIG. 15 and FIG. 16, “BbB” indicates a case where a bit-by-bit method not using Viterbi detection is used as a playback method, and “DF-VA-16” indicates 16 states as a playback method. The case of using the judgment feedback Viterbi detection process (FIG. 10B) is shown, and “DF_-“VA-8” indicates a case where 8-state decision feedback Viterbi detection processing (FIG. 13B) is used as a reproduction method.
[0181]
In addition, as the displacement direction, (r, d) indicates the case of the lower right, and (l, u) indicates the case of the upper left.
[0182]
Further, the direction of decision feedback is from top to bottom, and the Viterbi detection direction is from left to right.
[0183]
As the spread function, the spread function of Expression (5) is used, and the level of the input signal Sin is 200 for white (Sw) corresponding to bit “1” and black (Sb) corresponding to bit “0”. 50. Further, the noise is Gaussian noise with an average value of 0, and the noise after applying intersymbol interference to the liquid crystal spatial light modulator 1 having a black and white random pattern of 480 × 640 pixels by an optical transfer function of 4 rows and 4 columns. And quantizing with 8 bits from 0 to 256.
[0184]
Furthermore, the S / N (signal / noise) ratio is
[Equation 46]

Define in. Here, σ is the standard deviation of noise.
[0185]
Further, the simulation result shown in FIG. 15A shows the case where the positional deviation is (δx, δy) = (0.1, 0.1), and the simulation result shown in FIG. Is (δx, δy) = (0.2, 0.2), and the simulation result shown in FIG. 16 shows that the positional deviation is (δx, δy) = (0.3, 0.3). It shows about a case.
[0186]
As can be seen from FIGS. 15 and 16, in the Viterbi detection using the decision feedback method and the reproduction by the bit-by-bit without using this, the bit error rate is clearly lower when the decision feedback method is used. Moreover, this effect is more remarkable as the S / N ratio is higher.
[0187]
Further, comparing the case where the 16-state determination feedback Viterbi detection process is used with the case where the 8-state determination feedback Viterbi detection process is used, it can be seen that there is almost no difference. Therefore, there is almost no performance degradation due to the reduction of the number of states from 16 states to 8 states using the reduced transfer function of 3 rows and 3 columns, and in this respect, the effect of using the 8-state decision feedback method is great.
[0188]
Further, in FIGS. 15 and 16, the graphs denoted by (r, d) are cases where the misalignment direction corresponds to the judgment feedback direction and the state block shape, and the graphs denoted by (l, u). Indicates a case where the misalignment direction does not correspond to the determination feedback direction and the state block shape, and the determination direction is opposite. When these are compared, as the misalignment increases, (l, u) is indicated. In this graph, the bit error rate deteriorates more.
[0189]
On the other hand, the bit error rate can be improved up to the graph indicated by (r, d) by changing the direction of decision feedback and the state block shape in accordance with the direction of displacement.
[0190]
As described above, according to the operation of the optical memory system S of the embodiment, the determination set based on the relative displacement direction between the pixel of the liquid crystal spatial light modulator 1 and the corresponding pixel of the CCD element 6. Since the determination feedback Viterbi detection process is performed using the feedback direction and the state block shape, it is possible to prevent the reproduction accuracy from being deteriorated due to the difference between the misalignment direction and the determination feedback direction and the Viterbi detection direction.
[0191]
In addition, the reference light L with respect to the two-dimensional information recorded in the optical memory 3 by driving the liquid crystal spatial light modulator 1 based on the known signal Sr._rIs detected based on the optical transfer function calculated based on the CCD output signal Sp obtained by receiving the reproduction signal light Lp obtained by irradiating the light and the known signal Sr. It is possible to accurately set the determination feedback direction and the state block shape by detecting the shift direction.
[0192]
Therefore, two-dimensional information corresponding to the input signal Sin can be recorded on the optical memory 3, and the input signal Sin can be accurately reproduced.
[0193]
In the above-described embodiment, the known signal Sr corresponding to the known M array is used as the known signal Sr. However, for example, only one pixel of the liquid crystal spatial light modulator 1 is set to “white”. The optical system transfer function may be set using a so-called impulse pattern in which the surrounding pixels are “black”, or the optical system transfer function may be set using a known general random pattern. .
[0194]
【The invention's effect】
As described above, according to the first aspect of the present invention, the determination feedback direction and the state block shape set based on the relative displacement direction between the modulation pixel and the corresponding light receiving pixel are obtained. Since the determination feedback Viterbi detection process is performed by using this, it is possible to prevent the reproduction accuracy from being deteriorated due to the difference between the shift direction, the determination feedback direction, and the Viterbi detection direction.
[0195]
Therefore, the input signal can be reproduced more accurately based on the two-dimensional information recorded on the recording medium.
[0196]
According to the invention described in claim 2, in addition to the effect of the invention described in claim 1, the shift direction is recorded on the recording medium by driving the spatial light modulation means based on the known input signal. Since it is detected based on the transfer function calculated based on the output signal obtained by receiving the reflected light obtained by irradiating the reproduction reference light to the two-dimensional information and the known input signal, it is accurate. In addition, it is possible to accurately set the judgment feedback direction and the state block shape by detecting the shift direction.
[0197]
According to the invention described in claim 3, in addition to the effect of the invention described in claim 1 or 2, the input signal can be accurately reproduced.
[0198]
According to the invention described in claim 4, in addition to the effect of the invention described in claim 1 or 2, it is possible to record two-dimensional information corresponding to the input signal and accurately reproduce the input signal. The input signal can be reproduced.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a schematic configuration of an optical memory system according to an embodiment.
FIG. 2 is a flowchart showing an information recording / reproducing process of the embodiment, (a) is a flowchart showing the information recording process, and (b) is a flowchart showing the information reproducing process.
FIG. 3 is a block diagram showing a configuration of an M array generator.
FIG. 4 is an explanatory diagram showing positional deviation between a liquid crystal spatial light modulator and a CCD element.
FIG. 5 is a flowchart showing an operation of detecting a misalignment direction.
6 is a diagram for explaining one-dimensional Viterbi detection when the intersymbol interference length is 2. FIG.
FIG. 7 is a state transition diagram of Viterbi detection when the intersymbol interference length is 2.
8 is a trellis diagram when the intersymbol interference length is 2. FIG.
FIG. 9 is a diagram for explaining survival paths and merging.
10A and 10B are diagrams illustrating state definitions, in which FIG. 10A illustrates a state definition in Viterbi detection, and FIG. 10B illustrates a state definition in determination feedback Viterbi detection;
FIG. 11 is a diagram illustrating a transfer function when there is a displacement.
FIG. 12 is a diagram illustrating a state definition in determination feedback Viterbi detection using a transfer function of 4 rows and 4 columns.
FIGS. 13A and 13B are diagrams showing determination feedback Viterbi detection in the case where a positional deviation and a spread function coexist, FIG. 13A is a diagram showing a symbol alphabet, and FIG.
FIG. 14 is a diagram illustrating a relationship between a misalignment direction, a determination feedback direction, and a state block shape.
15A and 15B are diagrams illustrating the effects of the embodiment, in which FIG. 15A is a diagram illustrating a case where the positional deviation is (δx, δy) = (0.1, 0.1), and FIG. 15B is a positional deviation. Is a diagram showing a case where (δx, δy) = (0.2, 0.2).
FIG. 16 is a diagram showing the effect of the embodiment when the positional deviation is (δx, δy) = (0.3, 0.3).
[Explanation of symbols]
1 ... Liquid crystal spatial light modulator
2, 5 ... Lens
3 ... Optical memory
4 ... Yθ stage
6 ... CCD element
7, 18 ... Mirror
8 ... Anamorphic optical system
9 ... CPU
10 ... Viterbi detector
11 ... Memory
12 ... Known signal generator
13 ... Switch
14, 15 ... Shutter
16 ... Argon laser
17 ... Half mirror
22 ... Delay circuit
23 ... Exclusive OR circuit
S: Optical memory system
Scl1, Scl2, Sci, Scw, Scd ... control signals
Sz: Deviation direction signal
Sp: CCD output signal
Sout ... Output signal
Scs: Drive signal
Sin ... Input signal
Sr: Known signal
La, Lb ... Argon laser light
Ls ... Signal light
Lp: Reproduction signal light
Lr ... Reference light

Claims (4)

Two-dimensional information recorded on the recording medium is irradiated by spatially modulating the recording signal light based on the input signal and irradiating the recording medium based on the input signal by the spatial light modulation means including a plurality of modulation pixels. Receiving means for receiving the reflected light of the reproduced reference light from the recording medium and generating an output signal corresponding to the reflected light, comprising a plurality of light receiving pixels respectively corresponding to the plurality of modulation pixels A Viterbi detection device that detects the input signal by performing a determination feedback Viterbi detection process on the output signal from the light receiving means,
The determination feedback direction and state in the determination feedback Viterbi detection process based on the pre-detected relative position shift direction between the one modulation pixel and the light receiving pixel corresponding to the one modulation pixel Setting means for setting the block shape;
Detection means for performing determination feedback Viterbi detection processing on the output signal using the set determination feedback direction and state block shape, and detecting the input signal;
A Viterbi detection device comprising: The Viterbi detection device according to claim 1,
The deviation direction is irradiated with the reproduction reference light on the two-dimensional information recorded on the recording medium by driving the spatial light modulator based on a known input signal that is a preset input signal. A transfer function calculated based on the output signal obtained by receiving the reflected light and the known input signal, the spatial light modulation means, the recording medium, and the reproduction reference light. A Viterbi detector, which is detected in advance based on a transfer function of an optical system including at least a reproduction reference light irradiating unit for irradiating and a light receiving unit. Viterbi detection device according to claim 1 or 2,
A recording medium on which the two-dimensional information is recorded;
The reproduction reference light irradiation means;
The light receiving means;
An optical information reproducing apparatus comprising: Viterbi detection device according to claim 1 or 2,
The spatial light modulation means;
Recording signal light emitting means for emitting the recording signal light;
Recording reference light emitting means for emitting recording reference light to a position on the recording medium on which the two-dimensional information is to be recorded at the time of recording the two-dimensional information;
The recording medium;
Reproduction reference light irradiating means for irradiating the recording medium with the reproduction reference light when reproducing the input signal;
The light receiving means;
An optical information recording / reproducing apparatus comprising:

Images