JP3866081B2

JP3866081B2 - Adaptive equalization apparatus and method

Info

Publication number: JP3866081B2
Application number: JP2001336357A
Authority: JP
Inventors: 武史山田; 繁冨里; 正松本
Original assignee: NTT Docomo Inc
Current assignee: NTT Docomo Inc
Priority date: 2000-11-01
Filing date: 2001-11-01
Publication date: 2007-01-10
Anticipated expiration: 2021-11-01
Also published as: JP2002374191A

Description

【０００１】
【発明の属する技術分野】
この発明は例えば移動通信で用いることができ、干渉を除去する適応等化装置及びその方法に関し、特に適応等化処理を効果的に行わせるためのシンボル同期タイミング生成手段及びその方法に係わるものである。
【０００２】
【従来の技術】
移動通信において干渉を除去する技術として適応等化器がある。適応等化器は、時間軸上での信号処理により、自分の信号ではあるが時間的に遅れてくる遅延波の間で起る符号間干渉を合成、除去するために従来から用いられている。このような適応等化器で用いる受信信号のシンボル同期タイミングの検出方法について従来例を以下に説明する。従来の方法では、最も早く到着するパスである先行波にシンボルタイミングを同期させていた。この場合の構成例を図１に示す。図に示していないが送信側では受信側で送信シンボルパターンが既知であるような長い同期ワード信号をまず送信し、その後、送信しようとする情報内容を示すデータを送信する。受信側では図に示していないが送信側からの受信電波を増幅、復調してベースバンド信号とされ、かつデジタル系列とされ、このデジタル系列の受信信号が入力端子１１に入力されこの受信信号と、同期ワード生成器１２からの同期ワード信号との相関を相関器１３で求めることにより伝搬路応答を推定する。この伝搬路応答、つまり相関器１３の出力信号は例えば図２に示すように変化し、しきい値信号設定器１４からのしきい値信号Ｔｈｓと相関出力信号Ｓｉｇｃとをシンボル同期タイミング生成器１５で比較し、相関出力信号Ｓｉｇｃがしきい値信号Ｔｈｓを超える最初のタイミングｔ１を検出し、先行波のシンボル同期タイミングを得る。このシンボル同期タイミング信号で入力端子１１からのデジタル化された受信信号を、サンプラ１６において、サンプリングし、その出力であるサンプリング信号を適応等化器１７へ入力し、適応等化器１７は入力されたサンプリング信号を適応等化処理して、判定シンボル信号を出力端子１８に出力する。
【０００３】
【発明が解決しようとする課題】
この従来の先行波にシンボル同期タイミングを合わせる方法では、等化範囲が時間的に限られている適応等化器１７を用いる場合、その等化範囲外に電力が大きい相関出力信号が存在することがあり、この場合十分な等化が行えず、その結果特性が劣化するという問題があった。
【０００４】
【課題を解決するための手段】
この発明では、少くとも１シンボル長以上、好ましくは適応等化器の等化範囲にほぼ相当する長さより以上の一定時間幅内で相関器から得られた相関出力信号の電力和を、その一定時間幅を時間的にずらしながら算出し、この発明の１面ではこの電力和が最大となるタイミングを検出し、このタイミングに基づいてシンボル同期タイミングを生成する。この発明の他の面では上記電力和が最初にしきい値を超えたタイミングを検出し、このタイミングに基づいてシンボル同期タイミングを生成する。
【０００５】
この構成により、適応等化器の等化範囲内のパス電力和を最大とすることができるため、効果的に等化処理を行うことができる。この結果、良好な受信特性が期待できる。
【０００６】
【発明の実施の形態】
第１実施例
この発明の第１実施例の構成を図３に図１と対応する部分に同一参照番号を付けて示す。相関器１３では、同期ワード生成器１２で生成された同期ワード信号と入力端子１１からの受信信号との相関を出力する。この時、同期ワード信号としては、従来技術においても用いられているが、自己相関が高くかつ干渉波との相互相関は低いようなシンボルパターン、例えばＰＮ符号列を用いることにより、希望信号のシンボルタイミングを検出でき、かつ干渉波の影響を除去することができる。
【０００７】
一方、この実施例では電力測定部２２において、一定時間幅内の相関出力Ｓｉｇｃの和が、この時間幅を時間的に順次ずらされながら求められる。例えば窓信号生成器２１を設け、窓信号生成器２１から適応等化器１７の等化範囲と同じ時間幅の窓信号を出力する。例えば、等化範囲が４シンボル遅延の場合には、時間幅４Ｔ（Ｔはシンボル周期）の間で１となり、それ以外では０となるような窓信号を生成する。この窓信号と相関器１３よりの相関出力信号Ｓｉｇｃとを電力測定部２２に入力する。電力測定部２２では窓信号の範囲内における相関出力信号Ｓｉｇｃの電力和を、窓信号の時間的位置をずらして順次計算する。この様子を図４に示す。入力端子１１よりのベースバンド受信信号は、例えばシンボル周期Ｔの１／１６間隔でサンプリングされてデジタル化された系列であり、同期ワード生成器１２からの同期ワード信号も同一周期のデジタル化された系列であり、従って、相関出力信号Ｓｉｇｃも同一周期のデジタル系列である。図４Ａにこの相関出力信号Ｓｉｇｃの例を簡便的に連続波形として示した。図４Ｂに示すように等化範囲時間、この例では４Ｔの幅だけレベルが１、他は０の窓信号＃１を、相関出力信号のサンプル周期、この例ではＴ／１６だけ順次ずらして＃２，＃３，…とし、これら窓信号＃１，＃２，＃３，…のそれぞれがレベル１の間における相関出力信号Ｓｉｇｃの和をそれぞれ求める。つまり窓信号＃１に対しては、相関出力信号中の図４Ａに斜線を施した部分の和を求める。
【０００８】
電力測定部２２の具体例としては例えば一定時間幅が４Ｔで受信信号のサンプル周期がＴ／１６の前記例の場合は図５Ａに示すように４×１６個のシフト段をもつシフトレジスタ１１１の１端に相関出力信号Ｓｉｇｃを入力し、このシフトレジスタ１１１をＴ／１６のクロックＣｓでシフト動作させその入力と各シフト段の出力との総和を加算器１１２でとれば、Ｔ／１６ずつ順次ずれた窓信号内の電力和がＴ／１６ごとに得られる。あるいは図５Ｂに示すように相関器１３よりのＴ／１６ごとの相関出力信号をメモリ１１３に書き込み、例えば同期ワード信号の終と対応する相関出力信号の書き込みが終了すると読出しアドレス発生器１１４から４Ｔ分データに対応する４×１６のアドレスを発生させてメモリ１１３から記憶データを読出し、その読出した４×１６のデータ（相関出力信号）を加算器１１５で加算し、次に１アドレス分ずらして、読出しアドレスを４×１６発生させてメモリ１１３の記憶データを読出し加算器１１５で加算することを１アドレス分ずらしながら繰返してもよい。この場合、加算器１１５でその４×１６個のデータの加算値に対し、その最初のデータを減算すると共に次のアドレスのデータを加算することを順次行えばよい。このメモリ１１３への書込みを終了する前に読出しを並列的に行ってもよい。
【０００９】
なお、窓信号をずらす量は相関出力信号Ｓｉｇｃのサンプル周期に限らず、その整数倍、例えば２〜４倍、あるいはそれより大きくてもよい。しかし求めるシンボル同期タイミングの精度は窓信号のずれ量が小さい程高くなる。なお図３において、入力端子１１に入力されるデジタル系列の受信信号のデジタル化の際に用いたサンプリングクロックがサンプリング発生器１９から同期ワード生成器１２、窓信号生成器２１、電力測定器２２に入力され、相互のタイミングがとられる。
【００１０】
ただし電力和を図５Ａを参照し述べたようにシフトレジスタを用いて求める場合は窓信号発生器２１は不用である。
シンボル同期タイミング生成器２３では、電力測定部２２が出力する電力和の中で、最も電力が大きなものが得られるタイミングを用いて、シンボル同期タイミングを生成する。図４に示したような窓信号であれば、電力和が最大となる窓信号の立ち上がりタイミングをシンボル同期タイミングとすれば良い。
上述において窓信号の時間長（幅）、つまり電力測定部２２で電力和を求める一定時間幅は適応等化器１７の等化範囲と同じ時間的範囲が好ましいとしたが、窓信号の時間長は短かくても１シンボル周期程度以上とし、長い方はいくらでも長くしてもよいが、実用的には前記等化範囲の時間的範囲の２倍程度が考えられる。また電力測定部２２で求める電力和の数は最低は２個であり、最高は、同期ワード信号の発生から終了まで順次ずらした窓信号について電力和を求めればよい。しかし順次得られる電力和があるしきい値以下になり、この状態が所定回数連続すると、電力和を求めることを中止し、それまでに得られた電力和から最大のものを選択するようにしてもよい。なお電力和を求め始めた際も電力和がしきい値以上のものが得られるようになってからの電力和を有効なものとし、それ以後のものから最大のものを選択するようにしてもよい。
【００１１】
サンプラ１６では、シンボル同期タイミング信号を用いて受信信号をサンプリングする。適応等化器１７では、サンプラ１６からのサンプリング信号を入力とし、等化処理を行う。適応等化器１７の構成例を図６に示す。この構成はＭＬＳＥ（Maximum Likelihood Sequence Estimation）型と呼ばれ、最尤系列推定を行って受信信号を等化するものである（例えばJ.G. Proakis ：“Digital Communications 3^rd Edition，”McGraw-Hill pp. 249-254(1995)参照）。この適応等化器では、まず受信信号のシンボル系列候補信号を生成する。送信時に帯域制限や符号化のための処理が行われている場合には、それらを考慮してシンボル系列候補信号を生成する。次にシンボル系列候補信号に、パラメータ推定器３１が出力する適応等化手段重み付け係数を、レプリカ生成器３２で複素乗算し、その乗算結果を合成することにより受信信号のレプリカを生成する。レプリカ生成器３２の構成例を図７に示す、シンボル系列候補信号にそれぞれ適応等化器重み付け係数をそれぞれ複素乗算する複数の複素乗算器２４とこれら乗算結果を加算する複素加算器２５から構成されている。また、各複素乗算器２４に入力されるシンボル系列候補信号をタップ入力信号としてパラメータ推定器３１へ出力する。
【００１２】
このレプリカ信号と適応等化器の入力信号（サンプラ１６の出力信号）を加算器３３で差をとることにより推定誤差信号を生成する。通常はこの推定誤差信号が二乗器３４で二乗され、この二乗値が、最尤系列推定器（ＭＬＳＥ）３５で最尤系列推定を行うためのビタビアルゴリズムにおけるブランチメトリックとして用いられる。最後にビタビアルゴリズムで求められる最も確からしいシンボルパターンを判定シンボルとして出力端子１８に出力し、受信信号の復調が行われる。一方、パラメータ推定器３１では推定誤差信号とタップ入力信号を用いて、適応等化器重み付け係数を更新する。この更新には適応アルゴリズムを用いればよく、ＬＭＳ（Least Mean Square）アルゴリズムやＲＬＳ（Recursive Least Square）アルゴリズムを用いることができる（例えばJ.G. Proakis ：“Digital Communications 3^rd Edition，”McGraw-Hill pp. 639-644，pp. 654-660(1995)参照）。また、適応等化器重み付け係数の初期収束のためには、ＭＬＳＥ処理部３５よりのシンボル系列候補信号にかえてトレーニング信号と呼ばれる受信側で既知であるシンボルパターンを端子３６より、切替スイッチ３７を介してレプリカ生成器３２に入力して用いることができる。具体的には最初にシンボル系列候補信号の代わりにトレーニング信号を用いて、適応等化器重み付け係数を収束させ、その後スイッチ３７をＭＬＳＥ３５側に切替えてシンボル系列候補信号を用いて、シンボルの復調を行う。適応等化器については例えばJ.G. Proakis ：“Digital Communications 3^rd Edition，”McGraw-Hill pp. 636-676(1995)を参照されたい。
【００１３】
以上説明したように、この第１実施例ではシンボル同期タイミング生成器２３において一定時間幅（窓信号）内の電力が最大となるような同期タイミングを生成しているため、適応等化器１７において合成できるパス電力が最大となる。この結果、従来のシンボル同期タイミング生成器と比較して、希望信号電力対雑音電力比の面で有利となり、受信特性が改善できる。なお、シンボル同期タイミング生成器２３は必ずしも１シンボル周期の信号に限らない、例えば、適応等化器１７が１／２シンボル周期間隔のトランスバ−サルフィルタを持つ線形等化器の場合は１／２シンボル周期でサンプラ１６におけるサンプリングを行う。
第２実施例
この発明の第２実施例の構成と第１実施例との差異は、電力測定部２２における相関信号の電力測定方法の違いにある。適応等化器におけるレプリカ生成器３２では、図７に示したように、ある時間間隔でタップ入力信号を構成しているため、この時間間隔でサンプリングした受信信号に対して等化を行うこととなる。このため、レプリカ生成器３２のタップ間隔を考慮してシンボル同期タイミングを決定することにより、より効果的にパスダイバーシチ効果を得ることができる。適応等化器におけるタップ間隔としては、１シンボル間隔のものが多いが、この場合には、電力測定部２２における電力測定方法は以下のようになる。
【００１４】
窓信号の範囲内の相関出力信号をシンボル間隔毎にサンプルし、これらのサンプリング信号の電力和を計算する。この動作を窓信号の時間的位置をずらしながら繰り返し行う。例えば図４において窓信号の時間幅が４Ｔであれば、図４Ｂ中に示すように窓信号＃１，＃２,＃３,…における両端エッジと、３つの破線で示したタイミングにおける相関出力信号の各５つのサンプル値の和がそれぞれ計算される。図５Ａに示したシフトレジスタを用いて相関出力信号Ｓｉｇｃの電力和を求める場合は、図８に示すように、図５Ａと同一様に４×１６段シフトレジスタ１１１を用い、サンプリングクロックＣｓ（図３中のサンプリングクロック発生器１９の出力）によりシフト動作させて、相関出力信号Ｓｉｇｃを取込むが、入力信号と各１６シフト段ごとの出力を加算器１１２で加算する。この加算値（電力和）がＴ／１６ごとに出力されることになる。
【００１５】
シンボル同期タイミング生成器２３では、これらの電力和の中で最大となるタイミングを検出する。この方法により受信信号の到来パスがシンボル間隔毎だけでなく不規則な間隔で分布している環境においても、パスダイバーシチ効果を最大限得ることができるタイミングを検出できる。この第２実施例における相関出力信号の和をとる間隔は図３中のサンプラ１６によるサンプリング間隔にすればよく、必ずしも１シンボル間隔ではない。
第３実施例
この発明の第３実施例の構成を図９に図３と対応する部分に同一参照番号を付けて示す。この実施例では適応等化手段として、アダプティブアレーアンテナ処理と適応等化処理とを併用する時空等化構成としている。
【００１６】
アダプティブアレーアンテナは、自分の信号である希望波の方向にアンテナ利得が相対的に高いビームを向け、他ユーザの信号などの干渉波の方向に、相対的に利得が著しく低いヌルを向けるような指向特性ビームパターンを信号状況の変化に応じて適応的に生成するものである。このため同一の周波数（チャネル）を用いる干渉波、すなわち同一チャネル干渉を除去するために有効な技術である。このアダプティブアレーアンテナと適応等化器を融合させたものが時空等化器である。アダプティブアレーアンテナについては例えばR.T. Compton, Jr.：“Adaptive Antennas-Concepts and Performance,”Prentice-Hall, Englewood Cliffs(1988)を参照されたい。
【００１７】
時空等化器では、適応等化器の等化範囲を超えるため等化できないような長遅延波や同一チャネル干渉をアダプティブアレーアンテナにより除去し、適応等化器は等化範囲内にある遅延波について等化処理を行う。このように空間領域における等化処理を追加することにより、時間領域における信号処理のみでは演算量の増大からハードウェア化が困難となるような伝搬環境においても、合理的なハードウェア規模で等化処理が可能となる。
図９において、アダプティブアレーアンテナの各素子からの受信信号は入力端子１１−１〜１１−Ｎにベースバンド信号（デジタル化された）として入力され、これらＮ系統の受信信号はサンプラ１６−１〜１６−Ｎでそれぞれシンボル同期タイミング信号によりサンプリングされ、これら生成されるＮ系統のサンプリング信号は適応等化器４１に入力され、等化処理が行われる。この実施形態における適応等化器４１の構成例を図１０に示す。図６に示した適応等化器と異なる点は、アダプティブアレーアンテナ処理合成部５１を備えている点である。端子５２−１〜５２−ＮからのＮ系統のサンプリング信号はアダプティブアレーアンテナ処理合成部５１により合成され、この合成信号に対して等化処理が行われる。アダプティブアレーアンテナ処理合成部５１の構成例を図１１に示す。端子５２−１〜５２−ＮよりのＮ系統のサンプリング信号とＮ個のアダプティブアレーアンテナ（ＡＡＡ）重み付け係数とをそれぞれ複素乗算するＮ個の複素乗算器５３−１〜５３−Ｎと、これらＮ個の複素乗算器の出力が複素加算される複素加算器５４から構成されている。またアダプティブアレーアンテナ処理合成部５１で用いる重み係数はパラメータ推定器５５で求めており、推定誤差信号とアダプティブアレーアンテナ処理合成部５１における複素乗算器５３−１〜５３−Ｎの入力である第２タップ入力信号を用いて計算することができる。時空等化器について例えばRyuji Kohno：“Spatial and Temporal Communication Theory Using Adaptive Antenna Array”，IEEE Personal Communications, pp. 28-35, Feb. 1998，A.J. Paulraj and B.C. Ng, “Space-Time Modems for Wireless Personal Communications”，IEEE Personal Communications, pp. 36-48, Feb. 1998を参照されたい。
【００１８】
この実施例では、受信信号はＮ系統あるが、どれか１つの系統の受信信号を用いて相関器１３で同期ワード信号との相関を求めれば十分である。しかし、複数の受信信号と同期ワード信号との各相関を求め、これら相関出力信号を例えば加算、平均することにより、より精度良く相関出力信号を求めることが可能となる。
第４実施例
この発明の第４実施例の構成を図１２に示す。電力比較器６１及びシンボル同期タイミング生成器６２以外の各部は、図３に示した第１実施例及び第２実施例と同様である。電力比較器６１では、電力測定部２２で求められた電力和を比較して、大きいものから順にＬ個のタイミング点（Ｌは１以上の整数）選択する。Ｌが１の場合、この実施例は第１実施例と同一のものとなる。このＬの値の最大は、電力測定部２２で求めた電力和の数であるが、ハードウェア規模を考慮して設定すればよい。シンボル同期タイミング生成器６２では、選択されたＬ個の各電力和のタイミングに応じてシンボル同期タイミング信号を各々出力する。この実施例では、これらＬ個のシンボル同期タイミング信号に対応するサンプラ１６−１〜１６−Ｌを備えており、各々のサンプラ１６−１〜１６−Ｌでは入力端子１１からの受信信号をシンボル同期タイミング生成器６２よりのＬ個のシンボル同期タイミング信号によりそれぞれサンプリングする。適応等化器６３では、これらのＬ個のサンプリング信号を用いて等化処理を行う。
【００１９】
図１３にこの場合の適応等化器６３の構成例を示す。サンプラ１６−１〜１６−Ｌよりのサンプリング信号は端子６４−１〜６４−Ｌからそれぞれ推定誤差出力部６５−１〜６５−Ｌに入力される。推定誤差出力部６５−１に示すように図６に示したと同様の構成のパラメータ推定器３１−１、レプリカ生成器３２−１、加算器３３−１、二乗器３４−１を備え、他の推定誤差出力部６５−２〜６５−Ｌも各々のシンボル同期タイミングに対応するパラメータ推定部、加算器、二乗器及びレプリカ生成器を備えている。この例では、レプリカ生成及びパラメータ推定はＬ個のシンボル同期タイミング毎に行い、各々の加算器の推定誤差信号の二乗値を全て加算器６６で足し合わせてブランチメトリックとして、ＭＬＳＥ処理部３５でＭＬＳＥ処理を行っている。ＭＬＳＥ処理部３５の代わりに他の適応等化処理を行ってもよい。
【００２０】
また図１４に示すように、各推定誤差出力部６５−１〜６５−ＬにそれぞれＭＬＳＥ処理部３５−１〜３５−Ｌをも用意し、それぞれ各シンボル同期タイミングごとに独立に等化処理を行った後、品質対応処理部６７において各ＭＬＳＥ処理部３５−１〜３５−Ｌの出力に対して、通信品質に応じた処理を行って最終的な出力を得る、といった構成とすることも可能である。この場合通信品質に応じた処理としては、例えば得られた等化器出力（ＭＬＳＥ処理出力）を用いて、図中に破線で示すようにこれらの多数決判定を多数決判定器６９で行う方法や、ＭＬＳＥ処理部３５−１〜３５−Ｌがそれぞれ出力した時のその出力と対応する加算器３３−１〜３３−Ｌの各誤差信号の二乗の大きさに応じた重みを、ＭＬＳＥ処理部３５−１〜３５−Ｌの出力にそれぞれ付けて加算器６８で加算し、その加算信号をしきい値回路６８ａで２値化して出力する。この後者については、推定誤差がより小さいものを通信品質が良いとして大きな重み付けを行えば良い。
【００２１】
また第３実施例のようにアダプティブアレーアンテナ処理合成部を用いることも可能である。この場合の構成例を図１５に示す。入力端子１１−１〜１１−Ｎよりの各受信信号は、シンボル同期タイミング生成器６２からの第１シンボル同期タイミング信号によりサンプラ１−１〜１−Ｎでそれぞれサンプリングされ、第２シンボル同期タイミング信号によりサンプラ２−１〜２−Ｎでそれぞれサンプリングされ、…、第Ｎシンボル同期タイミング信号によりサンプラＬ−１〜Ｌ−Ｎでそれぞれサンプリングされ、適応等化器７１に入力される。
【００２２】
適応等化器７１は図１６に示すようにサンプラ１−１〜１−Ｎの出力が推定誤差出力部７２−１に入力され、推定誤差出力部７２−１は図１０に示した構成と同様に、レプリカ生成器３２−１、誤差算出用加算器３３−１、誤差二乗器３４−１、アダプティブアレーアンテナ処理合成部５１−１、パラメータ推定器５５−１より構成される。サンプラ２−１〜２−Ｎの出力〜サンプラＬ−１〜Ｌ−Ｎの出力がそれぞれ推定誤差出力部７２−２〜７２−Ｌに入力される。推定誤差出力部７２−２〜７２−Ｌはそれぞれ推定誤差出力部７２−１と同様な構成とされている。推定誤差出力部７２−１〜７２−Ｌよりの二乗誤差信号は加算器６６で加算されてＭＬＳＥ処理部３５に入力される。
【００２３】
アダプティブアレーアンテナを用いることで、適応等化器の等化範囲外のパスを除去することができ、受信特性をさらに向上させることが可能となる。
以上の説明では、必要な部分をＬ系統用意することを前提にしてきたが、各部を時分割的に利用することによりＬよりも小さい値で実現することができ、これによりハードウェア規模の低減が図れる。
以上説明したように、この実施例では複数のシンボル同期タイミングを用いることで、より多くのパスを合成できるようになるため、受信特性が改善される。
第５実施例
この発明の第５実施例では、図１２及び図１５における電力比較器６１において、第４実施例のようにシンボル同期タイミング生成時に、単純に電力和の大きい順に固定的にＬ個のタイミングを選択するのではなく、図１７Ａに示すようにしきい値設定器７０にしきい値Ｔｈを予め設定し、電力測定部２２からの電力和中の電力比較器６１でしきい値Ｔｈを超えるＱ個（Ｑは１＜Ｑ＜Ｌ＜Ｍを満たす整数）に基づき、-Ｑ個のシンボル同期タイミングを生成する。この操作により、通信状況に応じて適応的にＱの値が選択されることとなり、例えば電力和が上からＬ番目以内ではあるが、等化処理を行っても受信特性の改善効果が小さいような小さな電力和の窓信号から求めたシンボル同期タイミングに対する等化処理を省略することができ、この結果、等化処理全体の演算量を削減することが可能となる。つまりクレーム中でＬ個のタイミング選択はＬ個は所定値ではなく、しきい値以上の電力和の数のことも意味するものである。
【００２４】
また、選択されたＬ点のうち上位数点で十分な受信特性を発揮できる窓信号内の電力和が得られる場合においても、それ以下の電力和に基づくシンボル同期タイミングについての等化処理を省略することで、受信特性をほとんど劣化させることなく、等化処理における演算量の削減が可能となる。つまり、比較的大きなしきい値を超える電力和の数が予め決めた数（例えば１とか２など）あれば、そのしきい値を超えた電力和に基づくシンボル同期タイミングについてのみ等化処理をする。例えば図１７Ｂに示すように電力測定部２２からのＭ個の電力和はそれぞれ電力比較器６１ａ，６１ｂ，６１ｃへ供給されそれぞれしきい値Ｔｈａ，Ｔｈｂ，Ｔｈｃ（Ｔｈａ＞Ｔｈｂ＞Ｔｈｃ）と比較され、各電力比較器６１ａ，６１ｂ，６１ｃはそれぞれＭ個の入力電力和中のしきい値を越えたものが出力され、電力比較器６１ａからは受信状態が良好であっても１又は２個の電力和が出力される程度にしきい値Ｔｈａが設定され、電力比較器６１ｂからは３乃至４個が出力され、比較器６１ｃでは５〜７個が出力されるように他のしきい値が選定される。電力比較器６１ｂの出力は禁止ゲート１２１へ供給され、禁止ゲート１２１は電力比較器６１ａよりしきい値を超える出力により禁止され、電力比較器６１ｃの出力は禁止ゲート１２２へ供給され、禁止ゲート１２２は電力比較器６１ｂよりしきい値を超える出力により禁止される。電力比較器６１ａ、禁止ゲート１２１，１２２の各出力はシンボル同期タイミング生成器６２に入力され、その入力された各電力和と対応したタイミングのシンボル同期タイミング信号を出力する。
【００２５】
この実施例では、Ｑの値が変動するため、レプリカ生成器やパラメータ推定器等を何系統用意するのかという問題が生じるが、ハードウェア規模に余裕のあるときには十分な数を用意し、Ｑの値が小さい場合は、このうちの一部を利用するという方法がある。この場合は、全てを動作させる場合と比較して、消費電力を減らせるといったメリットがある。また、ハードウェア規模を削減したい場合は、第４実施例で述べたように、レプリカ生成器やパラメータ推定器などを時分割で利用すれば良い。
第６実施例
この発明の第６実施例の機能構成を図１８に示す。この実施例ではシンボル同期タイミング生成器を２つ備えた構成となる。つまり図３に示した電力測定部２２からの電力和が最大となるタイミングにシンボル同期タイミング信号を生成するシンボル同期タイミング生成器２３の他に、図１に示した従来のシンボル同期タイミング生成器１５を設け、相関出力信号がしきい値信号設定器１４からのしきい値を超えるタイミング、つまり受信波の先行波のタイミングからシンボル同期タイミングを生成している。
【００２６】
これらシンボル同期タイミング生成器１５と２３からの各シンボル同期タイミング信号をタイミング選択器８１で切替えて、サンプラ１６へ供給する。この実施例では、各シンボル同期タイミングでのサンプリング信号に対する適応等化処理を例えば時分割的に行い、通信品質に応じて何れのシンボル同期タイミング信号で復調処理を行うか決定する。通信品質としては、例えば、各々のシンボル同期タイミング信号による、トレーニング信号区間の最後のシンボルにおける推定誤差信号を用い、これら誤差信号の大きさを比較し、推定誤差電力が小さくなるタイミングで復調処理を行うように、タイミング選択器８１を制御すればよい。
第７実施例
この発明の第７実施例の機能構成を、図３と対応する部分に同一参照番号を付けて図１９に示す。この実施例においては電力測定部９１で、図４Ｂに示したように順次ずれた窓信号＃１，＃２，＃３,…内の各相関出力信号の電力和を、順次出力してシンボル同期タイミング生成器９２に入力し、シンボル同期タイミング生成器９２では、入力された電力和が、しきい値信号設定器９３に設定されているしきい値信号を、最初に超えた時の窓信号のタイミングに基づいてシンボル同期タイミング信号を生成する。その他は図３に示した構成と同様である。この第７実施例においても、その相関出力信号の電力和は、サンプラ１６のサンプリング間隔でサンプルしたものの和を用いてもよい。
第８実施例
図２０Ａに示すように、電波の到来パスが異なる方向のいくつかの群Ｇ１，Ｇ２，…,ＧＰに分かれ、かつ図２０Ｂに示すようにこれらパスよりの受信信号Ｇ１′，Ｇ２′，…,ＧＰ′の遅延時間が大きく異なる伝搬路環境では、第１及び第２実施例（関連して第３，４，５，６実施例も含む）で示した構成では、適応等化器のタップ数が少ない場合、十分良好な受信特性が得られない場合がある。この実施例では、このような伝搬環境においても効果的に等化処理が行えるようにする。
【００２７】
この第８実施例の構成を図２１に、図３と対応する部分に同一参照番号に付加番号を付けて示す。この実施例では第２実施例で用いた構成に加え、Ｎ系統のアレーアンナテ受信信号を入力端子１１−１,…,１１−Ｎより入力とし、マルチビーム生成器１０１によりＰ系統のビームの受信信号を出力し、これらＰ系統ビーム受信信号がＰ系統の相関器１３−１，…,１３−Ｐへ出力されると共に、Ｐ系統のサンプラ１６−１，…,１６−Ｐへ出力され、相関器１３−１,…，１３−Ｐではその入力された各ビームの受信信号と同期ワード生成器１２からの同期ワード信号との相関がとられ、これら各相関出力信号は電力測定部２２−１，…,２２−Ｐにおいて窓信号生成器２１からの窓信号の範囲内における各相関出力信号の電力和がそれぞれ計算され、これらの計算はその窓信号の時間位置が順次ずらされて行われる。電力測定部２２−１，…,２２−Ｐよりの各順次出力される電力和はそれぞれシンボル同期タイミング生成器２３−１，…,２３−Ｐに入力され、各シンボル同期タイミング生成器２３−１，…,２３−Ｐにおいて、それぞれ電力和の中で最も電力が大きくなる窓信号のタイミングを用いて、シンボル同期タイミングを生成し、そのタイミングで、サンプラ１６−１，…,１６−Ｐの対応するビームの受信信号をサンプリングする。これらサンプラ１６−１，…,１６−Ｐの各サンプリング出力は適応等化器１０５へ供給される。
【００２８】
マルチビーム生成器１０１の構成例を図２２に示す。図１１に示したアダプティブアレーアンテナと同様に、各アンテナ受信入力端子１１−１，…,１１−Ｎよりの入力信号に対し、乗算器１０２−１,…,１０２−Ｎでそれぞれ重み付け係数を乗じ、これら乗算結果を加算器１０３−１で足し合わせることでマルチビームの１つのビームの出力信号を得る。図１１に示したアダプティブアレーアンナテの信号の到来状況に応じて適応的に重み付け係数を更新することにより、１つのビーム出力信号を出力するのに対して、マルチビーム生成器１０１では信号の到来状況に関わらず予め求めた重み付け係数を用いて固定方向の１つのビームの受信信号を出力し、かつそのような固定方向のビーム形成器１０４−１,…,１０４−Ｐの複数個を設ける点が異なる。このようにマルチビーム生成器１０１では、Ｐ組の重み付け係数を用いてＰ系統のビームをビーム形成器１０４−１,…，１０４−Ｐにより生成するが、それらのビームは、それぞれメインビーム方向が異なり、例えば図２３に示すように通常は全てのビームＧ１，…，ＧＰを用いることにより全ての電波到来方向をカバーできるように生成される。このような性質を持つビームは例えばＢｕｔｌｅｒＭａｔｒｉｘ回路を用いることで容易に生成できる。図２３はＢｕｔｌｅｒＭａｔｒｉｘ回路を用いて形成したマルチビームの例である（例えば、J. Ltva and T.K. Lo：“Digital Beamforming in Wireless Communications,”Artech House, Boston London pp. 22-34, (1996)参照）。
【００２９】
マルチビーム生成器１０１によって得られたＰ系統のマルチビーム受信信号Ｇ１′，…，ＧＰ′は、先に述べたようにそれぞれ相関器１３−１，…,１３−Ｐ、電力測定部２２−１，…,２２−Ｐを通して、シンボル同期タイミング生成器２３−１，…,２３−Ｐに入力され、Ｐ個のシンボル同期タイミング信号が生成される。これらＰ個のシンボル同期タイミング信号を用いてサンプラ１６−１，…,１６−Ｐにおいて、Ｐ系統のビーム受信信号Ｇ１′，…，ＧＰ′からＰ系統のサンプリング信号を生成する。つまり各シンボル同期タイミング生成器２３−１，…，２３−Ｐにおいては図２４に示すように図４を参照して説明したと同様に対応する電力測定部２２−１，…,２２−Ｐにおける各電力和が最も大となるタイミングがシンボル同期タイミングとされる。図２１では各ビーム系統ごとに相関器、電力測定部、シンボル同期タイミング生成器及びサンプラを用意しているが、これらの部を各々の系統において時分割で使用しても良い。
【００３０】
このようにして生成されたＰ系統のサンプリング信号は適応等化器１０５に入力され、等化処理が行われる。この実施例における適応等化器１０５の構成例を図２５に示し、図１０と対応する部分に同一参照番号を付けてある。サンプラ１６−１，…,１６−ＰからのＰ系統のサンプリング信号は線形合成器１０６において重み付けされた後合成され、この合成信号に対して等化処理が行われる。等化処理の内容については第３実施例と同様である。つまりこの場合の線形合成器１０６は図１０中のアダプティブアレイアンテナ処理合成部５１と対応する。線形合成器１０６に対する重み付け係数は固定としてもよい。
【００３１】
このように、マルチビームの各異なる方向のビーム毎にシンボル同期タイミングを検出することにより、図２４に示したように時間領域だけでなく空間領域も用いてシンボル同期タイミングを検出することが可能となる。図２０に示した到来電波のパス例のように、到来パス群Ｇ１，…，ＧＰ間で受信信号の遅延時間が大きく異なり、かつ空間的にも離れている場合において、受信信号を空間領域で分離し、分離した受信信号毎にシンボル同期タイミングを検出することにより、適応等化器１０５の等化範囲が小さい場合でも、十分良好な受信特性を得ることができる。
【００３２】
この図２１に示したマルチビーム生成器１０１を用い、各ビームごとに受信信号に対し、この発明によるシンボル同期を行う手法は先に述べた第１乃至第７実施例に対しても適用することができる。図１２に示したように和電力が大きいＬ個のタイミングを用いる場合は、図２１中の各サンプラ１６−１，…,１６−Ｐのそれぞれについて、Ｌ個のサンプラを設け、つまり、図１５中の入力端子１１−１，…,１１−Ｎに、マルチビーム生成器１０１からのＰ個の出力信号をそれぞれ供給したと同様な構成とすればよく、適応等化器１０５の構成は例えば図１６に示したものと同様な構成とすればよい。
第９実施例
図２６Ａに示すように、電波が複数パスＧ１，…，ＧＰより到来しているが、図２６Ｂに示すようにそのうち受信電力が比較的大きいパスの受信信号Ｇ１′が他のパスの受信信号Ｇ２′，…，ＧＰ′より時間的に離れて到来している伝搬路環境を考える。この場合、第１乃至８実施例で示した構成では電力の大きい１つのパスの受信信号Ｇ１′が窓信号内に入るようにシンボル同期タイミングを合わせることになり、他のパスの受信信号Ｇ２′，…，ＧＰ′は等化範囲Ｔ_AER外に存在することになる。この結果等化範囲Ｔ_AER内の電力は最大となるが、パスが１つしかないためパスダイバーシチ効果を十分に得られず、フェージング環境下では良好な受信特性が得られない場合がある。この実施例ではこのような伝搬環境においても良好な受信特性を得られるようにする。
【００３３】
この第９実施例の機能構成を図２７に図１２と対応する部分に同一参照番号を付けて示す。この実施例では第１乃至８実施例で用いた電力測定部２２の代わりに電力分散値測定部１３１を用いることを特徴としている。電力分散値測定部１３１は、窓信号生成器２１よりの窓信号としきい値信号設定器１３２より出力される、しきい値信号Ｐthを入力とし、窓信号内における１サンプル毎の相関出力信号電力を測定し、その電力平均値Ｐ_Nを平均値計算部１３１ａで計算し、また電力分散値σ²を分散値計算部１３１ｂで算出する。算出された電力平均値Ｐ_Nが上記しきい値信号Ｐthより大きいと、比較部１３１ｃで判定したものと対応する電力分散値σ²を、例えば分散値が小さい順にＫ個（Ｋは１以上の整数）を分散値比較部１３１ｄで選択出力する。Ｋの値はハードウェア規模を考慮して決定すれば良い。なお窓信号内の各相関出力信号のサンプル値をｘ（ｎ）、窓信号内のサンプル数をＮとすると分散値σ²はσ²＝（１／Ｎ）Σ_n-1 ^N（ｘ（ｎ）−Ｐ_N)²により求められる。
【００３４】
上記電力分散値測定部１３１によって出力された電力分散値σ²は、シンボル同期タイミング生成器１３３に入力され、これらＫ個の各電力分散値σ²と対応するタイミング、例えば対応する窓信号の立上りタイミングに基づきシンボル同期タイミング信号を出力する。例えば窓信号内の分散値が０となる時は、窓信号内の各サンプル点にパスが等電力で存在している状態を示しており、適応等化処理によりパスダイバーシチ効果を最大限得ることができる。以上の過程によりシンボル同期タイミング信号が生成され、サンプラ１６−１〜１６−Ｋにより、受信信号のサンプリング信号が生成され、これらサンプリング信号は適応等化器６３に入力され、等化処理が行われる。この実施例における適応等化器６３の構成は図１３、図１４に示したものと同様の構成を使用することができる。
【００３５】
このようにこの第９実施例を用いることで、合成信号電力は最大とは限らないが、等化範囲内に多くのパスが存在するシンボル同期タイミングを検出することができ、パスダイバーシチ効果により、フェージング環境下においても良好な受信特性を期待できる。
第１０実施例
この第１０実施例は図２８に示すように、図１８に示した第６実施例に図２７に示した電力分散値測定部１３１を用いる構成を加えたものである。第６実施例では２つであったシンボル同期タイミング生成器１５及び２３（図１２中の６２）の他に、３つ目のシンボル同期タイミング生成器１３３を備える。これらシンボル同期タイミング生成器１５，６２及び１３３からの各シンボル同期タイミング信号をタイミング選択器８１で切り替えて、サンプラ１６−１〜１６−Ｌ（又は１６−Ｋ）へ供給する。サンプラとしてはＬとＫの大きい方の数だけ用意する。
【００３６】
第６実施例と同様に、各同期タイミングで適応等化処理を例えば時分割的に行い、通信品質を基準に何れのシンボル同期タイミング信号をとるかをタイミング選択器８１を制御して決定する。
この結果、より伝搬環境に適したシンボル同期タイミングを検出することが可能となり、受信特性の向上が期待できる。なお図２８において、電力測定器２２、電力比較器６１及びシンボル同期タイミング生成器６２と、しきい値信号設定器１４及びシンボル同期タイミング生成器１５との一方を省略してもよい。
【００３７】
これら第９実施例、第１０実施例により、電波が複数パスで到来し、しかも、そのうち受信電力が比較的大きいパスが他のパスより時間的に離れて到来している伝搬路環境においても、適応等化器においてパスダイバーシチ効果をより獲得できるシンボル同期タイミングを検出することが可能となり、受信特性を改善できる。
第７実施例乃至第９実施例のそれぞれに対しても第２実施例を適用し、電力測定部９１又は電力分散測定部１３１において、窓信号の範囲内の相関出力信号を、サンプラ１６又は１６−１〜１６−Ｐのサンプリング間隔毎にサンプルし、これらサンプリング信号の電力和又は電力分散値を計算してもよい。またこの第６実施例、第７実施例、第９実施例及び第１０実施例に対しても、第３実施例と同様に適応等化器１７をアダプティブアレーアンテナ処理と適応等化処理とを併用する時空等化構成としてもよい。更に第６実施例と同様に、第７実施例、第８実施例に図１に示した従来のシンボル同期タイミング生成器１５を併用し、両シンボル同期タイミング信号を通信品質に応じて選択的に使用するようにしてもよい。第９実施例及び第１０実施例においてＫが複数の場合に、図１３又は図１４に示す適応等化処理を行えばよい。第９実施例及び第１０実施例に対しても第８実施例に示したようなマルチビーム受信にも適用できる。
【００３８】
上述した各実施例はそれぞれコンピュータによりプログラムを実行させて機能させることもできる。
【００３９】
【発明の効果】
以上述べたようにこの発明により、適応等化器における等化処理をより効果的に行えるシンボル同期タイミングを検出することが可能となり、受信特性を改善できる。
【図面の簡単な説明】
【図１】従来の適応等化装置の機能構成を表す図。
【図２】従来のシンボルタイミング手段における相関出力信号と検出タイミングの例を表す波形図。
【図３】この発明の第１実施例の機能構成を表す図。
【図４】この発明に用いる相関出力信号と窓信号の関係例を示す波形図。
【図５】Ａは図３中の電力測定部２２の具体例を示す図、Ｂはその他の例を示す図である。
【図６】図３中の適応等化器１７（ＭＬＳＥ型適応等化器）の機能構成例を表す図。
【図７】図６中のレプリカ生成器３２の構成例を表わす図。
【図８】この発明の第２実施例における電力測定部２２の具体例を示す図。
【図９】この発明の第３実施例の機能構成例を表す図。
【図１０】図９中の適応等化器４１の機能構成例を表す図。
【図１１】図９中のアダプティブアレーアンテナ処理合成部５１の構成例を表す図。
【図１２】この発明の第４実施例の機能構成例を表す図。
【図１３】図１２中の適応等化器６３の機能構成例を表す図。
【図１４】図１２中の適応等化器６３の他の機能構成例を表す図。
【図１５】第４実施例においてアダプティブアレーアンテナ処理合成部を用いた場合の機能構成例を表す図。
【図１６】図１５中の適応等化器７１の機能構成例を表す図。
【図１７】Ａはしきい値以上の電力和に基づくタイミングを検出する例を示す機能構成図、Ｂは電力和の大きさに応じた数のタイミングを検出する例を示す機能構成図である。
【図１８】この発明の第６実施例の機能構成を表す図。
【図１９】この発明の第７実施例の機能構成を表す図。
【図２０】Ａは電波の到来パスが異なる方向の群に分かれている状態例を示す図、Ｂは異なる方向のパス間で受信信号の遅延時間が大きく異なる例を示す図である。
【図２１】この発明の第８実施例の機能構成を表す図。
【図２２】図２１中のマルチビーム生成器１０１の具体例を示す図。
【図２３】マルチビームの例を示す図。
【図２４】図２１に示した実施例におけるシンボル同期タイミングの生成動作を示す図。
【図２５】図２１中の適応等化器１０５の具体的機能構成例を示す図。
【図２６】Ａは電波の到来パスが異なる方向に分かれている状態例を示す図、Ｂは１つのパスの受信電力が大きく、これと遅延時間が大きく異なる小さい受信電力の複数のパスが生じている状態を示す図である。
【図２７】この発明の第９実施例の機能構成を示す図。
【図２８】この発明の第１０実施例の機能構成を示す図。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an adaptive equalization apparatus and method that can be used in, for example, mobile communication and removes interference, and more particularly to a symbol synchronization timing generation means and method for effectively performing adaptive equalization processing. is there.
[0002]
[Prior art]
There is an adaptive equalizer as a technique for removing interference in mobile communication. An adaptive equalizer is conventionally used for synthesizing and removing intersymbol interference that occurs between delayed signals that are delayed in time, although it is a signal, by signal processing on the time axis. . A conventional example of a method for detecting the symbol synchronization timing of the received signal used in such an adaptive equalizer will be described below. In the conventional method, the symbol timing is synchronized with the preceding wave that is the earliest arriving path. A configuration example in this case is shown in FIG. Although not shown in the figure, the transmission side first transmits a long synchronization word signal whose transmission symbol pattern is known on the reception side, and then transmits data indicating the content of information to be transmitted. Although not shown in the figure on the receiving side, the received radio wave from the transmitting side is amplified and demodulated to form a baseband signal and a digital series, and this digital series received signal is input to the input terminal 11 and this received signal and The channel response is estimated by obtaining the correlation with the synchronization word signal from the synchronization word generator 12 by the correlator 13. The propagation path response, that is, the output signal of the correlator 13 changes as shown in FIG. 2, for example, and the threshold signal Ths from the threshold signal setter 14 and the correlation output signal Sigc are converted into a symbol synchronization timing generator 15. The first timing t1 when the correlation output signal Sigc exceeds the threshold signal Ths is detected, and the symbol synchronization timing of the preceding wave is obtained. A sampler 16 samples the received signal digitized from the input terminal 11 with the symbol synchronization timing signal, and inputs the output sampling signal to the adaptive equalizer 17, which is input to the adaptive equalizer 17. The sampled signal is subjected to adaptive equalization processing, and a determination symbol signal is output to the output terminal 18.
[0003]
[Problems to be solved by the invention]
In this conventional method of matching the symbol synchronization timing with the preceding wave, when the adaptive equalizer 17 whose equalization range is limited in time is used, there is a correlation output signal having a large power outside the equalization range. In this case, sufficient equalization cannot be performed, and as a result, there is a problem that characteristics deteriorate.
[0004]
[Means for Solving the Problems]
In the present invention, the power sum of the correlation output signal obtained from the correlator within a fixed time width of at least one symbol length or more, preferably more than a length substantially corresponding to the equalization range of the adaptive equalizer, is made constant. The time width is calculated while being shifted in time, and in one aspect of the present invention, a timing at which this power sum is maximum is detected, and symbol synchronization timing is generated based on this timing. In another aspect of the present invention, a timing at which the power sum exceeds a threshold value for the first time is detected, and symbol synchronization timing is generated based on this timing.
[0005]
With this configuration, the path power sum within the equalization range of the adaptive equalizer can be maximized, so that equalization processing can be performed effectively. As a result, good reception characteristics can be expected.
[0006]
DETAILED DESCRIPTION OF THE INVENTION
First embodiment
The configuration of the first embodiment of the present invention is shown in FIG. 3 with the same reference numerals assigned to the portions corresponding to FIG. The correlator 13 outputs the correlation between the synchronization word signal generated by the synchronization word generator 12 and the received signal from the input terminal 11. At this time, as a synchronization word signal, a symbol pattern of a desired signal is used by using a symbol pattern having a high autocorrelation and a low cross-correlation with an interference wave, for example, a PN code string. The timing can be detected and the influence of the interference wave can be removed.
[0007]
On the other hand, in this embodiment, the power measurement unit 22 obtains the sum of correlation outputs Sigc within a certain time width while sequentially shifting the time width in terms of time. For example, a window signal generator 21 is provided, and a window signal having the same time width as the equalization range of the adaptive equalizer 17 is output from the window signal generator 21. For example, when the equalization range is a 4-symbol delay, a window signal is generated that becomes 1 during a time width 4T (T is a symbol period) and 0 otherwise. This window signal and the correlation output signal Sigc from the correlator 13 are input to the power measuring unit 22. The power measuring unit 22 sequentially calculates the power sum of the correlation output signal Sigc within the window signal range by shifting the temporal position of the window signal. This is shown in FIG. The baseband received signal from the input terminal 11 is a sequence that is sampled and digitized, for example, at 1/16 interval of the symbol period T, and the synchronization word signal from the synchronization word generator 12 is also digitized in the same period. Therefore, the correlation output signal Sigc is also a digital sequence having the same period. FIG. 4A simply shows an example of the correlation output signal Sigc as a continuous waveform. As shown in FIG. 4B, an equalization range time, in this example, a window signal # 1 having a level of 1 for the width of 4T, and 0 for the others is sequentially shifted by the sampling period of the correlation output signal, in this example, T / 16. 2, # 3,..., And the window signals # 1, # 2, # 3,. That is, for the window signal # 1, the sum of the hatched portions in FIG. 4A in the correlation output signal is obtained.
[0008]
As a specific example of the power measuring unit 22, for example, in the case of the above example in which the fixed time width is 4T and the sample period of the received signal is T / 16, the shift register 111 having 4 × 16 shift stages as shown in FIG. When the correlation output signal Sigc is input to one end, the shift register 111 is shifted by the clock Cs of T / 16, and the sum of the input and the output of each shift stage is taken by the adder 112, T / 16 is sequentially applied. A power sum in the shifted window signal is obtained every T / 16. Alternatively, as shown in FIG. 5B, the correlation output signal for each T / 16 from the correlator 13 is written into the memory 113. For example, when the writing of the correlation output signal corresponding to the end of the synchronization word signal is completed, the read address generators 114 to 4T A 4 × 16 address corresponding to the minute data is generated, the stored data is read from the memory 113, the read 4 × 16 data (correlation output signal) is added by the adder 115, and then shifted by one address. The generation of 4 × 16 read addresses and the addition of the data stored in the memory 113 by the read adder 115 may be repeated while shifting by one address. In this case, the adder 115 may sequentially perform the subtraction of the first data and the addition of the data at the next address with respect to the added value of the 4 × 16 data. Reading may be performed in parallel before the writing to the memory 113 is completed.
[0009]
Note that the amount by which the window signal is shifted is not limited to the sampling period of the correlation output signal Sigc, and may be an integer multiple thereof, for example, 2 to 4 times, or larger. However, the accuracy of the required symbol synchronization timing increases as the window signal shift amount decreases. In FIG. 3, the sampling clock used for digitizing the received digital signal input to the input terminal 11 is transferred from the sampling generator 19 to the synchronization word generator 12, the window signal generator 21, and the power measuring device 22. Are input and timed together.
[0010]
However, the window signal generator 21 is not necessary when the power sum is obtained using a shift register as described with reference to FIG. 5A.
The symbol synchronization timing generator 23 generates symbol synchronization timing using the timing at which the power with the highest power is obtained from the power sum output by the power measurement unit 22. In the case of the window signal as shown in FIG. 4, the rising timing of the window signal that maximizes the power may be set as the symbol synchronization timing.
In the above description, the time length (width) of the window signal, that is, the constant time width for obtaining the power sum by the power measuring unit 22 is preferably the same time range as the equalization range of the adaptive equalizer 17. May be at least about one symbol period, and the longer one may be as long as possible, but in practice it may be about twice the time range of the equalization range. The number of power sums obtained by the power measuring unit 22 is at least two, and the maximum may be obtained for the window signals that are sequentially shifted from the generation to the end of the synchronization word signal. However, if the power sum obtained sequentially falls below a certain threshold value and this state continues for a predetermined number of times, the power sum calculation is stopped and the maximum power sum obtained so far is selected. Also good. Even when starting to calculate the power sum, the power sum after the power sum is obtained more than the threshold value is made effective, and the maximum one after that is selected. Good.
[0011]
The sampler 16 samples the received signal using the symbol synchronization timing signal. The adaptive equalizer 17 receives the sampling signal from the sampler 16 and performs equalization processing. A configuration example of the adaptive equalizer 17 is shown in FIG. This configuration is called MLSE (Maximum Likelihood Sequence Estimation) type, and performs maximum likelihood sequence estimation to equalize the received signal (for example, J.G. Proakis: “Digital Communications 3”).^rd Edition, “McGraw-Hill pp. 249-254 (1995)). This adaptive equalizer first generates a symbol sequence candidate signal of a received signal. Processing for band limitation and encoding is performed at the time of transmission. If so, a symbol sequence candidate signal is generated in consideration of this, and then the adaptive equalization means weighting coefficient output from the parameter estimator 31 is complex-multiplied by the replica generator 32 to the symbol sequence candidate signal. Then, a replica of the received signal is generated by synthesizing the multiplication results, and a configuration example of the replica generator 32 shown in FIG. It comprises a multiplier 24 and a complex adder 25 for adding the multiplication results, and a symbol sequence candidate signal input to each complex multiplier 24 is used as a parameter as a tap input signal. And outputs it to the data estimator 31.
[0012]
The difference between this replica signal and the input signal of the adaptive equalizer (the output signal of the sampler 16) is taken by an adder 33 to generate an estimation error signal. Normally, this estimation error signal is squared by a squarer 34, and this square value is used as a branch metric in a Viterbi algorithm for performing maximum likelihood sequence estimation by a maximum likelihood sequence estimator (MLSE) 35. Finally, the most probable symbol pattern obtained by the Viterbi algorithm is output to the output terminal 18 as a determination symbol, and the received signal is demodulated. On the other hand, the parameter estimator 31 updates the adaptive equalizer weighting coefficient using the estimation error signal and the tap input signal. An adaptive algorithm may be used for this update, and an LMS (Least Mean Square) algorithm or an RLS (Recursive Least Square) algorithm can be used (for example, J.G. Proakis: “Digital Communications 3^rd Edition, “McGraw-Hill pp. 639-644, pp. 654-660 (1995)). In addition, for initial convergence of the adaptive equalizer weighting coefficient, the symbol sequence candidate signal from the MLSE processing unit 35 is used. Instead, a symbol pattern known on the receiving side called a training signal can be input from the terminal 36 to the replica generator 32 via the changeover switch 37. Specifically, first, instead of the symbol sequence candidate signal. Then, the adaptive equalizer weighting coefficient is converged by using the training signal, and then the switch 37 is switched to the MLSE 35 side to demodulate the symbol by using the symbol sequence candidate signal, for example, JG Proakis: “Digital Communications 3^rd Edition, “McGraw-Hill pp. 636-676 (1995).
[0013]
As described above, in the first embodiment, the symbol synchronization timing generator 23 generates a synchronization timing that maximizes the power within a certain time width (window signal). The path power that can be combined is maximized. As a result, compared with the conventional symbol synchronization timing generator, it is advantageous in terms of the ratio of desired signal power to noise power, and reception characteristics can be improved. Note that the symbol synchronization timing generator 23 is not necessarily limited to a signal of one symbol period. For example, when the adaptive equalizer 17 is a linear equalizer having a transversal filter with a 1/2 symbol period interval, it is 1/2. Sampling in the sampler 16 is performed at a symbol period.
Second embodiment
The difference between the configuration of the second embodiment of the present invention and the first embodiment is the difference in the power measurement method of the correlation signal in the power measurement unit 22. In the replica generator 32 in the adaptive equalizer, as shown in FIG. 7, since the tap input signal is configured at a certain time interval, equalization is performed on the received signal sampled at this time interval. Become. Therefore, by determining the symbol synchronization timing in consideration of the tap interval of the replica generator 32, the path diversity effect can be obtained more effectively. The tap interval in the adaptive equalizer is often one symbol interval. In this case, the power measurement method in the power measurement unit 22 is as follows.
[0014]
The correlation output signal within the range of the window signal is sampled every symbol interval, and the power sum of these sampling signals is calculated. This operation is repeated while shifting the time position of the window signal. For example, if the time width of the window signal in FIG. 4 is 4T, as shown in FIG. 4B, the correlation output signal at the timings indicated by the two end edges and the three broken lines in the window signals # 1, # 2, # 3,. The sum of each of the five sample values is calculated. When the power sum of the correlation output signal Sigc is obtained using the shift register shown in FIG. 5A, as shown in FIG. 8, the 4 × 16 stage shift register 111 is used as in FIG. The correlation output signal Sigc is fetched by the shift operation by the output of the sampling clock generator 19 in 3), and the adder 112 adds the input signal and the output of each 16 shift stages. This added value (power sum) is output every T / 16.
[0015]
The symbol synchronization timing generator 23 detects the maximum timing among these power sums. With this method, it is possible to detect the timing at which the maximum path diversity effect can be obtained even in an environment where the incoming paths of received signals are distributed not only at symbol intervals but also at irregular intervals. The interval for obtaining the sum of the correlation output signals in the second embodiment may be the sampling interval by the sampler 16 in FIG. 3, and is not necessarily one symbol interval.
Third embodiment
The configuration of the third embodiment of the present invention is shown in FIG. 9 with the same reference numerals assigned to the portions corresponding to FIG. In this embodiment, the adaptive equalization means has a space-time equalization configuration using both adaptive array antenna processing and adaptive equalization processing.
[0016]
An adaptive array antenna directs a beam having a relatively high antenna gain in the direction of a desired wave that is its own signal, and directs a null having a relatively low gain in the direction of an interference wave such as a signal of another user. A directional characteristic beam pattern is adaptively generated according to a change in signal condition. For this reason, it is an effective technique for removing interference waves using the same frequency (channel), that is, co-channel interference. A space-time equalizer is a combination of the adaptive array antenna and the adaptive equalizer. See, for example, R.T. Compton, Jr .: “Adaptive Antennas-Concepts and Performance,” Prentice-Hall, Englewood Cliffs (1988) for adaptive array antennas.
[0017]
The space-time equalizer removes long delay waves and co-channel interference that cannot be equalized because they exceed the equalization range of the adaptive equalizer, using an adaptive array antenna, and the adaptive equalizer delay waves that are within the equalization range. Equalization processing is performed for. By adding equalization processing in the spatial domain in this way, even in a propagation environment where hardware implementation is difficult due to an increase in the amount of computation with only signal processing in the time domain, equalization is performed with a reasonable hardware scale. Processing is possible.
In FIG. 9, received signals from each element of the adaptive array antenna are input as baseband signals (digitized) to input terminals 11-1 to 11-N, and these N systems of received signals are samplers 16-1 to 16-1. Each of the N-sampling signals generated is sampled at 16-N by the symbol synchronization timing signal, and the generated N systems of sampling signals are input to the adaptive equalizer 41 to perform equalization processing. A configuration example of the adaptive equalizer 41 in this embodiment is shown in FIG. The difference from the adaptive equalizer shown in FIG. 6 is that an adaptive array antenna processing combining unit 51 is provided. N types of sampling signals from the terminals 52-1 to 52-N are combined by an adaptive array antenna processing combining unit 51, and equalization processing is performed on the combined signals. A configuration example of the adaptive array antenna processing combining unit 51 is shown in FIG. N complex multipliers 53-1 to 53-N that complex-multiply N sampling signals from terminals 52-1 to 52-N and N adaptive array antenna (AAA) weighting coefficients, respectively. The complex adder 54 is configured to complex add the outputs of the complex multipliers. The weighting factor used in the adaptive array antenna processing combining unit 51 is obtained by the parameter estimator 55, and the estimation error signal and the second input which is the complex multipliers 53-1 to 53-N in the adaptive array antenna processing combining unit 51 are input. It can be calculated using the tap input signal. For example, Ryuji Kohno: “Spatial and Temporal Communication Theory Using Adaptive Antenna Array”, IEEE Personal Communications, pp. 28-35, Feb. 1998, AJ Paulraj and BC Ng, “Space-Time Modems for Wireless Personal Communications ", IEEE Personal Communications, pp. 36-48, Feb. 1998.
[0018]
In this embodiment, there are N received signals. However, it is sufficient to obtain the correlation with the synchronization word signal by the correlator 13 using any one of the received signals. However, it is possible to obtain a correlation output signal with higher accuracy by obtaining each correlation between a plurality of received signals and the synchronization word signal and adding and averaging these correlation output signals, for example.
Fourth embodiment
The configuration of the fourth embodiment of the present invention is shown in FIG. Each part other than the power comparator 61 and the symbol synchronization timing generator 62 is the same as in the first and second embodiments shown in FIG. The power comparator 61 compares the power sums obtained by the power measuring unit 22 and selects L timing points (L is an integer of 1 or more) in order from the largest. When L is 1, this embodiment is the same as the first embodiment. The maximum value of L is the number of power sums obtained by the power measuring unit 22, but may be set in consideration of the hardware scale. The symbol synchronization timing generator 62 outputs a symbol synchronization timing signal according to the timing of each of the selected L power sums. In this embodiment, samplers 16-1 to 16-L corresponding to these L symbol synchronization timing signals are provided. In each of the samplers 16-1 to 16-L, a received signal from the input terminal 11 is symbol-synchronized. Sampling is performed by L symbol synchronization timing signals from the timing generator 62. The adaptive equalizer 63 performs equalization processing using these L sampling signals.
[0019]
FIG. 13 shows a configuration example of the adaptive equalizer 63 in this case. Sampling signals from the samplers 16-1 to 16-L are input from the terminals 64-1 to 64-L to the estimation error output units 65-1 to 65-L, respectively. As shown in the estimation error output unit 65-1, a parameter estimator 31-1, a replica generator 32-1, an adder 33-1, and a squarer 34-1 having the same configuration as shown in FIG. The estimation error output units 65-2 to 65-L also include a parameter estimation unit, an adder, a squarer, and a replica generator corresponding to each symbol synchronization timing. In this example, replica generation and parameter estimation are performed every L symbol synchronization timings, and the square values of the estimation error signals of the respective adders are added together by the adder 66 as a branch metric, and the MLSE processing unit 35 performs MLSE. Processing is in progress. Instead of the MLSE processing unit 35, other adaptive equalization processing may be performed.
[0020]
Further, as shown in FIG. 14, MLSE processing units 35-1 to 35-L are also prepared for the respective estimation error output units 65-1 to 65-L, and equalization processing is performed independently at each symbol synchronization timing. After the processing, the quality correspondence processing unit 67 can perform a process corresponding to the communication quality on the output of each of the MLSE processing units 35-1 to 35-L to obtain a final output. It is. In this case, as processing according to the communication quality, for example, using the obtained equalizer output (MLSE processing output), a method of performing these majority determinations by the majority determination unit 69 as indicated by a broken line in the figure, When the MLSE processing units 35-1 to 35-L respectively output the weights corresponding to the squares of the error signals of the adders 33-1 to 33-L corresponding to the outputs, the MLSE processing unit 35- 1 to 35-L are added to each other and added by an adder 68. The added signal is binarized by a threshold circuit 68a and output. As for the latter, it is only necessary to give a large weight to a communication with a good estimation quality if the estimation error is smaller.
[0021]
It is also possible to use an adaptive array antenna processing combining unit as in the third embodiment. A configuration example in this case is shown in FIG. The received signals from the input terminals 11-1 to 11-N are sampled by the samplers 1-1 to 1-N by the first symbol synchronization timing signal from the symbol synchronization timing generator 62, respectively, and the second symbol synchronization timing signal Are sampled by the samplers 2-1 to 2-N, respectively, are sampled by the samplers L-1 to LN by the Nth symbol synchronization timing signal, and are input to the adaptive equalizer 71.
[0022]
As shown in FIG. 16, the adaptive equalizer 71 inputs the outputs of the samplers 1-1 to 1-N to the estimation error output unit 72-1, and the estimation error output unit 72-1 has the same configuration as that shown in FIG. In addition, the replica generator 32-1, the error calculating adder 33-1, the error squarer 34-1, the adaptive array antenna processing combining unit 51-1, and the parameter estimator 55-1. The outputs of the samplers 2-1 to 2-N to the outputs of the samplers L-1 to LN are input to the estimation error output units 72-2 to 72-L, respectively. The estimated error output units 72-2 to 72-L have the same configuration as the estimated error output unit 72-1. The square error signals from the estimation error output units 72-1 to 72-L are added by the adder 66 and input to the MLSE processing unit 35.
[0023]
By using an adaptive array antenna, a path outside the equalization range of the adaptive equalizer can be removed, and reception characteristics can be further improved.
In the above description, it has been assumed that the necessary parts are prepared in the L system. However, by using each part in a time-sharing manner, it can be realized with a value smaller than L, thereby reducing the hardware scale. Can be planned.
As described above, in this embodiment, by using a plurality of symbol synchronization timings, it becomes possible to synthesize more paths, so that reception characteristics are improved.
Example 5
In the fifth embodiment of the present invention, the power comparator 61 in FIGS. 12 and 15 simply selects L timings in the order of increasing power sum when generating symbol synchronization timing as in the fourth embodiment. Instead, as shown in FIG. 17A, a threshold value Th is preset in the threshold value setter 70, and the power comparator 61 in the power sum from the power measuring unit 22 exceeds the threshold value Th (Q Is 1<Q<L<-Q symbol synchronization timings are generated based on an integer satisfying M). By this operation, the value of Q is adaptively selected according to the communication status. For example, although the power sum is within the Lth from the top, the effect of improving the reception characteristics is small even if equalization processing is performed. It is possible to omit the equalization processing for the symbol synchronization timing obtained from the window signal with a small power sum, and as a result, it is possible to reduce the amount of calculation of the entire equalization processing. In other words, L timing selection in the claim means that L is not a predetermined value but the number of power sums equal to or greater than a threshold value.
[0024]
In addition, even when the power sum in the window signal that can exhibit sufficient reception characteristics at the top few points among the selected L points can be obtained, equalization processing for symbol synchronization timing based on the power sum less than that is omitted By doing so, it is possible to reduce the amount of calculation in the equalization processing with almost no deterioration in reception characteristics. That is, if the number of power sums exceeding a relatively large threshold is a predetermined number (for example, 1 or 2), equalization processing is performed only for symbol synchronization timing based on the power sum exceeding the threshold. . For example, as shown in FIG. 17B, the M power sums from the power measuring unit 22 are supplied to the power comparators 61a, 61b, 61c, respectively, and compared with the threshold values Tha, Thb, Thc (Tha> Thb> Thc), respectively. Each of the power comparators 61a, 61b, 61c outputs a value exceeding the threshold value in the M input power sums, and the power comparator 61a outputs one or two of them even if the reception state is good. The threshold value Tha is set to such an extent that the power sum is output, and 3 to 4 are output from the power comparator 61b, and other thresholds are selected so that 5 to 7 are output from the comparator 61c. Is done. The output of the power comparator 61b is supplied to the prohibition gate 121. The prohibition gate 121 is prohibited by the output exceeding the threshold value from the power comparator 61a. The output of the power comparator 61c is supplied to the prohibition gate 122. Is prohibited by the output exceeding the threshold value from the power comparator 61b. The outputs of the power comparator 61a and the prohibition gates 121 and 122 are input to the symbol synchronization timing generator 62, and a symbol synchronization timing signal having a timing corresponding to each input power sum is output.
[0025]
In this embodiment, since the value of Q varies, there arises a problem of how many systems of replica generators, parameter estimators, etc. are prepared. However, when the hardware scale is sufficient, a sufficient number is prepared, When the value is small, there is a method of using a part of these. In this case, there is a merit that power consumption can be reduced as compared with the case where all are operated. If it is desired to reduce the hardware scale, a replica generator, a parameter estimator, etc. may be used in a time division manner as described in the fourth embodiment.
Sixth embodiment
FIG. 18 shows a functional configuration of the sixth embodiment of the present invention. In this embodiment, a configuration including two symbol synchronization timing generators is provided. That is, in addition to the symbol synchronization timing generator 23 that generates the symbol synchronization timing signal at the timing at which the power sum from the power measurement unit 22 shown in FIG. 3 becomes maximum, the conventional symbol synchronization timing generator 15 shown in FIG. And the symbol synchronization timing is generated from the timing at which the correlation output signal exceeds the threshold value from the threshold signal setter 14, that is, the timing of the preceding wave of the received wave.
[0026]
The symbol synchronization timing signals from these symbol synchronization timing generators 15 and 23 are switched by a timing selector 81 and supplied to the sampler 16. In this embodiment, adaptive equalization processing for a sampling signal at each symbol synchronization timing is performed, for example, in a time division manner, and it is determined which symbol synchronization timing signal is to be demodulated according to communication quality. As the communication quality, for example, the estimated error signal in the last symbol of the training signal interval by each symbol synchronization timing signal is used, the magnitudes of these error signals are compared, and the demodulation process is performed at the timing when the estimated error power becomes small. The timing selector 81 may be controlled to perform.
Example 7
The functional configuration of the seventh embodiment of the present invention is shown in FIG. 19 with the same reference numerals assigned to the portions corresponding to those in FIG. In this embodiment, the power measuring section 91 sequentially outputs the power sum of each correlation output signal in the window signals # 1, # 2, # 3,... Sequentially shifted as shown in FIG. The signal is input to the timing generator 92, and the symbol synchronization timing generator 92 outputs the window signal when the input power sum exceeds the threshold signal set in the threshold signal setting unit 93 for the first time. A symbol synchronization timing signal is generated based on the timing. The other configuration is the same as that shown in FIG. Also in the seventh embodiment, the power sum of the correlation output signal may be the sum of those sampled at the sampling interval of the sampler 16.
Example 8
As shown in FIG. 20A, the incoming path of radio waves is divided into several groups G1, G2,..., GP in different directions, and received signals G1 ′, G2 ′,. In propagation path environments in which the delay times of GP ′ are greatly different, in the configuration shown in the first and second embodiments (including the third, fourth, fifth and sixth embodiments), the number of taps of the adaptive equalizer If there is a small amount, sufficiently good reception characteristics may not be obtained. In this embodiment, equalization processing can be effectively performed even in such a propagation environment.
[0027]
The configuration of the eighth embodiment is shown in FIG. 21, in which parts corresponding to those in FIG. In this embodiment, in addition to the configuration used in the second embodiment, N array antenna reception signals are input from the input terminals 11-1,..., 11-N, and the multi-beam generator 101 receives the P system beam reception signals. These P-system beam received signals are output to P-system correlators 13-1,..., 13-P, and are output to P-system samplers 16-1,. 13-P, 13-P correlates the received signal of each input beam with the synchronization word signal from the synchronization word generator 12, and these correlation output signals are supplied to the power measuring units 22-1, .., 22-P, the power sum of each correlation output signal within the range of the window signal from the window signal generator 21 is calculated, and these calculations are performed by sequentially shifting the time positions of the window signals. The power sums sequentially output from the power measuring units 22-1,..., 22-P are respectively input to the symbol synchronization timing generators 23-1,. ,..., 23-P, the symbol synchronization timing is generated using the timing of the window signal with the largest power in the power sum, and the correspondence of the samplers 16-1,. The received signal of the beam to be sampled is sampled. The sampling outputs of these samplers 16-1,..., 16-P are supplied to the adaptive equalizer 105.
[0028]
A configuration example of the multi-beam generator 101 is shown in FIG. Like the adaptive array antenna shown in FIG. 11, the multipliers 102-1,..., 102-N respectively multiply the input signals from the antenna reception input terminals 11-1,. These multiplication results are added by the adder 103-1, thereby obtaining an output signal of one beam of the multi-beam. While one beam output signal is output by adaptively updating the weighting coefficient according to the arrival state of the signal of the adaptive array antenna shown in FIG. 11, the multi-beam generator 101 receives the signal. Regardless of the situation, a received signal of one beam in a fixed direction is output using a weighting coefficient obtained in advance, and a plurality of beam formers 104-1,..., 104-P in such a fixed direction are provided. Is different. As described above, the multi-beam generator 101 generates P-system beams by the beam formers 104-1,..., 104-P using P sets of weighting coefficients. Differently, for example, as shown in FIG. 23, normally, all beams G1,..., GP are generated so as to cover all directions of arrival of radio waves. A beam having such properties can be easily generated by using, for example, a Butler Matrix circuit. FIG. 23 is an example of a multi-beam formed using a Butler Matrix circuit (see, for example, J. Ltva and TK Lo: “Digital Beamforming in Wireless Communications,” Artech House, Boston London pp. 22-34, (1996). ).
[0029]
The P-system multi-beam received signals G1 ′,..., GP ′ obtained by the multi-beam generator 101 are respectively correlated with the correlators 13-1,. ,..., 22-P are input to symbol synchronization timing generators 23-1,..., 23-P, and P symbol synchronization timing signals are generated. Using the P symbol synchronization timing signals, the samplers 16-1,..., 16-P generate P system sampling signals from the P system beam reception signals G1 ′,. That is, in each of the symbol synchronization timing generators 23-1,..., 23-P, as shown in FIG. 24, the corresponding power measuring units 22-1,. The timing at which each power sum is the largest is the symbol synchronization timing. In FIG. 21, a correlator, a power measurement unit, a symbol synchronization timing generator, and a sampler are prepared for each beam system, but these units may be used in a time division manner in each system.
[0030]
The P-system sampling signal generated in this way is input to the adaptive equalizer 105 and equalization processing is performed. A configuration example of the adaptive equalizer 105 in this embodiment is shown in FIG. 25, and the same reference numerals are given to the portions corresponding to FIG. P-system sampling signals from the samplers 16-1,..., 16-P are weighted and synthesized by the linear synthesizer 106, and equalization processing is performed on the synthesized signal. The contents of the equalization process are the same as in the third embodiment. That is, the linear combiner 106 in this case corresponds to the adaptive array antenna processing combiner 51 in FIG. The weighting coefficient for the linear synthesizer 106 may be fixed.
[0031]
As described above, by detecting the symbol synchronization timing for each beam in different directions of the multi-beam, it is possible to detect the symbol synchronization timing using not only the time domain but also the spatial domain as shown in FIG. Become. As shown in the example of the incoming radio wave path shown in FIG. 20, when the delay time of the received signal is greatly different between the incoming path groups G1,. By separating and detecting the symbol synchronization timing for each separated received signal, sufficiently good reception characteristics can be obtained even when the equalization range of the adaptive equalizer 105 is small.
[0032]
The multi-beam generator 101 shown in FIG. 21 and the symbol synchronization method according to the present invention for the received signal for each beam are also applied to the first to seventh embodiments described above. Can do. When using L timings with large sum power as shown in FIG. 12, L samplers are provided for each of the samplers 16-1,..., 16-P in FIG. .., 11-N may be configured in the same manner as the P output signals from the multi-beam generator 101 are supplied to each of the input terminals 11-1,..., 11-N. A configuration similar to that shown in FIG.
Ninth embodiment
As shown in FIG. 26A, radio waves arrive from a plurality of paths G1,..., GP, but as shown in FIG. 26B, a received signal G1 'of a path with relatively large received power is a received signal G2 of another path. Consider a propagation path environment arriving at a time distance from ', ..., GP'. In this case, in the configuration shown in the first to eighth embodiments, the symbol synchronization timing is adjusted so that the received signal G1 ′ of one path having high power falls within the window signal, and the received signal G2 ′ of the other path is set. , ..., GP 'is equalization range T_AERWill be outside. This result equalization range T_AERHowever, since there is only one path, a sufficient path diversity effect cannot be obtained, and good reception characteristics may not be obtained in a fading environment. In this embodiment, good reception characteristics can be obtained even in such a propagation environment.
[0033]
The functional configuration of the ninth embodiment is shown in FIG. 27 with the same reference numerals assigned to the portions corresponding to FIG. In this embodiment, a power dispersion value measuring unit 131 is used instead of the power measuring unit 22 used in the first to eighth embodiments. The power variance value measuring unit 131 receives the window signal from the window signal generator 21 and the threshold signal Pth output from the threshold signal setting unit 132 as input, and the correlation output signal power for each sample in the window signal. , And average power P_NIs calculated by the average value calculation unit 131a, and the power distribution value σ²Is calculated by the variance value calculation unit 131b. Calculated power average value P_NIs greater than the threshold signal Pth, the power variance value σ corresponding to that determined by the comparison unit 131c.²For example, K values (K is an integer equal to or greater than 1) are selected and output by the variance value comparison unit 131d in ascending order of variance values. The value of K may be determined in consideration of the hardware scale. If the sample value of each correlation output signal in the window signal is x (n) and the number of samples in the window signal is N, the variance σ²Is σ²= (1 / N) Σ_n-1 ^N(X (n) -P_N)²Is required.
[0034]
The power variance value σ output by the power variance value measuring unit 131²Is input to the symbol synchronization timing generator 133, and each of these K power distribution values σ²A symbol synchronization timing signal is output on the basis of the timing corresponding to, for example, the rising timing of the corresponding window signal. For example, when the variance value in the window signal is 0, this indicates that the path exists at equal power at each sample point in the window signal, and the path diversity effect can be maximized by adaptive equalization processing. Can do. The symbol synchronization timing signal is generated by the above process, and sampling signals of the reception signal are generated by the samplers 16-1 to 16-K. These sampling signals are input to the adaptive equalizer 63, and equalization processing is performed. . The configuration of the adaptive equalizer 63 in this embodiment can be the same as that shown in FIGS.
[0035]
As described above, by using the ninth embodiment, the combined signal power is not necessarily the maximum, but it is possible to detect the symbol synchronization timing in which many paths exist within the equalization range, and due to the path diversity effect, Good reception characteristics can be expected even in a fading environment.
10th embodiment
As shown in FIG. 28, the tenth embodiment is obtained by adding a configuration using the power distribution value measuring unit 131 shown in FIG. 27 to the sixth embodiment shown in FIG. In addition to the two symbol synchronization timing generators 15 and 23 (62 in FIG. 12), which is two in the sixth embodiment, a third symbol synchronization timing generator 133 is provided. The symbol synchronization timing signals from these symbol synchronization timing generators 15, 62 and 133 are switched by the timing selector 81 and supplied to the samplers 16-1 to 16-L (or 16-K). Prepare as many samplers as the larger of L and K.
[0036]
As in the sixth embodiment, adaptive equalization processing is performed, for example, in a time-sharing manner at each synchronization timing, and the symbol selector 81 is controlled to determine which symbol synchronization timing signal is taken based on the communication quality.
As a result, it becomes possible to detect symbol synchronization timing more suitable for the propagation environment, and improvement of reception characteristics can be expected. In FIG. 28, one of the power measuring unit 22, the power comparator 61, and the symbol synchronization timing generator 62, the threshold signal setting unit 14, and the symbol synchronization timing generator 15 may be omitted.
[0037]
According to the ninth and tenth embodiments, even in a propagation path environment where radio waves arrive in multiple paths, and among them, a path with relatively large received power arrives away from other paths in time. In the adaptive equalizer, it becomes possible to detect the symbol synchronization timing at which the path diversity effect can be obtained more, and the reception characteristics can be improved.
The second embodiment is applied to each of the seventh to ninth embodiments, and the power measurement unit 91 or the power dispersion measurement unit 131 outputs the correlation output signal within the window signal range to the sampler 16 or 16. Samples may be performed at sampling intervals of −1 to 16-P, and the power sum or power distribution value of these sampling signals may be calculated. Also in the sixth embodiment, the seventh embodiment, the ninth embodiment, and the tenth embodiment, the adaptive equalizer 17 performs adaptive array antenna processing and adaptive equalization processing in the same manner as in the third embodiment. It is good also as a space-time equalization structure used together. Further, similarly to the sixth embodiment, the conventional symbol synchronization timing generator 15 shown in FIG. 1 is used in combination with the seventh and eighth embodiments, and both symbol synchronization timing signals are selectively selected according to the communication quality. It may be used. In the ninth embodiment and the tenth embodiment, when there are a plurality of K, the adaptive equalization processing shown in FIG. 13 or FIG. 14 may be performed. The ninth embodiment and the tenth embodiment can also be applied to multi-beam reception as shown in the eighth embodiment.
[0038]
Each of the embodiments described above can be made to function by causing a computer to execute the program.
[0039]
【The invention's effect】
As described above, according to the present invention, it is possible to detect the symbol synchronization timing at which the equalization processing in the adaptive equalizer can be performed more effectively, and the reception characteristics can be improved.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a functional configuration of a conventional adaptive equalization apparatus.
FIG. 2 is a waveform diagram showing an example of a correlation output signal and detection timing in conventional symbol timing means.
FIG. 3 is a diagram showing a functional configuration of the first embodiment of the present invention.
FIG. 4 is a waveform diagram showing a relationship example between a correlation output signal and a window signal used in the present invention.
5A is a diagram showing a specific example of the power measurement unit 22 in FIG. 3, and B is a diagram showing another example. FIG.
6 is a diagram illustrating a functional configuration example of an adaptive equalizer 17 (MLSE type adaptive equalizer) in FIG. 3;
7 is a diagram illustrating a configuration example of a replica generator 32 in FIG.
FIG. 8 is a diagram showing a specific example of a power measuring unit 22 in the second embodiment of the present invention.
FIG. 9 is a diagram illustrating a functional configuration example of a third embodiment of the invention.
10 is a diagram illustrating a functional configuration example of an adaptive equalizer 41 in FIG. 9;
11 is a diagram illustrating a configuration example of an adaptive array antenna processing combining unit 51 in FIG. 9;
FIG. 12 is a diagram illustrating a functional configuration example of a fourth embodiment of the invention.
13 is a diagram illustrating a functional configuration example of an adaptive equalizer 63 in FIG.
14 is a diagram illustrating another functional configuration example of the adaptive equalizer 63 in FIG.
FIG. 15 is a diagram illustrating an example of a functional configuration when an adaptive array antenna processing combining unit is used in the fourth embodiment.
16 is a diagram illustrating a functional configuration example of an adaptive equalizer 71 in FIG.
FIG. 17 is a functional configuration diagram illustrating an example of detecting timing based on a power sum equal to or greater than a threshold, and B is a functional configuration diagram illustrating an example of detecting a number of timings corresponding to the magnitude of the power sum. .
FIG. 18 is a diagram showing a functional configuration of a sixth embodiment of the present invention.
FIG. 19 is a diagram showing a functional configuration of a seventh embodiment of the present invention.
20A is a diagram illustrating an example of a state in which radio wave arrival paths are divided into groups in different directions, and FIG. 20B is a diagram illustrating an example in which the delay time of a received signal differs greatly between paths in different directions.
FIG. 21 is a diagram showing a functional configuration of an eighth embodiment of the present invention.
22 is a diagram showing a specific example of the multi-beam generator 101 in FIG.
FIG. 23 is a diagram showing an example of a multi-beam.
FIG. 24 is a diagram showing a symbol synchronization timing generation operation in the embodiment shown in FIG. 21;
FIG. 25 is a diagram showing a specific functional configuration example of the adaptive equalizer 105 in FIG. 21;
FIG. 26A is a diagram showing an example of a state in which the arrival paths of radio waves are divided in different directions, and FIG. FIG.
FIG. 27 is a diagram showing a functional configuration of a ninth embodiment of the invention.
FIG. 28 is a diagram showing a functional configuration of a tenth embodiment of the present invention.

Claims

A synchronization word generator that outputs a synchronization word signal of the same signal sequence as the synchronization word transmitted by the transmitter;
A correlator that takes the synchronization word signal and the reception signal as input, calculates a correlation between the reception signal and the synchronization word signal, and outputs the correlation value as a correlation output signal;
A power measurement unit that takes a correlation output signal as an input, and calculates the power sum of the correlation output signal within a certain time width while shifting the time width in time,
Comparing the power sum obtained by the power measuring unit according to a certain standard, L timings (L is an integer equal to or greater than 1) are selected, and each timing is obtained, and based on the selected L timings A symbol synchronization timing generator for generating and outputting a symbol synchronization timing signal;
A sampler that samples each received signal using the L symbol synchronization timing signals;
An adaptive equalization apparatus comprising: an adaptive equalizer that receives the sampled received signal as input and performs adaptive equalization processing to output a decision symbol signal.

The apparatus of claim 1.
An adaptive equalization apparatus comprising: a power comparator that inputs a power sum from the power measurement unit and outputs a timing at which a power sum greater than a threshold is obtained to the symbol synchronization timing generator.

The apparatus of claim 2.
A plurality of threshold values are set, and a power comparator is provided that outputs to the symbol synchronization timing generator a timing at which only the power sum exceeding the largest value among the power sums exceeding the threshold values is obtained. A characteristic adaptive equalizer.

A synchronization word generator that outputs a synchronization word signal of the same signal sequence as the synchronization word transmitted by the transmitter;
A correlator that takes the synchronization word signal and the reception signal as inputs, calculates the correlation between the reception signal and the synchronization word signal, and outputs the correlation value as a correlation output signal;
A power measurement unit that takes a correlation output signal as an input, and calculates the power sum of the correlation output signal within a certain time width while shifting the time width in time,
A symbol synchronization timing generator that detects a timing at which the magnitude of the power sum determined by the power meter first exceeds a threshold, and generates and outputs a symbol synchronization timing signal based on the timing;
A sampler that receives the symbol synchronization timing signal and the received signal and samples the received signal;
An adaptive equalizer that takes a received signal sampled by the sampler as an input, performs adaptive equalization processing, and outputs a decision symbol signal;
An adaptive equalization apparatus comprising:

A synchronization word generator that outputs a synchronization word signal of the same signal sequence as the synchronization word transmitted by the transmitter;
A correlator that takes the synchronization word and the received signal as input, calculates a correlation between the received signal and the synchronization word, and outputs the correlation value as a correlation output signal;
Using the correlation output signal and the threshold signal as input, find the average value and variance of the correlation output signal within a certain time width while shifting the time width in time, and the variance value at the timing when the average value is greater than the threshold signal Power dispersion value measuring unit that outputs K (K is an integer of 1 or more),
A symbol synchronization timing generator for generating K symbol synchronization timing signals based on the K timings;
A sampler for sampling each received signal using the K symbol synchronization timing signals;
An adaptive equalization apparatus comprising: an adaptive equalizer that receives the sampled received signal as input and performs adaptive equalization processing to output a decision symbol signal.

The apparatus according to claim 1 or 5,
A second symbol synchronization timing generator for generating a symbol synchronization timing signal based on the timing of the preceding wave;
A symbol synchronization timing signal from the symbol synchronization timing generator;
An adaptive equalizer comprising: a timing selector that adaptively switches a symbol synchronization timing signal from the second symbol synchronization timing generator according to communication quality and supplies the symbol synchronization timing signal to the sampler.

The apparatus of claim 5.
A power measurement unit that receives the correlation output signal as input and obtains the power sum of the correlation output signal within the fixed time width while shifting the time, and L selected from among the power sums obtained by the power measurement unit (L is an integer equal to or greater than 1) obtaining each timing at which each power sum is obtained, and generating a symbol synchronization timing signal based on these L timings, and the symbol synchronization A timing selector that adaptively switches between a symbol synchronization timing signal from the timing generator and a symbol synchronization timing signal from the second symbol synchronization timing generator according to communication quality and supplies the symbol synchronization timing signal to the sampler. A characteristic adaptive equalizer.

The device according to claim 1, 4 or 5,
The power measurement unit or the power variance value measurement unit includes a configuration in which a correlation output signal within the fixed time width is sampled at a sampling period of the sampler and a power sum or a variance value of the sampled signal is obtained. An adaptive equalization apparatus.

The device according to claim 1, 4 or 5,
The adaptive equalizer is a space-time equalizer that performs signal processing in a spatial domain together with equalization processing in a time domain.

The apparatus according to claim 1 or 5,
L or K is a plurality, and the adaptive equalizer outputs an error signal power between each of the plurality of sampled received signals and each replica signal, and an error of these estimation error output units An adaptive equalization apparatus comprising: an adaptive equalization processing unit that adds signal power and inputs the signal power as error signal power.

The apparatus according to claim 1 or 5,
The L or K is plural, and the adaptive equalizer is adapted to the adaptive equalization processing unit that is performed on the plurality of sampled received signals and the output from the adaptive equalization processing unit according to the communication quality. An adaptive equalization apparatus comprising: a final processing unit configured to perform final processing and perform final processing.

The device according to claim 1, 4 or 5,
A multi-beam generator that weights received signals from a plurality of antennas to form a multi-beam antenna directivity and outputs a received signal from each beam of the multi-beam to the sampler,
For each received signal from the multi-beam generator, the correlator, the power measurement unit or power variance value measurement unit, and the symbol synchronization timing generator, the symbol synchronization timing signal from the symbol synchronization timing generator, An adaptive equalization apparatus characterized by being supplied to the sampler to which a corresponding beam reception signal is supplied.

Generating a synchronization word signal of the same signal sequence as the synchronization word transmitted by the transmitter;
Calculating the correlation between the received signal and the synchronization word signal, and using the correlation value as the correlation output signal;
A process for obtaining the power sum of the correlation output signals within a certain time width while shifting the time width in time.
Comparing the power sums according to a certain standard, selecting L (L is an integer of 1 or more), and generating a symbol synchronization timing signal based on the timing at which each of the selected power sums is obtained;
Sampling each of the received signals using the L symbol synchronization timing signals;
An adaptive equalization method including a step of obtaining a decision symbol signal by performing adaptive equalization processing on the sampled received signal.

14. The method of claim 13, wherein
The adaptive equalization method characterized in that the L selections are performed by selecting a power sum larger than a threshold value to obtain the L timings.

14. The method of claim 13, wherein
The adaptive equalization method characterized in that the L selections are selected in descending order of power sum.

14. The method of claim 13, wherein
The adaptive equalization method characterized in that the L selections are performed by selecting a power sum larger than the largest threshold among power sums larger than a plurality of threshold values to obtain timing.

Generating a synchronization word signal of the same signal sequence as the synchronization word transmitted by the transmitter;
Calculating the correlation between the received signal and the synchronization word signal, and using the correlation value as the correlation output signal;
A process of obtaining the power sum of the correlation output signals within a certain time width while shifting the time,
Detecting a timing at which the power sum when the magnitude of the power sum exceeds a threshold value first is obtained, and generating a symbol synchronization timing signal based on the timing;
Sampling the received signal using the symbol synchronization timing signal and the received signal;
A process of obtaining a determination symbol signal by performing adaptive equalization processing on the sampled received signal;
An adaptive equalization method comprising:

Generating a synchronization word signal of the same signal sequence as the synchronization word transmitted by the transmitter;
Calculating the correlation between the received signal and the synchronization word signal, and using the correlation value as the correlation output signal;
The average value and variance value of the correlation output signal within a certain time width are obtained by shifting the time width in time, and K variance values at the timing when the average value is larger than the threshold signal (K is an integer of 1 or more). The process of seeking
Generating a symbol synchronization timing signal based on the K timings;
Sampling each of the received signals using the K symbol synchronization timing signals;
An adaptive equalization method including a step of obtaining a decision symbol signal by performing adaptive equalization processing on the sampled received signal.

A method according to claim 13, 17 or 18,
The correlation power signal within the predetermined time width is sampled at a sampling period with respect to the received signal, and the sum of the sampled signals is set as the power sum, or the power variance value of the sampled signal is set as the variance value. An adaptive equalization method characterized by

A method according to claim 13 or 18,
Symbol synchronization used for the sampling of the received signal based on the communication quality of the process of generating the second symbol synchronization timing signal based on the timing of the preceding wave, the symbol synchronization timing signal, and the second symbol synchronization timing signal An adaptive equalization method comprising a process of adaptively switching as a timing signal.

The method of claim 18, wherein:
The process of obtaining the power sum of the correlation output signal within the fixed time width while shifting the time width in time, and L power sums selected from the power sum (L is an integer of 1 or more) Each of the obtained timings is obtained, and the process of generating the second symbol synchronization timing signal based on these L timings, the symbol synchronization timing signal, and the second symbol synchronization timing signal according to the communication quality. And adaptively switching as a symbol synchronization timing signal used for the sampling of the received signal.

The method of claim 18, wherein:
The process of obtaining the power sum of the correlation output signal within the fixed time width while shifting the time width in time, and L power sums selected from the power sum (L is an integer of 1 or more) Obtaining each timing, generating a second symbol synchronization timing signal based on the L timings, generating a third symbol synchronization timing signal based on the timing of the preceding wave, and the symbol synchronization A step of adaptively switching the timing signal, the second symbol synchronization timing signal, and the third symbol synchronization timing signal as a symbol synchronization timing signal used for the sampling of the received signal according to communication quality. An adaptive equalization method characterized by

A method according to claim 13 or 18,
The L or K is plural, and the adaptive equalization processing calculates error signal power between the plurality of sampled received signals and each replica signal, adds these error signal powers, and adds the error signal powers An adaptive equalization method characterized by performing adaptive equivalence processing based on the above and performing symbol determination.

A method according to claim 13 or 18,
The L or K is plural, and the adaptive equalization processing performs adaptive equalization processing for each of the plurality of sampled received signals, and processes the adaptive equalization processing according to the communication quality. An adaptive equalization method characterized by performing the final result.

A method according to claim 13, 17 or 18,
An adaptive equalization method characterized in that the adaptive equalization processing performs signal processing in a spatial domain together with equalization processing in a time domain.

A method according to claim 13, 17 or 18,
Weights are applied to received signals from a plurality of antennas to form multi-beam directivity characteristics, and the correlation calculation is performed for each received signal from each beam of the multi-beam, based on the power sum or the variance value. An adaptive equalization method characterized by obtaining a symbol synchronization timing signal and sampling a beam reception signal corresponding to each symbol synchronization timing signal.