JP2004096371A - Wireless transmission method and apparatus for digital information - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a wireless transmission method and apparatus for digital information capable of realizing 100Mbps high speed transmission in a sufficient communication area in the SISO transmission. <P>SOLUTION: Digital information is sampled at a 1/N interval (N is an integer of 2 or over) of a symbol period T of the digital information to extract a transmission pulse train, and the transmission method is configured such that the transmission pulse train is transmitted via a wireless transmission channel subjected to band limit wherein the transmission pulse train takes nearly the same impulse response at N sets of consecutive sampling points and takes an impulse response closer to zero more than the value of nearly the same impulse response at sampling points other than the N-sets of sampling points. <P>COPYRIGHT: (C)2004,JPO

Description

ここで、ρは受信ＣＮＲである。上式より、ＭＩＭＯ伝送を用いると、３ｄＢのＣＮＲ劣化を許容することにより、伝送速度を２倍に高めることが可能であることが分かる。
【００１３】
（２）式は、Ｎ多重伝送において、多重化に起因した通信品質の劣化が信号電力のＮ倍の増大のみである場合の通信容量の一般的な数式表現である。以上より、ＳＩＳＯ伝送においても、Ｎ倍の信号電力の増大以外には品質劣化を生じない形でＮ多重伝送を実現することができれば、十分な通信エリアでの１００Ｍｂｐｓ　高速伝送を実現する可能性があることが分かる。
【００１４】
本発明によるディジタル情報の無線伝送方式では、
ディジタル情報を該ディジタル情報のシンボル周期Ｔの１／Ｎ間隔（Ｎは２以上の整数）でサンプリングを行って伝送パルス列を取り出し、該伝送パルス列は連続するＮ個のサンプリング点でほぼ同一のインパルス応答の値をとり当該Ｎ個のサンプリング点以外のサンプリングでは前記ほぼ同一のインパルス応答の値よりも零に近いインパルス応答の値をとるような帯域制限された無線伝送チャネルを介して伝送（以下本願では、このディジタル情報の無線伝送の形式を、「分数タップのパーシャルレスポンス伝送」という）を行うことにより、信号電力がＮ倍に増大することを除く品質劣化はほとんど生じないＮ多重伝送を実現している。
【００１５】
本願発明による分数タップのパーシャルレスポンス伝送の性能について、通常のパーシャルレスポンス伝送の性質と対比して以下に詳述する。分数タップのパーシャルレスポンス伝送におけるインパルス応答の一例を図８に示す。参考のために、通常の伝送におけるインパルス応答の一例を図９に示す。また、従来のパーシャルレスポンス伝送におけるインパルス応答の一例を図１０に示す。図８，図９および図１０において、Ｔは情報変調のシンボル周期を表している。また、受信機における理想的な復調タイミングに対応したサンプリングタイミングを黒丸で示してある。
【００１６】
図９に示すように、通常の伝送においては、シンボル周期Ｔ毎にサンプリングが行われる。また、隣接シンボルからの符号間干渉がほぼ零となるようなインパルス応答を用いた帯域制限が一般に行われる。また、図１０に示すように、従来のパーシャルレスポンス伝送においては、信号成分とほぼ等しい大きさの隣接シンボルからの符号化間干渉が存在する。ただし、従来のパーシャルレスポンス伝送においては、サンプリングはシンボル周期Ｔ毎に行われる。
【００１７】
これに対して、本発明によるディジタル情報の無線伝送方法である分数タップのパーシャルレスポンス伝送では、シンボル周期Ｔの１／Ｎ間隔（Ｎは２以上の整数）でサンプリングが行われる。図８に示した例においては、シンボル周期Ｔの１／２間隔でのサンプリングが行われている。分数タップのパーシャルレスポンス伝送においては、連続するＮ個のサンプリング点でほぼ同一のインパルス応答の値をとり、それら以外のサンプリング点では、より零に近いインパルス応答の値をとるような帯域制限を行う。図８に示した例においては、インパルス応答は２個のサンプリング点でほぼ同一の値をとり、それ以外のサンプリング点ではほぼ零となっている。
【００１８】
図８は、本発明によるディジタル情報の無線伝送方法で使用するインパルス応答を模式的に表現したものである。このようなインパルス応答は、ディジタルフィルタを用いて近似的に実現することができる。本発明によるディジタル情報の無線伝送方法で使用するインパルス応答の近似的な実現例▲１▼を図１２に示す。図１２では、ｔ＝０について対称なインパルス応答のｔ＞０の部分のみを示してある。図１２において、シンボル周期はＴ_ｓであり、サンプリング点は、ｔ＝±Ｔ／４，±３Ｔ／４，±５Ｔ／４，…である。ｔ＝±Ｔ／４を除くインパルス応答の値は、ほぼ零となっている。参考のために、ロールオフ率５０％および１００％におけるナイキストフィルタのインパルス応答▲２▼，▲３▼も示す。
【００１９】
図１３には、図１２に示した本発明によるインパルス応答の周波数特性▲４▼が示してある。参考のため、現状の無線ＬＡＮにおいて、最も一般的に使用されている帯域制限フィルタの特性▲８▼およびベースバンド受信信号のスペクトル▲５▼およびロールオフ率５０％および１００％におけるナイキストフィルタの周波数特性▲６▼，▲７▼も示してある。現状の無線ＬＡＮと比較すると、伝送信号の帯域幅をわずかに拡げるだけで、２多重伝送が実現できることが分かる。
多重数Ｎに関する理論的な制約条件は存在しないが、実用上はＮ＝２〜４とすることが現実的である。
【００２０】
以下では、まず、本発明による分数タップのパーシャルレスポンス伝送において、連続するＮ個のサンプリング点が同一のインパルス応答値をとり、かつそれら以外のサンプリング点のインパルス応答値が零である場合の雑音特性，周波数特性について明らかにする。
同一インパルス応答値を有する連続するＮ個のサンプリング点以外のサンプリング点におけるインパルス応答の値が零でない場合の特性については、後で詳述する。
【００２１】
以下では、Ｎ＝２の場合の分数タップパーシャルレスポンス伝送と通常の伝送について、伝送速度，雑音特性およびスペクトル特性の比較を行う。これらの場合における帯域制限フィルタのインパルス応答は、図８および図９に示してある。以下では、情報変調がＢＰＳＫ（Ｂｉｎａｒｙ　Ｐｈａｓｅ−Ｓｈｉｆｔ　Ｋｅｙｉｎｇ）の場合について説明するが、本発明のディジタル情報の無線伝送方法は、情報変調が任意の線形変調である場合に適用可能である。
【００２２】
Ｎ＝２の場合に、Ｍをシンボル数としたとき、Ｍ・Ｔ期間内での分数タップパーシャルレスポンス伝送および通常の信号伝送における受信信号は、各々次式（３），（４）で表現される。
【数２】

ここで、ｒ_１＋Ｄ（ｔ）　およびｒ_１（ｔ）　は、各々分数タップパーシャルレスポンス伝送および通常の伝送における受信信号を表す。また、ｈ_１＋Ｄ（ｔ）　およびｈ_１（ｔ）　は、各々分数タップパーシャルレスポンス伝送および通常のＢＰＳＫ伝送における送受信を合わせた帯域制限フィルタのインパルス応答を表す。また、上式において、ａ_ｉはｉ番目の送信シンボルを表している。ＢＰＳＫ変調を行うので、ａ_ｉの取り得る値は＋１または−１の２値となる。上式（３），（４）において、ｎ（ｔ）　は付加的なガウス雑音を表している。
【００２３】
（４）式において表現されているように、通常のＢＰＳＫ伝送においては、情報シンボルを時間Ｔ間隔で配置し、Ｍ・Ｔ期間内にＭシンボルの情報を伝送する。一方、（３）式から分かるように、分数タップパーシャルレスポンス伝送においては、情報シンボルを時間Ｔ／２間隔で配置し、同一期間内に２Ｍシンボルの情報を伝送する。このように、分数タップパーシャルレスポンス伝送を用いると、通常の伝送と比較して２倍の伝送速度を実現することができる。ここではＮ＝２の場合について説明したが、同様に、連続するＮ個のサンプリング点が同一のインパルス応答値をとるような分数タップパーシャルレスポンス伝送を用いると、通常の伝送と比較してＮ倍の伝送速度を実現することができる。
【００２４】
次に、本願発明による分数タップパーシャルレスポンス伝送および通常の伝送における雑音特性の比較を行う。
簡単のため、復調タイミングは理想的であるものとする。すなわち、次式（５），（６）が成り立つものとする。
【数３】

【００２５】
ところで、サンプリングされた受信信号は次式（７），（８）で表現される。（７），（８）式においては、時間に関する連続関数の値とそのサンプリング値を、同一関数名で表現している。
【数４】

【００２６】
（７），（８）式に（５），（６）式を代入することにより、次式（９），（１０）が得られる。
【数５】

【００２７】
（１０）式は、ＢＰＳＫ伝送において符号干渉が存在しない場合の受信信号を表している。ａ_ｉの取り得る値は＋１または−１の２値であるから、｜ｎ（ｋ）｜＜１が成り立つ場合には、データ復調における判定誤りは発生しない。この場合におけるビット誤り率は、平均雑音電力をＰとすると、次式（１１），（１２）で表現される。
【数６】

【００２８】
一方、（９）式は、パーシャルレスポンス符号の一種であるデュオバイナリ符号を用いて信号伝送を行った場合の受信信号と同一表現となる。デュオバイナリ符号を用いた信号伝送においては、誤り伝播を防止する目的でプリコーディングが行われることが一般的である。デュオバイナリ符号に対するプリコーディングは、次式（１３）で表現される。
【数７】

【００２９】
ここで、ｃ_Ｋは送信２進データ系列であり、ｄ_Ｋはプリコーディング出力である。ｄ_ｋも２進データ系列となる。ｄ_ｋをＢＰＳＫ変調の信号点にマッピングしたものがａ_ｋとなる。
送信２進データ系列ｃ_ｋ，プリコーディング出力ｄ_ｋ，マッピング出力ａ_ｋおよび受信信号ｒ_１＋Ｄ（ｋ）の関数を表１に示す。表１においては、ガウス雑音の項は省略してある。
【００３０】
【表１】

【００３１】
表１より、送信２進データ系列ｃ_ｋが１のとき受信信号ｒ_１＋Ｄ（ｋ）は０となり、ｃ_ｋが０のときには受信信号ｒ_１＋Ｄ（ｋ）は−２または＋２となることが分かる。すなわち、受信信号ｒ_１＋Ｄ（ｋ）が０であるか±２であるかを識別することにより、送信データ系列ｃ_ｋの復調が可能である。
【００３２】
従って、この場合の信号点間ユークリッド距離は２であり、｜ｎ（ｋ）｜＜１が成り立てば、データ復調における判定誤りは発生しない。このときのビット誤り率は、平均雑音電力をＰとすると、次式（１４）で表現される。
【数８】

【００３３】
（１１）式および（１４）式より、（１０）式で表現されるＢＰＳＫ伝送と（９）式で表現される本発明による分数タップパーシャルレスポンス伝送のビット誤り率特性は同一であることが化分かる。
【００３４】
一方、（１０）式で表現されるＢＰＳＫ伝送の信号電力は１であるのに対して、（９）式で表現される本発明による分数タップパーシャルレスポンス伝送の信号電力は０または４となる。ａ_ｉに関して、＋１と−１の発生頻度が同じであるものと仮定すると、（９）式で表現される受信信号の平均電力は２となる。
【００３５】
ａ_ｉに関する＋１と−１の発生頻度は同じであるものとし、受信信号における平均信号対雑音電力比（受信ＳＮＲ）をρとすると、次式（１５），（１６）が成り立つ。
【数９】

【００３６】
以上より、ＢＰＳＫ伝送において、本発明によるＮ＝２の分数タップパーシャルレスポンス符号化を行うことにより、ビット誤り率特性の３ｄＢの劣化を許容すれば、２倍の高速化が実現できることが分かる。
同様に、ＱＰＳＫ伝送においても、Ｉ成分とＱ成分の両方に本発明によるＮ＝２の分数タップパーシャルレスポンス符号化を行うことにより、２倍の高速化を実現することができる。
【００３７】
ＱＰＳＫ伝送におけるサンプリングされた受信信号は、（１０）式で表現される。ただし、この場合には、ａ_ｉの取り得る値は＋１＋ｊ，＋１−ｊ，−１＋ｊ，−１−ｊの４値となる。｜ｎ（ｋ）｜＜１が成り立つ場合には、データ復調における判定誤りは発生しないので、平均雑音電力をＰとすると、ビット誤り率は次式（１７）で表現される。
【数１０】

【００３８】
一方、ＱＰＳＫ伝送において、本発明によるＮ＝２の分数タップパーシャルレスポンス符号化を行った場合のサンプリングされた受信信号は、（７）式で表現される。ただし、この場合にも、ａ_ｉの取り得る値は＋１＋ｊ，＋１−ｊ，−１＋ｊまたは−１−ｊの４値となる。
【００３９】
この場合においても、｜ｎ（ｋ）｜＜１が成り立つ場合には、データ復調における判定誤りは発生しないので、平均雑音電力をＰとすると、ビット誤り率は次式（１８）で表現される。
【数１１】

【００４０】
ＱＰＳＫ伝送において、（８）式で表現される受信信号の信号電力は２であるので、受信信号の平均ＳＮＲをρとすると、次式（１９）が成り立つ。
【数１２】

【００４１】
一方、ＱＰＳＫ伝送において、（７）式で表現される本発明による分数タップパーシャルレスポンス符号化を行った場合の受信信号の信号電力は０、４または８となる。ａ_ｉに関して、４通りの値（＋１＋ｊ，＋１−ｊ，−１＋ｊ，−１−ｊ）の発生頻度が同じであるものと仮定すると、（９）式で表現される受信信号の平均電力は４となる。このとき、次式が成り立つ。
【数１３】

【００４２】
以上より、ＱＰＳＫ伝送においても、本発明によるＮ＝２の分数タップパーシャルレスポンス符号化を行うことにより、ビット誤り率特性の３ｄＢの劣化を許容すれば、２倍の高速化が実現できることが分かる。
【００４３】
ＱＰＳＫ伝送およびＱＰＳＫ伝送に本発明によるＮ＝２の分数タップパーシャルレスポンス符号化を行った場合（以下では、ＦＰＲ（１，１）と略記する）に得られる通信容量を図１３に示す。
【００４４】
次に、ＱＰＳＫ伝送とＦＰＲ（１，１）伝送の周波数特性の比較を行う。図８と図９を比較すると、ＱＰＳＫ伝送における帯域制限フィルタのインパルス応答の時間幅は、ＦＰＲ（１，１）伝送の帯域制限フィルタと比較すると４／３倍であることが分かる。すなわち、同一特性の帯域制御フィルタ（例えば、ナイキストフィルタ）を使用した場合には、ＱＰＳＫ伝送と比較するとＦＰＲ（１，１）伝送の伝送帯域幅は４／３倍となる。
【００４５】
以上では、本発明による分数タップパーシャルレスポンス伝送において、連続するＮ個のサンプリング点が同一のインパルス応答値をとり、かつそれら以外のサンプリング点でのインパルス応答値が零である場合の特性について明らかにした。
【００４６】
次に、同一インパルス応答値を有する連続するＮ個のサンプリング点以外のサンプリング点におけるインパルス応答の値が零でない場合の特性について説明する。
本発明によるＮ＝２の分数タップパーシャルレスポンス符号化において、ＱＰＳＫ伝送と同一特性のフィルタで帯域制限を行った場合のインパルス応答を図１４に示す。この場合には、同一のインパルス応答値をとる連続する２個のサンプリング点の前後のサンプリング点において、インパルス応答が零ではない値をとっている。
【００４７】
インパルス応答が零ではない値をとる前後のサンプリング点のうち、過去側のサンプリング点からの符号間干渉については、パーシャルレスポンス符号の復号にＤＤＦＳＥ（Ｄｅｌａｙｅｄ　Ｄｅｃｉｓｉｏｎ　Ｆｅｅｄｂａｃｋ　Ｓｅｑｕｅｎｃｅ　Ｅｓｔｉｍａｔｏｒ）を用いることにより取り除くことが可能である。
【００４８】
以下では、この点について説明する。
一般に、パーシャルレスポンス符号の復号を、ビタビ復号法を用いて行うことが可能であることは知られている。ＰＲ（１，１）伝送のビタビ復号におけるメトリックは、次式（２１）により計算される。
【数１４】

【００４９】
（７）式および（１９）式より、真の伝送データに対応したトレリス状態におけるメトリックは、次式（２２）で表現される。
【数１５】

【００５０】
一方、過去側のサンプリング点からの符号間干渉ｐ（｜ｐ｜≪１）が存在する場合のサンプリングされた受信信号は次式（２３）で表現される。
【数１６】

ＤＤＦＳＥにおいては、トレリス状態毎に過去の判定データ（∧ａ_ｋ−２，∧ａ_ｋ−３，）を記憶しているので、∧ａ_ｋ−２を用いて過去側のサンプリング点からの符号間干渉をキャンセルすることができる。
【００５１】
【実施例】
図１は本発明の一実施例における、ディジタル情報を無線伝送するための装置の送信機のブロック構成である。
図１において、ディジタル情報１を送信データ分割手段２によりサブ系列３−１からサブ系列３−Ｎに分割し、前記サブ系列３−１からサブ系列３−Ｎそれぞれのディジタル情報に基づき、情報変調手段４−１から情報変調手段４−Ｎによってベースバンド変調信号５−１からベースバンド変調信号５−Ｎを生成している。
前記ベースバンド変調信号５−１からベースバンド変調信号５−Ｎをそれぞれの信号遅延手段６−１から信号遅延手段６−Ｎによってシンボルタイミングが互いに異なるように時間軸をずらして、送信ベースバンド信号加算手段８において算術加算することによりデータ期間の送信ベースバンド信号９を生成している。前記データ期間の送信ベースバンド信号９と、伝送路特性推定のためにプリアンブル生成手段１０において生成した送信プリアンブル変調信号１１とを、フレーム合成手段１２によって合成してフレーム変調信号１３を生成し、アップコンバート手段１４によって前記フレーム変調信号１３を無線周波数に周波数変換した無線周波信号１５を送信アンテナ１６から無線伝送路に送信している。
【００５２】
図２は本発明の一実施例における、ディジタル情報を無線伝送するための装置の受信機のブロック図である。
図２において、受信アンテナ２０から受信した無線周波数の受信信号２１を、ダウンコンバート手段２２によってベースバンド帯域周波数の受信ベースバンドＩＱ信号に周波数変換している。
伝送路特性推定手段２８によって、プリアンブル期間のベースバンド信号からベースバンド変調のシンボル長より短い時間間隔での伝送路特性の推定を行い、推定された前記無線伝送路の伝送路特性情報を用いて、波形等化手段３０によって波形等化を行うことにより、データ期間の受信ベースバンドＩＱ信号を復調して、前記送信側から伝送された前記ディジタル情報の受信データ信号を得ている。
【００５３】
図３は本発明の一実施例において、多重化数が２であり、かつ情報変調がシンボルレート１１Ｍｓｙｍｂｏｌ／ｓｅｃＱＰＳＫの場合の、ディジタル情報を無線伝送するための装置の送信機ブロック構成図である。図３において、送信しようとする４４Ｍｂｐｓ　のデータ１を、送信データ分割回路２において２２Ｍｂｐｓ　のデータ２系統、サブ系列３−１およびサブ系列３−２にシリアル・パラレル変換し、サブ系列３−１は情報変調回路４−１で、サブ系列３−２は情報変調回路４−２によってＱＰＳＫ変調され、ベースバンド変調信号５−１とベースバンド変調信号５−２が出力される。
前記ベースバンド変調信号５−１は、０信号遅延回路６−１で信号を０遅延させベースバンド遅延信号７−１を生成し、前記ベースバンド変調信号５−２はＴ／２信号遅延回路６−で信号をＴ／２遅延させベースバンド遅延信号７−２を生成している。ここで、Ｔはシンボルレートの逆数であり、この場合は約９１ｎｓである。
前記ベースバンド遅延信号７−１と前記ベースバンド遅延信号７−２は送信ベースバンド信号加算回路８において加算され、送信ベースバンド情報変調信号９が出力される。
【００５４】
データの分割と多重化についての時系列の詳細を図５を用いて説明する。ディジタル情報の４４Ｍｂｐｓ　のデータ（ｄ_０〜ｄ_ｎ）を、送信データ分割回路２において２２Ｍｂｐｓ　のデータ２系統，サブ系列３−１（ｄ_０，ｄ_２，ｄ_４，〜，ｄ_ｎ　−１）およびサブ系列２（ｄ_１，ｄ_３，ｄ_５，〜，ｄ_ｍ）にシリアル・パラレル変換し、サブ系列３−１（ｄ_０，ｄ_２，ｄ_４，〜，ｄ_ｎ−１）は情報変調回路４−１でベースバンド変調信号５−１（ａ_０〜ａ_ｎ／２）に、サブ系列３−２は情報変調回路４−２によってベースバンド変調信号５−２（ｂ_０〜ｂ_ｎ）にそれぞれＱＰＳＫ変調される。
前記ベースバンド変調信号５−１のタイミングに対して前記ベースバンド変調信号５−２のタイミングをずらせるために、前記ベースバンド変調信号５−１は０信号遅延回路６−１で信号を０遅延させてベースバンド遅延信号７−１を生成し、前記ベースバンド変調信号５−２はＴ／２信号遅延回路６−２で信号をＴ／２遅延させてベースバンド遅延信号７−２を生成している。
【００５５】
プリアンブル生成回路１０ａにより、受信機においてＴ／２間隔での伝送路特性を求めるための送信プリアンブル信号１１ａを生成し、プリアンブル変調回路１０ｂによりＱＰＳＫ変調されて送信プリアンブル変調信号１１が出力される。前記送信ベースバンド情報変調信号９と前記送信プリアンブル変調信号１１はフレーム合成回路１２によって合成されて、フレーム変調信号１３が出力される。前記フレーム変調信号１３は、帯域制限送信フィルタ１７によって帯域制限されたのち、アップコンバート回路によって無線周波に変調され被変調無線周波が伝送路に送信される。
【００５６】
図４は本発明の一実施例において、多重化数が２であり、かつ情報変調がシンボルレート１１Ｍｓｙｍｂｏｌ／ｓｅｃ　のＱＰＳＫの場合の、ディジタル情報を無線伝送するための装置の受信機ブロック構成図である。図４において、受信アンテナで受信した無線周波数帯の受信信号をダウンコンバート回路によってダウンコンバートして受信ベースバンド信号に変換し、帯域制限受信フィルタにより前記受信ベースバンド信号を帯域制限した帯域制限受信信号をＡ／Ｄ変換回路により受信ディジタル信号に変換される。前記受信ディジタル信号は、２分岐され、一方は伝送路特性推定回路へ、残りは波系等化回路へ各々印加される。伝送路特性推定回路には、例えばマッチドフィルタが含まれており、プリアンブル期間のマッチドフィルタの出力信号を用いて、伝送路のＴ／２間隔のインパルス応答推定を行ってる。前記波形等化回路は、前記伝送路特性推定回路により得られた伝送路のインパルス応答推定値を用いて、波形等化を行う。
【００５７】
等化低域系で表現したベースバンド受信信号をｒ（ｔ）　とする。データ期間において、ｒ（ｔ）　をＴ／２監視区でサセンプリングし、波形等化および復調処理を行う。
【００５８】
以下においては、ｔ＝ｍＴ／２（ｍは整数）におけるｒ（ｔ）　のサンプル値をｒ_ｍと表記する。
また、Ｔ／２間隔で求めた伝送路のインパルス応答をｈ_ｎと表記する。
【００５９】
伝送路のインパルス応答の例を図６に示す。
ｈ_０，ｈ_１以外は、ほぼ零となっている。以下では、ｈ_０を先行成分，ｈ_１を遅延成分とよぶ。
送信データａにおける第ｎ番目の変調信号をａ_ｎ、送信データｂにおける第ｎ番目の変調信号をｂ_ｎと表記する。
【００６０】
送信データａにおいて、最初の変調信号ａ_０の先行成分の受信タイミングをｒ_０とする。このとき、送信データａの第ｎ番目の変調信号の先行成分に対応した受信信号はｒ_２ｎであり、送信データｂの第ｎ番目の変調信号の先行成分に対応した受信信号はｒ_２ｎ＋１である。
【００６１】
送信データａにおける最初の変調信号ａ_０の先行成分の受信タイミングをｔ＝０とする。このとき、送信データａの第ｎ番目の変調信号の先行成分はｔ＝２ｎにおいて受信される。また、送信データｂの第ｎ番目の変調信号の先行成分はｔ＝２ｎ＋１において受信される。ｔ＝２ｎにおいて送信データａの波形等化および復調を行い、ｔ＝２ｎ＋１において送信データｂの波形等化および復調を行う。
【００６２】
このとき、受信信号ｒ_ｎは次式（２４）で表現される。
【数１７】

ここで、ｎ_ｎは雑音成分である。また、無線伝送路での遅延時間は有限であるものとし、ｈ_ｎ＝０（ｎ＜−１，ｎ＞Ｎ）とした。
（２４）式において、第１項目は信号成分を、第２項目は時系列からの符号間干渉を、第３項目はもう一方の系列からの干渉を表す。
【００６３】
ここで、次式（２５）により、変調信号ａ_ｎを表す。
【数１８】

【００６４】
（２４）式と（２５）式より、次式（２６）式が成り立つ。
【数１９】

【００６５】
これは、シンボル長Ｔ／２の場合の符号間干渉のある伝送路での受信信号の一般的な表現であり、ＭＬＳＥ（Ｍａｘｉｍｕｍ　Ｌｉｋｌｉｈｏｏｄ　Ｓｅｑｕｅｎｃｅ　Ｅｓｔｉｍａｔｏｒ），ＤＤＦＳＥ，ＤＦＥ（Ｄｅｃｉｓｉｏｎ　Ｆｅｅｄｂａｃｋ　Ｅｑｕａｌｉｚｅｒ）等の従来の等化器で復調することができる。
【００６６】
【発明の効果】
以上詳細に述べたように、本発明においては、送信機側で、送信データを分割して時間をずらして多重化することにより、多重しない場合に比べ、多重数倍の無線伝送速度を実現することができる。また、受信側でシンボル長より短い時間間隔での伝送路特性情報を用いて波形等化することにより、符号間干渉を精度良く抑圧することができるので、無線伝送帯域幅を変更することなく、高速高品質伝送が可能となる０従って、本発明によるディジタル情報の無線伝送に及ぼす効果は極めて大きい。
【図面の簡単な説明】
【図１】本発明によるディジタル情報の無線伝送方法を説明するためのブロック図である。
【図２】本発明によるディジタル情報の無線伝送方法により伝送された無線周波信号を受信するための受信機の構成例を示すブロック図である。
【図３】本発明によるディジタル情報の無線伝送方法を実施する送信機の構成例を示すブロック図である。
【図４】本発明方法により伝送された無線周波信号を受信する受信機の実施例を示すブロック図である。
【図５】図３に示す構成例の動作を説明するためのタイムチャートである。
【図６】本発明方法により伝送される無線周波信号の伝送路インパルス応答例を示すタイムチャートである。
【図７】従来の無線ＬＡＮ標準の通信容量を示す特性図である。
【図８】本発明に用いる分数タップのパーシャルレスポンス伝送におけるインパルス応答の１例を示す波形図である。
【図９】通常の伝送におけるインパルス応答を示す波形図である。
【図１０】従来のパーシャルレスポンス伝送をした場合のインパルス応答を示す波形図である。
【図１１】ＱＰＳＫ伝送及びＱＰＳＫ伝送に本発明によるＮ＝２の分数タップパーシャルレスポンス符号化を行った場合に得られる通信容量を示す特性図である。
【図１２】帯域制限フィルタのインパルス応答側の２シンボル長を示す特性図である。
【図１３】帯域制限フィルタのインパルス応答側の２／シンボル長を示す特性図である。
【図１４】本発明によるＮ＝２分数タップパーシャルレスポンス符号化においてＱＰＳＫ伝送と同一特性のフィルタで帯域制限を行った場合のインパルス応答を示す波形図である。
【符号の説明】
１　伝送されるディジタル情報
２　送信データ分割手段，送信分割回路
３−１，…３−Ｎ　サブ系列
４−１，…４−Ｎ　情報変調手段，情報変調回路
５−１，…５−Ｎ　ベースバンド変調信号
６−１，…６−Ｎ　信号遅延手段，遅延回路
７−１，７−２　ベースバンド遅延信号
８　ベースバンド信号加算手段，ベースバンド信号加算回路
９　送信ベースバンド情報変調信号
１０　プリアンブル生成手段
１０ａ　プリアンブル生成回路
１０ｂ　プリアンブル変調回路
１１　送信プリアンブル変調信号
１２　フレーム合成回路
１３　フレーム変調信号
１４　アップコンバート回路
１５　無線周波信号
１６　送信アンテナ
１７　帯域制限送信フィルタ
２０　受信アンテナ
２１　受信信号
２２　ダウンコンバート回路
２３　受信ベースバンド信号
２４　帯域制限受信フィルタ
２５　受信帯域制限信号
２６　Ａ／Ｄ変換器
２７　ベースバンドＩ／Ｑ信号
２７−１，２７−２　受信ディジタル信号
２８　伝送路特性推定回路
２９　伝送路特性情報
３０　波形等化手段
３１　受信データ
３２　同期回路[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method and apparatus for wirelessly transmitting digital information.
[0002]
[Prior art]
IEEE 802.11, which was substantially the only international standard for wireless LANs, was established in 1997. Initially, it was not popular due to its reputation for high price and slow operation (up to 2 Mbps), but low-priced products appeared in 1999, and the high-speed standard IEEE 802.11b supporting up to 11 Mbps was formulated, It has been rapidly spreading in Japan since 2000.
However, even with the 11 Mbps standard 11b, the effective throughput is about 4 Mbps, and considering the current situation where 100 base-T Ethernet is becoming widespread, it cannot be said that a sufficient transmission rate is obtained (100 base-T). The -T Ethernet throughput is said to be about 30 Mbps). In a system in which a wired LAN and a wireless LAN coexist, the wireless LAN section becomes a bottleneck at present.
IEEE 802.11 has also formulated a higher-speed standard 11a using a new frequency of 5 GHz (standards 802.11 and 11b are 2.4 GHz systems), and recently 11a-compliant products have appeared. I've been. The transmission speed of 11a is 24 Mbps at maximum, and an effective throughput of 18 Mbps can be theoretically obtained. According to the option 11a, a maximum transmission rate of 54 Mbps can be obtained. However, since high-speed transmission is realized by adopting multi-level modulation of 64 QAM, it is susceptible to noise and multipath, and the communicable range is in a narrow area that can be seen. It has been limited.
[0003]
[Problems to be solved by the invention]
In a wired network in an office, the speed is increasing, and 100base-T is becoming widespread. On the other hand, in a wireless network, the actual effective throughput is about 18 Mbps even in the fastest wireless LAN, and in a mixed system of a wired LAN and a wireless LAN, the wireless LAN becomes a bottleneck.
In the United States and other countries, products with a maximum transmission speed exceeding 100 Mbps are also sold, but they are an extension of the 54 Mbps option of 11a, and the communication range is limited to a narrow area that can be seen. In addition, they cannot be used in Japan due to problems in the Radio Law.
That is, there is a need for a high-speed wireless LAN transmission system that can secure a sufficient communication area and achieve an effective throughput of 30 Mbps or more.
In fact, MMAC, a wireless LAN standardization organization in Japan, and IEEE 802.11 in the United States are studying a high-speed transmission method capable of realizing 100 Mbps or more.
[0004]
FIG. 7 shows the communication capacity of the existing wireless LAN standard. In FIG. 7, Es / No (signal energy per symbol divided by noise power density) and transmission speed (Mbps) when the packet error rate is 1% in the packet transmission with a data length of 1000 bytes. The relationship is shown. For reference, the communication capacity of Shannon in a SISO (Single Input / Single Output) transmission line and a MIMO (Multiple Input / Multiple Output) transmission line is also shown. As can be seen from FIG. 7, the communication capacity of the existing wireless LAN standard schemes Barker, CCK, PBCC and OFDM is smaller than the Shannon capacity of SISO transmission, and even if this is increased to 100 Mbps by a method such as high efficiency modulation, The required Es / No is greater than 25 dB. The communicable distance at this time is several meters or less, and it is not possible to secure a sufficient communication area as a wireless LAN.
[0005]
On the other hand, FIG. 7 shows that when performing MIMO transmission, high-speed transmission of 100 Mbps may be realized in a sufficient communication area. However, considering that the capacity of the actual system is smaller than the Shannon capacity, it is necessary to perform at least 4 × 4 or more MIMO transmission in order to secure a sufficient communication area. There is a problem that becomes. Particularly, in a PC card type wireless LAN terminal for a notebook personal computer, it is almost impossible to mount four independent antennas.
[0006]
SUMMARY OF THE INVENTION The present invention has been made to solve the problems described above, and it is an object of the present invention to provide a method and an apparatus for wirelessly transmitting digital information capable of realizing 100 Mbps high-speed transmission in a sufficient communication area in SISO transmission. Aim.
[0007]
[Means for Solving the Problems]
In order to achieve this object, a method for wirelessly transmitting digital information according to the present invention comprises the steps of: take out,
The transmission pulse train takes substantially the same impulse response value at successive N sampling points, and takes an impulse response value closer to zero than the substantially identical impulse response value at samplings other than the N sampling points. It is configured to be transmitted through such a band-limited wireless transmission channel.
The transmission can be configured to suppress the intersymbol interference component caused by the tail of the impulse response by the decision feedback type waveform equalization.
Also, the transmission can be configured to suppress the intersymbol interference component caused by the impulse response head by subtracting the interference replica generated in advance.
Trellis-modulated digital information can be used as the digital information.
[0008]
Also, the wireless transmission method of digital information according to the present invention, on the transmitting side,
Dividing digital information into a plurality of sub-sequences,
Generate a plurality of baseband modulation signals based on the digital information for each sub-sequence,
Generate a transmission baseband signal of the data period by arithmetically adding the plurality of baseband modulation signals by shifting the time axis so that the symbol timings are different from each other,
A preamble signal for transmission characteristic estimation and a signal obtained by frequency-converting the transmission baseband signal of the subsequent data period into a radio frequency are transmitted to a radio transmission path,
On the receiving side,
Frequency conversion of the received radio frequency signal into a baseband band,
Using the received baseband signal in the preamble period, the transmission path characteristics are estimated at time intervals shorter than the symbol length of the baseband modulation,
By performing waveform equalization using the transmission path characteristic information of the estimated wireless transmission path, to demodulate the received baseband signal in the data period,
It is configured to obtain a reception data signal of the digital information transmitted from the transmission side.
Since the transmission baseband signal in the data period is a signal obtained by arithmetically adding a plurality of baseband modulation signals while shifting the time axis so that the symbol timings are different from each other, intersymbol interference occurs.
Therefore, the receiver uses the received baseband signal in the preamble period to estimate the transmission path characteristics at time intervals shorter than the symbol length of the baseband modulation, and performs waveform equalization using the estimated impulse response. This demodulates the received baseband signal in the multiplexed data period.
[0009]
BEST MODE FOR CARRYING OUT THE INVENTION
Next, the theoretical basis that the wireless transmission method of digital information according to the present invention can achieve 100 Mbps high-speed transmission in a sufficient communication area while performing SISO transmission will be described below.
[0010]
As described above, the upper limit of the communication capacity of the existing wireless LAN standard system is suppressed by the communication capacity of Shannon in the SISO transmission path. The above-mentioned existing wireless LAN standard systems include a trellis modulation system (a combination of high-efficiency modulation and trellis coding to increase the distance between signals and improve noise characteristics), which is a type of coded modulation system, and OFDM. (Orthogonal Frequency Division Multiplexing). In order to realize 100 Mbps high-speed transmission that can provide a sufficient communicable area, it is necessary to develop a communication method that exceeds the above-mentioned Shannon capacity.
If a communication method using a MIMO transmission path such as space-time transmission is used, the conditions of high speed and a sufficient communication area can be satisfied, but four or more antennas are required for both the transmitter and the receiver. Therefore, it is extremely difficult to realize a practical device using this communication method.
[0011]
By the way, the communication capacity in an ideal MIMO transmission line where the correlation between the SISO transmission line and the antenna is all zero is expressed by the following equations (1) and (2), respectively.
[0012]
(Equation 1)

Here, ρ is the received CNR. From the above equation, it can be seen that when MIMO transmission is used, the transmission speed can be doubled by allowing 3 dB of CNR degradation.
[0013]
Equation (2) is a general mathematical expression of the communication capacity in the case of the N-multiplex transmission where the deterioration of the communication quality due to the multiplexing is only N-times increase of the signal power. As described above, even in SISO transmission, if N-multiplex transmission can be realized in a form that does not cause quality deterioration except for an N-fold increase in signal power, it is possible to realize 100 Mbps high-speed transmission in a sufficient communication area. You can see that there is.
[0014]
In the wireless transmission system for digital information according to the present invention,
The digital information is sampled at intervals of 1 / N (N is an integer of 2 or more) of the symbol period T of the digital information to extract a transmission pulse train, and the transmission pulse train has substantially the same impulse response at N consecutive sampling points. Is transmitted through a band-limited radio transmission channel that takes an impulse response value closer to zero than the substantially same impulse response value at the sampling other than the N sampling points (hereinafter, referred to as the present application). This type of wireless transmission of digital information is referred to as "fractional tap partial response transmission", thereby realizing N multiplex transmission with almost no quality deterioration except that the signal power is increased N times. I have.
[0015]
The performance of fractional tap partial response transmission according to the present invention will be described in detail below in comparison with the characteristics of ordinary partial response transmission. FIG. 8 shows an example of the impulse response in the fractional tap partial response transmission. For reference, an example of an impulse response in normal transmission is shown in FIG. FIG. 10 shows an example of an impulse response in a conventional partial response transmission. 8, 9, and 10, T represents the symbol period of information modulation. The sampling timing corresponding to the ideal demodulation timing in the receiver is indicated by a black circle.
[0016]
As shown in FIG. 9, in normal transmission, sampling is performed every symbol period T. In addition, band limitation using an impulse response such that intersymbol interference from adjacent symbols becomes almost zero is generally performed. Further, as shown in FIG. 10, in the conventional partial response transmission, there is inter-coding interference from an adjacent symbol having a size substantially equal to a signal component. However, in the conventional partial response transmission, sampling is performed every symbol period T.
[0017]
In contrast, in fractional tap partial response transmission, which is a digital information wireless transmission method according to the present invention, sampling is performed at 1 / N intervals (N is an integer of 2 or more) of the symbol period T. In the example shown in FIG. 8, sampling is performed at a half interval of the symbol period T. In partial response transmission with fractional taps, band limitation is performed such that substantially the same value of impulse response is taken at N consecutive sampling points and the value of impulse response is closer to zero at other sampling points. . In the example shown in FIG. 8, the impulse response has almost the same value at two sampling points, and is almost zero at the other sampling points.
[0018]
FIG. 8 schematically illustrates an impulse response used in the digital information wireless transmission method according to the present invention. Such an impulse response can be approximately realized using a digital filter. FIG. 12 shows an approximate example (1) of the impulse response used in the wireless transmission method of digital information according to the present invention. FIG. 12 shows only a part of the impulse response symmetric about t = 0, where t> 0. In FIG. 12, the symbol period is T _s And the sampling points are t = ± T / 4, ± 3T / 4, ± 5T / 4,... The value of the impulse response except for t = ± T / 4 is almost zero. For reference, impulse responses (2) and (3) of the Nyquist filter at roll-off rates of 50% and 100% are also shown.
[0019]
FIG. 13 shows the frequency characteristic (4) of the impulse response according to the present invention shown in FIG. For reference, in the current wireless LAN, the characteristic (8) of the band-limiting filter most commonly used, the spectrum (5) of the baseband received signal, and the frequency of the Nyquist filter at the roll-off rates of 50% and 100% Characteristics (6) and (7) are also shown. Compared to the current wireless LAN, it can be seen that two-multiplex transmission can be realized only by slightly increasing the bandwidth of the transmission signal.
Although there is no theoretical constraint on the multiplex number N, it is practical to set N = 2 to 4 in practical use.
[0020]
In the following, first, in the fractional tap partial response transmission according to the present invention, noise characteristics when N consecutive sampling points have the same impulse response value and the impulse response values of the other sampling points are zero And frequency characteristics.
The characteristics when the value of the impulse response at non-consecutive N sampling points having the same impulse response value is not zero will be described later in detail.
[0021]
In the following, the transmission rate, noise characteristics, and spectrum characteristics of the fractional tap partial response transmission and the normal transmission when N = 2 are compared. The impulse response of the band limiting filter in these cases is shown in FIGS. Hereinafter, the case where the information modulation is BPSK (Binary Phase-Shift Keying) will be described. However, the wireless transmission method of digital information of the present invention is applicable when the information modulation is any linear modulation.
[0022]
When N = 2 and M is the number of symbols, the received signals in the fractional tap partial response transmission and the normal signal transmission within the MT period are expressed by the following equations (3) and (4), respectively. You.
(Equation 2)

Where r _{1 + D} (T) and r ₁ (T) represents a received signal in fractional tap partial response transmission and normal transmission, respectively. Also, h _{1 + D} (T) and h ₁ (T) represents the impulse response of the band-limiting filter that combines transmission and reception in fractional tap partial response transmission and normal BPSK transmission. In the above equation, a _i Represents the i-th transmission symbol. Since BPSK modulation is performed, a _i Is a binary value of +1 or -1. In the above equations (3) and (4), n (t) represents additional Gaussian noise.
[0023]
As expressed in equation (4), in normal BPSK transmission, information symbols are arranged at time T intervals, and information of M symbols is transmitted within an MT period. On the other hand, as can be seen from equation (3), in fractional tap partial response transmission, information symbols are arranged at intervals of time T / 2, and information of 2M symbols is transmitted within the same period. As described above, when the fractional tap partial response transmission is used, a transmission speed twice as high as that of the normal transmission can be realized. Here, the case of N = 2 has been described. Similarly, when fractional tap partial response transmission in which N consecutive sampling points have the same impulse response value is used, N times as much as normal transmission is used. Transmission speed can be realized.
[0024]
Next, the noise characteristics of the fractional tap partial response transmission according to the present invention and the noise characteristics of the normal transmission will be compared.
For simplicity, it is assumed that the demodulation timing is ideal. That is, it is assumed that the following equations (5) and (6) hold.
[Equation 3]

[0025]
By the way, the sampled received signal is expressed by the following equations (7) and (8). In equations (7) and (8), the value of a continuous function with respect to time and its sampling value are represented by the same function name.
(Equation 4)

[0026]
By substituting equations (5) and (6) into equations (7) and (8), the following equations (9) and (10) are obtained.
(Equation 5)

[0027]
Equation (10) represents a received signal when there is no code interference in BPSK transmission. a _i Is a binary value of +1 or -1. Therefore, when | n (k) | <1, a decision error in data demodulation does not occur. The bit error rate in this case is expressed by the following equations (11) and (12), where P is the average noise power.
(Equation 6)

[0028]
On the other hand, Expression (9) is the same expression as a received signal when signal transmission is performed using a duobinary code, which is a kind of partial response code. In signal transmission using a duobinary code, precoding is generally performed for the purpose of preventing error propagation. The precoding for the duobinary code is expressed by the following equation (13).
(Equation 7)

[0029]
Where c _K Is a transmission binary data sequence, d _K Is the precoding output. d _k Is also a binary data sequence. d _k Is mapped to a signal point of BPSK modulation. _k It becomes.
Transmission binary data sequence c _k , Precoding output d _k , Mapping output a _k And the received signal r _{1 + D} Table 1 shows the function of (k). In Table 1, the term of Gaussian noise is omitted.
[0030]
[Table 1]

[0031]
From Table 1, it can be seen that the transmission binary data sequence c _k Is 1 when the received signal r _{1 + D} (K) becomes 0 and c _k Is 0, the received signal r _{1 + D} It can be seen that (k) is -2 or +2. That is, the received signal r _{1 + D} By identifying whether (k) is 0 or ± 2, the transmission data sequence c _k Can be demodulated.
[0032]
Therefore, in this case, the Euclidean distance between signal points is 2, and if | n (k) | <1, a determination error in data demodulation does not occur. The bit error rate at this time is represented by the following equation (14), where P is the average noise power.
(Equation 8)

[0033]
From the equations (11) and (14), it can be seen that the bit error rate characteristics of the BPSK transmission expressed by the equation (10) and the fractional tap partial response transmission according to the present invention expressed by the equation (9) are the same. I understand.
[0034]
On the other hand, the signal power of the BPSK transmission expressed by the equation (10) is 1, whereas the signal power of the fractional tap partial response transmission according to the present invention expressed by the equation (9) is 0 or 4. a _i Assuming that the occurrence frequencies of +1 and −1 are the same, the average power of the received signal expressed by the equation (9) is 2.
[0035]
a _i Assuming that the frequencies of occurrence of +1 and −1 are the same, and the average signal-to-noise power ratio (received SNR) in the received signal is ρ, the following equations (15) and (16) hold.
(Equation 9)

[0036]
From the above, it can be seen that in the BPSK transmission, by performing the fractional tap partial response coding of N = 2 according to the present invention, if the degradation of the bit error rate characteristic by 3 dB is allowed, the speedup can be doubled.
Similarly, in QPSK transmission, by performing N = 2 fractional tap partial response encoding according to the present invention on both the I component and the Q component, it is possible to achieve twice as high speed.
[0037]
The sampled received signal in QPSK transmission is expressed by equation (10). However, in this case, a _i Are four values of + 1 + j, + 1-j, -1 + j, and -1-j. When | n (k) | <1, a decision error in data demodulation does not occur, and if the average noise power is P, the bit error rate is expressed by the following equation (17).
(Equation 10)

[0038]
On the other hand, in QPSK transmission, a sampled received signal when N = 2 fractional tap partial response encoding according to the present invention is performed is expressed by equation (7). However, also in this case, a _i Are four values of + 1 + j, + 1-j, -1 + j or -1-j.
[0039]
Also in this case, if | n (k) | <1, a decision error in data demodulation does not occur, and if the average noise power is P, the bit error rate is expressed by the following equation (18). .
[Equation 11]

[0040]
In QPSK transmission, the signal power of the received signal expressed by the equation (8) is 2, and the following equation (19) holds when the average SNR of the received signal is ρ.
(Equation 12)

[0041]
On the other hand, in the QPSK transmission, the signal power of the received signal when the fractional tap partial response encoding according to the present invention expressed by Expression (7) is performed is 0, 4, or 8. a _i Assuming that the occurrence frequencies of the four values (+ 1 + j, + 1-j, -1 + j, -1-j) are the same, the average power of the received signal expressed by the equation (9) is 4. . At this time, the following equation is established.
(Equation 13)

[0042]
From the above, it can be seen that even in QPSK transmission, by performing the fractional tap partial response encoding of N = 2 according to the present invention, if the degradation of the bit error rate characteristic by 3 dB is allowed, the speedup can be doubled.
[0043]
FIG. 13 shows communication capacities obtained when N = 2 fractional tap partial response encoding according to the present invention is performed on QPSK transmission and QPSK transmission (hereinafter, abbreviated as FPR (1, 1)).
[0044]
Next, the frequency characteristics of QPSK transmission and FPR (1, 1) transmission are compared. 8 and 9 that the time width of the impulse response of the band-limiting filter in QPSK transmission is 4/3 times that of the band-limiting filter of FPR (1, 1) transmission. That is, when a band control filter (for example, a Nyquist filter) having the same characteristics is used, the transmission bandwidth of FPR (1, 1) transmission is 4/3 times that of QPSK transmission.
[0045]
In the above, in the fractional tap partial response transmission according to the present invention, the characteristics when the N consecutive sampling points have the same impulse response value and the impulse response values at the other sampling points are zero are evident. did.
[0046]
Next, a description will be given of a characteristic in a case where the value of the impulse response at a sampling point other than the N consecutive sampling points having the same impulse response value is not zero.
FIG. 14 shows an impulse response when band limitation is performed with a filter having the same characteristics as that of QPSK transmission in N = 2 fractional tap partial response encoding according to the present invention. In this case, the impulse response has a non-zero value at the sampling points before and after two consecutive sampling points having the same impulse response value.
[0047]
Among the sampling points before and after the impulse response takes a value other than zero, intersymbol interference from the past sampling point can be removed by using DDFSE (Delayed Decision Feedback Back Estimator) for decoding the partial response code. It is possible.
[0048]
Hereinafter, this point will be described.
It is generally known that decoding of a partial response code can be performed using a Viterbi decoding method. A metric in Viterbi decoding of PR (1, 1) transmission is calculated by the following equation (21).
[Equation 14]

[0049]
From the equations (7) and (19), the metric in the trellis state corresponding to the true transmission data is expressed by the following equation (22).
[Equation 15]

[0050]
On the other hand, when there is intersymbol interference p (| p | ≪1) from the past sampling point, the sampled received signal is expressed by the following equation (23).
(Equation 16)

In DDFSE, past judgment data ($ a _k-2 , ∧a _k-3 ,), So ∧a _k-2 Can be used to cancel intersymbol interference from a past sampling point.
[0051]
【Example】
FIG. 1 is a block diagram of a transmitter of a device for wirelessly transmitting digital information according to an embodiment of the present invention.
In FIG. 1, digital information 1 is divided into sub-sequences 3-1 to 3-N by transmission data dividing means 2, and information modulation is performed based on the digital information of each of the sub-sequences 3-1 to 3-N. The baseband modulation signal 5-N is generated from the baseband modulation signal 5-1 by the information modulation unit 4-N from the unit 4-1.
The transmission baseband signal is shifted from the baseband modulation signal 5-1 to the baseband modulation signal 5-N by the signal delay means 6-1 to the signal delay means 6-N so that the symbol axes are different from each other. The transmission baseband signal 9 in the data period is generated by arithmetic addition in the adding means 8. The transmission baseband signal 9 in the data period and the transmission preamble modulation signal 11 generated by the preamble generation means 10 for estimating the channel characteristics are synthesized by the frame synthesis means 12 to generate a frame modulation signal 13 and A radio frequency signal 15 obtained by frequency-converting the frame modulated signal 13 into a radio frequency by a converting means 14 is transmitted from a transmission antenna 16 to a radio transmission path.
[0052]
FIG. 2 is a block diagram of a receiver of an apparatus for wirelessly transmitting digital information according to an embodiment of the present invention.
In FIG. 2, a reception signal 21 of a radio frequency received from a reception antenna 20 is frequency-converted by a down-conversion means 22 into a reception baseband IQ signal of a baseband frequency.
The channel characteristics estimating means 28 estimates the channel characteristics at a time interval shorter than the symbol length of the baseband modulation from the baseband signal in the preamble period, and uses the estimated channel characteristic information of the radio channel. By performing waveform equalization by the waveform equalizing means 30, the reception baseband IQ signal in the data period is demodulated to obtain a reception data signal of the digital information transmitted from the transmission side.
[0053]
FIG. 3 is a block diagram of a transmitter of an apparatus for wirelessly transmitting digital information when the number of multiplexing is 2 and the information modulation is a symbol rate of 11 Msymbol / sec QPSK in one embodiment of the present invention. In FIG. 3, 44 Mbps data 1 to be transmitted is serial-parallel-converted into two systems of 22 Mbps data, a sub-sequence 3-1 and a sub-sequence 3-2 in a transmission data division circuit 2, and the sub-sequence 3-1 is In the information modulation circuit 4-1, the sub sequence 3-2 is QPSK-modulated by the information modulation circuit 4-2, and a baseband modulation signal 5-1 and a baseband modulation signal 5-2 are output.
The baseband modulation signal 5-1 is delayed by 0 in a 0 signal delay circuit 6-1 to generate a baseband delay signal 7-1, and the baseband modulation signal 5-2 is converted to a T / 2 signal delay circuit 6. The signal is delayed by T / 2 to generate a baseband delay signal 7-2. Here, T is the reciprocal of the symbol rate, which in this case is about 91 ns.
The baseband delay signal 7-1 and the baseband delay signal 7-2 are added in a transmission baseband signal addition circuit 8, and a transmission baseband information modulation signal 9 is output.
[0054]
The details of the time series for data division and multiplexing will be described with reference to FIG. 44 Mbps data of digital information (d ₀ ~ D _n ) Are transmitted in the transmission data division circuit 2 by two systems of 22 Mbps data and a sub-sequence 3-1 (d ₀ , D ₂ , D ₄ , ~, D _{n -1} ) And subsequence 2 (d ₁ , D ₃ , D ₅ , ~, D _m ) Is converted to serial / parallel, and the sub sequence 3-1 (d ₀ , D ₂ , D ₄ , ~, D _n-1 ) Is a baseband modulated signal 5-1 (a) in the information modulation circuit 4-1. ₀ ~ A _{n / 2} ), The sub-sequence 3-2 is converted by the information modulation circuit 4-2 into a baseband modulated signal 5-2 (b ₀ ~ B _n ) Is QPSK modulated.
In order to shift the timing of the baseband modulation signal 5-2 with respect to the timing of the baseband modulation signal 5-1, the baseband modulation signal 5-1 delays the signal by zero by a zero signal delay circuit 6-1. Then, a baseband delay signal 7-1 is generated, and the baseband modulation signal 5-2 is delayed by T / 2 in a T / 2 signal delay circuit 6-2 to generate a baseband delay signal 7-2. ing.
[0055]
The preamble generation circuit 10a generates a transmission preamble signal 11a for obtaining transmission path characteristics at T / 2 intervals in the receiver, and performs QPSK modulation by the preamble modulation circuit 10b to output the transmission preamble modulation signal 11. The transmission baseband information modulation signal 9 and the transmission preamble modulation signal 11 are combined by a frame combination circuit 12, and a frame modulation signal 13 is output. After the band modulation signal 13 is band-limited by the band-limiting transmission filter 17, it is modulated into a radio frequency by an up-conversion circuit, and the modulated radio frequency is transmitted to the transmission path.
[0056]
FIG. 4 is a block diagram of a receiver of an apparatus for wirelessly transmitting digital information when the number of multiplexing is 2 and the information modulation is QPSK having a symbol rate of 11 Msymbol / sec in one embodiment of the present invention. is there. In FIG. 4, a band-limited reception signal in which a reception signal in a radio frequency band received by a reception antenna is down-converted by a down-conversion circuit and converted into a reception baseband signal, and the reception baseband signal is band-limited by a band-limited reception filter. Is converted into a received digital signal by an A / D conversion circuit. The received digital signal is divided into two, one of which is applied to a transmission path characteristic estimating circuit, and the other is applied to a wave equalizer. The transmission path characteristic estimating circuit includes, for example, a matched filter, and estimates an impulse response of the transmission path at T / 2 intervals using an output signal of the matched filter in the preamble period. The waveform equalization circuit performs waveform equalization using the impulse response estimation value of the transmission path obtained by the transmission path characteristic estimation circuit.
[0057]
Let the baseband received signal expressed in the equalized low-pass system be r (t). In the data period, r (t) is susceptible in the T / 2 monitoring section to perform waveform equalization and demodulation processing.
[0058]
In the following, the sample value of r (t) at t = mT / 2 (m is an integer) is represented by r _m Notation.
Further, the impulse response of the transmission line obtained at the T / 2 interval is represented by h _n Notation.
[0059]
FIG. 6 shows an example of the impulse response of the transmission line.
h ₀ , H ₁ Other than that, it is almost zero. In the following, h ₀ Is the preceding component, h ₁ Is called a delay component.
Let the n-th modulated signal in the transmission data a be _n , The n-th modulated signal in the transmission data b is represented by b _n Notation.
[0060]
In the transmission data a, the first modulated signal a ₀ The reception timing of the preceding component of ₀ And At this time, the received signal corresponding to the leading component of the n-th modulated signal of the transmission data a is r _2n And the received signal corresponding to the leading component of the n-th modulated signal of the transmission data b is r _{2n + 1} It is.
[0061]
First modulated signal a in transmission data a ₀ Is set to t = 0. At this time, the leading component of the n-th modulated signal of the transmission data a is received at t = 2n. The leading component of the n-th modulated signal of the transmission data b is received at t = 2n + 1. At t = 2n, waveform equalization and demodulation of transmission data a are performed, and at t = 2n + 1, waveform equalization and demodulation of transmission data b are performed.
[0062]
At this time, the received signal r _n Is expressed by the following equation (24).
[Equation 17]

Where n _n Is a noise component. Further, the delay time in the wireless transmission path is assumed to be finite, and h _n = 0 (n <-1, n> N).
In equation (24), the first item represents the signal component, the second item represents the intersymbol interference from the time series, and the third item represents the interference from the other series.
[0063]
Here, according to the following equation (25), the modulation signal a _n Represents
(Equation 18)

[0064]
From the expressions (24) and (25), the following expression (26) is established.
[Equation 19]

[0065]
This is a general expression of a received signal on a transmission path having intersymbol interference in the case of the symbol length T / 2, and is a conventional expression such as MLSE (Maximum Likelihood Sequence Estimator), DDFSE, DFE (Decision Feedback Equalizer), or the like. Demodulation can be performed by an equalizer.
[0066]
【The invention's effect】
As described in detail above, according to the present invention, the transmitter side divides the transmission data and multiplexes the data at staggered times, thereby realizing a wireless transmission speed that is a multiple of the multiplexing number as compared with the case without multiplexing. be able to. In addition, by performing waveform equalization using transmission path characteristic information at a time interval shorter than the symbol length on the receiving side, it is possible to accurately suppress intersymbol interference, without changing the wireless transmission bandwidth. Therefore, the effect of the present invention on wireless transmission of digital information is extremely large.
[Brief description of the drawings]
FIG. 1 is a block diagram for explaining a wireless transmission method of digital information according to the present invention.
FIG. 2 is a block diagram showing a configuration example of a receiver for receiving a radio frequency signal transmitted by a digital information radio transmission method according to the present invention.
FIG. 3 is a block diagram showing a configuration example of a transmitter for implementing the digital information wireless transmission method according to the present invention.
FIG. 4 is a block diagram showing an embodiment of a receiver for receiving a radio frequency signal transmitted by the method of the present invention.
FIG. 5 is a time chart for explaining the operation of the configuration example shown in FIG. 3;
FIG. 6 is a time chart showing an example of a transmission path impulse response of a radio frequency signal transmitted by the method of the present invention.
FIG. 7 is a characteristic diagram showing a communication capacity of a conventional wireless LAN standard.
FIG. 8 is a waveform diagram showing an example of an impulse response in a fractional tap partial response transmission used in the present invention.
FIG. 9 is a waveform diagram showing an impulse response in normal transmission.
FIG. 10 is a waveform diagram showing an impulse response when a conventional partial response transmission is performed.
FIG. 11 is a characteristic diagram showing a communication capacity obtained when performing QPSK transmission and N = 2 fractional tap partial response encoding according to the present invention.
FIG. 12 is a characteristic diagram showing two symbol lengths on the impulse response side of the band-limiting filter.
FIG. 13 is a characteristic diagram showing 2 / symbol length on the impulse response side of the band limiting filter.
FIG. 14 is a waveform diagram showing an impulse response when band limitation is performed by a filter having the same characteristics as that of QPSK transmission in N = 2 fractional tap partial response encoding according to the present invention.
[Explanation of symbols]
1 Digital information transmitted
2 Transmission data division means, transmission division circuit
3-1 ... 3-N sub-series
4-1... 4-N information modulation means and information modulation circuit
5-1 ... 5-N baseband modulation signal
6--1,... 6-N signal delay means and delay circuit
7-1, 7-2 baseband delay signal
8 Baseband signal addition means, baseband signal addition circuit
9 Transmission baseband information modulation signal
10 Preamble generation means
10a Preamble generation circuit
10b preamble modulation circuit
11 Transmit preamble modulation signal
12 Frame synthesis circuit
13 frame modulation signal
14. Up-conversion circuit
15 Radio frequency signal
16 transmitting antenna
17 Band-limited transmission filter
20 receiving antenna
21 Received signal
22 Down conversion circuit
23 Received baseband signal
24 Band Reception Filter
25 Reception band limited signal
26 A / D converter
27 Baseband I / Q signal
27-1, 27-2 Received digital signal
28 Transmission line characteristic estimation circuit
29 Transmission line characteristic information
30 Waveform equalization means
31 Received data
32 Synchronous circuit

Claims (7)

The digital information is sampled at intervals of 1 / N (N is an integer of 2 or more) of the symbol period T of the digital information, and a transmission pulse train is taken out. At the sampling points other than the N sampling points, the signal is transmitted via a band-limited radio transmission channel that takes an impulse response value closer to zero than the substantially identical impulse response value. Characteristic wireless transmission method for digital information. 2. The digital information wireless transmission method according to claim 1, wherein transmission is performed in which an intersymbol interference component caused by a tail of an impulse response is suppressed by a decision feedback type waveform equalization. A wireless transmission method of digital information, wherein an intersymbol interference component caused by an impulse response head is suppressed by subtracting a previously generated interference replica. 2. The method according to claim 1, wherein the digital information is trellis-modulated digital information. On the sending side,
Dividing digital information into a plurality of sub-sequences,
Generate a plurality of baseband modulation signals based on the digital information for each sub-sequence,
Generate a transmission baseband signal of the data period by arithmetically adding the plurality of baseband modulation signals by shifting the time axis so that the symbol timings are different from each other,
A preamble signal for transmission characteristic estimation and a signal obtained by frequency-converting the transmission baseband signal of the subsequent data period into a radio frequency are transmitted to a radio transmission path,
On the receiving side,
Frequency conversion of the received radio frequency signal into a baseband band,
Based on the received baseband signal in the preamble period, the transmission path characteristics are estimated at time intervals shorter than the symbol length of the baseband modulation,
By performing waveform equalization using the transmission path characteristic information of the estimated wireless transmission path, to demodulate the received baseband signal in the data period,
A wireless transmission method of digital information for obtaining a reception data signal of the digital information transmitted from the transmission side. Transmission data dividing means for dividing digital information to be transmitted into a plurality of sub-sequences;
A plurality of information modulation means for generating a baseband modulation signal for each sub-sequence, a plurality of signal delay means for making the symbol timings of the plurality of baseband modulation signals different from each other,
Transmission baseband signal addition means for adding a plurality of baseband modulation signals respectively output from the plurality of signal delay means,
Preamble generating means for generating a baseband modulated signal of a preamble for estimating a transmission path characteristic on a receiving side,
Frame combining means for combining the transmission baseband information modulation signal output from the transmission baseband signal addition means and the transmission preamble modulation signal output from the preamble generation means to obtain a frame modulation signal,
Up-conversion means for up-converting the frame modulated signal to a radio frequency,
A transmitter for transmitting radio waves of digital information, comprising a transmission antenna for transmitting an up-converted radio frequency signal to a radio transmission path. In order to receive a radio frequency signal transmitted by combining a baseband information modulation signal and a preamble modulation signal obtained by adding a plurality of baseband signals having different symbol timings,
A receiving antenna for receiving the transmitted radio frequency signal;
Down-converting means for down-converting a reception signal received by the reception antenna and converting it into a reception baseband IQ signal;
Channel characteristics estimating means for estimating channel characteristics at a time interval shorter than the symbol length of baseband modulation from the received baseband IQ signal in a preamble period;
Waveform equalization means for performing waveform equalization of the received baseband IQ signal in a data period using the transmission path characteristic information from the transmission path characteristic estimation means and demodulating data at the conversion point of the symbol is provided. A receiver for wireless transmission of digital information.

Images