JP4122952B2 - Downlink receiver for code division multiple access - Google Patents

Description

【００１１】
受信信号サンプルは、時間ダイバーシチサンプル及び／又は空間ダイバーシチサンプルである。空間ダイバーシチサンプルは、Ｎを送信機の数とし、Ｍ＞Ｎとして、受信アンテナ配列のＭ本の異なるアンテナにより受信される。サンプリングレートは１チップである。
【００１２】
一具体例においては、Ｍ＝Ｎ＋１である。なお、アンテナ素子の本数Ｍが多くなるほど、再生性能も高くなる。
【００１３】
複数の空間的に離間したアンテナ素子に代えて、又はこれに加えて、サンプリングレートを高めて、単一の又はより少ない本数のアンテナ素子を使用してもよい。Ｍ本のアンテナ素子を１本のアンテナ素子に置換するためには、時間ダイバーシチサンプルは、ｔ＝（チップ期間）／Ｍ毎にサンプリングされる。
【００１４】
この具体例では、必要なアンテナ素子の本数を減らすことができ、したがって移動端末装置における受信アンテナ配列のサイズを小型化することができる。この利点は、技術及び市場価値の観点から小型化が重要となるハンドヘルド型の端末装置において特に重要である。
【００１５】
チャンネル行列の推定は、推定器を用いて、周知の手法により実行される。ここでは、既知のパイロットトーンを用いて、送信アンテナと受信アンテナとの間の振幅及び移相伝達関数が推定される。
【００１６】
希望信号の検出は、この信号に対し、例えばウォルシュアダマール符号等の固有の符号を乗算することにより実行される。この処理は、当分野で周知であり、特に逆拡散として知られている。
【００１７】
行列演算子は、好ましくは、チャンネル行列の共役転置行列に、チャンネル行列とチャンネル行列の共役転置行列との積の逆行列を乗算したものとする。
【００１８】
端末装置において必要とされる演算パワーを小さくするために、行列演算子は、好ましくは繰り返し算出される。
【００１９】
本発明の第２の側面である受信装置は、複数の基地局から送信されてくる、符号分割多元接続の希望信号と符号分割多元接続の他の信号とが組み合わされた信号から符号分割多元接続の希望信号を受信する受信装置において、受信アンテナ配列からの受信信号サンプルの数を判定し、受信信号ベクトルを定義する受信信号サンプル数判定手段と、基地局と受信アンテナ配列との間のマルチパスフェージングを推定し、チャンネル行列を定義するマルチパスフェージング推定手段と、受信信号ベクトルに対し、チャンネル行列に基づく行列演算子を乗算し、各基地局からの合成信号を再生する合成信号再生手段と、合成信号の１つから希望信号を検出する検出手段とを備える。そして、行列演算子は、合成信号再生手段により、以下の式に基づいて繰り返し算出されることを特徴とする。
【００２０】
【数４】

【００２１】
【発明の実施の形態】
図１は、マルチパス伝搬環境における基地局１及び移動端末装置７を示している。マルチパス効果は、特に、移動端末装置７が建物や乗り物等の信号拡散オブジェクトによって取り囲まれている都市の中心部やそれに類似する環境において生じやすい。これらの信号拡散オブジェクトを図１では「Ｘ」により示している。これらのパスの特性は、移動端末装置７及びこの移動端末装置７の周囲のオブジェクトが動き回ることにより、時刻毎に変化する。この所謂「フェージング」の影響は、基地局１及び移動端末装置７の一方又は両方に２本以上のアンテナを設けることにより改善される。これにより、少なくとも空間ダイバーシチが実現され、更に、これらのアンテナは、例えば時空間符号化等のより高度な無線通信技術に利用することもできる。
【００２２】
基地局１の送信アンテナＴ_１と移動端末装置７の受信アンテナＲ_１との間には、２つの信号伝搬パスＰ_１１ａ，Ｐ_１１ｂが示されている。実際には、一対の送信アンテナと受信アンテナの間には、これより多くのパスが存在することが多い。図１に示すように、信号伝搬パスＰ_１１ａは、信号伝搬パスＰ_１１ｂより短く、したがって、送信アンテナＴ_１から送信された希望信号の２つのコピーが時間的にずれて受信アンテナＲ_１で受信される。信号伝搬パスＰ_１２ａは、基地局１の送信アンテナＴ_１と移動端末装置７の第２の受信アンテナＲ_２との間のパスを示している。移動端末装置７の第２のアンテナＲ_２は、希望信号の受信信号パワーを更に高めるための受信機空間ダイバーシチを実現している。ここで、上述のような時間的にずれた希望信号のコピーを可干渉的に結合するためには、何らかのメカニズムが必要である。
【００２３】
図２は、移動端末装置７においてＲａｋｅ受信機８を用いた従来の無線通信システムを示している。上述のように、ＣＤＭＡシステムにおいて、セル内の各データチャンネル又は音声チャンネルは、より高い周波数の符号で拡散されており、セル内で使用されている各符号は、互いに直交している。移動端末装置７が備えるＲａｋｅ受信機８は、複数のＲａｋｅフィンガ（Rake finger）を有し、各Ｒａｋｅフィンガは、アンテナＲ_１及びアンテナＲ_２が受信した入力合成信号に対し、移動端末装置７に対応する希望チャンネル固有の符号を乗算する乗算器９を備えている。更に、各Ｒａｋｅフィンガは、乗算器９における符号の乗算の前に、それぞれ異なる遅延量を与える遅延素子１０を備えている。各遅延は、伝搬環境のマルチパス特性に起因する、希望信号の各コピー間の時間的なずれに対応するように設定される。Ｒａｋｅフィンガの各出力信号は、結合器１１において、希望信号の各コピーの遅延を考慮して可干渉的に結合され、希望信号が生成される。
【００２４】
最終的な信号を生成するために、希望信号の複数のコピー又はサンプルを用いているので、最終的な信号の信号対雑音比は、この手法を用いない場合に比べて向上する。しかしながら、マルチパス伝搬環境においては、基地局１からの希望信号のコピーと他の信号のコピー間ではチップ同期（chip alignment）が維持されないことがある。これにより、他のチャンネルの成分が希望信号のサンプルに重なり、チャンネル間の直交性が完全でなくなる。この結果、図２においてＩｃとして示すセル内多元接続干渉（intra-cell multiple access interference：以下、ＭＡＩという。）が生じる。更に、マルチパス伝搬環境では、時間的にずれた他の基地局１からの信号の複数のコピーが希望信号に重なり、これらの信号間の直交性が損なわれるという現象も生じる。この結果、図２においてＩｂとして示すセル間（inter-cell）ＭＡＩが生じる。セル間ＭＡＩは、セル内ＭＡＩとは異なるマルチパス伝搬によって引き起こされるものであり、Ｒａｋｅ受信機８によって希望信号の信号対雑音比は改善されるが、ＭＡＩの問題は解消されない。
【００２５】
図３は、セル内ＭＡＩを最小化する適応等化（adaptive equalization）を用いる従来の無線通信システムの構成を示している。この無線通信システムは、基地局１’から送信されてくるパイロットトーン３を用いる。パイロットトーン３は、システム内の移動端末装置７にとって既知である。移動端末装置７は、チャンネル推定器６を備え、チャンネル推定器６は、受信したパイロットトーン３を用いて、一対の送信アンテナと受信アンテナ間の合成チャンネル特性（composite channel characteristic）を決定する。合成チャンネル特性は、例えばアンテナＴ_１とアンテナＲ_１との間の伝搬パスＰ_１１ａ，Ｐ_１１ｂによって定義される。次に、チャンネル行列Ｈが推定される。チャンネル行列Ｈの要素Ｈ_ｔｒは、送信アンテナ（アンテナＴ_１）と受信アンテナ（アンテナＲ_１，Ｒ_２）との間で伝送される信号の伝達関数を定義する。チャンネル推定器６によってこのチャンネル行列Ｈが推定されると、その逆行列Ｈ^−１が決定される。この機能（Ｈ^−１）は、等化器１２を用いて、受信アンテナＲ_１，Ｒ_２で受信された入力信号のサンプルｒ_１，ｒ_２に逆チャンネル行列Ｈ^−１を乗算することにより実行される。この処理により、基地局１’のアンテナＴ１から送信された元の信号（ｓ）が結合され、時間的に整合（time aligned）されたサンプル（ｒ_３）が算出される。これにより、チャンネル信号は直交し、図３に示すように合成信号サンプル（ｒ_３）に希望チャンネル固有の符号を乗算することにより、希望チャンネル（４）を容易に抽出することができる。この処理により、セル内ＭＡＩ（Ｉｃ）を実質的に低減することができる。しかしながら、ここでは、セル間ＭＡＩ（Ｉｂ）は取り除かれない。セル間ＭＡＩは、全体のＭＡＩの１／３に達することもあり、これが重大な干渉源となってシステムの性能を劣化させることがある。
【００２６】
セルラＣＤＭＡシステムのダウンリンクを図４に示す。この具体例では、端末装置２７は、３つの基地局２０からの信号（ｓ）を強く受信している。端末装置２７は、更に遠く離れた基地局（図示せず）からの信号も受信していることがある。セル間信号（ｓ_１〜ｓ_Ｎ）は、ＭＡＩの原因となり、このＭＡＩは既知のシステムでは抑圧されない。
【００２７】
図５ａは、本発明に基づくセルラＣＤＭＡシステムを示している。また、図５ｂは、本発明に基づく受信機の具体的構成を示している。端末装置２７は、複数の基地局又は送信機（２０_１〜２０_Ｎ）からの信号（ｓ_１〜ｓ_Ｎ）を受信する。各基地局からの信号（ｓ）は、各基地局に固有の疑似ランダム符号によって符号化されている。多くの場合、受信機に強い影響を与える基地局の数は、少なく、通常２〜３個である。基地局が受信機から十分遠ければ、その信号は十分弱く、ＭＡＩに大きな影響を与えることはない。
【００２８】
端末装置２７は、複数の（Ｍ本の）アンテナ素子と、それぞれ基地局（２０_１〜２０_Ｎ）に対応する複数の（Ｎ個の）チャンネル推定器が組み込まれた信号処理回路２６と、行列算出器２５とを備える。更に、端末装置２７は、希望信号を得るためのデスクランブラ又は逆拡散器を備える。
【００２９】
アンテナ素子（Ｒ_１〜Ｒ_Ｍ）は、空間的に離間して配設されており、これにより入力合成信号（ｓ_１＋ｓ_２＋・・・＋ｓ_Ｎ）の空間ダイバーシチサンプル（ｒ_１〜ｒ_Ｍ）が得られる。サンプリングレートは１チップ又は拡散符号内の各サンプル分に相当する。アンテナ素子は、搬送波の波長の１／２以上離間して配設する必要がある。アンテナ素子の本数は、ＭＡＩに強い影響を有する送信機２０の数をＮとして、Ｎ＋１以上設けられる。一般的なセルラシステムの場合、アンテナ素子の本数は、例えば４である。
【００３０】
これに代えて、単一のアンテナ（Ｒ）を用い、異なるサンプルを異なる時刻に受信するようにしてもよい。ここでは、空間的に離間して配設された複数のアンテナ素子が受信するサンプルに代えて、時間ダイバーシチサンプルが使用される。ここで、サンプリングレートは、時間ダイバーシチを用いない場合に必要とされるアンテナ素子の本数をＭとして、（１チップ）／Ｍである。また、空間ダイバーシチサンプルと時間ダイバーシチサンプルとを組み合わせて使用してもよい。
【００３１】
複数のチャンネル推定器を備える信号処理回路２６は、チャンネル行列Ｈ’を生成する。チャンネル行列Ｈ’の各要素Ｈ’_ｎｍは、それぞれ基地局の送信アンテナ（Ｔ_１〜Ｔ_Ｍ）の１つと、受信アンテナ（Ｒ_１〜Ｒ_Ｍ）の１つとの間の合成チャンネル特性を表している。チャンネル推定の手法は周知であり、ここでは、適切ないかなる手法を用いてもよい。
【００３２】
所望のユーザの受信機のｍ番目のアンテナ素子（Ｒ_ｍ）によって受信される信号は、下記式（１）のように表される。
【００３３】
【数５】

【００３４】
ここで、Ｓ_ｎは、各基地局（２０_ｎ）からの合成信号を表す。所望のユーザの受信機のＭ本のアンテナによって受信される信号全体は、下記式（２）のように表される。
【００３５】
【数６】

【００３６】
所望のユーザに送信された信号は、基地局からの合成信号Ｓ_１，Ｓ_２，・・・，Ｓ_Ｎのうちの１つの信号に含まれる。一般性を損なうことなく、希望信号が合成信号Ｓ_１に含まれるとする。すなわち、合成信号Ｓ_２〜Ｓ_Ｎは、所望のユーザにとってセル間ＭＡＩであり、合成信号Ｓ_１内の希望信号以外の部分は、所望のユーザにとってセル内ＭＡＩである。
【００３７】
この具体例におけるＭＡＩ抑圧法の手順は、以下の通りである。
１．全てのアンテナで受信された信号を、式（２）で示すｒとして整理する。
２．それぞれ異なる基地局（セル）からのパイロット信号を用いて、Ｈ_１１〜Ｈ_ＭＮを推定する。
３．式（２）に示す所望のユーザに対して整理された信号に対して、式（３）に示す行列Ｈ^＋を生成する。
【００３８】
【数７】

【００３９】
ここで、（ ）^ｈは、共役転置行列を表し、（ ）^−１は逆行列を表す。
４．式（４）に示すように、Ｈ^＋に対し、式（２）に示す受信信号ｒを右から乗算し、拡散合成信号Ｓ_１，Ｓ_２，・・・，Ｓ_Ｎを再生する。
【００４０】
【数８】

【００４１】
ここで，合成信号Ｓ_２，・・・，Ｓ_Ｎを無視し、合成信号Ｓ_１のみを抽出する。これにより、セル間ＭＡＩは、完全にキャンセルされる。
５．合成信号Ｓ_１を直接デスクランブル及び逆拡散する。合成信号Ｓ_１内の全ての全ての信号は互いに直交するため、セル内ＭＡＩは、デスクランブル及び逆拡散の処理によってキャンセルされる。
【００４２】
端末装置のアンテナ（Ｒ_１〜Ｒ_Ｍ）で受信されたサンプル信号（ｒ_１〜ｒ_Ｍ）には、チャンネル行列演算子Ｈ^＋が左からから乗算され、これにより、各基地局又は送信機２０からの個々の、時間的及び空間的に整合された合成信号（ｓ_１〜ｓ_Ｎ）が再生される。各基地局と端末装置の受信アンテナ（Ｒ）との間の合成チャンネル特性又はフェージングＨ_ｍｎを補正することにより、基地局からの信号は、時間的に整合し、各基地局からの信号の符号の直交性が維持される。
【００４３】
所望の基地局（２０_１）からの全ての信号（ｓ_１）は直交するため、この合成信号（ｓ_１）に単に所定の符号を乗算するのみで、希望信号が算出される。このようにして、希望信号からセル間ＭＡＩ及びセル内ＭＡＩの両方が取り除かれる。
【００４４】
信号処理回路２６におけるチャンネル推定処理及び行列算出器２５における行列の算出は、特定用途向け集積回路（application specific integrated circuit：ＡＳＩＣ）等のハードウェアにより実行してもよく、端末装置内に設けられたプロセッサによりソフトウェアとして実行してもよい。
【００４５】
行列の演算は、簡単にするために、例えば反復アルゴリズムを用いて実現するとよい。疑似逆チャンネル行列（pseudo-inverse channel matrix）Ｈ^＋のための好ましい反復演算は、以下の通りである。
【００４６】
Ｈ_ｎ＝［Ｈ_ｎ−１ｈ_ｎ］をＭ×ｎ行列とする。ここで、Ｍ≧ｎであり、Ｈ_ｎ−１はＨ_ｎの第１のｎ−１行を含む行列であり、ｈ_ｎはＨ_ｎのｎ番目の行であり、ランク（Ｈ_ｎ）はｎである。
【００４７】
【数９】

【００４８】
これらに代えて、受信信号サンプルの行列（ｒ_１〜ｒ_Ｍ）に作用して、各基地局又は送信機（Ｔ）からの合成信号（ｓ_１〜ｓ_ｎ）を算出する他の行列演算子を用いてもよい。すなわち、希望信号を含む合成信号の時間的及び空間的に整合（aligned）されたサンプルは、個別に導出された後、単純に希望信号の符号が乗算され、通常の手法で希望信号が得られる。行列演算子は「高い（tall）」必要があり、すなわち、１つの拡散データ用の非相関サンプルの数をＭとし、送信機の数をＮとして、行数Ｍは列数Ｎより大きな数である必要がある。
【００４９】
図６は、図５に示す受信機における信号対雑音比（Ｅｂ／Ｎ０）に対するビット誤り率（Bit-error-rate：ＢＥＲ）と、Ｒａｋｅ受信機（図２）を用いた従来の受信機におけるそれとを比較して示すグラフである。図６から分かるように、Ｒａｋｅ受信機のＢＥＲは、信号対雑音比が高くなっても余り小さくならないが、図５に示す受信機では、信号対雑音比が高くなると、ＢＥＲは良くなる。これは、図５に示す受信機では、ＭＡＩが排除されるためである。
【００５０】
図５に示すシステムは、上述のように、単に受信機空間ダイバーシチモードで使用してもよく、又は各基地局において複数の送信アンテナＴが同じ信号を送信する送信機ダイバーシチに使用してもよい。更なる変形例として、基地局の各アンテナＴから異なる複素シンボルが送信される時空間符号化技術を用いてもよい。但し、この手法では受信機側で更なるアンテナが必要となり、例えば、送信アンテナの本数を２倍にすると、受信アンテナの本数も２倍にする必要がある。
【００５１】
以上、本発明を好ましい実施の形態を用いて説明した。当業者にとって明かな変形及び修正は、本発明の範囲内にある。
【図面の簡単な説明】
【図１】 マルチパス伝搬環境において動作する移動通信システムを示す図である。
【図２】 Ｒａｋｅ受信機を組み込んだ従来の受信機の構成を示す図である。
【図３】 適応等化を用いる従来の受信機の構成を示す図である。
【図４】 複数の基地局を備えるセルラＷＣＤＭＡシステムを示す図である。
【図５ａ】 本発明に基づくセルラＷＣＤＭＡシステムを示す図である。
【図５ｂ】 本発明に基づく受信機の構成を示す図である。
【図６】 図５ｂに示す受信機のＢＥＲ対Ｅｂ／Ｎ０を図２に示す受信機のそれと比較して示す図である。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to cellular communication systems, and more particularly to a method for receiving a code division multiple access signal in a multipath propagation environment. Further, the present invention relates to a mobile receiver such as a mobile phone or a laptop computer in a code division multiple access system such as a third generation cellular communication system, but is not limited thereto.
[0002]
[Prior art]
Code division multiple access (hereinafter referred to as CDMA) signals overlap in time and frequency and are distinguished from each other by a unique spreading code such as a Walsh-Hadamard code. In principle, the desired signal is reproduced by multiplying the received composite signal including the desired signal by a unique spreading code. By this processing, a desired signal is obtained and other components are easily excluded. This process assumes that each code used in a cell is orthogonal to other codes used by base stations for other users in that cell. In a complete environment in which this orthogonality is maintained, the desired signal can be reproduced without any interference from other signals used in the cell. The signal used in the cell is also superimposed with a further code to distinguish it from signals from other cells.
[0003]
[Problems to be solved by the invention]
Here, in an actual environment where multipath propagation exists, the orthogonality of the received signal is significantly impaired. As a result, the desired signal includes interference components from other signals in the cell and signals from other cells. Such a phenomenon is called multiple access interference (hereinafter referred to as “MAI”), which includes intra-cell MAI (interference from signals used in the cell) and inter-cell MAI (others). Interference from signals used in other cells).
[0004]
Rake receivers are used to increase the signal-to-noise ratio of the desired signal, but are not specifically designed to reduce MAI. The Rake receiver has a plurality of input terminals, and multiplies a plurality of samples of an input composite signal (including a desired signal) by a spreading code of the desired signal to reproduce the desired signal better. Each input terminal of the Rake receiver is connected to an independent antenna element, and a copy of a desired signal received by each antenna element is correlated with a desired signal received by another antenna element, Noise received by different antenna elements is uncorrelated and thereby provides spatial diversity. Alternatively or in addition, the Rake receiver input signal is delayed in time to compensate for the operation of the communication system in a multipath environment where copies of the desired signal arrive at the receiver at different times. Is done. The output signal of the Rake receiver is coherently combined so that the desired signal is reproduced more accurately. Specific examples of Rake receivers are disclosed, for example, in US Pat. Nos. 5,305,349 and 6,026,115. Although the bit error rate is improved to some extent by increasing the signal-to-noise ratio of the desired signal, MAI is a significant signal degradation factor in these types of CDMA receivers.
[0005]
Various in-cell MAI suppression methods have been proposed so far, but these methods only suppress a part of MAI, and in particular, these methods do not target inter-cell MAI. Specific examples of MAI suppression methods that rely heavily on adaptive equalization include IEEE Vehicular Technology Conference (1999), Markku J. Heikkila, "CDMA Downlink with Adaptive Channel Equalization". "InterferenceSuppression in CDMA Downlink through adaptive channel equalization", IEEE Vehicular Technology Conference 1999 (IEEE Vehicular Technology Conference, 1999), Stefan Werner, Jorma Lilleberg "Downlink channel decorrelation in CDMA systems with long codes", IEEE Vehicular Technology Conference 1998 (IEEE Vehicular Technology Conference, 1998), Irfan Ga "Linear Receivers for the DS-CDMA Downlink Exploiting Orthogonality of Spreading Sequences" by Irfan Ghauri and Dirk TM slock ”,“ Globecom '99 ”, Hooli, K, Latva-aho, M and Junti, M“ In WCDMA downlink receivers ” Multiple access interference Suppression with linearclip equalizer in WCDMA downlink receivers, Vehicular Technology Globecom '99, Hooli, K, Latva -Inter-path interference suppression in WCDMA systems with low spreading factor by Latva-aho, M and Juntti, M (Inter-path Interference Suppression in WCDMA Systems with Low Spreading Factors).
[0006]
[Means for Solving the Problems]
The present invention provides an improved or at least alternative receiver apparatus and method for code division multiple access signals. In general terms, the present invention proposes to recover a predetermined number of spatially and temporally aligned composite signal samples, each containing a signal transmitted from a single transmitter. This reduces both intra-cell MAI in signals transmitted from one transmitter of a plurality of base stations and inter-cell MAI caused by signals from transmitters of other base stations in a multipath propagation environment. can do. The synthesized signal samples are regenerated by receiving a plurality of signal samples including synthesized signals from a plurality of base stations and applying a mathematical process that separates the synthesized signal components received from each base station. Each composite signal component is spatially and temporally matched, so that the desired signal can be easily obtained by multiplying the composite signal component of the desired signal from the base station by the sign of the desired signal.
[0007]
This approach has various advantages, including the ability to reduce or eliminate both inter-cell MAI and intra-cell MAI. Inter-cell MAI is removed by separating the combined signal from each base station. The separated combined signal is a plurality of time-aligned or orthogonal signals from a single base station, and only the desired signal can be extracted by despreading the combined signal. MAI is also removed.
[0008]
With the reception method according to the present invention, it is possible to suppress inter-cell MAI and intra-cell MAI without performing soft hand over.
[0009]
The reception method according to the first aspect of the present invention is a code division multiple from a signal that is transmitted from a plurality of base stations and is a combination of a desired signal of code division multiple access and another signal of code division multiple access. In a receiving method for receiving a desired signal for connection, a step of determining a number of received signal samples from a receiving antenna arrangement and defining a received signal vector, and a multi-channel between the base station and the receiving antenna array Estimating path fading and defining a channel matrix; multiplying a received signal vector by a matrix operator based on the channel matrix to regenerate a combined signal from each base station; and one of the combined signals Detecting a desired signal. The matrix operator is repeatedly calculated based on the following expression.
[0010]
[Equation 3]

[0011]
The received signal sample is a time diversity sample and / or a spatial diversity sample. Spatial diversity samples are received by M different antennas in the receive antenna array, where N is the number of transmitters and M> N. The sampling rate is 1 chip.
[0012]
In one specific example, M = N + 1. The reproduction performance increases as the number M of antenna elements increases.
[0013]
Instead of or in addition to a plurality of spatially spaced antenna elements, a single or fewer antenna elements may be used with increased sampling rate. In order to replace M antenna elements with one antenna element, time diversity samples are sampled every t = (chip period) / M.
[0014]
In this specific example, the number of necessary antenna elements can be reduced, and thus the size of the receiving antenna array in the mobile terminal apparatus can be reduced. This advantage is particularly important in a handheld terminal device in which miniaturization is important from the viewpoint of technology and market value.
[0015]
The estimation of the channel matrix is performed by a known method using an estimator. Here, the amplitude and phase shift transfer function between the transmitting antenna and the receiving antenna are estimated using known pilot tones.
[0016]
The detection of the desired signal is performed by multiplying this signal by a unique code such as a Walsh Hadamard code. This process is well known in the art and is particularly known as despreading.
[0017]
The matrix operator is preferably obtained by multiplying the conjugate transpose matrix of the channel matrix by the inverse matrix of the product of the channel matrix and the conjugate transpose matrix of the channel matrix.
[0018]
In order to reduce the computing power required in the terminal device, the matrix operator is preferably calculated repeatedly.
[0019]
A receiving apparatus according to a second aspect of the present invention provides a code division multiple access from a signal that is transmitted from a plurality of base stations and is a combination of a desired signal of code division multiple access and another signal of code division multiple access. Multi-path between the base station and the receiving antenna array, and a receiving signal sample number determining means for determining the number of received signal samples from the receiving antenna array and defining a received signal vector Multipath fading estimation means for estimating a fading and defining a channel matrix; a composite signal reproduction means for multiplying a received signal vector by a matrix operator based on the channel matrix and reproducing a composite signal from each base station; from one of the combined signal and a detection means that detect the desired signal. The matrix operator is repeatedly calculated based on the following expression by the combined signal reproducing means .
[0020]
[Expression 4]

[0021]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 shows a base station 1 and a mobile terminal apparatus 7 in a multipath propagation environment. The multipath effect is particularly likely to occur in the center of a city where the mobile terminal device 7 is surrounded by signal diffusion objects such as buildings and vehicles, and similar environments. These signal spreading objects are indicated by “X” in FIG. The characteristics of these paths change for each time as the mobile terminal device 7 and objects around the mobile terminal device 7 move around. The influence of this so-called “fading” is improved by providing two or more antennas in one or both of the base station 1 and the mobile terminal device 7. This realizes at least space diversity, and these antennas can also be used for more advanced wireless communication technologies such as space-time coding.
[0022]
Between the transmitting antennas _{T 1} of the base station 1 and the receiving antenna _{R 1} of the mobile terminal 7, the two signal propagation paths _{P 11a, P 11b} is shown. In practice, there are often more paths between a pair of transmit and receive antennas. As shown in FIG. 1, the signal propagation path P _11a is shorter than the signal propagation path P _11b , so that two copies of the desired signal transmitted from the transmission antenna T ₁ are received at the reception antenna R ₁ with a time lag. Is done. A signal propagation path P _{12 a} indicates a path between the transmission antenna T ₁ of the base station 1 and the second reception antenna R ₂ of the mobile terminal device 7. The second antenna R ₂ of the mobile terminal device 7 realizes receiver space diversity for further increasing the received signal power of the desired signal. Here, some mechanism is necessary for coherently coupling the copy of the desired signal shifted in time as described above.
[0023]
FIG. 2 shows a conventional radio communication system using a Rake receiver 8 in the mobile terminal device 7. As described above, in a CDMA system, each data channel or voice channel in a cell is spread with a higher frequency code, and the codes used in the cell are orthogonal to each other. Rake receiver 8 the mobile terminal device 7 is provided has a plurality of Rake fingers (Rake finger), each Rake finger, the input composite signal antenna _{R 1} and antenna _{R 2} is received, the mobile terminal apparatus 7 A multiplier 9 for multiplying the corresponding code specific to the desired channel is provided. Further, each Rake finger is provided with a delay element 10 that gives a different delay amount before the multiplication of the code in the multiplier 9. Each delay is set to correspond to a time lag between each copy of the desired signal due to the multipath characteristics of the propagation environment. The output signals of the Rake fingers are combined in a combiner 11 in a coherent manner in consideration of the delay of each copy of the desired signal, and a desired signal is generated.
[0024]
Since multiple copies or samples of the desired signal are used to generate the final signal, the signal-to-noise ratio of the final signal is improved compared to not using this technique. However, in a multipath propagation environment, chip alignment may not be maintained between a copy of a desired signal from the base station 1 and another signal copy. As a result, the components of the other channels overlap with the sample of the desired signal, and the orthogonality between the channels is not perfect. As a result, intra-cell multiple access interference (hereinafter referred to as MAI) shown as Ic in FIG. 2 occurs. Furthermore, in a multipath propagation environment, a phenomenon occurs in which a plurality of copies of signals from other base stations 1 that are shifted in time overlap with a desired signal, and orthogonality between these signals is lost. This results in an inter-cell MAI shown as Ib in FIG. Inter-cell MAI is caused by multipath propagation different from intra-cell MAI, and the Rake receiver 8 improves the signal-to-noise ratio of the desired signal, but the problem of MAI is not solved.
[0025]
FIG. 3 shows a configuration of a conventional wireless communication system using adaptive equalization that minimizes intra-cell MAI. This wireless communication system uses a pilot tone 3 transmitted from the base station 1 ′. The pilot tone 3 is known to the mobile terminal device 7 in the system. The mobile terminal apparatus 7 includes a channel estimator 6, and the channel estimator 6 uses the received pilot tone 3 to determine a composite channel characteristic between a pair of transmission antennas and reception antennas. The combined channel characteristics are defined by, for example, propagation paths P _11a and P _11b between the antenna T ₁ and the antenna R ₁ . Next, the channel matrix H is estimated. The element H _tr of the channel matrix H defines a transfer function of a signal transmitted between the transmitting antenna (antenna T ₁ ) and the receiving antenna (antennas R ₁ and R ₂ ). When this channel matrix H is estimated by the channel estimator 6, its inverse matrix H ⁻¹ is determined. This function (H ⁻¹ ) is executed by using the equalizer 12 to multiply the samples r ₁ and r ₂ of the input signals received by the receiving antennas R ₁ and R ₂ by the inverse channel matrix H ^−1. Is done. By this process, the original signal (s) transmitted from the antenna T1 of the base station 1 ′ is combined, and a time-aligned sample (r ₃ ) is calculated. Thus, the channel signals are orthogonal, and the desired channel (4) can be easily extracted by multiplying the composite signal sample (r ₃ ) by the code specific to the desired channel as shown in FIG. By this process, the in-cell MAI (Ic) can be substantially reduced. However, the inter-cell MAI (Ib) is not removed here. Inter-cell MAI can reach 1/3 of the total MAI, which can be a significant source of interference and degrade system performance.
[0026]
The downlink of a cellular CDMA system is shown in FIG. In this specific example, the terminal device 27 strongly receives signals (s) from the three base stations 20. The terminal device 27 may also receive a signal from a base station (not shown) further away. Inter-cell signals (s _{1 to} s _N ) cause MAI, which is not suppressed in known systems.
[0027]
FIG. 5a shows a cellular CDMA system according to the present invention. FIG. 5b shows a specific configuration of the receiver according to the present invention. Terminal device 27 receives the signal _(s 1 ~s _N) from multiple base stations or transmitters ₍₂₀ 1 _{~20 N).} The signal (s) from each base station is encoded by a pseudo-random code unique to each base station. In many cases, the number of base stations that have a strong influence on the receiver is small, usually two to three. If the base station is far enough from the receiver, the signal is weak enough and does not significantly affect the MAI.
[0028]
The terminal device 27 includes a signal processing circuit 26 in which a plurality of (M) antenna elements, a plurality of (N) channel estimators corresponding to the base stations (20 _{1 to} 20 _N ) are incorporated, and a matrix And a calculator 25. Further, the terminal device 27 includes a descrambler or a despreader for obtaining a desired signal.
[0029]
The antenna elements (R _{1 to} R _M ) are spatially separated from each other, whereby the spatial diversity sample (r _{1 to} r _M ) of the input composite signal (s ₁ + s ₂ +... + S _N ). ) Is obtained. The sampling rate corresponds to one sample or each sample in the spreading code. The antenna elements need to be arranged at a distance of 1/2 or more of the carrier wave wavelength. The number of antenna elements is N + 1 or more, where N is the number of transmitters 20 that have a strong influence on MAI. In the case of a general cellular system, the number of antenna elements is, for example, 4.
[0030]
Alternatively, a single antenna (R) may be used to receive different samples at different times. Here, time diversity samples are used in place of samples received by a plurality of antenna elements arranged spatially apart. Here, the sampling rate is (1 chip) / M, where M is the number of antenna elements required when time diversity is not used. Moreover, you may use combining a space diversity sample and a time diversity sample.
[0031]
A signal processing circuit 26 including a plurality of channel estimators generates a channel matrix H ′. Each element H ′ _nm of the channel matrix H ′ represents a combined channel characteristic between one of the transmission antennas (T _{1 to} T _M ) of the base station and one of the reception antennas (R _{1 to} R _M ), respectively. Yes. The method of channel estimation is well known, and any appropriate method may be used here.
[0032]
The signal received by the m-th antenna element (R _m ) of the desired user's receiver is expressed as the following equation (1).
[0033]
[Equation 5]

[0034]
Here, S _n represents a combined signal from each base station (20 _n ). The entire signal received by the M antennas of the desired user's receiver is expressed as the following equation (2).
[0035]
[Formula 6]

[0036]
The signal transmitted to the desired user is included in one of the combined signals S ₁ , S ₂ ,..., S _N from the base station. Without loss of generality, the desired signal included in the composite signal S _1. That is, the combined signals S _{2 to SN} are inter-cell MAI for a desired user, and the portion other than the desired signal in the combined signal S ₁ is an intra-cell MAI for the desired user.
[0037]
The procedure of the MAI suppression method in this specific example is as follows.
1. The signals received by all the antennas are arranged as r shown in Expression (2).
2. H _{11 to} H _MN are estimated using pilot signals from different base stations (cells).
3. A matrix H ⁺ shown in Equation (3) is generated for the signal arranged for the desired user shown in Equation (2).
[0038]
[Expression 7]

[0039]
Here, () ^h represents a conjugate transpose matrix, and () ⁻¹ represents an inverse matrix.
4). As shown in Expression (4), H ^{+ is} multiplied by the received signal r shown in Expression (2) from the right to reproduce the spread synthesized signals S ₁ , S ₂ ,..., S _N.
[0040]
[Equation 8]

[0041]
Here, the synthesized signals S ₂ ,..., S _N are ignored, and only the synthesized signal S ₁ is extracted. As a result, the inter-cell MAI is completely canceled.
5. Synthesis signals S ₁ directly descramble and despread. Since all all signals of the composite signal S ₁ is orthogonal to each other, the MAI cell is canceled by the processing of the descrambling and despreading.
[0042]
The antenna of the terminal device _(R 1 _{~R M)} received by the sample signal _(r 1 ~r _M), the channel matrix operator ^{H +} is multiplied from left, thereby, each base station or transmitter 20 The individual, temporally and spatially aligned composite signals (s _{1 to} s _N ) from are reproduced. By correcting the composite channel characteristics or fading H _mn between each base station and the receiving antenna (R) of the terminal device, the signals from the base stations are matched in time, and the sign of the signal from each base station The orthogonality is maintained.
[0043]
Since all the signals from a desired base station (20 ₁₎ (s ₁₎ is orthogonal, only simply multiplied by a predetermined code in the synthesis signal (s _1), the desired signal is calculated. In this way, both inter-cell MAI and intra-cell MAI are removed from the desired signal.
[0044]
The channel estimation processing in the signal processing circuit 26 and the matrix calculation in the matrix calculator 25 may be executed by hardware such as an application specific integrated circuit (ASIC), and are provided in the terminal device. It may be executed as software by a processor.
[0045]
For the sake of simplicity, the matrix operation may be realized using, for example, an iterative algorithm. A preferred iterative operation for the pseudo-inverse channel matrix H ⁺ is as follows:
[0046]
_Let H _n = [H _n−1 h _n ] be an M × n matrix. Here, an M ≧ _{n, H n-1} is a matrix comprising a first n-1 rows of _{H n, h n} is the n-th row of _{H n,} rank _{(H n)} is n It is.
[0047]
[Equation 9]

[0048]
Instead of these, other matrix operators that operate on a matrix (r _{1 to} r _M ) of received signal samples to calculate a composite signal (s _{1 to} s _n ) from each base station or transmitter (T). May be used. In other words, the temporally and spatially aligned samples of the synthesized signal including the desired signal are derived individually and then simply multiplied by the sign of the desired signal to obtain the desired signal in the usual manner. . The matrix operator needs to be “tall”, that is, the number of uncorrelated samples for one spread data is M, the number of transmitters is N, and the number of rows M is greater than the number of columns N. There must be.
[0049]
FIG. 6 shows a bit-error-rate (BER) with respect to the signal-to-noise ratio (Eb / N0) in the receiver shown in FIG. 5 and a conventional receiver using a Rake receiver (FIG. 2). It is a graph which compares and shows it. As can be seen from FIG. 6, the BER of the Rake receiver does not become very small even if the signal-to-noise ratio increases, but in the receiver shown in FIG. 5, the BER improves as the signal-to-noise ratio increases. This is because MAI is eliminated in the receiver shown in FIG.
[0050]
The system shown in FIG. 5 may simply be used in receiver space diversity mode, as described above, or may be used for transmitter diversity where multiple transmit antennas T transmit the same signal at each base station. . As a further modification, a space-time coding technique in which different complex symbols are transmitted from each antenna T of the base station may be used. However, this method requires an additional antenna on the receiver side. For example, if the number of transmission antennas is doubled, the number of reception antennas also needs to be doubled.
[0051]
It has been described by way of exemplary preferred meaning facilities the present invention. Variations and modifications apparent to those skilled in the art are within the scope of the invention.
[Brief description of the drawings]
FIG. 1 shows a mobile communication system operating in a multipath propagation environment.
FIG. 2 is a diagram showing a configuration of a conventional receiver incorporating a Rake receiver.
FIG. 3 is a diagram illustrating a configuration of a conventional receiver using adaptive equalization.
FIG. 4 shows a cellular WCDMA system comprising a plurality of base stations.
FIG. 5a shows a cellular WCDMA system according to the present invention.
FIG. 5b shows the structure of a receiver according to the present invention.
6 shows a comparison of the BER vs. Eb / N0 of the receiver shown in FIG. 5b with that of the receiver shown in FIG.

Claims (6)

In a receiving method of receiving a desired signal of code division multiple access from a signal obtained by combining a desired signal of code division multiple access and another signal of code division multiple access transmitted from a plurality of base stations,
Determining the number of received signal samples from the receive antenna array and defining a received signal vector;
Estimating multipath fading between the base station and the receive antenna array and defining a channel matrix;
Multiplying the received signal vector by a matrix operator based on the channel matrix to regenerate a combined signal from each base station;
Possess and detecting a desired signal from one of said combined signal,
The reception method according to claim 1, wherein the matrix operator is repeatedly calculated based on the following equation .

  The reception method according to claim 1, wherein the received signal sample is a time diversity sample and / or a spatial diversity sample.   3. The receiving method according to claim 2, wherein the spatial diversity sample is received by M different antennas of the receiving antenna array, where N is the number of transmitters and M> N.   4. The reception method according to claim 2, wherein the time diversity sample is sampled every t = (chip period) / M, where N is the number of transmitters and M> N.   The matrix operator is obtained by multiplying a conjugate transpose matrix of the channel matrix by an inverse matrix of a product of a channel matrix and a conjugate transpose matrix of the channel matrix. The receiving method described. In a receiving apparatus that receives a desired signal of code division multiple access from a signal obtained by combining a desired signal of code division multiple access and another signal of code division multiple access transmitted from a plurality of base stations,
A reception signal sample number determination means for determining the number of reception signal samples from the reception antenna array and defining a reception signal vector;
Multipath fading estimating means for estimating multipath fading between the base station and the receiving antenna array and defining a channel matrix;
A composite signal reproducing means for multiplying the received signal vector by a matrix operator based on the channel matrix and reproducing a composite signal from each base station;
And a detecting means that detect a desired signal from one of said combined signal,
The receiving apparatus according to claim 1, wherein the matrix operator is repeatedly calculated based on the following equation by the combined signal reproducing means .

Images