TW200412092A - Channel estimation in orthogonal frequency-division multiplexing (OFDM) systems - Google Patents

Channel estimation in orthogonal frequency-division multiplexing (OFDM) systems Download PDF

Info

Publication number
TW200412092A
TW200412092A TW92122187A TW92122187A TW200412092A TW 200412092 A TW200412092 A TW 200412092A TW 92122187 A TW92122187 A TW 92122187A TW 92122187 A TW92122187 A TW 92122187A TW 200412092 A TW200412092 A TW 200412092A
Authority
TW
Taiwan
Prior art keywords
channel
symbol
period
training
symbols
Prior art date
Application number
TW92122187A
Other languages
Chinese (zh)
Other versions
TWI239179B (en
Inventor
Kuo-Hui Li
Mao-Ching Chiu
Chao-Ming Chang
Charles Huang
Hung-Kun Chen
Original Assignee
Integrated Programmable Communications Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US10/308,513 external-priority patent/US7272108B2/en
Application filed by Integrated Programmable Communications Inc filed Critical Integrated Programmable Communications Inc
Publication of TW200412092A publication Critical patent/TW200412092A/en
Application granted granted Critical
Publication of TWI239179B publication Critical patent/TWI239179B/en

Links

Landscapes

  • Mobile Radio Communication Systems (AREA)
  • Radio Transmission System (AREA)

Abstract

A system and method for estimating channel characteristics in orthogonal frequency-division multiplexing (OFDM) systems with transmitter diversity is presented. The disclosed approach is compatible with the Institute of Electrical and Electronics Engineers (IEEE) "Wireless Local Area Network (LAN) Medium Access Control (MAC) and Physical Layer (PHY) Specification." In the disclosed system and method, an additional training symbol is transmitted during the data period. This provides additional information that may be used to more accurately estimate channel characteristics.

Description

200412092 玖、發明說明: 【發明所屬之技術領域】 本發明疋一種通祝糸統和方法,且特別關於一種估測通道特徵值 之系統和方法,應用於採傳輸分集式架構之正交分頻多工系統。 【先前技術】 射頻區域網路(Radio-frequency local area network)系統運作 受到美國聯邦機構嚴格管制。舉例來說,5.15 — 5. 25 GHz、5. 25_5ί 35 GHz和5.725-5.825 GHz免執照國家資訊基礎頻帶(unlicensed national informati〇n structure,u—nh)的使用受到美國聯邦管制 法(CFR) 15.407節47條的規範。不僅美國聯邦管制法對射頻網路的 使用有明文規範,其他組織,如··美國電子電機工程師協會(丨阳讨忉柁 of Electrical and Electronics Engineers,IEEE)也對無線網路 制叮技術規格’藉以確保不同製造商生產的無線系統的互通性。例如, 美國電子電機工程師協會制訂的『無線區域網路媒介接取控制層和實 體層規格:5Ghz頻帶之高速實體層』(wireless lAN Medium Access200412092 发明 Description of the invention: [Technical field to which the invention belongs] The present invention provides a general system and method, and particularly relates to a system and method for estimating channel characteristic values, which are applied to orthogonal frequency division using a transmission diversity architecture. Multiplexed system. [Previous technology] Radio-frequency local area network system operation is strictly controlled by the US federal agencies. For example, the use of 5.15-5. 25 GHz, 5. 25_5ί 35 GHz, and 5.725-5.825 GHz unlicensed national information structure (u-nh) is subject to US Federal Regulations (CFR) 15.407 Section 47 specifications. Not only does the US Federal Regulations regulate the use of radio frequency networks, other organizations, such as the Institute of Electrical and Electronics Engineers (IEEE) also regulate the technical specifications of wireless networks. This ensures interoperability of wireless systems from different manufacturers. For example, "Wireless LAN Network Access Control Layer and Physical Layer Specification: High-Speed Physical Layer in 5Ghz Band" (wireless lAN Medium Access)

Control (MAC) and Physical Layer (PHY) Specifications* High-Speed Physical Layer in the 5 GHz Band),以下簡稱為 IEEE 5Ghz 6 200412092 標準,就對5Ghz頻帶運作的系統訂定一些規範。 在IEEE 5Ghz標準中,其中的一部分是關於一正交分頻多工 (orthogonal frequency division multiplexing,0FDM)的實體層 收敛私序(physical layer convergence procedure,PLCP)子層。 第一 A圖繪示IEEE 5Ghz標準之一展現協定資料單元(Presentati〇n Protocol Data Unit,PPDU)的訊框格式。如第一 A圖所示,該展現 協疋:貝料單元訊框包括:一短訓練時段(sh〇rt_training peri〇(j)H〇 ; 一長训練時段(long—training period) 120,接於該短訓練時段ho 之後,一信號時段(signaling peri〇d) 13Q,接在該長訓練時段12〇 之後’以及多個資料時段(data period) 140、142、144,接在該信 5虎時段130之後。依據IEEE 5Ghz標準,該長訓練時段120、信號時段 130和多個資料時段14〇、142、144都包含一保護時間間距(guard interval , GI)。 該短訓練時段110内包含十個符元(纪6 · · . A,“),功用 為"ί口號4欢’則(signal detection)、初步頻率獲取(coarse—freqUenCyControl (MAC) and Physical Layer (PHY) Specifications * High-Speed Physical Layer in the 5 GHz Band, hereinafter referred to as the IEEE 5Ghz 6 200412092 standard, establishes some specifications for systems operating in the 5Ghz band. In the IEEE 5Ghz standard, a part of it is about a physical layer convergence procedure (PLCP) sublayer of an orthogonal frequency division multiplexing (OFDM). Figure A shows the frame format of a Presentation Protocol Data Unit (PPDU), one of the IEEE 5Ghz standards. As shown in the first figure A, the display cooperation: the frame of the shell material unit includes: a short training period (sh〇rt_training peri〇 (j) H〇; a long training period (long-training period) 120, then After the short training period ho, a signal period (signaling period) 13Q, followed by the long training period 120 ′ and multiple data periods 140, 142, and 144, followed by the letter 5 After period 130. According to the IEEE 5Ghz standard, the long training period 120, signal period 130, and multiple data periods 14, 40, 142, and 144 all include a guard interval (GI). The short training period 110 includes ten A symbol (Ji 6 · ·. A, "), whose function is " ί slogan 4 huan '(signal detection), preliminary frequency acquisition (coarse—freqUenCy

acquisition)、分集式接收挑選(diversity seiecti〇n),以及 IEEE 5Ghz私準所制叮的其他功能。關於短訓練時段Η。的完整規範詳細定 義於IEEE 5Ghz標準,在此就不在對其做更深入討論。acquisition), diversity reception selection (diversity seiection), and other features of the IEEE 5Ghz private standard. About short training periods Η. The full specification of is defined in detail in the IEEE 5Ghz standard and will not be discussed in more detail here.

該長訓練時段120包含一保護時間間距⑽),和二長訓練符元(7J 與冗)。依據iEEE 5Ghz標準,每個長刻練符元由53個子載波的符元 所組成,其直流電壓值為零,而其内容由_串數字序列I經過運算而 產生。該數字序列j如下: Z = {U,-1,-U,l,- [Eq. 13. 此外,臓5Ghz標準定義了長訓練符元之產生運算函數為: 53 χ(ί)=w{t Λ xijz^^^~^)^F{t~TGI2) ㈣ , [Eq. 2], 其中灿是該長訓練符元之—時_符元;心)是-加權因子, 功用為訊號之頻譜細驗(spe咖麵ing);々是—子載波引數; 双々)是Eq. 1所定義的訓練符元之一來 "’而是保護時間間距,在 IEEE 5Ghz標準中該數值為l 6 s。 IEEE 5Ghz標準除了定義該長訓練 的内各外(依據Eq. 2),同 時也規定長訓練符元的數目為二(yj遍 -、幻’藉以增加通道估測的精準 度0 又’ IEEE 5Ghz標準規定第一長訓練符元 之内容相同。因此,設計相同的長訓練符元^ Z與第二長訓練符元Z 第一長訓練符元J 155 和第 >一長糾I練付元H 65依序傳送於兮 長訓練時段12〇。 所以,對一二 支傳輸分集式正交分頻多 傳送如下内容: 工系統(如:第二圖所示), 〜第一傳輸器260 (1)於長訓練時段 165a於一第一通道忍; 之時間内,傳送二長訓練符 元/155a和/ S 170a於該第一通 (2)於k號時段13〇之& 之時間内,傳送信號資料 道忍;又, 傳送資料/¾ 180a和 ⑶在後續資料時段叫142之時間内 /¾ 190a於该第一通道忍。 類似地’ 一第二傳輸器265傳送如下内容·· ⑴於長t)i|練時段⑽之時間内,傳送二長觀符元^娜和^ 165b於一第二通道怂; ⑵於信號時段130之時間内,傳送信號資訊snob於該第二通 道从;又, (3)在後續資料時段140和142之時間内,傳送資料Λ 180b和 众190b於該第二通道总。 一接收器205接收來自第一、第二傳輸器之傳輸訊號,而該接收訊 '疋傳送付元和通道特徵值(channel charaCteriStiC)的函數。妙 過移除賴時_輯,每健收符元辦—反向侧輯換運算。 200412092 如此,對於一如第二圖所示的二支傳輸分集式正交分頻多工系統,頻 率軸上該接收訊號K可表示為: Υ^(Η^Χ) + (ΗΒ·Χ) + Ζη ’ [Eq. 3]. 其中Z是接收雜訊,忍和你是頻率軸上的通道特徵值,並且假設 在訊框時間(frame duration)内維持不變,而此二通道的傳輸延遲 也假設是相同。因為同一長訓練符元/傳送於此二支傳輸分集式系統 的二傳輸器上,Eq. 3可簡化為: y^{ha^hb)-x^z,. [Eq. 4]. 類似地,後續接收資料區塊可表示為: y2=(ha+hb)^x + z2j [Eq· 5], Y3={Ha+Hb)^S + Z3, [Eq. 6], Y4=HA.DAl+HB.DB,Z4, [Eq. 7], 和 y5 = ha.da1 + hb.dB2 + z5 [Eq. 8]. 合併Eqs. 4和5,可獲得下列方程式: (Y^Y2)-r =(ΗΑ+ΗΒ)\2\χ\2)^(ζ^ζ2)·χ\ [Eq. 9], 經過移項後,又可改寫為·· 10 200412092 A+圮(z 丨+ζ2)·χ 2 ~2, 或更精確的表示為: HA(k) + HR(k) = ^hm)) ·X{k) (Z{(k)H-Z2(k)).X(k) 2 2 ,灸= 1,L,", [Eq· l〇], 其中丨是正交分頻多工系統的子載波數目,而々是子載波的引數。 在Eq· 1中,對於所有的々值,x(々)e{±1},因此傳輪符元/α)的 複數共梃符it /⑷也是和傳輸符元相同。又,因為雄)e{±1},該傳輸 符元/U)的平方項丨/⑷|2都會是卜此外,因為z(/:)e㈣,(Z⑷+ 么U))/U)和(/(々)+么(々))的統計值,一般來說是相同的。 略過雜訊項,此二通道的綜合特徵值d从+你)可用下列方程式 來估測得知: k = hLjN [Eq. 11], 雖然Eq· 11提供一種計算綜合通道特徵值从的方法,很明顯的, 重複傳送符元/卻無法協助區別個別通道(总和你)的通道特徵值。 換句話說,因為二支傳輸的通道(忍和你)上都傳輸單一符元/,因 此該系統的通道特徵值就可用一二元單一方程式來表示,只有綜合通 道特徵值展可以計算得知。又,雖然重複傳送符元j可增加訊號雜訊 11 200412092 比(s獅Ho-n〇ise ratl0,SNR)可是訊號雜訊比的提升卻無法協 助計算出個別通道的特徵值。 、、、雖然目前有—些複雜演算法可以從综合通道特增值中,分隔出個別 k的特&值’但疋⑨麵算法都需要先對通道雜做額外的假設, 才可能正確估測個職道的特徵值。因此,這些通道估測演算法僅適 用於這些假設紅的條件下。又,因為這些通道侧演算法的高運算 雜度虽此—支傳輸分集式系統擴充為更多支傳輸架構時(如:三 支或四支傳輸),此計算複雜度將會成指數倍數成長。 因此在此-工業領域存在著_個尚未有解決方案的需求,來解決 上述的問題,以補足現有演算法的不足與不適切性。 、 【發明内容】 加本發明是一種估測通道特徵值之系統和方法,應用於採傳輸分集式 蜋構之正父分頻多工系統環境。 簡言之’本發明之系統之一實施例在架構上包括一傳輸單元,功能 ^第巧段傳送―訓練符胁+通道,同時,傳送該訓練符元 二通道;於第二時段傳送該訓練符元之—複數共輛符元於該第 通道,同時’傳送該训練符元之一負健數共輛符元於該第二通道。 么月提ί、通道特徵值估測方法,翻於採傳輸分集式架構之 12 200412092 正交分頻多工系統環境。 本發明之此方法之一實施例的操作步驟包括:於第一時段傳送一訓 練符元於一第一通道;同時傳送該訓練符元於一第二通道;於第二時 段傳送該訓練符元之-複數共輛符元於該第一通道;同時,傳送該訓 練符元之一負值複數共軛符元於該第二通道。 習此藝者在審視下文之圖示說明和實施方式後,應可輕易暸解本發· 明之其他系統、方法、功能和優點。而且,舉凡此系統、方法、功能 與優點皆屬本發明之專利範圍。 【實施方式】 下文將詳細敘述圖示所繪示之實施例。雖然上列圖示已繪示部分實 施例,但本發明之實施例並不限於此;相反的,本發明範圍涵蓋其他 相關、局部更動或等效之實施例。 w 本發明的一些實施例將在下文中敘述。在這些實施例中,利用傳送 額外的訓練符元,以達到精確估測通道特徵值的目的。不同於先前系 統和方法必須耗費龐大運算能力或是對一多傳輸分集式系統有額外的 操作假設,本發明之實施例僅需簡單計算能力,並對多傳輸分集式系 統做較少的操作假設。 第二A圖和第三β圖繪示一系統之一實施例,用以估測通道特徵 13 200412092 值。第三A圖繪示一第一傳輸器所傳送之符元,而第三B圖繪示一第 二傳輸器所傳送之符元。第三A圖和第三b圖繪示本發明之一實施例 之一實體層收斂程序(physical layer convergence procedure,PLCP) 檔頭欄位(preamble field),其功用為訊號同步。如第三A圖和第三 B圖所示,該實體層收斂程序檔頭欄位包含:一短訓練時段31〇 ; 一長 訓練時段320 ,接於該短訓練時段31〇之後;一信號時段,接於該 長訓練時段320之後;以及一複數個資料時段34〇、342、3仏等,接 於該信號時段之後。依據臓5Ghz鮮,該綱料段31〇、長 訓練時段320、信號時段330和多資料時段34〇、342、344都包含一保 護時間間距。 丨兮达付兀於該短訓練的The long training period 120 includes a guard interval ⑽) and two long training symbols (7J and redundant). According to the iEEE 5Ghz standard, each long-cut symbol consists of 53 subcarrier symbols. Its DC voltage value is zero, and its content is generated by the _ string of digital sequences I after operation. The number sequence j is as follows: Z = {U, -1, -U, l,-[Eq. 13. In addition, the 臓 5Ghz standard defines the operation function for generating long training symbols as: 53 χ (ί) = w { t Λ xijz ^^^ ~ ^) ^ F {t ~ TGI2) ㈣, [Eq. 2], where Chan is the long training symbol—time_symbol; heart) is a weighting factor, the function of which is the signal Spectral inspection (speca-ing); 々 is-the subcarrier argument; double 々) is one of the training symbols defined by Eq. 1 to "'but the guard interval, which is the value in the IEEE 5Ghz standard Is l 6 s. The IEEE 5Ghz standard not only defines the internals of the long training (in accordance with Eq. 2), but also specifies the number of long training symbols to be two (yj times-, "magic 'to increase the accuracy of channel estimation 0 and' IEEE 5Ghz The standard stipulates that the content of the first long training symbol is the same. Therefore, the same long training symbol ^ Z and the second long training symbol Z are designed. The first long training symbol J 155 and the > H 65 is sequentially transmitted at a long training period of 120. Therefore, the diversity orthogonal transmission of one or two transmissions is as follows: industrial system (as shown in the second figure), ~ the first transmitter 260 ( 1) During the long training period 165a on a first channel; within the time period, send the two long training symbols / 155a and / S 170a within the first pass (2) within the time period of 30th & , The transmission of data and data; and, the transmission data / ¾ 180a and ⑶ in the subsequent data period called 142 / ¾ 190a on the first channel. Similarly, a second transmitter 265 transmits the following content ... · During the time period of long t) i |, the second long view symbol ^ na and ^ 165b are transmitted on a second pass. (1) within the signal period 130, transmit the signal information snob from the second channel; and, (3) within the subsequent data periods 140 and 142, transmit the data Λ 180b and the public 190b to the second channel Channel total. A receiver 205 receives transmission signals from the first and second transmitters, and the reception signal is a function of transmitting a payment element and a channel charaCteriStiC. It is better to remove the time-lapse series, and do it every time the symbol is received—the reverse-side sequence conversion operation. 200412092 So, for a two-branch transmission diversity orthogonal frequency division multiplexing system as shown in the second figure, the received signal K on the frequency axis can be expressed as: : ^ (Η ^ ×) + (ΗΒ · χ) + Zη '[Eq. 3]. Where Z is the received noise, and you are the channel characteristic value on the frequency axis, and it is assumed to remain unchanged within the frame duration, and the transmission delay of the two channels is also The assumption is the same. Because the same long training symbol / transmission is on the two transmitters of the two transmission diversity systems, Eq. 3 can be simplified as: y ^ {ha ^ hb) -x ^ z ,. [Eq. 4]. Similarly , The subsequent received data blocks can be expressed as: y2 = (ha + hb) ^ x + z2j [Eq · 5], Y3 = {Ha + Hb) ^ S + Z3, [Eq. 6], Y4 = HA.DAl + HB.DB, Z4, [Eq. 7], and y5 = ha.da1 + hb.dB2 + z5 [Eq. 8]. Combining Eqs. 4 and 5 gives the following equation: (Y ^ Y2) -r = (ΗΑ + ΗΒ) \ 2 \ χ \ 2) ^ (ζ ^ ζ2) · χ \ [Eq. 9], after shifting the term, it can be rewritten as ... 10 200412092 A + 圮 (z 丨 + ζ2) · χ 2 ~ 2, or more precisely: HA (k) + HR (k) = ^ hm)) · X {k) (Z {(k) H-Z2 (k)). X (k) 2 2 , Moxibustion = 1, L, ", [Eq · l0], where 丨 is the number of subcarriers of the orthogonal frequency division multiplexing system, and 々 is the argument of the subcarrier. In Eq · 1, for all 々 values, x (々) e {± 1}, so the complex number it / ⑷ of the round symbol / α) is also the same as the transmission symbol. In addition, because the male) e {± 1}, the square term of the transmission symbol / U) 丨 / ⑷ | 2 will be Bu. Furthermore, because z (/ :) e㈣, (Z⑷ + Mod U)) / U) and The statistical value of (/ (々) + Mod (值)) is generally the same. Ignoring the noise term, the integrated eigenvalues d of the two channels (from + you) can be estimated by the following equation: k = hLjN [Eq. 11], although Eq · 11 provides a method for calculating the integrated channel eigenvalues from Obviously, repeated transmission of symbols / can't help distinguish the channel characteristic values of individual channels (total you). In other words, because a single symbol / is transmitted on the two transmission channels (forbearance and you), the channel characteristic value of the system can be expressed by a single binary equation, and only the comprehensive channel characteristic value extension can be calculated. . In addition, although repeated transmission of symbol j can increase signal noise 11 200412092 ratio (shoho-noise ratl0, SNR), but the improvement of signal noise ratio does not help calculate the characteristic value of individual channels. Although there are some complex algorithms that can separate the special & value of individual k from the integrated channel special added value, all the surface algorithms need to make additional assumptions on the channel miscellaneous before they can be correctly estimated. Eigenvalues of each job. Therefore, these channel estimation algorithms are only suitable for these hypothetical conditions. Also, because of the high computational complexity of these channel side-effect algorithms—when the branch transmission diversity system is expanded to more branches (such as three or four branches), the computational complexity will grow exponentially. . Therefore, there is a need in the industrial field for which there is no solution to solve the above problems, so as to make up for the shortcomings and inadequacy of existing algorithms. [Summary of the Invention] The present invention is a system and method for estimating channel characteristic values, and is applied to the environment of a positive-parent frequency-division multiplexing system using a transmission diversity method. In short, 'an embodiment of the system of the present invention includes a transmission unit in the architecture, and the function ^ paragraph transmission-training symbol threat + channel, at the same time, the training symbol channel is transmitted; the training is transmitted in the second period The number of symbols—a total number of symbols in the second channel, and a negative number in the training symbol—a total number of symbols in the second channel. Mo Yue mentioned that the channel characteristic value estimation method is based on the transmission diversity architecture 12 200412092 orthogonal frequency division multiplexing system environment. The operation steps of an embodiment of the method of the present invention include: transmitting a training symbol on a first channel in a first period; transmitting the training symbol on a second channel; and transmitting the training symbol in a second period. Among them, a plurality of complex symbols are transmitted on the first channel; at the same time, a negative complex conjugate symbol of one of the training symbols is transmitted on the second channel. After reviewing the illustrations and implementations below, the artist should be able to easily understand other systems, methods, functions, and advantages of the present invention. Moreover, all such systems, methods, functions and advantages are within the patent scope of the present invention. [Embodiment] The embodiment shown in the drawings will be described in detail below. Although some of the embodiments have been shown in the above figure, the embodiments of the present invention are not limited to this. On the contrary, the scope of the present invention covers other related, partially modified or equivalent embodiments. w Some embodiments of the invention will be described below. In these embodiments, extra training symbols are transmitted to achieve the purpose of accurately estimating channel characteristic values. Different from previous systems and methods that must consume huge computing power or have additional operating assumptions for a multi-transmission diversity system, the embodiments of the present invention only require simple computing power and make fewer operating assumptions for multi-transmission diversity systems. . The second graph A and the third beta graph illustrate an embodiment of a system for estimating the channel characteristic 13 200412092 value. The third diagram A shows the symbols transmitted by a first transmitter, and the third diagram B shows the symbols transmitted by a second transmitter. Figures 3A and 3b show a physical layer convergence procedure (PLCP) preamble field, which is an embodiment of the present invention, and its function is signal synchronization. As shown in Figures 3A and 3B, the header field of the entity layer convergence program contains: a short training period of 31; a long training period of 320, which follows the short training period of 31; a signal period , After the long training period 320; and a plurality of data periods 34, 342, 3, etc., after the signal period. According to 臓 5Ghz, the program segment 31, long training period 320, signal period 330, and multi-data period 34, 342, 344 all include a protection time interval.丨 Xida pays attention to the short training

310 ’依據臓5Ghz標準。t該短訓練符元傳送後,長訓練符元u 和/ 365a傳送於該長訓練時段32()。此處,大寫^代表—正交分资 工系統之-組頻率軸符心何視為是—向量(赠⑷,包含㈣ 素⑽園t),㈣是該正交分頻㈣統之子紐數目。符元^ 元素JU)是由第1個子紐所傳送。必須知道的是,在經由一制 線發射之前’該航|會級過反向傅利葉轉換㈤⑽. 一)運算’變成—時間轴訊號,並且添加―循環_ 14 200412092 (cycllc preflx) ’接著利用一射頻模組轉換為—射頻類比訊號。 緊接於符元/重複傳送之後,信號資料U7〇a傳送於訓練時段 330。傳送該錢資料^施後,長訓練符元之_複數共減 傳送於-第-資料時段34Q。如上文所述,因為〇每侃素都為實數 (™ber)自然的,丫中每個元素也都為實數。同時,因為I中 每個元素都為實數,所以岛z是相同的。然而,必須知道的是,在 臓5Ghz標準之外,^不—定必定是實數,可以是複數,擁有虛 部元素。 類似地如第圖所不,該第二傳輸器依據臟驗標準,傳 送訓練符元於該短訓練日後31Q。#該短訓練符元傳送後,長訓練符元 / 355b和/ 365b傳送_練時段緊接於符元{重複傳送 之後,信號資料纖b傳送於訓、練時段33〇。傳送該信號資料謂b 後,長訓練符元之-負值複數共概符元_傳送於—第一資料時 段340。如上文所述,因為z中每個元素都為實數,自然的,—f中每 個元素也都為實數。同樣的’必須知道的是,在臟論標準之外, /不一定必須限定是實數,可以是複數,擁有虛部元素。基於此一認知, 如果/是-複數,傳送於時段385a和3咖的符元可以是(U、 H味賴W -顯符元軸。為簡化說明,下文敘述 200412092 將採用符元對次)為例作說明。 如上文中所敘述,本發明之實施例不僅是重複傳送訓練符元/,第 三A圖和第三B圖之系統還在該第一通道增添傳送共軛符元/,以及 在第二通道增添傳送負值共軛符元—/。增添傳送符元^和—/的好處 將於下文中敘述,並搭配第四圖說明之。 第四圖繪示一二支傳輸分集式正交分頻多工系統,包括一無線裝置 0 (wireless device) 470和一接收器405。該無線裝置470可以是一 無線區域網路的接取點(access p〇int)單元、一無線區域網路卡、 蜂巢式電活、一無線個人數位助理(pers〇nal assista时, PDA)或是一具無線傳輸功能的可攜式電腦等。如第四圖所示,該無線 裝置470包括二傳輸器460和465,傳送資料於一正交分頻多工環境。 該接收器405接收由此二傳輸器侧和465所傳送之訊號。如第四圖 所示’ 一第一通道轉移函數(channel transfer function)总對第_ 馨 傳輸器侧所傳送之訊號造成改變,而-第二通道轉移函數簡第二 傳輸器465所傳送之訊號造成改變。此通道轉移函數就是通道特徵值。 因此,當該第一傳輸器460和該第二傳輸器465傳送符元^時(其 步驟為:先對符元,進行反向傅利葉轉換運算,產生一時間軸訊號!; 接著加入-循環前置位元組,產生伽;透過一射頻模組,將如轉換為 16 一射頻^ °就’並且利用傳輸天線傳送;^),在頻率軸上的接收符元 Κ可表示為: 、1疋弟一接收符元之雜訊。因為相同訓練符元/傳送於此_ 專輸刀集式系統之二傳輸器上,助· 12可簡化為:310 'according to the 5Ghz standard. After the short training symbol is transmitted, the long training symbols u and / 365a are transmitted in the long training period 32 (). Here, capital ^ represents-what is the meaning of the group frequency axis of the orthogonal frequency division system is a vector (gift, including ㈣ prime unit t), where ㈣ is the number of children of the orthogonal frequency division system . The symbol ^ element JU) is transmitted by the first child button. It must be known that, before launching through a line, 'the flight | will pass the inverse Fourier transform. 一. A) The operation' becomes the time axis signal, and add ―Cycle_ 14 200412092 (cycllc preflx) ‖ and then use a The RF module is converted into an RF analog signal. Immediately after the symbol / repeated transmission, the signal data U70a is transmitted in the training period 330. After sending the money data ^, the long complex symbol _ complex number is subtracted. It is sent at -Q-data period 34Q. As mentioned above, because every prime is a real number (™ ber), every element in Ya is also a real number. At the same time, because every element in I is a real number, the island z is the same. However, it must be known that outside the 臓 5Ghz standard, ^ No—it must be a real number, it can be a complex number, and it has imaginary elements. Similarly, as shown in the figure, the second transmitter transmits training symbols 31Q after the short training day according to the dirty test standard. #After the short training symbol is transmitted, the long training symbol / 355b and / 365b are transmitted _ training period immediately after the symbol {repeated transmission, the signal data fiber b is transmitted at the training and training period 33 °. After the signal data is transmitted as b, the long training symbol-negative complex number common symbol_ is transmitted in the first data period 340. As mentioned above, because each element in z is a real number, naturally, every element in -f is also a real number. It is also necessary to know that outside of the dirty standard, / does not have to be limited to real numbers, it can be complex numbers, and has imaginary elements. Based on this knowledge, if / is-plural, the symbols transmitted in the periods 385a and 3c can be (U, H Weilai W-explicit symbol axis. To simplify the description, the following description 200412092 will use symbol pairs) As an example. As described above, the embodiment of the present invention is not only to repeatedly transmit training symbols /, the system of the third graph A and the third graph B adds transmission conjugate symbols / on the first channel, and Send negative conjugate symbol-/. The benefits of adding the teleportation symbols ^ and-/ will be described below and illustrated with the fourth figure. The fourth figure shows a two-transmit diversity orthogonal frequency division multiplexing system including a wireless device 0 (wireless device) 470 and a receiver 405. The wireless device 470 may be an access point unit of a wireless local area network, a wireless local area network card, a cellular electrical activity, a wireless personal digital assistant (PDA), or It is a portable computer with wireless transmission function. As shown in the fourth figure, the wireless device 470 includes two transmitters 460 and 465 for transmitting data in an orthogonal frequency division multiplexing environment. The receiver 405 receives signals transmitted by the two transmitter sides and 465. As shown in the fourth figure, a first channel transfer function always changes the signal transmitted by the __th transmitter side, and the second channel transfer function is the signal transmitted by the second transmitter 465 Cause change. This channel transfer function is the channel characteristic value. Therefore, when the first transmitter 460 and the second transmitter 465 transmit the symbol ^ (the steps are: first perform inverse Fourier transform operation on the symbol to generate a time axis signal !; then add-before the loop Set the tuple to generate gamma; through a radio frequency module, it will be converted into 16 radio frequency ^ ° 'and transmitted using a transmission antenna; ^), the received symbol κ on the frequency axis can be expressed as:, 1 疋Brother Yi receives noise from Fu Yuan. Because the same training symbols / transmissions are used on the second transmitter of the special-set tool set system, the help · 12 can be simplified to:

恥㈨ Α).χ + Ζι [Eq. 13]. :、、的因為相同訓練符元又再度被傳送,此二傳輸器46〇和处 所送出之第二接收符^可表示為: Y^+Hb).X+Z2 [Eq. 14]. 又,如果信號資料370a接著傳送,為一第三傳送符元Γ3 , 收符元則是: ^⑼ + [Eq. 15], ’、中S是頻率軸上的信號資料。在—實施例中,當傳送信號資料^ 之後’«訓練符元之複數共姆元/3阪傳送於該第—傳輸器侧, 成為第四^ 7; ; *,該長訓練符元之貞值複數共婦元傳送於該 第-傳輸為465 ’成為第四符元$。如上文所提及,因為,是實數,該 複數共姆元作負值複數共_元—Λ都騎數。此外,基於相同 原因’符元間有以下關係式·· 17 200412092 [Eq. 16], -Ϊ--Χ [Eq. 17], 和 \X(k)\2=l 〇 [Eq. 18]· 所以,依據IEEE 5Ghz標準,該第四接收符元可表示為: Y4 = (HA-xy(HBi-X)yz4, [Eq. 19], _ 或,更進一步簡化為·· Υ4 = (ΗΑ-ΗΒ}Χ^Ζ4 [Eq· 20]· 結合Eqs. 13和20,就可提供一方法,單獨計算出忍和怂值。 換句話說,不同於先前技術僅能計算綜合通道特徵值(从-似总),本發 明可分離計算出从和怂個別的數值,因為: (Υ^Υ4)·Χ*=({ΗΑ+ΗΒ)·Χ + Ζ^Χ^((ΗΑ-ΗΒ)·Χ + Ζ4)·Χ* = 2Hx\X^{Z^Z,\r · [Eq. 21]. 在此必須注意的是,Eq. 21式中每個變數項目都是一正交分頻多 工符元之一頻率軸訊號。若是由子載波的角度來敘述,Eq. 21可改寫 為: 18 ,412〇92 [Eq. 22], 其中少是正交分頻多工系統中子載波之數目,而々是子載波之引數。 计异Eq· 13和20後’通道轉移函數嚴々)就可用下列方程式、· 野卜剛+聊婦) 2 [Eq. 23]· 又,依據 Eq. 23, Ha 可估 涓if 為: 好⑴一剛+聊.雄) k = \,1 ,Ν [Eq. 24], 或,簡化表示為: HaJY^'x 2 [Eq· 25]· 在此必須注意的是,Eqs. 24和25方程式中, 一估測誤差 (estimation error)與雜訊項(说幻//2有關。一般而言,該雜訊項 之估測誤差之平均值(mean)為0 (j?((z丨+Z4)x/2) = 〇,f是統計期望值 函數)。而且,該估測誤差之變異值(variance )是σ〗/2 · (var((Z+Z〇J/2)= var(U+Z4)/2)二醫((ζ!+ζ4)/2卜σ】/2,其中 var(;) 疋統計變異值函數,而Z和%之變異值假設為4)。 该第一通道之特徵值总也可以類似方式計算得之: (Υ,-Υ^-Χ + ~^ηα-Ηβ)·Χ^-Ζ4)·Χ* = 2Ηβ·\Χ\2+(Ζ^Ζ4)·Χ* 19 200412092 [Eq. 26], 又更進一步簡化為:Disgrace Α) .χ + Zι [Eq. 13].: ,,, because the same training symbols are transmitted again, the second receiver ^ sent by the two transmitters 46 and the place can be expressed as: Y ^ + Hb) .X + Z2 [Eq. 14]. Also, if the signal data 370a is transmitted next, it is a third transmission symbol Γ3, and the received symbol is: ^ ⑼ + [Eq. 15], ', where S is Signal data on the frequency axis. In the embodiment, when the signal data ^ is transmitted, the "common number of training symbols / 3/3" is transmitted on the first transmitter side and becomes the fourth ^ 7; *, the long training symbol The value of the plural number is transmitted at the first-transmission of 465 'to become the fourth symbol $. As mentioned above, because it is a real number, the complex common number element is a negative complex number total element — Λ is the riding number. In addition, for the same reason, there are the following relations between the symbols: 17 200412092 [Eq. 16], -Ϊ--Χ [Eq. 17], and \ X (k) \ 2 = l 〇 [Eq. 18] · Therefore, according to the IEEE 5Ghz standard, the fourth receiving symbol can be expressed as: Y4 = (HA-xy (HBi-X) yz4, [Eq. 19], _ or, further simplified to ·· Υ4 = (ΗΑ -ΗΒ} χ ^ Z4 [Eq · 20] · Combined with Eqs. 13 and 20, a method can be provided to calculate the tolerance and counseling values separately. In other words, unlike the previous technology, only the integrated channel characteristic values (from -Similarly), the present invention can separate and calculate individual values from the following, because: (Υ ^ Υ4) · × * = ({ΗΑ + ΗΒ) · × + ^^^ ((ΗΑ-ΗΒ) · χ + Z4) · Χ * = 2Hx \ X ^ {Z ^ Z, \ r · [Eq. 21]. It must be noted here that each variable item in Eq. 21 is an orthogonal frequency division multiplexing One of the symbols is the frequency axis signal. If it is described from the angle of the subcarrier, Eq. 21 can be rewritten as: 18,412〇92 [Eq. 22], where the number of subcarriers in the orthogonal frequency division multiplexing system is at least, And 々 is the argument of the subcarrier. After calculating the difference between Eq · 13 and 20, the channel transfer function is strict.) The following equation can be used (, Ye Bugang + Liao Fu) 2 [Eq. 23] · Also, according to Eq. 23, Ha can be estimated as: Any good Yigang + Liao. Xiong) k = \ , 1, N [Eq. 24 ], Or, the simplified expression is: HaJY ^ 'x 2 [Eq · 25] · It must be noted here that in Eqs. 24 and 25, an estimation error and noise terms (say magic / / 2 is related. In general, the mean (mean) of the estimation error of this noise term is 0 (j? ((Z 丨 + Z4) x / 2) = 〇, f is a statistical expectation value function). Moreover, The variation of the estimation error is σ〗 / 2 · (var ((Z + Z〇J / 2) = var (U + Z4) / 2) Second Doctor ((ζ! + Ζ4) / 2 Bu σ] / 2, where var (;) 疋 statistical variation value function, and the variation values of Z and% are assumed to be 4). The characteristic value of the first channel can always be calculated in a similar manner: (方式, -Υ ^ -Χ + ~ ^ ηα-Ηβ) · × ^ -Z4) · × * = 2Ηβ · \ × \ 2+ (Z ^ Z4) · × * 19 200412092 [Eq. 26], which is further simplified as:

Ht 或 [Eq· 27].Ht or [Eq · 27].

[Eq. 28]. 因此怂可估測為: Η. ,〇W4)l 2 [Eq. 29].[Eq. 28]. Therefore, it can be estimated as: Η., 〇W4) l 2 [Eq. 29].

如同於Eqs. 24和25之估測值,E · 和29式中之一估測誤 差也與雜訊項(^U)272有關。因此,兮“ 、 。亥雜訊項之估測誤差之平均值為 〇 (双UW/2H)) ’而且該估 秩差之變異值是4/2 (var((U〇J/2> ν8Γ((^Κ4)/2)=σ】/2 )。 由Eqs· 12 S 29之方%式推導可上乂看出,藉由在資料時段之第一 ^刀傳送叙/和—/,每個單獨通道之特徵值皆可精確計算。所以, 不僅侷限於估測綜合通道特徵值,本發明之方法也可用以估測單獨通 道之特徵值。 本發明之另一實施例,藉由結合Eqs. 13、14和20,可獲得較佳 20 200412092 訊號完整性和較低估測誤差。因為Eqs· 13和14是重複傳送相同訓練 符元之後的接收訊號,所以結合此二方程式可視為進一步的訊號平 均運算(signal averaging)。因此,經由利用重複傳送訓練符元,所 帶來訊號雜訊比提升的好處,個別的通道特徵值可依照下列方程式計 算: W + r2 + 2r4 )r = | 外 + (Zi + Z2 + 2ζ4 ) · X* [Eq. 30], 和 H _(^ + y2 + 2y4)-x (z,+z,+2z,Vx " 4 ~ 4 或,等效表示為: [Eq· 31], 幻+2[⑽.錄)hlL ^ 4 [Eq. 32], 因此,从可估測為:卜迎)ι +明)+邛_•雄),[Eq. 33] 4 ? * 而且,不同於Eqs· 24、25、28和29之計算,Ecl· 32中雜訊X員所 導致之估測誤差是(2+2+2/)//4。所以,估測誤差之平均值是〇 (汉(说说2/4)^74)=0 ),而估測誤差之變異值是3$/8 21 200412092 (var((/^+Z+2Z)//4)=var((^I+石+2石)/4)=3σ】/8,其中假設/、石和 Z之變異值為σ〗)。 由Eq. 32可看出,估測誤差之變異值減少,連帶提升估測精準度。 類似的,第二通道之特徵值怂也可由下列方程式得之: [Eq. 34], 於上式中,估測誤差之平均值是〇 (風(汾么―2Z4)J/4>〇)而估測 决差之變異值是 W/8 ( var( (^+^2-2^)//4)= var((Z+Z2-2^)/4)=34/8,其中假設z、石和%之變異值為)。 總體而言,藉由傳送額外的長訓練符元火、或是該符元之複數共軛 符元/和負值複數共軛符元—/,估測誤差之變異值可進一步降低。 對於上述之多傳輸分集式系統,本發明之另一實施例可當成一方 法’估測通道特徵值。第五圖和第六圖繪示此方法之實施例。 第五圖是-流程圖,緣示本方法之實施步驟,運作於一二支傳輸分 集式正交分頻多工系統之無線裝置·。依據臓5Ghz標準,在第五 圖傳輸之訊號都包含保護時間間距。如第五圖所示,於一第一時段時, 一訓練符元傳送於第-和第二通道(步驟·)。在—實施例中,該無 線錢470包括第-通道傳輸單元(transmitiQgic) 555和第二通道 22 200412092 傳輸單元565 在傳送資訊於第—通道和第二通道。緊接於第— 時段之傳送訓練符元之後(步驟52G),在第二時段中,該符元之… 數共輛符元傳送於該第-通道(步驟咖)。於此同時,在第二時段中是 該符元之-負值複數共歸元傳送於該第二通道(步驟⑽)。. 若通道估測方法依據臟5Ghz標準,該第—時段是長訓練時段之As with the estimated values of Eqs. 24 and 25, the estimated error of one of E · and 29 is related to the noise term (^ U) 272. Therefore, the average value of the estimation errors of the noise terms is “0 (double UW / 2H))” and the variation of the estimated rank difference is 4/2 (var ((U〇J / 2 > ν8Γ) ((^ Κ4) / 2) = σ] / 2). It can be seen from the derivation of Eqs · 12 S 29% formula that by sending the // — / in the first ^ knife of the data period, each The characteristic values of the individual channels can be accurately calculated. Therefore, the method is not limited to estimating the characteristic values of the integrated channels. The method of the present invention can also be used to estimate the characteristic values of the individual channels. In another embodiment of the present invention, Eqs is combined with 13, 14, and 20, you can get better 20 200412092 signal integrity and lower estimation error. Because Eqs · 13 and 14 are received signals after repeatedly transmitting the same training symbols, combining these two equations can be considered as further Signal averaging. Therefore, by using repeated transmission of training symbols, the benefits of signal-to-noise ratio improvement can be obtained. The individual channel characteristic values can be calculated according to the following equation: W + r2 + 2r4) r = | outside + (Zi + Z2 + 2ζ4) · X * [Eq. 30], and H _ (^ + y2 + 2y4) -x (z, + z, + 2z, Vx " 4 ~ 4 or, equivalently expressed as: [Eq · 31], magic +2 [⑽. 录) hlL ^ 4 [Eq. 32], so it can be estimated from: Bu Ying) ι + 明) + 邛 _ • Male), [Eq. 33] 4? * Moreover, different from the calculation of Eqs · 24, 25, 28, and 29, the estimation error caused by the noise X member in Ecl · 32 is (2 + 2 + 2 / ) // 4. So, the average value of the estimation error is 0 (Han (say 2/4) ^ 74) = 0), and the variation value of the estimation error is 3 $ / 8 21 200412092 (var ((// ^ + Z + 2Z) // 4) = var ((^ I + 石 +2 石) / 4) = 3σ] / 8, which assumes that the variation value of /, stone, and Z is σ). It can be seen from Eq. 32 As a result, the variation of the estimation error is reduced, and the estimation accuracy is improved. Similarly, the characteristic value of the second channel can also be obtained by the following equation: [Eq. 34], In the above formula, the average of the estimation error The value is 〇 (Wind (Fen Mo-2Z4) J / 4 > 〇) and the variance of the estimated difference is W / 8 (var ((^ + ^ 2-2 ^) // 4) = var ((Z + Z2-2 ^) / 4) = 34/8, where z, stone, and% are assumed to be variability values.) In general, by transmitting additional long training symbol fires, or complex conjugates of the symbols Symbol / and negative complex conjugate symbol— /, miscalculation The variation value can be further reduced for as much as the above-described transmission diversity system, a further embodiment of the present invention may be one as method 'channel characteristic estimated value. The fifth and sixth figures show embodiments of this method. The fifth figure is a flowchart showing the implementation steps of this method. It operates on one or two transmission diversity orthogonal frequency division multiplexing systems. According to the 臓 5Ghz standard, the signals transmitted in the fifth picture all include the guard time interval. As shown in the fifth figure, during a first period, a training symbol is transmitted on the first and second channels (step ·). In the embodiment, the wireless money 470 includes a first channel transmission unit (transmitQgic) 555 and a second channel 22 200412092 a transmission unit 565 is transmitting information to the first channel and the second channel. Immediately after training symbols are transmitted in the first period (step 52G), in the second period, a total of ... symbols of the symbols are transmitted in the-channel (step ca). At the same time, in the second period, the -negative complex co-reversion of the symbol is transmitted to the second channel (step ⑽). If the channel estimation method is based on the dirty 5Ghz standard, the first period is a long training period

一’是實體層收斂程序之檔頭攔位的時框,而第二時段是後續資料時 段之一時段。 第六圖是一流程圖,繪示一方法,估測通道特徵值,實施於接收器 4〇5。如第六圖所示,一接收器接收符元(步驟咖)。當該符元接收完 成後’個別通道效應可由該接收符元中計算得之(步驟63〇)。此個別 通道效應可用以估測個別通道之特徵值(步驟640)。在-實施例中, 该接收器405包括接收單元625、個別通道效應之計算單元635和估測 皁元645,分別用於第六圖中接收(步驟62〇)、個別通道效應計算(步 驟630)和估測(步驟64〇)等步驟。又,在一範例實施例中,該接收 符70是第五圖中傳輸訊號之類比訊號。又,對於一 /7支傳輸分集式 branch)系統,該接收器405於步驟620接收/7個符元,其中每 個符元都是相同訓練符元經過不同通道效應轉換後之訊號,形成一 /7 個77兀一次聯立方程組之系統。參考Eqs. 12至34的方程式,該;7個 23 200412092 通道特徵值就可個別計算得之。 由第五圖和第六圖巾可看出,本方法之實_可獲得更精確估測的 個別通道職值,而非僅是估職合通道特徵值,或必彡㈣通道特徵 值做額外假設。 雖然上文已提及本發明之示範實施例,習此藝者應瞭解,本發明之 上述實施例可㈣做改變及更動。例如:為方便說明之故_二支傳 輸分集式系驗祕上文巾,然而f此藝者應瞭解,上文提及之方法 可擴及多支傳輸分集式系統(三、四或更多支傳輪 > 此外,雖然第四 圖緣不-無線裝置47G之天線,習此#者應瞭解,該傳輸器可以是一 無線區域祕接取點單元、—無線區域網路卡、—蜂巢式電話、一無 線個人數位助理或是其他可收送資料之類似無線裝置。又,雖然在本 發明之-實施例中’-額外的訓練槪傳送於特㈣段…習此藝者 應瞭解,該額外的訓練符元也可傳送於後續資料時段中任—時段。又, 雖然第三A圖和第u圖册示—額外的崎符元(如‘該長訓練符 元的複數共婦元或是負值複數共_元),f此藝者應瞭解,更多額 外的訓練符元也可應用於此系統,藉提 * 捉幵通道估測時的訊號雜訊 比,以及用於計算多支傳輸錢式系統( ^ ^一支傳輸)的個別通道 特微值。又,雖然本發明之數個實施例都依循ffiEE咖標準,習此 24 200412092 藝者應瞭解’本發明之方法㈣财獻任意正交分㈣^充之傳 輸環境。又,雖細E 5Ghz標準用以敘述本發明之部分概念内容, 習此藝者膽解’本發H細村也義於ffiEE 2條標準 (IEEE觀.llg)或是其他類似無線標準,不管是採用何種運作頻帶, 舉凡此等改變或更動都應是在本發明之專利範圍内。 【圖式簡單說明】 為讓本文說明更容純解,特舉下觸示,並搭配簡單說明。 第- A圖和第- B圖是封包訊框圖,綠示職標準之展現協 定資料單元(P廳ntation P购c〇1 Data _,麵)之訊框架 第二圖是-架構圖’繪示_二支傳輸分集式正交分頻多工系統,運 作方式依循IEEE 5Ghz標準。 第三A圖和第三B圖繪示本發明之一系統之一實施例,用以估測通 道特徵值。 第四圖是-架翻,依據本發明之—實施例所繪示之一通道特徵值 估測系統,用於一二支傳輸分集式正交分頻多工系統。 第五圖是-流程圖,緣示本發明之一方法之一實施例,應用於一二 支傳輸分集式正父分齡統之-傳輸H端,用以估測通道特徵值。 25 200412092 第六圖是一流程圖,繪示本發明之一方法之一實施例,應用於一二 支傳輸分集式正交分頻多工系統中之一接收器端,用以估測通道特徵 值。 【圖式標記說明】 110 短訓練時段 120 長訓練時段 130 信號時段 140 資料時段 142 資料時段 144 資料時段 150 第一長訓練符元 155 第一長訓練符元/ 160 第二長訓練符元 165 第二長訓練符元yf 170 信號資料S 180 傳送資料Θ 190 傳送資料及 155a 長訓練符元yT 155b 長訓練符元J 165a 長訓練符元/ 165b 長訓練符元J 170a 信號資料夕 170b 信號資料5 180a 傳送資料/¾ 180b 傳送資料Λ 190a 傳送資料/¾ 190b 傳送資料及 205 接收器 260 第一傳輸器 265 第二傳輸器 26 200412092 310 短訓練時段 320 長訓練時段 330 信號時段 340 資料時段 342 資料時段 344 資料時段 355a長訓練符元J 355b長訓練符元J 365a長訓練符元J 365b長訓練符元J 370a信號資料51 370b信號資料 385a長訓練符元/之一複數共輛符元/ 385b長訓練符元J之一負值複數共軛符元-/ 390a傳送資料及 390b傳送資料/¾ 405 接收器 460 第一傳輸器 465 第二傳輸器 470 無線裝置 555 第一通道傳輸單元 565 第二通道傳輸單元 625 接收單元 635 個別通道效應之計算單元 645 估測單元 27One is the time frame of the header block of the convergence process at the physical layer, and the second period is one of the subsequent data periods. The sixth figure is a flow chart showing a method for estimating channel characteristic values, which is implemented in the receiver 405. As shown in the sixth figure, a receiver receives symbols (step C). When the symbol is received, the individual channel effect can be calculated from the received symbol (step 63). This individual channel effect can be used to estimate the characteristic value of the individual channel (step 640). In the embodiment, the receiver 405 includes a receiving unit 625, an individual channel effect calculation unit 635, and an estimated saponin 645, which are respectively used for receiving in the sixth figure (step 62) and the individual channel effect calculation (step 630). ) And estimation (step 64). Furthermore, in an exemplary embodiment, the receiver 70 is an analog signal such as the transmission signal in the fifth figure. In addition, for a 7-branch diversity branch system, the receiver 405 receives / 7 symbols in step 620, where each symbol is a signal after the same training symbols are transformed by different channel effects to form a / 7 System of 77 simultaneous simultaneous equations. With reference to the equations of Eqs. 12 to 34, this; 7 23 200412092 channel characteristic values can be calculated individually. As can be seen from the fifth chart and the sixth chart, the practicality of this method can obtain more accurate estimates of the individual channel job values, instead of just estimating the combined channel feature values, or the required channel feature values as additional Assumptions. Although the exemplary embodiments of the present invention have been mentioned above, those skilled in the art should understand that the above-mentioned embodiments of the present invention can be changed and modified. For example, for the sake of explanation, the two-transmit diversity system is a secret check, but the artist should understand that the method mentioned above can be extended to multi-transmit diversity systems (three, four, or more). Support wheel> In addition, although the fourth picture is not the antenna of the wireless device 47G, those who are familiar with it should know that the transmitter can be a wireless area secret access point unit,-wireless LAN card,-honeycomb Phone, a wireless personal digital assistant, or other similar wireless device that can send and receive data. Also, although in the embodiment of the present invention-additional training is transmitted to the special section ... This additional training symbol can also be transmitted in any subsequent period of the data period. Also, although the third A and u albums show the additional saki symbol (such as 'the complex number of this long training symbol') Or negative complex numbers total _ yuan), the artist should understand that more additional training symbols can also be applied to this system, by mentioning the signal-to-noise ratio in the * capture channel estimation, and used for calculation Individual channel characteristics of a multi-transmission money system (^^ one transmission) Also, although several embodiments of the present invention follow the EEEE standard, those skilled in the art should understand that the method of the present invention can provide arbitrary orthogonal distributions and sufficient transmission environment. Also, although the fine E 5Ghz The standard is used to describe part of the concept of the present invention. Learners of this art will be savvy. 'Benfa H Seimura also means ffiEE 2 standards (IEEE concept.llg) or other similar wireless standards, regardless of the operating frequency band. All such changes or changes should be within the scope of the patent of the present invention. [Simplified description of the drawings] In order to make the description of this article more pure, the following touches are provided with simple descriptions. Figure-A and Figure- Figure B is a block diagram of the packet message, the message framework of the green agreement standard presentation protocol data unit (P Hall ntation P purchase c〇1 Data _, surface). The second figure is-the architecture diagram 'drawing _ two transmission diversity The orthogonal frequency division multiplexing system operates in accordance with the IEEE 5Ghz standard. Figures 3A and 3B show an embodiment of a system of the present invention for estimating channel characteristic values. The fourth figure is a rack According to the embodiment of the present invention, a channel characteristic value is shown. The measurement system is used for one or two transmission diversity orthogonal frequency division multiplexing systems. The fifth figure is a flowchart showing an embodiment of a method of the present invention, which is applied to one or two transmission diversity positive parent division Age-of-transmission H-end, used to estimate channel characteristic values. 25 200412092 The sixth diagram is a flowchart showing one embodiment of a method of the present invention, which is applied to one or two transmission diversity orthogonal frequency divisions. One of the receivers in the multiplexing system is used to estimate the characteristic value of the channel. [Schematic mark description] 110 Short training period 120 Long training period 130 Signal period 140 Data period 142 Data period 144 Data period 150 First long training symbol Yuan 155 first long training symbol / 160 second long training symbol 165 second long training symbol yf 170 signal data S 180 transmission data Θ 190 transmission data and 155a long training symbol yT 155b long training symbol J 165a long Training symbols / 165b Long training symbols J 170a Signal data Xi 170b Signal data 5 180a Transmission data / ¾ 180b Transmission data Λ 190a Transmission data / ¾ 190b Transmission data and 205 Receiver 260 First transmission 265 second transmitter 26 200412092 310 short training period 320 long training period 330 signal period 340 data period 342 data period 344 data period 355a long training symbol J 355b long training symbol J 365a long training symbol J 365b long training symbol Yuan J 370a signal data 51 370b signal data 385a long training symbol / one complex number of car symbols / 385b long training symbol J one of negative value complex conjugate symbol- / 390a transmission data and 390b transmission data / ¾ 405 Receiver 460 First transmitter 465 Second transmitter 470 Wireless device 555 First channel transmission unit 565 Second channel transmission unit 625 Receiving unit 635 Calculation unit for individual channel effects 645 Estimation unit 27

Claims (1)

zuuHizuyz 拾、申請專利範圍: •一估測通道特徵值之方法, 多工無線通訊系統,其步驟包括: 應用於一二支傳輪分集式正交分頻 〇丨練顿,倾,—第„而該長 網zuuHizuyz's patent application scope: • A method for estimating channel characteristic values, a multiplexed wireless communication system, the steps of which include: Application to one or two wheel diversity diversity orthogonal frequency divisions. And the long net 日,&的規格相符於美„子電駐程師學會所制訂的『無線區域 之媒介接取控制層和實體層標準』; 道; λ長礼束時段内’傳輸該長觀符元於—第二通道; 在貝枓時段内,傳輸該長訓練符元之一複數共婦元於該第 在該資料時段内 一通道; ’傳輸#長訓練符元之—負值複數共締元於該第 妓麵該長訓練符元、以及該長.符元之複數共輛符元和 /、車厄付元;及 負值複數On the day, the specifications of & are in line with the "Medium Access Control Layer and Physical Layer Standards of the Wireless Area" formulated by the Institute of Electrical and Electronics Engineers of the United States; Road; — The second channel; during the Behr period, a complex number of the long training symbols is transmitted in the first data period; 'transmit #long training symbols—the negative complex numbers are associated with The first protagonist faces the long training symbol and the plural of the long symbol. There are a total of symbols and / or a car, and a negative complex number. 由所純到的符元'該㈣練符元之複婦_元 長蛛練符元之負值«共_元,經過運算,獲得通道特徵值。 2·—估術道特徵值之方法,其步驟包括: 在第一時段内,傳送1練符元於-第-通道; 在該第—時如’軸·符擔—第二通道; 28 200412092 在一第二時段内,傳送該訓練符元之一複數共扼符元於該第一通 道;及 在4第二時段内,傳送該訓練符元之一負值複數共幸厄符元於該第 —通道。 3·如申請專利範圍第2項所述之方法,其中在第一時段傳送訓 練符元於第一通道之步驟,包括: 在—實體層收斂程序封包之前置檔頭攔位中之長訓練時段内,傳 送该訓練符元於該第一通道。 4.如申請專利範圍第3項所述之方法,更包括: 在該實體綠練序封包之前置檔·財之訓練時段 内’再次傳送該訓練符元於該第一通道。 練符 >.如申請專職㈣2賴述之枝,其巾在第—時段傳送訓 几於第二通道之步驟,包括 、、/―實體層收絲序封包之前頭攔位中之長訓練時段内, 送该勤丨練符元於該第二通道。 更包括: 6·如申請專利範圍第5項所述之方法, 内 在該實體層收斂程序封包之前置槽 , 罝怕碩攔位中之該長訓練時段 ’再次傳送該訓練符元於該第二通道。 29 200412092 7.如_請專利範圍第2項所述之方法, /、T在弟二時段傳送該 礼東符元之複數共婦元於第―通道之步驟,包括: 在一實體層收斂程序封包之一資料時 奴内,傳送該訓練符元之複 數共軛符元於該第一通道。 8. 如申請專利範圍第2項所述 = 友具中在弟二時段傳送該 训練符元之負賴數共婦元於第二通道之步驟,包括·· 在一實體層收斂程序封包之一资袓 匕之貝枓時段内,傳送該訓練符元之該 負值複數共軛符元於該第二通道。 9. 如申請專利範圍第2項所述之方法,更包括: 在其他時段内,傳送額外的訓練符元於該第一通道。 •々中°H專她15第9項所述之方法,其中該其他時段之數 目正比於傳輸通道之數目。 11·如申請專利範圍第2項所述之方法,更包括: 在其他日成内’傳铜外的訓練符元於該第二通道。 12. -估測通道特徵值之方法,包括: 接收-第-符元,而該第一符元包括: 口卞東符70在一第—時段内傳送於一第一通道,及 該訓練符元,在分每 w弟一日寸段内傳送於一第二通道; 30 200412092 接收一第二符元, 該訓練符元之一 一通道,及 而該第二符元包括: 複數共軛符元,在一第二時段内傳From the pure symbol, the compound character of the trained symbol element _yuan, the negative value of the long spider symbol element «total_element, is calculated to obtain the channel characteristic value. 2 · —The method of estimating the characteristic value of the art path, the steps include: transmitting 1 symbol element in the first period in the first channel; in the first period, such as' axis · fudan—the second channel; 28 200412092 Transmitting a complex conjugate symbol of the training symbol to the first channel in a second period; and transmitting a negative complex symbol of the training symbol to the first channel in the second period; The first channel. 3. The method as described in item 2 of the scope of patent application, wherein the step of transmitting training symbols on the first channel in the first period includes:-Long training in the header block before the entity layer convergence program packet During the period, the training symbol is transmitted to the first channel. 4. The method as described in item 3 of the scope of patent application, further comprising: within the training period of archiving and wealth before the entity's green training sequence packet ', transmitting the training symbol to the first channel again. Exercises> If you apply for a full-time job, the second step is to transfer the training to the second channel in the first period, including the long training period in the first block before the physical layer receives the sequence packet. Within, send the training symbol to the second channel. The method further includes: 6. The method described in item 5 of the scope of patent application, which is slotted before the physical layer convergence program packet, fearing that the long training period in the master block 'retransmits the training symbol in the first Two channels. 29 200412092 7. As described in the method of item 2 of the patent scope, /, T The step of transmitting the plural of the Li Dong Fu Yuan in the second channel during the second period, including: Convergence procedure at a physical layer When a packet of data is contained in the slave, the complex conjugate symbol of the training symbol is transmitted to the first channel. 8. As described in item 2 of the scope of the patent application = The step of transmitting the negative number of the training symbol in the second channel by Youguzhong during the second period, including ... During a time period of one asset, the negative complex conjugate symbol of the training symbol is transmitted to the second channel. 9. The method according to item 2 of the patent application scope, further comprising: transmitting additional training symbols to the first channel during other periods. • The method described in item 9 of Langzhong ° H15, wherein the number of other periods is proportional to the number of transmission channels. 11. The method as described in item 2 of the scope of patent application, further comprising: transmitting training symbols outside the copper channel to the second channel in other Japanese products. 12.-A method for estimating a channel characteristic value, including: receiving a -th-symbol, and the first symbol includes: the mouth 70 is transmitted to a first channel within a first-period, and the training symbol Yuan, transmitted on a second channel within one inch of each minute; 30 200412092 receives a second symbol, one channel of the training symbol, and the second symbol includes: a complex conjugate symbol Yuan, pass in a second period 傳送於 該訓練符元之—負值複數共輛符元,在該第—時段内 該第二通道;及 利用該接收之第-符元和第二符元,估測通道特徵值 13. 道特徵 如申請專利範圍第12項所述之方法,其中估剩通 值之步驟包括: 由該第-符元和第二符以’分離出第—通道效應;及 由該第-符元和第二符元中’分離出第二通道效應。 14.如申請專利範圍第13項所述之方法,其中更包括: 由該分離出來的第-通道效應,估測一第一通道之特徵值;及 由該分離出來的第二通道效應,估測一第二通道之特徵值。 15· —估測通道特徵值之系統,包括: 在-第-時段内,傳送1練符元於一第一通道的傳輸單元; 在該第-時段内,傳送該訓練符元於一第二通道的傳輸單元; 在-第二時段内’傳送該訓練符元之一複數共軏符元於該第一通 道的傳輸單元;及 31 200412092 _值複數共軛符元於該第 在該第二時段内,傳送該訓練符元之一 •通道的傳輪單元。 16. 如申請專利範圍第15項所述之系純,其中於第二時段内 傳送該訓練符元之複數共輛符元於第—通道崎傳輸單元包括: 傳送該訓練符元之複 在一實體層收斂程序封包之一資料時段内 數共輛付元於該第一通道的傳輪單元。 17. 如申請專利範圍第15項所述之系 疵,其中於第二時段内 •通道的該傳輪單元包 傳迗該訓練符元之負值複數共輛符元於第二; 括: 旻數/、輕付70於該第二通道的傳輪單元。 18.如申請專利範圍第15項所述之系統,更包括· Γ時助’傳物卜辑—第—峨傳輪單元。 .一估測通道特徵值之系統,包括: 接收-第-符㈣_元,㈣—符元包括: 在日_送於卞通道;及 •丨練符元’在該第一時段傳送於—第二通道; 接收一第二符元的接收單元^㈣& 32 200412092 該傳送符元之一複數共軛符元,在一第二時段内傳送於該第一通 道;及 該傳送符元之一負值複數共軛符元,在該第二時段内傳送於該第 二通道;及 利用該接收之第一符元和第二符元,估測該第一與第二通道特徵 值的接收單元。 20. 如申請專利範圍第19項所述之系統,其中該估測通道特 徵值之接收單元,包括: 由該第一符元和第二符元,分離出第一通道效應的接收單元;及 由該第一符元和第二符元,分離出第二通道效應的接收單元。 21. 如申請專利範圍第20項所述之系統,更包括: 由該分離出來的第一通道效應,估測一第一通道特徵值的接收單 元;及 由該分離出來的第二通道效應,估測一第二通道特徵值的接收單 元0 33Transmitted to the training symbol—a negative complex number of symbols, the second channel in the first period; and using the received first and second symbols to estimate the channel characteristic value 13. Road The method is characterized in the method described in item 12 of the scope of patent application, wherein the step of estimating the pass value includes: separating the first channel effect from the first symbol and the second symbol by '; and the second channel symbol and the second symbol The second channel effect is separated in the two symbols. 14. The method according to item 13 of the patent application scope, further comprising: estimating the characteristic value of a first channel from the separated first-channel effect; and estimating the second channel effect from the separated, estimating Measure a characteristic value of the second channel. 15 · —A system for estimating channel characteristic values, including: transmitting a training symbol to a first channel during a period of the first period; transmitting the training symbol to a second channel during the period of the first period; A transmission unit of the channel; a transmission unit that transmits a complex conjugate symbol of one of the training symbols to the first channel in the second period; and 31 200412092 _value a complex conjugate symbol at the first in the second During the period, one of the training symbols is transmitted as the channel's pass unit. 16. As described in item 15 of the scope of patent application, wherein the transmission of the plurality of training symbols in the second period to the total number of symbols in the first channel-channel transmission unit includes: transmitting the plurality of training symbols in one During the data period of the physical layer convergence program packet, a total of vehicles are paid to the transfer unit of the first channel. 17. The defect as described in item 15 of the scope of patent application, wherein during the second period, the transmission unit of the channel transmits a negative plural number of the training symbols to the second symbol; including: : Count / pay lightly to the wheel transfer unit on the second channel. 18. The system described in item 15 of the scope of patent application, further comprising: Γ 时 助 ’Biography Series-No.-E Chuanlun unit. A system for estimating channel characteristic values, including: receive-the-symbol __ yuan, ㈣-symbols include: on the day _ sent to the 卞 channel; and • 丨 practice symbol 'is transmitted in- A second channel; a receiving unit receiving a second symbol ^ ㈣ & 32 200412092 a complex conjugate symbol of one of the transmission symbols transmitted to the first channel within a second period; and one of the transmission symbols A negative complex conjugate symbol is transmitted to the second channel in the second period; and the receiving unit for estimating the characteristic value of the first and second channels is received using the received first symbol and second symbol. . 20. The system according to item 19 of the scope of patent application, wherein the receiving unit for estimating the channel characteristic value comprises: a receiving unit for separating the first channel effect from the first symbol and the second symbol; and A receiving unit of the second channel effect is separated from the first symbol and the second symbol. 21. The system according to item 20 of the scope of patent application, further comprising: a receiving unit for estimating a characteristic value of a first channel from the separated first channel effect; and a separated second channel effect from the separated first channel effect, Receiving unit estimating a second channel characteristic value 0 33
TW92122187A 2002-12-03 2003-08-13 Channel estimation in orthogonal frequency-division multiplexing (OFDM) systems TWI239179B (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US10/308,513 US7272108B2 (en) 2002-08-01 2002-12-03 Channel estimation in orthogonal frequency-division multiplexing (OFDM) systems

Publications (2)

Publication Number Publication Date
TW200412092A true TW200412092A (en) 2004-07-01
TWI239179B TWI239179B (en) 2005-09-01

Family

ID=37001220

Family Applications (1)

Application Number Title Priority Date Filing Date
TW92122187A TWI239179B (en) 2002-12-03 2003-08-13 Channel estimation in orthogonal frequency-division multiplexing (OFDM) systems

Country Status (1)

Country Link
TW (1) TWI239179B (en)

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9198084B2 (en) 2006-05-26 2015-11-24 Qualcomm Incorporated Wireless architecture for a traditional wire-based protocol
US8667144B2 (en) 2007-07-25 2014-03-04 Qualcomm Incorporated Wireless architecture for traditional wire based protocol
US8811294B2 (en) 2008-04-04 2014-08-19 Qualcomm Incorporated Apparatus and methods for establishing client-host associations within a wireless network
US9398089B2 (en) 2008-12-11 2016-07-19 Qualcomm Incorporated Dynamic resource sharing among multiple wireless devices
US8102849B2 (en) 2009-02-12 2012-01-24 Qualcomm, Incorporated Association procedure to enable multiple multicast streams
US9264248B2 (en) 2009-07-02 2016-02-16 Qualcomm Incorporated System and method for avoiding and resolving conflicts in a wireless mobile display digital interface multicast environment
US9582238B2 (en) 2009-12-14 2017-02-28 Qualcomm Incorporated Decomposed multi-stream (DMS) techniques for video display systems
US9065876B2 (en) 2011-01-21 2015-06-23 Qualcomm Incorporated User input back channel from a wireless sink device to a wireless source device for multi-touch gesture wireless displays
US8964783B2 (en) 2011-01-21 2015-02-24 Qualcomm Incorporated User input back channel for wireless displays
US9413803B2 (en) 2011-01-21 2016-08-09 Qualcomm Incorporated User input back channel for wireless displays
US9787725B2 (en) 2011-01-21 2017-10-10 Qualcomm Incorporated User input back channel for wireless displays
US10135900B2 (en) 2011-01-21 2018-11-20 Qualcomm Incorporated User input back channel for wireless displays
US20130013318A1 (en) 2011-01-21 2013-01-10 Qualcomm Incorporated User input back channel for wireless displays
US10108386B2 (en) 2011-02-04 2018-10-23 Qualcomm Incorporated Content provisioning for wireless back channel
US9503771B2 (en) 2011-02-04 2016-11-22 Qualcomm Incorporated Low latency wireless display for graphics
US8674957B2 (en) 2011-02-04 2014-03-18 Qualcomm Incorporated User input device for wireless back channel
US9525998B2 (en) 2012-01-06 2016-12-20 Qualcomm Incorporated Wireless display with multiscreen service

Also Published As

Publication number Publication date
TWI239179B (en) 2005-09-01

Similar Documents

Publication Publication Date Title
TW200412092A (en) Channel estimation in orthogonal frequency-division multiplexing (OFDM) systems
CN103152312B (en) A kind of optical OFDM system clock synchronization system of power adjustable and method
CN101888351B (en) Channel estimating apparatus and method thereof
CN103873397B (en) A kind of new joint time domain and frequency domain OFDM receive channel estimation methods
CN104604325A (en) System and method for a collaborative service set
TWI333759B (en) Apparatus and method for inter-carrier interference self-cancellation and inter-carrier interference reconstruction and cancellation
CN102404268B (en) Method for estimating and compensating doppler frequency offset in Rician channels in high-speed mobile environment
CN106789813A (en) The generation method of leading symbol
CN106534002B (en) A kind of compressed sensing based power line channel estimation method
CN101783781A (en) Information transmission method for lowering peak to average power ratio of OFDM system signal
CN102299894A (en) Superimposed-periodic-sequence-based channel estimation method for asymmetrically clipped optical orthogonal frequency division multiplexing (ACO-OFDM) wireless optical communication system
CN1893409B (en) Method for eliminating interaction of pseudo-random series fill in OFDM modulation system
CN105187352B (en) A kind of integer frequency bias method of estimation leading based on OFDM
CN100521665C (en) Iterative decomposition method for fixed training sequence stuffing modulation system
CN106059979A (en) Carrier synchronization method in UFMC system
Athreya et al. Beyond 5G: Leveraging cell free TDD massive MIMO using cascaded deep learning
CN102769599B (en) A kind of novel signal of orthogonal frequency division multiplexing system processing method and device
Shi et al. Blind OFDM systems parameters estimation for software defined radio
CN107171988A (en) OMP condition of sparse channel methods of estimation based on compressed sensing in visible light communication
CN103428154B (en) The transform domain multiplexing method of the dual-selection channel based on Vector OFDM
CN104253772B (en) The channel estimation methods of ofdm system
CN106534030A (en) Channel estimation method based on joint training sequence and pilot in 802.11n multi-antenna OFDM system
CN115941405A (en) SNR estimation method and device of 5G small base station system based on SRS
CN101729479A (en) Blind channel estimation method based on cyclostationarity of OFDM signals
CN108834435A (en) Centralized multinode flow control for the more connectivity of 5G

Legal Events

Date Code Title Description
MM4A Annulment or lapse of patent due to non-payment of fees