[go: up one dir, main page]

TWI379561B - A method to adjust ofdm-symbol window functions and a windowing apparatus to use this method - Google Patents

A method to adjust ofdm-symbol window functions and a windowing apparatus to use this method Download PDF

Info

Publication number
TWI379561B
TWI379561B TW96146574A TW96146574A TWI379561B TW I379561 B TWI379561 B TW I379561B TW 96146574 A TW96146574 A TW 96146574A TW 96146574 A TW96146574 A TW 96146574A TW I379561 B TWI379561 B TW I379561B
Authority
TW
Taiwan
Prior art keywords
unit
frame
coefficient
factor
boxing
Prior art date
Application number
TW96146574A
Other languages
Chinese (zh)
Other versions
TW200926706A (en
Inventor
Yung Szu Tu
Original Assignee
Ite Tech 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
Application filed by Ite Tech Inc filed Critical Ite Tech Inc
Priority to TW96146574A priority Critical patent/TWI379561B/en
Publication of TW200926706A publication Critical patent/TW200926706A/en
Application granted granted Critical
Publication of TWI379561B publication Critical patent/TWI379561B/en

Links

Landscapes

  • Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
  • Noise Elimination (AREA)

Description

AFA-P06001 21493twf.doc/n 九、發明說明: 【發明所屬之技術領域】 本發明是有關於一種框化技術,且特別是有關於一 種調整框化函數的方法與裝置。 【先前技術】 近幾年來由於資訊科技的發達,通訊技術也迅速的 發展起來,各種通訊技術中尤其是以運用了多個子載波 (sub-carrier)平行傳輸的正交分頻多 工(Orthogonal Frequency-Division Multiplexing, OFDM)技術最被看好, 並且在每個子載波維持正交性(orthogonality)的情況下, 多個子載波的頻譜是容許重疊的,又能夠避免載波間的干 擾。 在大部分的應用中,每個符元(symbol)前面會插入一 個保s蒦區間’來防止〇fdM符元間的干擾(Inter-SymbolAFA-P06001 21493 twf.doc/n IX. Description of the Invention: [Technical Field] The present invention relates to a frame technique, and more particularly to a method and apparatus for adjusting a frame function. [Prior Art] In recent years, due to the development of information technology, communication technology has also developed rapidly. In various communication technologies, Orthogonal Frequency is used in particular for parallel transmission using multiple sub-carriers. The -Division Multiplexing (OFDM) technology is most favored, and in the case where orthogonality is maintained for each subcarrier, the spectrum of a plurality of subcarriers is allowed to overlap, and interference between carriers can be avoided. In most applications, a guard interval is inserted in front of each symbol to prevent interference between 〇fdM symbols (Inter-Symbol

Interference,ISI) ’ 並且,保護區間(Guard Interva卜 GI) 内會放入循環前置(cyclic prefix)訊號,也就是,重覆一段 原始符元内的信號。 在正交分頻多工系統中,接收端的接收信號將會在 時域上以符元長度L被區分出每個〇fdM符元,以進行後 端的信號處理,並還原出原始的資料。而針對加入一個循 環前置的OFDM符元而言,只要是所擷取出信號介於保 護區間加上符元長度(GI+L)之間’任意符元長度L的信 號,皆被視為良好的符元時序(Symb〇itiming)。 1379561 AFA-P06001 21493twf.doc/n 而在多重路徑的通道環境下,傳送端所傳送的信號 可能經由不同的路徑到達接收端,並且由於到達接收端的 時間不同,而發生接收端在不同的時間接收到兩個以上的 相同信號。以圖1為例(圖1繪示為在多重路徑的通道環境 下接收信號的示意圖),接收端的接收信號為在不同的時 間接收到兩個相同的信號(;信號110與信號120)的合成信Interference, ISI) ’ Also, a cyclic prefix signal is placed in the guard interval (GI), that is, the signal in the original symbol is repeated. In the orthogonal frequency division multiplexing system, the received signal at the receiving end will be distinguished from each 〇fdM symbol by the symbol length L in the time domain to perform signal processing at the rear end and restore the original data. For the OFDM symbol added to a loop preamble, as long as the signal of the extracted signal is between the guard interval plus the symbol length (GI+L), the signal of any symbol length L is regarded as good. Symbol timing (Symb〇itiming). 1379561 AFA-P06001 21493twf.doc/n In a multipath path environment, the signal transmitted by the transmitting end may arrive at the receiving end via a different path, and the receiving end receives at different times due to the different time of reaching the receiving end. More than two identical signals. Taking FIG. 1 as an example (FIG. 1 is a schematic diagram of receiving signals in a channel environment of multiple paths), the received signal at the receiving end is a composite of receiving two identical signals (signal 110 and signal 120) at different times. letter

號,並假設通道長度為信號110與信號120接收時相差的 時間Tch。 〇以圖1中的中間OFDM符元為例,只要是在有效擷取 區間Tav内,擷取出符元長度為L的一段信號,就可視為 良好的符兀時序,也就是說,在有效擷取區間内所擷 取出的信號,料會包含任何前後符元内的信號,也就不 會產生符元間的干擾。The number is assumed to be the time Tch at which the signal 110 differs from the signal 120 when it is received. Taking the intermediate OFDM symbol in FIG. 1 as an example, as long as it is within the effective acquisition interval Tav, a segment of the symbol whose length is L is taken as a good symbol timing, that is, effective. Take the signal taken out in the interval, and it will contain any signal in the preceding and succeeding symbols, so there will be no interference between the symbols.

在目前的技術中,接收端常用一框化函數來由接收 k號中擷取料個otDM符元。關丨為例,接收信號 (½號no與㈣120的合成信號)乘上一框化錄13〇,將 3 =取出t間的OFDM符元。而接收端所_的_ 二1了,用來操取信號外,還可以調整接收信號在頻 ^上的頻寬以及降低接收信號在旁辦(sideiQbe)上的能 二’以減少接收信號在通道中所受到的 ⑷嫩fe_,聰),並且當框化函數中 =\=1與下降的斜率越小時,將更能夠降低 接收在通道中的窄頻帶干擾。 …、而’在實際傳輸時,通道的脈衝響應隨時間變 1379561 A^A-POSOOl 2l493twf.doc/nIn the current technology, the receiving end often uses a boxing function to retrieve an otDM symbol from the receiving k number. For example, the received signal (the composite signal of No. 1 and No. 120) is multiplied by a frame of 13 〇, and 3 = OFDM symbols between t are taken. The _ _ 2 of the receiving end is used to manipulate the signal, and the bandwidth of the received signal on the frequency can be adjusted and the energy of the received signal on the sideiQbe can be reduced to reduce the received signal. The (4) tender fe_, Cong) received in the channel, and the smaller the slope of =\=1 and falling in the boxing function, will be able to reduce the narrow-band interference received in the channel. ..., and 'in the actual transmission, the impulse response of the channel changes with time 1379561 A^A-POSOOl 2l493twf.doc/n

無法有效地降低接收信號在通道中的窄頻帶干擾。It is not possible to effectively reduce the narrow band interference of the received signal in the channel.

化,通道長度Tch也隨著時間而改變,因而使得每個符元 能夠掏取的有效擷取區間Tav(如圖1所示)會隨著時間變長 或變短。但是,由於目前接收端皆使用固定的框化函數, 而當通道長度過大,有效擷取區間Tav(如圖1所示)變的過 短時’固定的框化函數將可能會擷取出包含前後符元的信 號,而造成符元間的干擾。或者,當有效操取區間Tav(如 圖1所示)較長時,接收端仍使用固定的框化函數,因此 目前常用的框化函數例如為梯型框化函數、Mw框 化函數(由Stefan H. Muller-Weinfurtner所提出之框化函數) 以及漢寧框化函數(Hanning windw)等等,並於圖2中績示 出上述的框化函數,圖2中橫座標為取樣時間,縱座^為 係數的大小。由圖2可觀察出,上述的框化函數其中^的 取樣時間内係數皆為1,而其左右兩邊皆具有一段滾邊, 二調整接收信號的頻譜,並且其滾邊左右對稱。而當接收 端在使用上述的框化函數時’需要預先儲存上述框化函 數也就疋接收%必需健存每個取樣時間上框化函數的係 數大小。而由於框化函數之中間皆為丨的特性,因此接收 端只需要儲存在每個取樣時間上滾邊内的係數。 然而,接收端所使用的框化函數的滾邊所對應的取 間間隔的數目通常約為⑽,因此,接收端仍然需要 化、許夕§己憶體來儲存約100個滾邊内的係數。並且當 =用,框化函數之滾邊所對應的取樣時間點越多在接收 端也就需要越龐大的記憶體來滾邊内的係數。 1379561 AFA-P06001 21493twf.doc/n 目前在美國公開專利US2004022175、US53575〇2、 歐盟公開專利EP1043875與電機電子工程師協會(压The channel length Tch also changes with time, so that the effective capture interval Tav (shown in Figure 1) that each symbol can take will become longer or shorter over time. However, since the current receiving end uses a fixed boxing function, and when the channel length is too large, the effective capturing interval Tav (as shown in FIG. 1) becomes too short, the 'fixed framed function may be extracted before and after the inclusion. The signal of the symbol, which causes interference between the symbols. Or, when the effective operation interval Tav (as shown in FIG. 1) is long, the receiving end still uses a fixed boxing function, so the currently used boxing functions are, for example, a ladder boxing function and a Mw boxing function (by The boxing function proposed by Stefan H. Muller-Weinfurtner) and the Hanning windw function, etc., and the above-mentioned boxing function is shown in Fig. 2. The abscissa in Fig. 2 is the sampling time, vertical The seat ^ is the size of the coefficient. It can be observed from Fig. 2 that the above-mentioned framed function has a coefficient of 1 in the sampling time, and both sides have a section of the edge, and the spectrum of the received signal is adjusted, and the edge is symmetric. When the receiving end uses the above-mentioned boxing function, it is necessary to store the above-mentioned boxing function in advance, and the receiving % must store the coefficient size of the boxing function at each sampling time. Since the middle of the framing function is 丨, the receiver only needs to store the coefficients in the rim at each sampling time. However, the number of interleaving intervals corresponding to the edge of the framing function used by the receiving end is usually about (10). Therefore, the receiving end still needs to reproduce the coefficients in about 100 rims. And when =, the more sampling time points corresponding to the rolling edge of the boxing function, the larger the memory is needed at the receiving end to the coefficients in the piping. 1379561 AFA-P06001 21493twf.doc/n is currently published in the US US2004022175, US53575〇2, EU public patent EP1043875 and Institute of Electrical and Electronics Engineers (pressure

Institute of Electrical and Electronic Engineers)的通訊會刊 (Transactions on communications)(如下列[l]所述)中,已揭 露各種框化的技術,然而,上述之專利與論文並未提及當 有效擷取區間Tav改變時,如何調整框化函數以預防符元 間的干擾與有效地減緩窄•頻帶干擾,並且,上述之專利與 論文所提出之框化函數在實際應用時,接收端仍需要花費 許多的記憶體來儲存每個取樣時間上之框化函數的係數。 [1]D. S. Muller-Weinfurtner, 'Optimum Nyquist windowing in OFDM receivers, IEEE Trans, on Communications, vol. 49, no. 3, pp. 417-420, 2001 【發明内容】 本發明的目的就是在提供一種調整框化函數的方 法,用以依據通道長度與保護區間的差值,來調整框化函 數之滾邊,以防止符元間干擾,並將低降低窄頻帶干擾。 本發明的再一目的是提供一種框化裝置,利用一互 補運昇,來產生多個滚邊係數,以節省裝置中之記憶體。 本發明提出一種調整框化函數的方法包括:偵測通 道長度,找出通道長度與保護區間的差值;以及利用差 值’調整框化函數之滾邊。 本發明再提出一種框化裝置,利用一框化函數對一 1379561 AFA-P06001 21493twf.doc/n 接收訊號進行框化運算,此框化函數具有一第一側之框化 係數與一第二侧之框化係數,而第一側之框化係數與第二 側之框化係數中各自包括多個第一半邊框化係數與多個第 二半邊框化係數。此框化裝置包括框化係數儲存單元、滾 邊振幅產生單元與框化運算單元。其中,框化係數儲存單 7G儲存第一侧之框化係數中之第一半邊框化係數。滚邊振 幅產生單元藉由第一半邊框化係數作一互補運算,產生第 一側之框化係數中之第二半邊框化係數以及第二側之框化 係數。框化運算單元利用第一側之框化係數中之第二半邊 框化係數以及滾邊振幅產生單元所產生的框化係數,對接 收信號進行框化。 本發明根據通道長度與保護區間的差值,來調整框化 函數之滾邊。當通道長度很長,使得上述差值很小時,便 縮小滾邊長度,以避免擷取到相鄰的符元。當通道長度很 短,使得上述差值很大時,便放大滾邊長度’以有效率地 降低接收信號的窄頻帶干擾。因此,本發明能夠防止符元 間干擾。另外’當本發明應用於多载波系統,例如:正交 分頻多工系統,本發明更能避免子載波間干擾 (Inter-Carrier Interference,ICI)。 為讓本發明之上述和其他目的、特徵和優點能更明 顯易懂’下文特舉較佳實施例,並配合所附圖式,作詳細 呑兄明如下。 【實施方式】 在習知技術中,接收端使用固定的框化函數來對接 收信號進行框化,但在實際應用時,由於通道長度的改 變,固定的框化函數將可能造成符元間的干擾,或是無法 有效地減緩通道中的窄頻帶干擾《並且,接收端需要預先 儲存框化函數來對接收信號進行框化,因而在硬體實現 時接收端必須預留大量b記憶體來儲存框化係數。因 此’本發明提出一種調整框化函數的方法與框化裝置,以 適時地調整框化函數,並且減少接收端儲存框化係數所需 的記憶體。 以下將提出本發明的第一實施例,以說明本發明之 框化裝置。在說明本實施之前,先假設本實施例所使用之 框化函數為梯形框化函數。圖3繪示為本發明實施例之框 化函數。請參照圖3,其橫座標為取樣時間n,其縱座標 為框化係數。在此,定義一滾邊長度,其值為框化函數之 滾邊對應的取樣時間長度,並表示為κ,而在圖3的例子 中’滾邊長度Κ=8。另外,假設欲進行框化之接收信號的 符元長度為L,因此,框化函數的總長度為K+L。 圖3中框化函數之滾邊為—線性函數,並且此框化函 數具有左右_之雖。而圖3巾之框化函數表示如下: 心[«]= (i)The Institute of Electrical and Electronic Engineers' Transactions on Communications (as described in [l] below) has revealed various framed techniques. However, the above patents and papers do not mention effective capture. When the interval Tav changes, how to adjust the framing function to prevent the interference between the symbols and effectively slow down the narrow band interference, and the above-mentioned patents and the proposed framed function in the paper still need to cost a lot at the receiving end. The memory stores the coefficients of the boxing function at each sampling time. [1] DS Muller-Weinfurtner, 'Optimum Nyquist windowing in OFDM receivers, IEEE Trans, on Communications, vol. 49, no. 3, pp. 417-420, 2001 SUMMARY OF THE INVENTION The object of the present invention is to provide an adjustment The method of boxing function is used to adjust the edge of the framed function according to the difference between the channel length and the guard interval to prevent inter-symbol interference and reduce the narrow-band interference. It is still another object of the present invention to provide a framing apparatus that utilizes a complementary lift to create a plurality of shirring coefficients to save memory in the device. The present invention provides a method for adjusting a framed function, including: detecting a channel length, finding a difference between a channel length and a guard interval, and adjusting a edge of the frame function by using a difference value. The present invention further provides a framer device for performing a frame operation on a 13759621 AFA-P06001 21493 twf.doc/n received signal by using a frame function, the framed function having a first side frame factor and a second side The framed coefficient, and the frame factor of the first side and the frame factor of the second side each include a plurality of first half frameization coefficients and a plurality of second half frameization coefficients. The framer includes a frame factor storage unit, a roll amplitude generation unit, and a frame operation unit. The boxing coefficient storage unit 7G stores the first half of the frame factor of the first side frame factor. The edge amplitude generating unit performs a complementary operation by the first half framering coefficient to generate a second half framering coefficient of the frame coding coefficient of the first side and a frame factor of the second side. The framed operation unit frames the received signal by using the second half of the frame factor of the first side frame factor and the frame factor generated by the edge amplitude generating unit. The invention adjusts the edge of the boxing function according to the difference between the channel length and the guard interval. When the length of the channel is so long that the difference is small, the length of the edge is reduced to avoid capturing adjacent symbols. When the length of the channel is short so that the above difference is large, the length of the edge is amplified to effectively reduce the narrow band interference of the received signal. Therefore, the present invention can prevent inter-symbol interference. Further, when the present invention is applied to a multi-carrier system, for example, an orthogonal frequency division multiplexing system, the present invention can more effectively avoid Inter-Carrier Interference (ICI). The above and other objects, features, and advantages of the present invention will become more apparent <RTIgt; </ RTI> <RTIgt; </ RTI> <RTIgt; [Embodiment] In the prior art, the receiving end uses a fixed framed function to frame the received signal, but in actual application, due to the change of the channel length, the fixed framed function may cause inter-symbol Interference, or can not effectively slow down the narrow-band interference in the channel. "And the receiving end needs to store the framed function in advance to frame the received signal. Therefore, when the hardware is implemented, the receiving end must reserve a large amount of b memory to store. Boxing factor. Therefore, the present invention proposes a method and a framer for adjusting a frame function to adjust the frame function in a timely manner and reduce the memory required for the receiving end to store the frame factor. The first embodiment of the present invention will be described below to explain the framed device of the present invention. Before explaining this embodiment, it is assumed that the frame function used in this embodiment is a trapezoidal frame function. FIG. 3 illustrates a block function of an embodiment of the present invention. Referring to Figure 3, the abscissa is labeled as the sampling time n, and its ordinate is labeled as the frame factor. Here, a roll length is defined, which is the sampling time length corresponding to the edge of the boxing function and is expressed as κ, and in the example of Fig. 3, the 'roll length Κ=8. Further, assuming that the symbol length of the received signal to be framed is L, the total length of the framing function is K + L. The edge of the boxing function in Figure 3 is a linear function, and this boxing function has left and right _. The boxing function of Figure 3 is as follows: Heart [«] = (i)

C, K&lt;.n&lt;,L RK[K+L-ri\, L + \&lt;n&lt;,L+K~\ 〇&gt; 其他 上述第(l)式中,c為 常數,且其值為1,AM為第《 取樣時T框化係數。為了方便㈣以下實關,在圖3 應,框化係數標*為3(H,w = 2對應的框化係 以此類推’將框化係數標示為 中 數在圖3中標示為.302 AFA-P06001 21493twf.doc/n 301〜314。而在本發明之實施例所提出的框化裝置,只需 儲存4個框化係數的值,就可得到所有的框化係數 301〜314之值,並進行信號的框化。然而,以習知的框化 技術來說,卻必須儲存14個框化係數,才能夠進行框化 之運算。在之後的段落中,將會描述本實施例為何只需儲 存4個框化係數。 另外,為了方便說明本發明實施例,以下預先定義 圖3中每個框化係數301-314之值,其中,左側的框化係 數 301 〜307 之值依序為 0.875、0.75、0.625、0.5、0.375、 0.25、0.125。右側的框化係數308〜314之值依序為 0.875、0.75、0.625、0·5、0.375、0.25、0.125。並且,為 了方便說明本發明實施例,將左側之框化係數3〇1〜307定 義為兩個群組’其中框化係數301〜303定義為左半邊框化 係數’而框化係數304〜307定義為右半邊框化係數;同樣 地’右側之框化係數3〇8〜314分別被定義為左半邊框化係 數308〜310與右半邊框化係數311〜314。 圖4繪示為本發明第一實施例之框化裝置方塊圖。請 參照圖4 ’在此’假設框化裝置400包含於一接收端中。 並且,假設接收端應用於OFDM系統,而來自於通道中 的接收信號包含多個OFDM符元,並在每個符元中包含 一保護區間’且保護區間内為循環前置信號,而未加保護 區間之原始符元長度為L,也就是,原始符元之實際長度 為L個單位取樣時間。另外,接收端將接收信號經過取樣 後’產生取樣信號S[n]輸入至框化裝置400,而框化裝置 1379561 AFA-P06001 21493twf.doc/n 400將對取樣信號s[n]進行框化,以擷取出每個〇1;1)1^符 元。C, K&lt;.n&lt;,L RK[K+L-ri\, L + \&lt;n&lt;,L+K~\ 〇&gt; In the above other formula (l), c is a constant and its value For 1, AM is the "T boxing factor at the time of sampling. In order to facilitate (4) the following, in Figure 3, the boxing factor is *3 (H, w = 2 corresponds to the boxing system, and so on.) The boxing factor is marked as the median number in Figure 3. AFA-P06001 21493 twf.doc/n 301~314. In the framed device proposed by the embodiment of the present invention, only the values of the four framed coefficients are stored, and all the values of the framed coefficients 301 to 314 can be obtained. And the signal is framed. However, in the conventional frame technique, it is necessary to store 14 frame coefficients to be able to perform the frame operation. In the following paragraphs, the description will be given. In order to facilitate the description of the embodiments of the present invention, the values of each of the frame factors 301-314 in FIG. 3 are defined in advance, wherein the values of the frame factors 301 307 to 307 on the left side are sequentially The values of the frame factors 308 to 314 on the right side are 0.875, 0.75, 0.625, 0.55, 0.375, 0.25, and 0.125, respectively. In the embodiment of the invention, the boxing coefficients 3〇1 to 307 on the left side are defined as two groups' The frame factors 301 to 303 are defined as the left half frame factor and the frame coefficients 304 to 307 are defined as the right half frame coefficients; likewise, the frame coefficients 3 〇 8 314 to the right are defined as the left half frame. Figure 4 is a block diagram of a framed device according to a first embodiment of the present invention. Please refer to Figure 4 'herein' assuming that the framer 400 is included in a receiving Furthermore, it is assumed that the receiving end is applied to the OFDM system, and the received signal from the channel contains a plurality of OFDM symbols, and includes a guard interval in each symbol and a loop preamble in the guard interval. The original symbol length of the unprotected interval is L, that is, the actual length of the original symbol is L unit sampling time. In addition, the receiving end samples the received signal and generates a sampling signal S[n] to the frame. The device 400 is assembled, and the framed device 1379561 AFA-P06001 21493 twf.doc/n 400 will frame the sampled signal s[n] to extract each 〇1; 1) 1^ symbol.

圖4中框化裝置400包括框化係數儲存單元41〇、滾 邊振幅產生單元420與框化運算單元43〇β其中,滾邊振 幅產生單元420包括一第一加法單元422與與多工單元 425。而框化運异單元430包括乘法單元432、第二加法單 元434、緩衝單元436與第三加法單元438。而框化係數儲 存單元410僅儲存四個框化係數304〜3〇7。以下便開始說 明框化裝置400如何指利用框化係數儲存單元41〇所儲存 四個框化係數304〜307來對取樣信號s[n]進行框化。 首先,滾邊振幅產生單元420依序由四個框化係數 304〜307取出一特定框化係數〜㈨。其中,與框 化函數 (2)〇 冰H,χΛη] = &lt; wkW,The framing device 400 of FIG. 4 includes a framing coefficient storage unit 41, a roll amplitude generating unit 420, and a framing operation unit 43A, wherein the rim amplitude generating unit 420 includes a first adding unit 422 and a multiplex unit 425. The frame division unit 430 includes a multiplication unit 432, a second addition unit 434, a buffer unit 436, and a third addition unit 438. The frame factor storage unit 410 stores only four frame coefficients 304 to 3〇7. In the following, it is explained how the framer 400 refers to the frame signal s[n] being framed by the four frame coefficients 304 to 307 stored by the frame factor storage unit 41. First, the edge amplitude generating unit 420 sequentially takes out a specific frame factor ~ ~ by a plurality of framing coefficients 304 to 307. Where, with the boxing function (2) 〇 ice H, χΛη] = &lt; wkW,

一 κ 1 τι 一 —l· 1 〜 2One κ 1 τι one —l· 1 〜 2

接著,滾邊振幅產生單元420中的第一加法單元將 〜[«]與一常數C相減後,輸出差值心w = c 一。其 =’,數c為實數’在本實施例中,常數匸例如為卜而 多工單兀425耦接至第一加法單元422與框化係數儲存單 元410,同時接收框化係數儲存單元41〇所輸出的〜㈣與 加法單元422所輸出的X、[η],並且,多工單元425在每 個取樣時間w,依據一選擇信號選擇輸出4[«]或之Μ 以作為第《取樣時間之滾邊振幅係數々问。 在此,假設框化裝置400要先對取樣時間《 = 1的取樣 11 1379561 AFA-P06001 21493twf.doc/n 信號S[l]進行框化’也就是說需要產生框化係數3〇卜並 由上述第(2)可知,〜[1]=〜[7] = 0.125,也就是說,在 W = 1時’滾邊振幅產生單元420將讀取出框化係數3〇7。 並且,由於框化係數301之值為與框化係數3〇7與丨之差 值,因此,多工單元425選擇作為滾邊振幅, 而滾邊振幅z/cW之值為0.875。 在本實施例中,多工單元425所接收的選擇信號可例 如為0或1,當需要產生之滾邊振幅為左半邊框化係數 (301〜303以及308〜310)時,選擇信號將為〇,而多工單元 425將來自第一加法單元422的差值分尺卜]輸出至框化運 异單元430,當需要產生之滾邊振幅為右半邊框化係數 (304〜307以及311〜314)時,選擇信號為丨’而多工單元 425將直接輸出由框化係數儲存單元41〇所讀取出之特定 框化係數至框化運算單元430。然本領域具有通常知識者 應可推知,選擇信號的產生方式可利用一計數器計數多 工單元425所輸出的滾邊振幅數目或取樣時間等等,來決 定選擇信號為0或1。 ' 接下來,框化運算單元430接收取樣時間„ = 1的取樣 信號s[i],並同時輸入於乘法單元432與第二加法單元 434。然後,乘法單元432將取樣信號s⑴與滚邊振幅產 生=元420所輸出之滾邊振幅zJ1]相乘後,輸出乘積 々[fxstiHKwssn]至第二加法單元434。接著,第二加 法單元434將取樣信號S[1]與乘法單元432之輸出 〇.875S[l]相減後,將輸出差值〇1258[1]輸出並儲存至緩 12 1379561 AFA-P06001 21493twf.doc/n 衝單元436。 . 之後,依照同樣的硬體之操作之後,滾邊振幅產生單 元420將依序讀取出框化係數儲存單元41〇内所儲存之框 化係數305與306,並依序產生滾邊振幅ζ〖[2] = 〇·75與 Zj3] = 〇·625 ’再透過框化運算單元430之運算後,將^ 積〇.25S[2]與0.375S[3]儲存至緩衝單元436。 接下來,框化裝置400要對取樣時間n = 4的取樣信號 • S[4]進行運算時,滾邊振幅產生單元420讀取框化係^儲° 存單元410中之框化係數304,其值為〇 5。並由於此時需 要的框化係數304為右半邊框化係數,因此,此時的選^ 信號為1,而多工單元425直接輸出所讀取之特定框化係 數〜[4]至框化運算單元43〇,來作為取樣時間” 邊振幅〜[4]。 &lt; 框化運算單元430接收取樣時間為《 = 4的取樣信號 S[4] ’而。其内部之乘法單元432將取樣信號s⑷與滾邊振 生單元420所輸出之滾邊振幅〜[4]相乘後,輸出 • =[4]xS[4H).剛至第二加法單元434,再經由第二加法 單元,434將取樣&lt;5號S[4]與乘法單元432之輸出〇.5S[4] 相減後,輸出〇.5S[4]並儲存至緩衝單元436。 之後依,¾同樣的硬體操作,滾邊振幅產生單元樣 框丄匕係數儲存單元410内讀取出框化係數 305〜307 ’再透過框化運算單元伽内部 0.7·與咖[7],並健存至緩衝單元伙。 接下來,由於取樣信號S[8]、S[9]、...、S[L]所對應 13 1379561 AFA-P06001 21493twf.doc/n 之框化係數為1 ’因此,取樣信號S[8]、S[9:|、 、S[L:| 將直接輸入至後端處理,並未輸入至框化裝置400。 之後’由於取樣信號S[L+1]〜S[L+7]要經由框化裝置 400之運算’因此,滾邊振幅產生單元420將繼續讀取框 化係數儲存單元410中之框化係數。在此,假設框化裝置 400先對取樣信號s[L+l]進行運算,因而需要產生框化係 數308。由於框化係數308之值相同於框化係數301之 值,且為框化係數307與1之差值。因此,滚邊振幅產生 單元420所讀取之特定框化係數xJZ + 1]為框化係數3〇7, 其值為0.125。 接著’加法單元422計算出特定框化係數々[1 + 1]與 1之差值後’輸出差值4[尤+ 1]至多工單元425,其值為 0.875。由於此時滾邊振幅產生單元420需要產生的框化 係數308為左半邊滾邊振幅,因此,選擇信號為〇 ^之 後,多工單元425將依據選擇信號將差值x’JL + 1]作為滾 邊振幅〜[[+ 1],並輸出至框化運算單元430。 框化運算單元430中之乘法單元432將接收取樣信號 S[L+1] ’並將滾邊振幅zJZ + Ι]與取樣信號S[L+1]相乘 後,將乘積^[1 + 1]父8[1&gt;1]=0.8758[1^1]輸出至第三加法 單元438。 由於框化裝置400所要框化的〇fdM符元中包含循環 前置信號,可推知取樣信號S[l]〜S[7]為取樣信號 S[L+1]〜S[L+7]相同之重複信號。因此,框化裝置4〇〇將 利用取樣信號S[l]〜S[7]與取樣信號s[L+l]〜S[L+7]為相同 14 ί S &gt; 1379561 AFA-P06001 21493twf.doc/n 信號的特性,來還原出原始的信號。 因此,在第二加法單元438接收到乘法單元432所輪 出之0.875S[L+1]後,第三加法單元438將讀取出之前儲 存於緩衝單元436的〇.125S[1],並將〇 8758[]1+1]與 0.125S[1]相加後輸出至後端處理,以讓後端之裝置還原出 原始之信號。 之後,依照同樣的硬體操作,框化裝置4〇〇將利用框 化係數儲存單元410内所儲存之框化係數3〇6與3〇5,依 序產生滾邊振幅〜[^ + 2]與2^[尤+ 3],其值分別為〇75與 0.625,再依序將〇.75S[L+2]加上之前儲存於緩衝單元436 的〇.25S[2]以及〇.625S[L+3]加上之前儲存於緩衝單元436 的〇.375S[3]後輸出至後端處理。 由上述的框化運算單元43〇之操作可知,在取樣時間 « 1 7日^,框化運异單元430中之乘法單元432將其取樣 信號與滾邊振幅之乘積輸出至第二加法單元434。在取樣 時間《 = 1 + 1〜Z + 7時,框化運算單元43〇中之乘法單元 432將其取樣信號與滾邊振幅之乘積輸出至第三加法單元 438。故本領域具有通常知識者應可推知,乘法單元432 還可包括一多工器或切換器等等,來決定取樣信號與滾邊 振幅之乘積輪出至第二加法單元434與第三加法單元438 其中之一。 ^接下來,框化裝置400將要對取樣信號S[L+4]進行運 算,而此時需要產生框化係數311,而框化係數311之值 相同於框化係數之值,因此,滾邊振幅產生單元42〇 15 1379561 AFA-P06001 21493twf.doc/n 將讀取出框化係數304,並作為特定框化係數义[I + 4] • 框化係數307,其值為〇.5。由於此時所需的框化係數^&quot;^ • 為右半邊框化係數,故此時選擇信號為丨,而多工單元Next, the first adding unit in the edge amplitude generating unit 420 subtracts ~[«] from a constant C, and outputs a difference heart w = c1. In the present embodiment, the constant 匸 is, for example, the multiplexer 425 is coupled to the first adding unit 422 and the framing coefficient storage unit 410, and receives the framing coefficient storage unit 41. 〜 output (4) and X, [η] output by the addition unit 422, and the multiplex unit 425 selects the output 4[«] or the 依据 according to a selection signal at each sampling time w as the "sampling" The roll amplitude coefficient of time is asked. Here, it is assumed that the framing device 400 first frames the sampling time "1 1 sample 1 1379561 AFA-P06001 21493 twf.doc/n signal S[l]", that is, it is necessary to generate a frame factor of 3 并As described in the above (2), ~[1]=~[7] = 0.125, that is, when W = 1, the 'rolling edge amplitude generating unit 420 reads out the frame factor 3〇7. Further, since the value of the framing coefficient 301 is different from the framing coefficient 3 〇 7 and 丨, the multiplex unit 425 is selected as the rim amplitude, and the value of the rim amplitude z/cW is 0.875. In this embodiment, the selection signal received by the multiplexing unit 425 can be, for example, 0 or 1. When the edge amplitude to be generated is the left half frame factor (301 to 303 and 308 to 310), the selection signal will be 〇. And the multiplex unit 425 outputs the difference ruler from the first addition unit 422 to the frame division operation unit 430, and the roll edge amplitude to be generated is the right half frameization coefficient (304 to 307 and 311 to 314). At this time, the selection signal is 丨' and the multiplex unit 425 directly outputs the specific framed coefficient read by the framing coefficient storage unit 41 to the framing operation unit 430. However, those skilled in the art should be able to infer that the selection signal can be generated by using a counter to count the number of edge amplitudes or sampling times output by the multiplexing unit 425, etc., to determine whether the selection signal is 0 or 1. Next, the frame operation unit 430 receives the sampling signal s[i] of the sampling time „=1, and simultaneously inputs it to the multiplication unit 432 and the second addition unit 434. Then, the multiplication unit 432 compares the sampling signal s(1) with the edge amplitude. After multiplying the edge amplitude zJ1] outputted by the = element 420, the product 々[fxstiHKwssn] is outputted to the second adding unit 434. Next, the second adding unit 434 outputs the sampling signal S[1] and the output of the multiplying unit 432. After the 875S[l] is subtracted, the output difference 〇1258[1] is output and stored to the buffer 12 1379561 AFA-P06001 21493 twf.doc/n punch unit 436. After that, after the same hardware operation, the edge amplitude The generating unit 420 will sequentially read the framed coefficients 305 and 306 stored in the framed coefficient storage unit 41, and sequentially generate the edge amplitude ζ [[2] = 〇·75 and Zj3] = 〇·625 ' After the operation of the framing operation unit 430, the memory 〇.25S[2] and 0.375S[3] are stored in the buffer unit 436. Next, the framing device 400 is required to sample the sampling time n=4. When S[4] performs an operation, the edge amplitude generating unit 420 reads the framed memory unit 4 The frame factor 304 in 10 has a value of 〇 5. Since the frame factor 304 required at this time is the right half frame coefficient, the selection signal at this time is 1, and the multiplex unit 425 directly outputs the The specific frame factor to read from [4] to the frame operation unit 43A is used as the sampling time "edge amplitude ~ [4]. &lt; The frame operation unit 430 receives the sampling signal S[4] ' whose sampling time is "=4". The internal multiplication unit 432 multiplies the sampling signal s(4) by the edge amplitude ~[4] outputted by the rimming unit 420, and outputs •=[4]xS[4H). It is immediately after the second adding unit 434, and then The second adding unit 434 subtracts the sample &lt;5 S[4] from the output 〇.5S[4] of the multiplying unit 432, and outputs 〇.5S[4] and stores it in the buffer unit 436. Then, according to the same hardware operation, the edge amplitude generating unit sample frame coefficient storage unit 410 reads out the frame factor 305~307' and then passes through the framed operation unit gamma internal 0.7· with the coffee [7], and Save to the buffer unit. Next, since the sampling signal S[8], S[9], ..., S[L] corresponds to 13 1379561 AFA-P06001 21493twf.doc/n, the frame factor is 1 ', therefore, the sampling signal S[8 ], S[9:|, , S[L:| will be directly input to the backend processing, and is not input to the framing device 400. Thereafter, since the sampling signals S[L+1] to S[L+7] are to be operated via the framing device 400, the edge amplitude generating unit 420 will continue to read the framing coefficients in the framing coefficient storage unit 410. Here, it is assumed that the framing device 400 first operates on the sampling signal s[L+l], and thus it is necessary to generate the frame factor 308. Since the value of the framing coefficient 308 is the same as the value of the framing coefficient 301, and is the difference between the framing coefficients 307 and 1. Therefore, the specific frame factor xJZ + 1] read by the edge amplitude generating unit 420 is a frame factor of 3〇7, and its value is 0.125. Next, the adding unit 422 calculates the difference between the specific frame factor 々[1 + 1] and 1 and outputs the difference 4 [especially + 1] to the multiplex unit 425, which has a value of 0.875. Since the framed coefficient 308 that needs to be generated by the edge amplitude generating unit 420 is the left half edge amplitude, after the selection signal is 〇^, the multiplex unit 425 will use the difference value x'JL + 1] as the edge amplitude according to the selection signal. ~[[+ 1], and output to the frame operation unit 430. The multiplication unit 432 in the frame operation unit 430 will receive the sampling signal S[L+1] ' and multiply the edge amplitude zJZ + Ι] by the sampling signal S[L+1], and then multiply the product ^[1 + 1] The parent 8[1&gt;1]=0.8758[1^1] is output to the third addition unit 438. Since the 前fdM symbol to be framed by the framing device 400 includes a cyclic preamble signal, it can be inferred that the sampling signals S[l]~S[7] are the same as the sampling signals S[L+1]~S[L+7]. Repeat signal. Therefore, the framing device 4 〇〇 will use the sampling signals S[l]~S[7] and the sampling signals s[L+l]~S[L+7] to be the same 14 ί S &gt; 1379561 AFA-P06001 21493twf. The characteristics of the doc/n signal to restore the original signal. Therefore, after the second adding unit 438 receives 0.875S [L+1] rotated by the multiplying unit 432, the third adding unit 438 will read out the 〇.125S[1] previously stored in the buffer unit 436, and Add 〇8758[]1+1] to 0.125S[1] and output it to the backend processing to let the device at the back end restore the original signal. Then, according to the same hardware operation, the framing device 4 利用 will use the framing coefficients 3 〇 6 and 3 〇 5 stored in the framing coefficient storage unit 410 to sequentially generate the rim amplitude ~ [^ + 2] and 2^[+3], the values are 〇75 and 0.625, respectively, and then 〇.75S[L+2] is added to 〇.25S[2] and 〇.625S[L] previously stored in buffer unit 436. +3] is added to the backend processing after the 〇.375S[3] previously stored in the buffer unit 436. As is apparent from the above operation of the frame operation unit 43, the multiplication unit 432 in the frame division unit 430 outputs the product of the sample signal and the edge amplitude to the second addition unit 434 at the sampling time «17. At the sampling time " = 1 + 1 to Z + 7, the multiplication unit 432 in the frame operation unit 43A outputs the product of the sampling signal and the edge amplitude to the third addition unit 438. Therefore, those skilled in the art should be able to infer that the multiplication unit 432 may further include a multiplexer or switcher or the like to determine the product of the sampling signal and the edge amplitude to be rotated to the second adding unit 434 and the third adding unit 438. one of them. Next, the framing device 400 is to calculate the sampling signal S[L+4], and at this time, it is necessary to generate the framing coefficient 311, and the value of the framing coefficient 311 is the same as the value of the framing coefficient, therefore, the edge amplitude The generating unit 42〇15 1379561 AFA-P06001 21493twf.doc/n will read out the boxing coefficient 304 and as a specific boxing coefficient meaning [I + 4] • The boxing coefficient 307 has a value of 〇.5. Since the required frame factor ^&quot;^ is the right half frame factor, the selection signal is 丨, and the multiplex unit

425直接將所讀取之〜[1 + 4]作為滾邊振幅〜[1 + 4]轸^ 至框化運算單元430。 S 接著,框化運算單元430接收取樣信號s[L+4],而乘 法單το 432將取樣信號S[L+4]與滾邊振幅產生單元42〇所 φ 輸出之滾邊振幅、[L + 4]相乘後,將乘積 z/f[Z + 4]xS[L+1]= 〇 5S[L+4]輸出至第三加法單元视。 在第三加法單元438接收到乘法單元432所輸出之 0.5S[L+4]後’第三加法單元438將讀取出之前儲存於緩衝 單元436的〇.5S[4] ’並將〇.5S[L+4]與〇.5S[4]相加後輸出 至後端處理,以讓後端之裝置還原出原始之信號。 之後,依照同樣的硬體操作,框化裝置400將利用框 化係數儲存單元410内所儲存之框化係數305、306與 307,依序產生滾邊振幅〜[Z + 5]、々[1 + 6]與々[Z + 7], 鲁其值分別為0.375、0.25與0.125,再依序將〇.375S[L+5] 加上之前儲存於緩衝單元436的0.625S[5]、將0.25S[L+6] 加上之前儲存於緩衝單元436的0.75S6]以及將0.125S[L+7] 加上之前儲存於緩衝單元436的〇.875S[7]後,輸出至後端 處理,以讓後端之裝置還原出原始之信號。 由上述圖4中之框化裝置400的操作說明,可歸納出 一方法步驟,如圖5所示。圖5繪示為本發明第一實施例 之框化方法的步驟流程圖。請同時參照圖4與圖5,首 16 1379561 AFA-P06001 21493twf.doc/n 先’滾邊振幅產生單元420依序讀取框化係數儲存單元 410中之框化係數(步驟S510)。之後,滾邊振幅產生單元 420計算所讀取出之框化係數與i的差值(步驟S52〇)。而 多工單元425利用一選擇信號選擇差值或所讀取出之框化 係數其中之一作為所使用的滾邊振幅,並輸出至框化運算 單元430(步驟S530)。425 directly reads ~[1 + 4] as the edge amplitude ~ [1 + 4] 轸 ^ to the frame operation unit 430. S, the frame operation unit 430 receives the sample signal s[L+4], and the multiplication unit το 432 sets the sample signal S[L+4] and the edge amplitude of the φ output of the edge amplitude generation unit 42 [, [L + 4] After multiplication, the product z/f[Z + 4]xS[L+1]= 〇5S[L+4] is output to the third addition unit. After the third addition unit 438 receives 0.5S[L+4] output by the multiplication unit 432, the third addition unit 438 will read out the 〇.5S[4] ' previously stored in the buffer unit 436 and will 〇. 5S[L+4] is added to 〇.5S[4] and output to the back-end processing to let the device at the back end restore the original signal. Then, according to the same hardware operation, the framing device 400 will sequentially generate the rim amplitudes [[Z + 5], 々 [1 +] by using the framing coefficients 305, 306, and 307 stored in the framing coefficient storage unit 410. 6] and 々[Z + 7], the values are 0.375, 0.25 and 0.125, respectively, and then 〇.375S[L+5] is added to 0.625S[5] previously stored in buffer unit 436, 0.25 S[L+6] is added to 0.75S6] previously stored in buffer unit 436 and 0.125S[L+7] is added to 〇.875S[7] previously stored in buffer unit 436, and output to the back end processing. In order to restore the original signal to the device at the back end. From the above description of the operation of the framing device 400 in Fig. 4, a method step can be summarized, as shown in Fig. 5. FIG. 5 is a flow chart showing the steps of the method of framing according to the first embodiment of the present invention. Referring to FIG. 4 and FIG. 5 simultaneously, the first 16 1379561 AFA-P06001 21493 twf.doc/n first 'roll edge amplitude generating unit 420 sequentially reads the framed coefficients in the framing coefficient storage unit 410 (step S510). Thereafter, the edge amplitude generating unit 420 calculates the difference between the read frame factor and i (step S52A). The multiplex unit 425 selects one of the difference values or the read frame factor as a used edge amplitude by a selection signal, and outputs it to the frame operation unit 430 (step S530).

接下來,框化運算單元430接收取樣信號(步驟 S540)。乘法單元432將取樣信號與來自多工單元425的滾 邊振幅相乘後,輪出一乘積(步驟S55〇)。然後,依據取樣 時間《,乘法單元432決定乘積輸出至第二加法單元434Next, the frame operation unit 430 receives the sampling signal (step S540). The multiplying unit 432 multiplies the sampling signal by the edge amplitude from the multiplex unit 425, and then multiplies a product (step S55). Then, based on the sampling time, the multiplying unit 432 determines the product output to the second adding unit 434.

與第二加法單元438其中之一,也就是判斷此時來自於滾 邊振幅產生單元420的滾邊振幅是對應至左側之框化係數 ,是右側之框化係數(步驟S56〇)。若來自於滾邊振幅產生 單元420的滾邊振幅對應至左侧之框化係數,則乘法單元 432將輸出其乘積至第二加法單元434 ’並經由加法單元 434將所接收之乘積與取樣信號相減後儲存至緩衝單元 436(步驟S565)。絲自於滾邊振幅產生單元42〇的滚邊 振幅對應至右側之框化魏,職法單元432將其 出至第三加法單元438 ’經由加法單元438將所接收之乘 Ϊ(力 436内所對應之乘積後’輪出至後端的裝 以下將提ίϋ本發_第二實❹卜㈣明本發 知例之調麵化的方法與裝置。圖6繪示為本發明第 施例之調整框化裝置方塊圖。請參照圖6,調整框^置 1379561 AFA-P06001 21493twf.doc/n 600包括框化係數儲存單元610、滾邊振幅產生單元620 與框化運算單元630。而框化運算單元630還包括乘法單 元632、緩衝單元636與加法單元638。 在說明本實施之前,先假設本實施例使用圖3中之框 化函數。而為了方便說明本發明實施例,將左側之框化係 數301〜307之值依序定義為0.125、0.25、0.375、0.5、 0.625、0.75與0.875,而右侧之框化係數308〜314之值依 序定義為 0.875、0.75、0.625、0.5、0.375、0.25 與 0.125。並且,假設框化係數儲存單元61〇所儲存圖3中之 框化係數為301〜307 ’而滾邊振幅產生單元620將能夠利 用框化係數301〜307,來調整框化函數之滾邊。 接著,再假設調整框化裝置600包含於一接收端,而 此接收端具有通道估測(channel estimation)模組(未繪示), 而此通道估測模組利用來自通道之接收信號估測出通道脈 衝響應’並由通道脈衝響應計算出通道長度。此外,假設 接收應用於OFDM糸統,而來自於通道_的接收信號 包含多個OFDM符元,並在每個符元中包含一保護區 間’且保護區間内為循環前置信號,而未加保護區間之原 始付元長度為L,也就是’原始符元之實際長度為乙個單 位取樣時間。而通道長度與保護區間的關係例如為圖1所 示。 另外,接收端將接收信號經過取樣後,產生取樣信號 S[n]輸入至調整框化的裝置6〇0,而調整框化的農置6〇〇 將利用由通道估測模組所計算出之通道長度,來對調整框 18 AFA-P06001 21493twf.doc/n 化函數之滚邊,並對取樣信號S[n]進行框化,以正確地擷 取出每個OFDM符元,以及有效地防止通道中的窄頻帶 干擾。 請參照圖3與圖6,開始時’滾邊振幅產生單元620 接收一通道長度,而此通道長度例如由接收端中的通道估 測模組所提供。 在接收到通道長度後,滚邊振幅產生單元620將找出 通道長度與保護區間的差值,此差值例如為保護區間的長 度減去通道長度。而由上述之圖1可觀察出,當保護區間 與通道長度的差值變大時,圖1中之有效擷取區間Tav也 將隨之變長;反之,當保護區間與通道長度的差值變小 時,圖1中之有效操取區間Tav也將隨之變短。 因此’接下來,滾邊振幅產生單元620將依據保護區 間與通道長度的差值,決定一縮減取樣因數 (down-sampling factor),表示為乃,以調整框化函數之滾 邊。舉例來說’當保護區間與通道長度的差值縮小時,滚 邊振幅產生單元620將縮減取樣因數£)變大,以縮短框化 函數之滾邊長度K,並防止框化函數之滾邊將會擷取到鄰 近之符元。在此,為了方便說明本實施例,假設滾邊振幅 產生單元620決定出縮減取樣因數£&gt;=2。 在滾邊振幅產生單元620決定縮減取樣因數Z)之後, 在每間隔縮減取樣因數Z),排除所對應的框化係數。在 此,由於縮減取樣因數£)=2,因此,滾邊振幅產生單元 620將排除滚邊振幅產生單元“ο中的框化係數、 303、 305、307、308、310、312 與 314。 而滾邊振幅產生單元620利用剩餘的框化係數302、 304、 306、309、311與313 ’作為一調整框化函數,以調 整滾邊斜率與長度,其中,剩餘的滚邊振幅3〇2、3〇4與 306之值分別為0.25、0.5與0.75。由剩餘的框化係數將可 組成一調整框化函數,如圖7所示。圖7繪示為本發明實 施例之調整框化函數。請參照圖7,其橫座標為取樣時間 w,其縱座標為框化係數。 圖7中之左側的框化係數7〇1 ' 702與703之值分別為 圖3中之框化係數302、304與306 ’右側的框化係數 704、705與706之值分別為圖3中之框化係數309、311 與313。而圖7中之滾邊長度t = 4,符元長度為L與圖3 中之符元長度相同,因此,調整框化函數的總長度為 K,+ L 〇 圖7中調整框化函數之滾邊為一線性函數,並且此調 整框化函數具有左右對稱之特性。而由圖3中之框化函數 的第(1)式’將可推出圖7中之調整框化函數的一通式,並 表示如下: 'RK[Dn], \&lt;,n^K/D-\One of the second addition units 438, that is, the edge amplitude from the roll amplitude generating unit 420 at this time is determined to be the framed coefficient corresponding to the left side, and is the frame factor of the right side (step S56). If the edge amplitude from the edge amplitude generating unit 420 corresponds to the frame factor of the left side, the multiplication unit 432 outputs the product to the second adding unit 434' and subtracts the received product from the sampled signal via the adding unit 434. It is then stored in the buffer unit 436 (step S565). The edge amplitude of the wire from the edge amplitude generating unit 42 is corresponding to the framed Wei on the right side, and the job unit 432 sends it out to the third adding unit 438' to multiply the received one via the adding unit 438. After the corresponding product, the following is the method and device for adjusting the surface of the present invention. The method and apparatus for adjusting the surface of the present invention are shown in Fig. 6 Blocking device block diagram. Referring to FIG. 6, the adjustment frame is set to 1379561 AFA-P06001 21493 twf.doc/n 600 includes a framed coefficient storage unit 610, a roll edge amplitude generating unit 620, and a framed operation unit 630. 630 further includes a multiplication unit 632, a buffer unit 636, and an addition unit 638. Before the present embodiment is described, it is assumed that the present embodiment uses the frame function in Fig. 3. For the convenience of the embodiment of the present invention, the frame factor on the left side is used. The values of 301 to 307 are sequentially defined as 0.125, 0.25, 0.375, 0.5, 0.625, 0.75, and 0.875, and the values of the frame factors 308 to 314 on the right side are sequentially defined as 0.875, 0.75, 0.625, 0.5, 0.375, 0.25. With 0.125. And, assuming boxing The number storage unit 61 储存 stores the frame factor in FIG. 3 as 301 to 307 ′ and the edge amplitude generation unit 620 can adjust the frame of the frame function by using the frame coefficients 301 to 307. Next, it is assumed that the frame is adjusted. The device 600 is included at a receiving end, and the receiving end has a channel estimation module (not shown), and the channel estimation module estimates the channel impulse response by using the received signal from the channel. The channel impulse response calculates the channel length. Furthermore, it is assumed that the reception is applied to the OFDM system, and the received signal from the channel_ contains a plurality of OFDM symbols, and includes a guard interval in each symbol and the guard interval is The preamble signal is looped, and the original unpaid length of the unprotected interval is L, that is, the actual length of the original symbol is the unit sampling time. The relationship between the channel length and the guard interval is as shown in Fig. 1. After the receiving end samples the received signal, the sampling signal S[n] is input to the adjusting frame device 6〇0, and the framed agricultural device 6〇〇 is utilized by the channel estimating module. Calculate the length of the channel, and trim the adjustment frame 18 AFA-P06001 21493twf.doc/n function, and frame the sampled signal S[n] to correctly extract each OFDM symbol, and valid To prevent narrowband interference in the channel. Referring to Figures 3 and 6, the starting edge amplitude generating unit 620 receives a channel length, which is provided, for example, by the channel estimation module in the receiving end. After the channel length, the edge amplitude generating unit 620 will find the difference between the channel length and the guard interval, for example, the length of the guard interval minus the channel length. It can be observed from the above FIG. 1 that when the difference between the guard interval and the channel length becomes larger, the effective capture interval Tav in FIG. 1 will also become longer; otherwise, when the difference between the guard interval and the channel length When it becomes smaller, the effective operation interval Tav in Fig. 1 will also become shorter. Therefore, next, the edge amplitude generating unit 620 determines a down-sampling factor based on the difference between the protected area and the channel length, and indicates that it is a roll to adjust the frame function. For example, when the difference between the guard interval and the channel length is reduced, the edge amplitude generating unit 620 increases the downsampling factor £ to shorten the edge length K of the boxing function and prevent the edge of the boxing function from being rolled. Capture to the adjacent symbol. Here, for convenience of explanation of the present embodiment, it is assumed that the piping amplitude generating unit 620 determines the downsampling factor £&gt;=2. After the roll amplitude generating unit 620 decides to reduce the sampling factor Z), the sampling factor Z) is reduced at each interval, and the corresponding frame factor is excluded. Here, since the downsampling factor £)=2, the edge amplitude generating unit 620 will exclude the frame factor, 303, 305, 307, 308, 310, 312, and 314 in the edge amplitude generating unit ο. The amplitude generating unit 620 uses the remaining framing coefficients 302, 304, 306, 309, 311, and 313' as an adjustment frame function to adjust the edge slope and length, wherein the remaining edge amplitudes are 3〇2, 3〇4 The values of 306 and 306 are respectively 0.25, 0.5 and 0.75. The remaining framed coefficients can form an adjustment frame function, as shown in Fig. 7. Fig. 7 shows an adjustment frame function according to an embodiment of the present invention. Figure 7, the abscissa is the sampling time w, and its ordinate is the boxing coefficient. The values of the boxing coefficients 7〇1' 702 and 703 on the left side in Fig. 7 are the frame factors 302, 304 and The values of the boxing coefficients 704, 705, and 706 on the right side of 306' are the frame factors 309, 311, and 313 in Fig. 3, respectively, and the length of the piping in Fig. 7 is t = 4, and the length of the symbol is L and is in Fig. 3. The symbol length is the same, so the total length of the adjustment box function is K, + L 调整 the adjustment box function in Figure 7 The rolling edge is a linear function, and the adjustment boxing function has the characteristics of left-right symmetry. However, the formula (1) of the boxing function in FIG. 3 can introduce a general formula of the adjustment boxing function in FIG. And expressed as follows: 'RK[Dn], \&lt;,n^K/D-\

1, KIO&lt;.n&lt;,L wjc/i&gt;L«]=· L+\^n&lt;,L+M)-\ (3) .〇, 其他 上述第(3)式中,ΑπΜ為第《取樣時間的框化係數。 接下來,請同時參考圖3、圖6與圖7,在排除部分 框化係數之後’滾邊振幅產生單元620將利用剩餘的框二 係數,產生滾邊振幅〜AD[W],並輸出至框化運算單元 630。在本實施例中’由於縮減取樣因數ιχ,因:,滚 1379561 AFA-P06001 21493twf.doc/n 邊振幅表示為。在此假設框化裝置600要先對取樣 時間《 = 1的取樣信號S[l]進行框化,因此,滾邊振幅產生 單元620將由框化係數儲存單元410讀取出剩餘的框化係 數302作為滾邊振1)¾ z/:/2[l],並輸出至框化運瞀單元 630,也就是,將圖7中之框化係數701作為滾邊振幅 輸出至框化運算單元630,其中,滾邊振幅\/2[η1, KIO&lt;.n&lt;,L wjc/i&gt;L«]=· L+\^n&lt;,L+M)-\ (3) .〇, in the above other formula (3), ΑπΜ is the first sampling The boxing factor of time. Next, please refer to FIG. 3, FIG. 6 and FIG. 7, at the same time, after the partial framing coefficient is excluded, the 'rolling edge generation unit 620 will generate the edge amplitude ~AD[W] using the remaining frame two coefficients, and output to the frame. The arithmetic unit 630. In the present embodiment, due to the reduction of the sampling factor ιχ, the roll amplitude is expressed as: 1379561 AFA-P06001 21493twf.doc/n. It is assumed here that the framing device 600 first frames the sampling signal S[l] of the sampling time "=1". Therefore, the edge amplitude generating unit 620 reads the remaining framing coefficient 302 from the framing coefficient storage unit 410 as The edge vibration 1)3⁄4 z/:/2[l] is output to the frame operation unit 630, that is, the frame factor 701 in FIG. 7 is output as the edge amplitude to the frame operation unit 630, wherein the edge is rolled. Amplitude\/2[η

之值為0.25。而框化運异單元630也將接收到取揭作缺 s[l],並輸入至乘法單元632。 ’ σ A 然後,乘法單元632將取樣信號s[l]與滾邊振幅產生 單元620所輸出之滾邊振幅〜&quot;[I]相乘後,輸出乘積 ~2[l]xS[l]=0.25S[l] ’並將乘積0.25S[1]儲存至緩衝單元 636。 之後,依照相同的硬體操作,滾邊振幅產生單元62〇 依序輸出剩餘的框化係數304與3〇6,來作為滾邊振幅 ,並透過乘法單元632分別與取樣信號$⑵ 與S[3]相乘後,儲存乘積〇 58[2]與〇 75S[3]至緩衝 636。 接下來,由於取樣信號s[4]、s[5]、…、S[L]所對應 之框化係數為1,因此’取樣信號S⑷、s[5]、 、s[L] 將直接輸入至後端處理,並未輸入至調整框化的 600。 接著,滾邊振幅產生單元62〇將讀取出剩餘的框化係 數306作為滾邊振幅z&amp;JL + i],並輸出至框化運算單元 630,其中,滾邊振幅z〜2[£ + 1]之值為〇乃。而框化運算 21 1379561 AFA-P06001 21493twf.doc/n 單元630也將接收到取樣信號S[L+1],並輸入至乘法單元 632。 然後,乘法單元632將取樣信號S[L+1]與滾邊振幅 產生單元620所輸出之滾邊振幅+ 相乘後,輸出乘 積〜/2[l+i]xS[l+1]=:〇 75S[L+1]至加法單元⑽。 由於調整框化的裝置600所要框化的OFDM符元中包 含循環前置信號,可推知取樣信號sphsp]為取樣信號 S[L+1卜S[L+3]相同之重複信號。因此,調整框化的裝置 600將利用取樣信號S[1]〜s[3]與取樣信號s[l+1]〜s[l+3] 為相同信號的特性’來還原出原始的信號。 因此,在加法單元638接收到乘法單元632所輸出之 〇.75S[L+l]後,加法單元638將讀取出之前儲存於缓衝單 元636的0.25S[1],並將〇.75S[L+l]與〇.25S[l]相加後輸出 至後端處理,以讓後端之裝置還原出原始之信號。 依照上述相同的硬體操作,滾邊振幅產生單元62〇將 剩餘的框化係數304、302分別作為滾邊振幅々/2[Z + 2]與 + ,並依序輪出至框化運算單元。再透過乘法 632依序將滾邊振幅々/2[£ + 2]與〜/2[1 + 3]乘以取樣 信號S[2]與S[3]後,將乘積〇.5S[L+2]與〇 2雖+3]輸出至 加法單元638。再透過加法單元638將〇 5S[L+2]加上之前 健存於緩衝單元636的〇.5S[2]後,輸出至後端處理,並且 將〇.25S[L+3]加上之前儲存於緩衝單元636的〇75 後’輸出至後端處理。 之後’在滾邊振幅產生單元62〇接收到通道估測模組 22 1379561 ΑΡΑ-ΡΟβοοι 21493twf.doc/n 所提供之通道長度時,仍依據保護區間與通道長度的差 值,決定縮減取樣因數乃,以調整框化函數之滾邊。舉例 * 來說,當保護區間與通道長度的差值很小時,滾邊振幅產 生單元620可採用縮減取樣因數Z)=1,也就是,調整框化 的裝置600直接使用圖3之框化函數,來對取樣函數s[n] 進行框化。而當保§蒦區間與通道長度的差值變大時,滚邊 振幅產生單元620可採用縮減取樣因數£&gt;=2,以縮減取樣 • (d〇Wn_sam沖叩)圖3之框化函數,並輪出剩餘的框化係 數’而讓調整框化的裝置600使用圖7中之框化函數,來 對取樣函數S[n]進行框化。依照此模式,適時地調整框化 函數之滾邊,將能夠避免擷取符元上的信號時,造成符元 間的干擾,也可以有效地降低接收信號在通道中的窄頻 干擾。 由上述圖6中調整框化的裝置600之操作說明,可歸 納出一方法步驟,如圖8所示。圖8繪示為本發明第二實 施例之調整框化的方法步驟流程圖。請同時參照圖6與圖 8,首先,滾邊振幅產生單元620接收一通道長度(步驟 S810)。接著,滾邊振幅產生單元62〇找出通道長度鱼保 護區間的差值(步驟S815),並決定一縮減取樣因數驟 S820)。之後,滾邊振幅產生單元62〇在每間隔縮減取樣 因數D,排除所對應的框化係數(步驟S825),並利用剩餘 的框化係數產生滾邊振幅,依序輸出至框化運算單元 630(步驟S830) ’以調整框化之滚邊。 接下來,框化運算單元630接收取樣信號(步驟 23 1379561 AFA-P06001 21493twf.doc/n S840)。然後,乘法單元632將取樣信號與來自滾邊振幅 產生單元620的浪邊振幅相乘後,輸出一乘積(步驟 • S850)。接者,依據取樣時間《,乘法單元632決定乘積輪 出至緩衝單元636與加法單元638其中之一,也就是判斷 來自於滾邊振幅產生單元620的滾邊振幅是對應至左側之 框化係數還是右側之框化係數(步驟S86〇p若來自於滾邊 振巾w產生單元620的滾邊振幅對應至左側之框化係數,則 . 乘法單元632將其乘積輸出並儲存至緩衝單元636(步驟 S870)。若來自於滾邊振幅產生單元62〇的滾邊振幅對應 至右側之框化係數,則乘法單元632將其乘積加上緩衝單 元636内所對應之乘積後輪出(步驟S880)。 以下結合上述的第一與第二實施例,提出本發明的 第三實施例,以說明本發明實施例之框化裝置,而本實施 例所提出之框化裝置能夠同時重複利用框化係數,又具有 調整框化之能力。圖9繪示為本發明第三實施例之框化裝 置方塊圖。請參照圖9 ’框化裝置900包括框化係數儲存 單元910、滾邊振幅產生單元920與框化運算單元930。其 中’滾邊振幅產生早元920包括一第一加法單元922與與 多工單元925。而框化運算單元930包括乘法單元932、第 —加法單元934、缓衝單元936與第三加法單元938。 在說明本實施之前,為了方便說明本發明實施例, 先假設本實施例使用圖3中之框化函數〜[«]。並且,假 设圖3中左側的框化係數301〜307之值依序為0.875、 〇·75、0.625、0.5、0.375、0.25、0.125。右側的框化係數 24 1379561 AFA-P06001 21493twf.doc/n 308〜314 之值依序為 0 875、0.75、0.625、0.5、0.375、 0.25、0.125。另外,為了方便說明本實施例,將左側之 框化係數301〜307定義為兩個群組,其中框化係數 301〜303定義為為左半邊框化係數,而框化係數304〜307 定義為右半邊框化係數;同樣地,右側之框化係數 308〜314分別被定義為左半邊框化係數308〜310與右半邊 框化係數311〜314。由於本實施例所提出之框化裝置900 能夠重複利用框化係數,因此,假設本實施例的框化係數 儲存單元910只儲存四個框化係數304〜307 接著,再假設框化裝置900包含於一接收端,而此接 收端具有通道估測模組(未繪示)’而此通道估測模組利用 來自通道之接收信號估測出通道脈衝響應,並由通道脈衝 響應計算出通道長度。此外,假設接收端應用於OFDM 系統,而來自於通道中的接收信號包含多個OFDM符 元,並在每個符元中包含一保護區間,且保護區間内為循 環前置信號,而未加保護區間之原始符元長度為L,也就 是,原始符元之實際長度為L個單位取樣時間。而通道長 度與保護區間的關係例如為圖1所示。 另外,接收端將接收信號經過取樣後,產生取樣信號 S[n]輸入至框化裝置900 ’而框化裝置900可相同於第二 實施例中的調整框化裝置600 ’能夠利用由通道估測模組 所計算出之通道長度,來對調整框化函數之滾邊,並對取 樣信號S[n]進行框化’以正確地擷取出每個OFDM符元, 以及有效地防止通道中的窄頻帶干擾。 25 1379561 AFA-P06001 21493twf.doc/n 請同時參照圖3、圖7與圖9,開始時,淚邊振幅產 生單元920接收一通道長度,而此通道長度例如由端 中的通道估測模組所提供。之後,滾邊振幅產生單元92〇 將找出通道長度與保護區間的差值,此差值例如為保講區 間的長度減去通道長度》 °° 接下來’滾邊振幅產生單元920將依據保護區間與通 道長度的差值’決定一縮減取樣因數,表示為D,以調整 框化函數之滾邊。在本實施例中’縮減取樣因數例如為 2的次方數’也就是說’縮減取樣因數D可表示為2,·,其 中,丨為正整數。在此,為了方便說明本實施例,假設滾 邊振幅產生單元920決定出縮減取樣因數D=2。 在滾邊振幅產生單元920決定縮減取樣因數之後, 在每間隔縮減取樣因數D,排除所對應的框化係數。在 此,由於縮減取樣因數£&gt;==2,因此,滾邊振幅產生單元 920將排除圖3中之框化係數30卜303、305、307、308、 310、312與314。並由剩餘的框化係數302、304、306、 309、311與313組成一調整框化函數’如圖7所示。其 中’剩餘的框化係數302、304與306之值分別為0.75、 0.5 與 0.25。 換句話說,此時滾邊振幅產生單元920只會利用到框化係 數儲存單元910中所儲存的框化係數304與306,來產生 出圖7中之框化係數701〜706,以對取應信號S[n]進行框 化。 在滾邊振幅決定了縮減取樣因數Z)之後,將依序由 26 四個框化係數3G4〜3G7取出—特定框化係數〜»,其 中’ 與框化函數々[«]的關係為如下: xjst/d[w] (4) \y&gt;K[K-Dnl n=:K..,K/D/2 ί w^Dn^ « =(尺/β/2) + 1,._·χ/£»-ΐ 在本實施例中,由於已決定出縮減取樣因數D = 2,=2,因 此’特雜化係數= [”],而上述第⑷可轉變為 χκη\.η\: 1 κ rt = 1,…,一 (5) 4 _K 、 Κ w =--1-1,..., —1 4 2 ,著,滾邊振幅產生單元侧中的第—加法單元將 與C 一 f q數^相減後,輸出差值 :::Γ 其中,常數c為實數,在本實施例 中,吊數例如為卜而多工單元925祕至第一加法單 =922與框化係數儲存單元41〇,同時接收框化係數儲存 早7L 910所輸出的%/2W與加法單元422所輸出的差值 ,並且,多工單元425在每個取樣時間”,依據一 選擇信號卿輸Α々Μ V士]財騎„取間之 滾邊振幅係數〜/2[w]。 在此,假設框化裝置900要先對取樣時間„ = 1的取樣 #號S[l]進行框化,也就是說需要產生圖7中之框化係數 70卜並由上述第(5)可知,滚邊振幅產生單元92〇取出的 特定框化係數〜/2[1]=〜[6],也就是說,在„ = 1時,滾邊 振幅產生單元620將讀取出框化係數306。並且,由^框 化係數701之值為與框化係數3〇6與j之差值,因此.,、多 ,單,925選擇^^作為滾邊振幅〜/2[1],而滚邊振巾1 尺/21」之值為0. 75。 1379561 AFA-P0600 1 21493twf.doc/n 多工單元925所接收的選擇信號例如為圖4中的選擇 信號’依據欲產生之框化係數為左半邊框化係數或右半邊 框化係數來改變其狀態。在上述滾邊振幅產生單元920的 操作中’由於需要產生的框化係數為框化係數701,而框 化係數701即為剩餘的框化係數3〇2,故框化係數7〇1為 左半邊框化係數’因此,選擇信號為〇,而多工單元925 選擇將第一加法單元922所輸出之差值作為滾邊振 幅〜/2[1] ’並輸出至框化運算單元930。 接下來,框化運算單元930接收取樣信號S[l],並輸 入至乘法單元932與第二加法單元934。之後,乘法單元 932將取樣信號S[1]與多工單元925所輸出之滾邊振幅 〜/2[1]相乘’輸出乘積〜/2[l]xS[1]=〇75S⑴至第二加法單 元934。接著,第二加法單元934將取樣信號s[1]與乘法 單元932所輸出的乘積〇.75S[1]相減後,將輸出並儲存差 值0.25S[1]至緩衝單元936。 之後,依照上述相同的硬體操作,滚邊振幅產生單元 920將依序取出框化係數儲存單元91〇内所儲存之框化係 數304與306,以產生滾邊振幅〜/2[2] = 0.5與 〜/2[3] = ()·25。在透過框化運算單元930產生並儲存差值 0.5S[2]與 〇.75S[3]至緩衝單元 936。 接下來,由於取樣信號S[4]、S[5]、…、S[L]所對應 之框化係數為1,因此’取樣信號S[4]、S[5]、…、S[L] 將直接輸入至後端處理,並未輸入至框化裝置9〇(^而在 接下來的步驟,將要對取樣信號S[L+1]〜s[L+3]進行框化 28 1379561 AFA-P06001 21493twf.doc/n 運鼻。在此,假設框化裝置400先對取樣信號S[L+1]進行 運算,因而需要產生框化係數704,並且,框化係數7〇4 之值為框化係數306與1之差值。 因此,接下來,滾邊振幅產生單元920讀取出框 化係數儲存單元910所儲存之框化係數306,作為特定框 化係數+ ,其值為〇.25 ^然後,加法單元922將計 异出特定框化係數〜/2[^ + 1]與1之差值,並輸出差值 + 至多工單元925,其中,差值心/2[1 + 1]為 0.75。由於此時需要產生之框化係數7〇4即為剩餘的框化 係數309,故框化係數704為左半邊滚邊振幅,因此,多 工單元925所接收的選擇信號為〇。之後,多工單元425 將依據選擇信號將差值+ 作為框化係數7〇4對應 的滾邊振幅Z/^2[L + 1],並輸出至框化運算單元930中之乘 法單元932。框化運算單元930中之乘法單元932接收取 樣信號S[L+1]。並由乘法單元932將滾邊振幅+ η與 取樣信號S[L+1]相乘後,輸出乘積 〜/2[1 + 1]xS[L+1]=〇.75S[L+l]至第三加法單元 938 〇 貝 由於框化裝置900所要框化的0Fdm符元中包含循環 前置信號’可推知取樣信號S⑴〜s[3]為取樣信號 S[L+l]~S[L+3]相同之重複信號。因此,框化裝置9〇〇將 利用取樣信號S[l]〜S[3]與取樣信號S[L+1]〜S[L+3]為相同 信號的特性,來還原出原始的信號。 在第三加法單元938接收到乘法單元932所輸出之乘 積0.75S[L+1]後,加法單元938將讀取出之前儲存於緩衝The value is 0.25. The framed transport unit 630 will also receive the squirrel s[l] and input it to the multiplying unit 632. 'σ A Then, the multiplication unit 632 multiplies the sampling signal s[l] by the edge amplitude ~&quot;[I] output by the edge amplitude generating unit 620, and outputs the product ~2[l]xS[l]=0.25S [l] 'Save the product 0.25S[1] to the buffer unit 636. Thereafter, according to the same hardware operation, the edge amplitude generating unit 62 sequentially outputs the remaining frame coefficients 304 and 3〇6 as the edge amplitude, and passes through the multiplying unit 632 and the sampling signals $(2) and S[3], respectively. After multiplication, the product 〇58[2] and 〇75S[3] are stored to the buffer 636. Next, since the framed coefficients corresponding to the sampling signals s[4], s[5], ..., S[L] are 1, the sampling signals S(4), s[5], and s[L] will be directly input. To the backend processing, it is not input to the adjustment frame 600. Next, the edge amplitude generating unit 62 读取 reads out the remaining frame factor 306 as the edge amplitude z&JL + i], and outputs it to the frame operation unit 630, where the edge amplitude z 2 2 [£ + 1] The value is 〇. The frame operation 21 1379561 AFA-P06001 21493 twf.doc/n unit 630 will also receive the sampling signal S[L+1] and input it to the multiplication unit 632. Then, the multiplication unit 632 multiplies the sampling signal S[L+1] by the edge amplitude + output by the edge amplitude generating unit 620, and outputs the product ~/2[l+i]xS[l+1]=:〇75S [L+1] to the addition unit (10). Since the OFDM symbol to be framed by the apparatus 800 for adjusting the frame includes the cyclic preamble, it can be inferred that the sampling signal sphsp] is a repetitive signal of the same sampling signal S[L+1, S[L+3]. Therefore, the apparatus for adjusting the frame 600 will restore the original signal by using the sampling signals S[1]~s[3] and the sampling signals s[l+1]~s[l+3] as the characteristics of the same signal. Therefore, after the addition unit 638 receives the 〇.75S[L+l] output by the multiplication unit 632, the addition unit 638 will read out the 0.25S[1] previously stored in the buffer unit 636, and the 〇.75S [L+l] is added to 〇.25S[l] and output to the backend processing to let the device at the back end restore the original signal. According to the same hardware operation as described above, the edge amplitude generating unit 62 作为 takes the remaining frame coefficients 304 and 302 as the edge amplitudes 々/2 [Z + 2] and +, respectively, and sequentially rotates to the frame operation unit. After multiplying the edge amplitude 々/2[£ + 2] and ~/2[1 + 3] by the sampling signal S[2] and S[3] by multiplication 632, the product 〇.5S[L+2 ] and 〇 2 although +3] are output to the addition unit 638. Then, the 〇5S[L+2] is added to the 〇.5S[2] of the buffer unit 636 by the addition unit 638, and then output to the back end processing, and 〇.25S[L+3] is added before Stored in the buffer unit 636 after the 〇 75 'output to the back end processing. Then, when the channel amplitude generating unit 62 receives the channel length provided by the channel estimation module 22 1379561 ΑΡΑ-ΡΟβοοι 21493twf.doc/n, the sampling factor is determined according to the difference between the guard interval and the channel length. To adjust the edge of the boxing function. For example, when the difference between the guard interval and the channel length is small, the edge amplitude generating unit 620 can use the downsampling factor Z)=1, that is, the framed device 600 directly uses the frame function of FIG. To frame the sampling function s[n]. When the difference between the interval and the channel length becomes larger, the edge amplitude generating unit 620 can use the reduced sampling factor £>=2 to reduce the sampling. (d〇Wn_sam) The frame function of FIG. And rotating the remaining framed coefficients' and letting the framed device 600 frame the sampling function S[n] using the boxing function in FIG. According to this mode, timely adjustment of the edge of the framed function will avoid interference between symbols when the signal on the symbol is captured, and can also effectively reduce the narrowband interference of the received signal in the channel. From the operational description of the apparatus 600 for adjusting the frame in Fig. 6, a method step can be summarized, as shown in Fig. 8. FIG. 8 is a flow chart showing the steps of the method for adjusting the frame according to the second embodiment of the present invention. Referring to FIG. 6 and FIG. 8, at the same time, first, the edge amplitude generating unit 620 receives a channel length (step S810). Next, the edge amplitude generating unit 62 finds the difference of the channel length fish protection section (step S815), and determines a downsampling factor (S820). Thereafter, the edge amplitude generating unit 62 reduces the sampling factor D at each interval, excludes the corresponding frame factor (step S825), and generates the edge amplitude by using the remaining frame factor, and sequentially outputs the result to the frame operation unit 630 (step S830) 'To adjust the framed roll. Next, the frame operation unit 630 receives the sampling signal (step 23 1379561 AFA-P06001 21493 twf.doc/n S840). Then, the multiplying unit 632 multiplies the sampling signal by the amplitude of the wave edge from the edge amplitude generating unit 620, and outputs a product (step S850). According to the sampling time, the multiplication unit 632 determines that the product is rotated out to one of the buffer unit 636 and the addition unit 638, that is, whether the edge amplitude from the edge amplitude generating unit 620 corresponds to the frame factor of the left side or the right side. The frame factor (step S86〇p) if the edge amplitude from the edge ringer w generating unit 620 corresponds to the frame factor on the left side, the multiplication unit 632 outputs the product and stores it in the buffer unit 636 (step S870). If the edge amplitude from the edge amplitude generating unit 62A corresponds to the frame factor on the right side, the multiplication unit 632 adds the product to the product corresponding to the buffer unit 636 and then rotates (step S880). A third embodiment of the present invention is provided to illustrate a framed device according to an embodiment of the present invention. The framed device of the present embodiment can simultaneously use the framed coefficient and have an adjustment frame. Figure 9 is a block diagram of a framed device according to a third embodiment of the present invention. Referring to Figure 9, the frame device 900 includes a frame factor storage unit 910, The edge amplitude generating unit 920 and the boxing operation unit 930. The 'rolling edge amplitude generating element 920 includes a first adding unit 922 and an multiplex unit 925. The framed operation unit 930 includes a multiplying unit 932 and a first adding unit 934. The buffer unit 936 and the third adding unit 938. Before explaining the present embodiment, in order to facilitate the description of the embodiment of the present invention, it is assumed that the present embodiment uses the boxing function ~[«] in Fig. 3. Also, assume that in Fig. 3 The values of the frame factors 301 to 307 on the left are 0.875, 〇·75, 0.625, 0.5, 0.375, 0.25, and 0.125. The value of the frame factor 24 1379561 AFA-P06001 21493 twf.doc/n 308~314 on the right side depends on The order is 0 875, 0.75, 0.625, 0.5, 0.375, 0.25, 0.125. In addition, for convenience of description of the present embodiment, the frame factors 301 to 307 on the left side are defined as two groups, wherein the frame factors 301 to 303 are defined. In order to frame the left half frame, the frame coefficients 304 to 307 are defined as right half frame coefficients; similarly, the frame coefficients 308 to 314 on the right side are defined as left half frame coefficients 308 to 310 and right half borders, respectively. The coefficients are 311 to 314. The framed device 900 proposed in this embodiment can reuse the framed coefficients. Therefore, it is assumed that the framed coefficient storage unit 910 of the present embodiment stores only four framed coefficients 304 to 307. Next, the framer device 900 is assumed to be included. At a receiving end, the receiving end has a channel estimation module (not shown), and the channel estimation module estimates the channel impulse response by using the received signal from the channel, and calculates the channel length from the channel impulse response. . In addition, it is assumed that the receiving end is applied to the OFDM system, and the received signal from the channel contains a plurality of OFDM symbols, and a guard interval is included in each symbol, and the loop pre-signal is not included in the guard interval. The original symbol length of the guard interval is L, that is, the actual length of the original symbol is L unit sampling time. The relationship between the channel length and the guard interval is as shown in Fig. 1. In addition, after the receiving end samples the received signal, the sampling signal S[n] is input to the framing device 900', and the framing device 900 can be the same as the adjusting frame device 600' in the second embodiment. The length of the channel calculated by the module is used to adjust the edge of the framed function and frame the sampled signal S[n] to correctly extract each OFDM symbol and effectively prevent narrowing in the channel. Band interference. 25 1379561 AFA-P06001 21493twf.doc/n Please refer to FIG. 3, FIG. 7 and FIG. 9 at the same time. At the beginning, the tear edge amplitude generating unit 920 receives a channel length, and the channel length is, for example, the channel estimation module in the end. Provided. Thereafter, the edge amplitude generating unit 92 will find the difference between the channel length and the guard interval, for example, the length of the guard interval minus the channel length. ° ° Next, the 'roll edge amplitude generating unit 920 will depend on the guard interval and The difference in channel length 'determines a reduced sampling factor, denoted D, to adjust the edge of the boxing function. In the present embodiment, the 'downsampling factor is, for example, the power of 2', that is, the 'downsampling factor D' can be expressed as 2, ·, where 丨 is a positive integer. Here, for convenience of explanation of the present embodiment, it is assumed that the edge amplitude generating unit 920 determines the downsampling factor D = 2. After the roll amplitude generating unit 920 decides to reduce the sampling factor, the sampling factor D is reduced at each interval, and the corresponding frame factor is excluded. Here, since the sampling factor £&gt;==2 is reduced, the edge amplitude generating unit 920 will exclude the frame coefficients 30, 303, 305, 307, 308, 310, 312, and 314 in Fig. 3. And an adjustment boxing function ' is composed of the remaining frame coefficients 302, 304, 306, 309, 311 and 313 as shown in FIG. The values of the remaining 'boxing coefficients 302, 304 and 306 are 0.75, 0.5 and 0.25, respectively. In other words, at this time, the edge amplitude generating unit 920 only uses the framed coefficients 304 and 306 stored in the framed coefficient storage unit 910 to generate the framed coefficients 701 to 706 in FIG. The signal S[n] is framed. After the roll amplitude determines the downsampling factor Z), it will be sequentially extracted from the 26 four frame factors 3G4~3G7 - the specific frame factor ~», where 'the relationship with the box function 々[«] is as follows: xjst /d[w] (4) \y&gt;K[K-Dnl n=:K..,K/D/2 ί w^Dn^ « =(尺/β/2) + 1,._·χ/ £»-ΐ In this embodiment, since the downsampling factor D = 2, = 2 has been determined, the 'special hybrid coefficient = ["], and the above (4) can be converted to χκη\.η\: 1 κ Rt = 1,...,1(5) 4 _K , Κ w =--1-1,..., —1 4 2 ,, the first-addition unit in the edge amplitude generating unit side will be the same as C-fq ^ After subtraction, the output difference is :::: Γ where the constant c is a real number. In the present embodiment, the hang number is, for example, the multiplex unit 925 secret to the first addition list = 922 and the framing coefficient storage unit 41. 〇, receiving the framed coefficient to store the difference between the %/2W outputted by the 7L 910 and the adding unit 422, and the multiplex unit 425 is at each sampling time, according to a selection signal.士]财骑 „ 之 之 之 振幅 振幅 振幅 〜 〜 〜 〜 〜 〜 〜 〜 〜 〜 〜 〜 〜 〜 〜 〜 900 is to first frame the sampling time #[1] of the sampling time „=1, that is, it is necessary to generate the frame factor 70 in FIG. 7 and it is known from the above (5) that the edge amplitude generating unit 92 The specific frame factor of 〇 is taken out~/2[1]=~[6], that is, when „=1, the edge amplitude generating unit 620 will read out the frame factor 306. The value of 701 is the difference between the boxing coefficient 3〇6 and j, so ., multi, single, 925 select ^^ as the edge amplitude ~/2[1], and the rolling vibrating 1 ft / 21" The value is 0.75. 1379561 AFA-P0600 1 21493twf.doc/n The selection signal received by the multiplex unit 925 is, for example, the selection signal in FIG. 4, which is changed according to the frame factor to be generated as the left half frame factor or the right half frame factor. status. In the operation of the above-described edge amplitude generating unit 920, 'the frame factor 701 is the remaining frame factor 〇2 because the frame factor to be generated is the frame factor 701, so the frame factor 7〇1 is the left half. The frame factor ', therefore, the selection signal is 〇, and the multiplex unit 925 selects the difference value output by the first adding unit 922 as the edge amplitude ~/2[1] ' and outputs it to the frame operation unit 930. Next, the frame operation unit 930 receives the sampling signal S[1] and inputs it to the multiplication unit 932 and the second addition unit 934. Thereafter, the multiplying unit 932 multiplies the sampling signal S[1] by the edge amplitude ~/2[1] output by the multiplex unit 925 'output product ~/2[l]xS[1]=〇75S(1) to the second addition Unit 934. Next, the second adding unit 934 subtracts the sampling signal s[1] from the product 〇.75S[1] output from the multiplying unit 932, and outputs and stores the difference value 0.25S[1] to the buffer unit 936. Then, in accordance with the same hardware operation as described above, the edge amplitude generating unit 920 sequentially extracts the framed coefficients 304 and 306 stored in the framed coefficient storage unit 91 to generate a roll amplitude of ~/2 [2] = 0.5. With ~/2[3] = ()·25. The difference values 0.5S[2] and 〇.75S[3] are generated and stored in the buffering operation unit 930 to the buffer unit 936. Next, since the framed coefficients corresponding to the sampling signals S[4], S[5], ..., S[L] are 1, the sampling signals S[4], S[5], ..., S[L ] will be directly input to the back-end processing, and will not be input to the framer 9 (^ and in the next step, the sampling signal S[L+1]~s[L+3] will be framed 28 1379561 AFA -P06001 21493 twf.doc/n. In this case, it is assumed that the framing device 400 first operates on the sampling signal S[L+1], so that a framed coefficient 704 needs to be generated, and the value of the framing coefficient 7 〇 4 is The difference between the framed coefficients 306 and 1. Therefore, the edge amplitude generating unit 920 reads the framed coefficient 306 stored by the framed coefficient storage unit 910 as a specific framed coefficient +, and its value is 〇.25. ^ Then, the addition unit 922 will calculate the difference between the specific frame factor ~/2[^ + 1] and 1 and output the difference + to the multiplex unit 925, where the difference heart/2[1 + 1] It is 0.75. Since the frame factor 7〇4 which needs to be generated at this time is the remaining frame factor 309, the frame factor 704 is the left half edge amplitude, and therefore, the selection signal received by the multiplex unit 925 is 〇. Thereafter, the multiplex unit 425 will use the difference + as the edge amplitude Z/^2[L + 1] corresponding to the frame factor 7〇4 according to the selection signal, and output it to the multiplication unit 932 in the frame operation unit 930. The multiplication unit 932 in the arithmetic operation unit 930 receives the sampling signal S[L+1], and multiplies the edge amplitude + η by the multiplication unit 932 by the sampling signal S[L+1], and outputs the product ~/2 [1 + 1] xS[L+1]=〇.75S[L+l] to the third adding unit 938. The mussels contain the cyclic preamble signal in the 0Fdm symbol to be framed by the framing device 900. The sampling signal S(1) can be inferred~ s[3] is the same repetitive signal as the sampling signal S[L+l]~S[L+3]. Therefore, the framer 9〇〇 will use the sampling signals S[l]~S[3] and the sampling signal S [L+1]~S[L+3] are characteristics of the same signal to restore the original signal. After the third addition unit 938 receives the product of the multiplication unit 932 output 0.75S[L+1], the addition Unit 938 will be stored in the buffer before being read out

29 AFA-P06001 21493twf.doc/n 單凡936的0.25S[1],並將〇.75S[L+1]與(^^⑴相加後輸 出至後端處理,以讓後端之裝置還原出原始之信號。 之後依照上述之硬體的操作,滾邊振幅產生單元92〇 將利用框化係數304與306,依序產生出框化係數7〇5與 . . 巾田 / 2 [ ] 5¾ ZK/2 [L + 3],並輸出至框 化運算單元930。其中,滾邊振幅、/2[1/ + 2]與、/2[1 + 3] 之值分別為0.5與0.25。之後,框化運算單元930依序將 〇.5S[L+2]加上之前儲存於緩衝單元936的〇 5S[2]後,輸 出至後端處理,並且將〇.25S[L+3]加上之前儲存於緩衝單 元936的0.75S[3]後,輸出至後端處理,以讓後端之裝置 還原出原始之信號。 在之後的操作中,滾邊振幅產生單元920仍會繼續接 收到通道估測模組所提供之通道長度,並依據保護區間與 通道長度的差值,決定縮減取樣因數D,以調整框化函數 之滾邊。因此,依照上述的操作方式,框化裝置9〇〇能夠 適時地調整框化函數之滾邊,以避免擷取符元上的信號 時’造成符元間的干擾,也可以有效地降低接收信號在通 道中的窄頻帶干擾。 綜上所述’本發明根據通道長度與保護區間的差 值,來調整樞化之滾邊。當通道長度很長,使得上述差值 很小時’便縮小滾邊長度,以避免擷取到相鄰的符元。當 通道長度很短’使得上述差值很大時,便放大滚邊長度, 以有效率地降低接收信號的窄頻帶干擾。因此,本發明能 夠防止符元間干擾。另外,當本發明應用於多載波系統, 1379561 AFA-P06001 21493twf.doc/n 例如:正交分頻多工系統,本發明更能避免子載波間干擾 (Inter-Carrier Interference,ICI)。 雖然本發明已以較佳實施例揭露如上,然其並非用 以限定本發明,任何熟習此技藝者,在不脫離本發明之精 神和範圍内,當可作些許之更動與潤飾,因此本發明之= 護範圍當視後附之申請專利範圍所界定者為準。 μ 【圖式簡單說明】 圖1繪示為在多重路徑的通道環境下接收信號的示音 圖。 /、 圖2繪示為多種型態之框化函數。 圖3纟會不為本發明實施例之框化函數。 圖4繪示為本發明第一實施例之框化裝置方塊圖。 圖5、纟會示為本發明第一實施例之框化方法的步驟流程 圖。 圖6繪示為本發明第二實施例之調整框化裝置方塊 圖。 圖7繪示為本發明實施例之調整框化函數。 圖8繪示為本發明第二實施例之調整框化的方法步驟 流程圖。 圖9繪示為本發明第三實施例之框化裝置方塊圖。 【主要元件符號說明】 =通道長度29 AFA-P06001 21493twf.doc/n 0.25S[1] of 936, and 〇.75S[L+1] is added to (^^(1) and output to the backend processing to restore the device at the back end. The original signal is output. Then, according to the above-mentioned hardware operation, the edge amplitude generating unit 92 will sequentially generate the frame factor 7〇5 and . /2 [L + 3], and output to the frame operation unit 930. The values of the edge amplitude, /2[1/ + 2], and /2[1 + 3] are 0.5 and 0.25, respectively. The arithmetic unit 930 sequentially adds 〇.5S[L+2] to the 〇5S[2] previously stored in the buffer unit 936, outputs it to the backend processing, and adds 〇.25S[L+3] before After being stored in 0.75S[3] of buffer unit 936, the output is output to the back end processing to allow the device at the back end to restore the original signal. In the subsequent operation, the edge amplitude generating unit 920 will continue to receive the channel estimation. The length of the channel provided by the module, and according to the difference between the protection interval and the length of the channel, the sampling factor D is determined to be adjusted to adjust the edge of the framed function. Therefore, according to the above operation mode, The framed device 9〇〇 can adjust the edge of the framed function in time to avoid the interference between the symbols when the signal on the symbol is captured, and can also effectively reduce the narrow band interference of the received signal in the channel. According to the above description, the invention adjusts the pivoting edge according to the difference between the length of the channel and the protection interval. When the length of the channel is long, the difference is small, the length of the edge is reduced to avoid the adjacent symbol. When the channel length is short 'to make the above difference large, the edge length is enlarged to effectively reduce the narrow band interference of the received signal. Therefore, the present invention can prevent inter-symbol interference. In addition, the present invention Applied to multi-carrier systems, 1379561 AFA-P06001 21493twf.doc/n For example: orthogonal frequency division multiplexing system, the present invention can more effectively avoid Inter-Carrier Interference (ICI). Although the present invention has been better The embodiments are disclosed above, but are not intended to limit the invention, and those skilled in the art can make some modifications and refinements without departing from the spirit and scope of the invention. The scope of protection = the scope of the patent application is subject to the definition of the patent application scope. μ [Simple diagram of the diagram] Figure 1 shows the diagram of the signal received in the channel environment of multi-path. /, Figure 2 3 is a block diagram of the embodiment of the present invention. FIG. 4 is a block diagram of a framed device according to a first embodiment of the present invention. FIG. 6 is a block diagram of an apparatus for adjusting a frame according to a second embodiment of the present invention. FIG. 7 illustrates an adjustment boxing function according to an embodiment of the present invention. FIG. 8 is a flow chart showing the steps of the method for adjusting the frame according to the second embodiment of the present invention. FIG. 9 is a block diagram of a framer according to a third embodiment of the present invention. [Main component symbol description] = channel length

31 1379561 AFA-P06001 2I493twf.doc/a31 1379561 AFA-P06001 2I493twf.doc/a

Tav:有效擷取區間 GI :保護區間 L:符元長度 110、120 :信號 130 :框化函數 K、K’ :滚邊長度 301〜314、701〜706 :框化係數 S[n]:取樣信號 400、900框化裝置 600 :調整框化的裝置 410、610、910 :框化係數儲存單元 420、620、920 :滾邊振幅產生單元 422、922 :第一加法單元 425、925 :多工單元 430、630、930 :框化運算單元 432、632、932 :乘法單元 434、934 :第二加法單元 436、636、936 :緩衝單元 438、938 :第三加法單元 638 :加法單元 S510〜S570 :本發明第—實施例之框化方法的各步驟 S810-S880 :本發明第二實施例之調整框化方法的各 步驟 32Tav: effective capture interval GI: guard interval L: symbol length 110, 120: signal 130: boxing function K, K': edge length 301~314, 701~706: boxing coefficient S[n]: sampling Signal 400, 900 framing device 600: framing device 410, 610, 910: framing coefficient storage unit 420, 620, 920: rim amplitude generating unit 422, 922: first adding unit 425, 925: multiplex unit 430, 630, 930: boxing operation units 432, 632, 932: multiplication units 434, 934: second addition units 436, 636, 936: buffer units 438, 938: third addition unit 638: addition units S510 to S570: Each step S810-S880 of the framing method of the first embodiment of the present invention: each step 32 of the framing method of the second embodiment of the present invention

Claims (1)

13795611379561 101年5月18曰修正替換頁 十、申請專利範圍: 1. 一種調整框化的方法,包括下列步驟: 偵測一通道長度; 找出該通道長度與一保護區間的一差值;以及 定義多個框化係數; 依據該差值,決定一縮減取樣因數; 框化係數中’每間隔該縮減取樣随,留下對 應的框化係數,以得到一調整框化函數;以及 由該調整框化函數,對接收訊號進行框化。 中定i Hr專鄉圍第1項所述之輕框化的方法, 中疋義一框化函數,表示如下: 其 RAn], C. ρ Γτ K^n&lt;L y^C + Z-n], L + l&lt;n^L + K_l η 其他 滾邊:長度,數二、二為實數’ ”為取樣時間’尤為 化係數、其對應第時為:常數,〜w為第,框 邊對應之取樣時料度樣時間’射滾邊長度k義為滾 3.如申清專利範圍 亥縮減取樣因數表 ^ =調整框化的方法’其 整框化函數,表示如τ: 為―正整數,定義-調 wK[n] 33 1379561 101年5月18日修正替換頁 RADn], C, RK[D(K/D + L^n)h l&lt;n&lt;K/D-l K / D &lt;n&lt; L L + l&lt;n&lt;L + K/D-l 其他 上述〜/J«]為第《取樣時間的調整框化係數。 4. 一種框化裝置’利用一框化函數對一接收訊號進行 ,化運算’其中雜化函數具有—第__侧之框化係數與一 第二側之框化係數,而該第一側之框化係數與該第二側之 鲁 框化係數中各自包括多個第一半邊框化係數與多個第二半 邊框化係數,該框化裝置包括: 一框化係數儲存單元,用以儲存該第一側之框化係數 中之該些第一半邊框化係數; 一滚邊振幅產生單元,藉由該些第一半邊框化係數作 一互補運异,產生該第一側之框化係數中之該些第二半邊 框化係數以及該第二側之框化係數;以及 一框化運算單元,利用該些第一側之框化係數中之該 鲁些第二半邊框化係數以及該滾邊振幅產生單元所產生的框 化係數,對該接收訊號進行框化。 5. 如申請專利範圍第4項所述之框化裝置,其中定義 一滾邊長度表示為尤,定義一取樣時間表示為《,上述《、 足為自然數,該滾邊長度尤表示有X個單位取樣時間,而 在第π時間的框化係數表示為%[«],而該滾邊振幅產生單 元包括: 一第一加法單元,搞接於該框化係數儲存單元,其 34 1379561May 18, 2011 Correction Replacement Page 10, Patent Application Scope: 1. A method of adjusting the frame, including the following steps: detecting a channel length; finding a difference between the channel length and a guard interval; and defining a plurality of framed coefficients; determining a reduced sampling factor according to the difference; in the boxing coefficient, the reduced sampling is performed every interval, leaving a corresponding framed coefficient to obtain an adjustment frame function; and the adjustment frame is The function is to frame the received signal. The light boxing method described in Item 1 of Zhongding i Hr Township, the Zhongyiyi boxing function, is expressed as follows: Its RAn], C. ρ Γτ K^n&lt;L y^C + Zn], L + l&lt;n^L + K_l η Other piping: length, number two, two is real number ' ” is the sampling time 'especially the coefficient, the corresponding time is: constant, ~w is the first, the sampling time corresponding to the frame edge The sample time 'shooting edge length k is the roll 3. If Shen Qing patent scope cuts down the sampling factor table ^ = adjusts the boxing method' its entire boxing function, such as τ: is "positive integer, definition - adjust wK [n] 33 1379561 Modified the replacement page RADn], C, RK[D(K/D + L^n)h l&lt;n&lt;K/Dl K / D &lt;n&lt; LL + l&lt;n&lt;L + K/Dl The other above ~/J«] is the "adjustment frame factor of the sampling time. 4. A framed device" uses a boxing function to perform a received signal, which is a hybrid function. Having a frame factor of a __ side and a frame factor of a second side, and the frame factor of the first side and the squaring coefficient of the second side each include a plurality of first half frame coefficients a plurality of second half frame coding factors, the frame formation device includes: a frame factor storage unit for storing the first half frameization coefficients of the frame coefficients of the first side; And performing, by the first half frame coding coefficients, a complementary operation, generating the second half frameization coefficients of the first side frame coefficients and the frame coding coefficient of the second side; and The operation unit framing the received signal by using the second half frameization coefficient of the frame coefficients of the first side and the frame factor generated by the edge amplitude generating unit. The framing device of the fourth aspect, wherein a length of the flange is defined as a definition, and a sampling time is defined as “, the above is a natural number, and the length of the edge particularly indicates that there are X unit sampling times, and The boxing coefficient of the πth time is expressed as %[«], and the edge amplitude generating unit comprises: a first adding unit, which is connected to the boxing coefficient storage unit, 34 1379561 101年5月18曰修正替換頁 中,該框化係數儲存單元儲存的該些第一半邊框化係數表 示為/2,該框化係數儲存單元依照取樣時 間”輸出一待定框化係數,表示為 ΧΚ [^]= wKlnl n = l,...,K/2 ^ΛΚ-nl n = K/2 +1,..., Κ -1 該第一加法單元將該特定框化係數〜[„]與一常數相減, 得到八[n] = C-χ&gt;]’其中,該常數表示為c,c為實數; 夕工單元,耦接至該第一加法單元與該框化係數儲 2元,並在每個取樣時間„,選擇輸出心㈤或咖 為第”取樣時間之滾邊振幅係數乂[„],其中, 6·如申請專利範圍第 5亥框化函數’表示如下: η=Λ”..,ΚΙ2 ”=尤 / 2 +1,···,足 一 1 〇 5項所述之框化褒置,其中定義 w尺[η] Rk^&gt; \&lt;n&lt;K/2 €~^ΛΚ~η], K/2 + l&lt;n&lt;K-l C, K&lt;n&lt;LWK[^ + L~n], L + l&lt;n&lt;i + x_i 〇, 其他 上述Γ二气度,,為第,樣時間的框化係數。 化運算單^包月括利乾圍第6項所述之框化裝置,其中該樞 第乘法單几,具有一第一輸入端、-第二輸入端 35 1379561 101年5月18日修正替換頁 與一輸出端’其第一輸入端接收一接收信號’表示為*, 其第二輸入端耦接至該多工單元,用以依據取樣時間w, 將該接收信號巩〃]乘以滚邊振幅係數〜[w]後輸出,表示為 zK[n]-S[n]; 一第二加法單元,具有一第一輸入端、一第二輸入端 與一輸出端,其第一輸入端接收該接收信號*,其第二 輸入端接收〜,用以依據取樣時間w,將該接收信 號*減去kW.sW後輸出,表示為(1-々W).%]; • 一緩衝單元’用以儲存該第二加法單元所輸出之 (1 一〜|&gt;]).%];以及 一第二加法單元,具有一第一輸入端、一第二輸入端 與一輸出端’用以當取樣時間《介於Z + 1S«幺Z +欠—1, 其第一輸入瑞接收,其第二輸入端讀取該緩衝單 元内所儲存之0 —,而其輸出端輸出 〜[«].风《]與(1 -〜h -1]).-幻之總和。 8. 如申請專利範圍第4項所述之框化裝置,其中該滾 籲邊振幅產生單元用以接收一通道長度,並計算該通道長度 與一保護區間之一差值,並依據該差值決定一縮滅取樣因 數。 7 9. 如申請專利範圍第8項所述之框化裝置,其中該縮 減取樣因數表示為乃= 2*’,另外定義一滾邊長度表^為 尺’定義一取樣時間表示為《,上述/、《、尺為自1數^ 該滚邊長度&amp;表示有尺個單位時間,而在第《時間框化 係數表示為%[«],而該滾邊振幅產生單元包括: 36 1379561 1〇1年5月18日修正替換頁 二第一加法單元,耦接於該框化係數儲存單元,其 二C存單^儲存的該些第—半邊框化係數 ,,違框化係數儲存卓元依照取樣時間?2於 出一特定框化係數,表示為 則 ΧΚ!〇[η\ wKWn], n = 1,..., A:/D/2 ^K[K~Dn], n = K/D/2 + \y...,K/D~\ 该第一加法單元將該特定框化係數〜㈨盥一 ^以得到、W=c-W^t’該倾表^數^目 L為實數;以及 六夕工單元耦接至該第一加法單元與該框化係數儲 存單元’並在每個取樣時間”,選擇輸出心⑺Μ或%//)[«]以 作為第《取樣時間之滾邊振幅係數,其中, ΖΚ!〇[η^ x kid ln], n = 1,...,^/2^/2 χκι〇[γι], n = KIDIl + \,..”KID-\。 10.如申請專利範圍第9項所述之框化裝置,其中該 框化函數經過該縮減取樣因數公調整後,可得一調整框化 函數’表示如下: ^αγΡ«], 1&lt;η&lt;Κ/2 C~RK[K-Dn], K/2 + \&lt;n&lt;K-\ Λ1; KID [n] = &quot;lC, K&lt;n&lt;L wkid[^ID + L~n], L + \&lt;n&lt;L + K-l 0, 其他 上述L為符元長度’ 表示第”取樣時間的框化係數。 U·如申請專利範圍第1〇項所述之框化裝置,其中該 37 1379561 框化運算單元包括: ιοί 年 5 月18 曰修正替換頁 第一乘法單元 • / X n j ^hii / \ ir — »· ».« 第〜輪入端 其第二輸人端_至該多卫單元,f咖], 將該接收信號啊乘以滾邊振幅係數於之間” ’ 為‘W·啊 ㈣ά ^ 一第二加法單元,具有一第一輸入端、— 矛一輸入墙In the correction replacement page of May 18, 101, the first half frameization coefficients stored in the frame factor storage unit are represented as /2, and the frame factor storage unit outputs a pending frame factor according to the sampling time. For ΧΚ [^]= wKlnl n = l,...,K/2 ^ΛΚ-nl n = K/2 +1,..., Κ -1 The first addition unit singulates the specific framing coefficient~[ „] subtracts from a constant to obtain eight [n] = C-χ&gt;]', where the constant is expressed as c, c is a real number; the evening unit is coupled to the first adding unit and the boxing coefficient is stored 2 yuan, and at each sampling time „, select the output heart (5) or the coffee is the “sampling time of the rolling edge amplitude coefficient 乂[„], where 6·such as the patent scope 5th boxing function' is expressed as follows: η =Λ"..,ΚΙ2 ”=尤/ 2 +1,···, the framed device described in item 1 〇5, where w rule [η] Rk^&gt;\&lt;n&lt;K /2 €~^ΛΚ~η], K/2 + l&lt;n&lt;Kl C, K&lt;n&lt;LWK[^ + L~n], L + l&lt;n&lt;i + x_i 〇, other above Γ2 ,, is the boxing coefficient of the first, sample time. The framed device of the sixth aspect, wherein the pivot multiplication method has a first input end, a second input end 35 1379561, a modified replacement page and an output end of the first input end of May 18, 101 Receiving a received signal 'represented as *, the second input end is coupled to the multiplex unit for multiplying the received signal by the sampling time w, multiplied by the edge amplitude coefficient 〜[w], and outputting zK[n]-S[n]; a second adding unit having a first input end, a second input end and an output end, the first input end receiving the received signal*, and the second input end thereof Receiving ~ for outputting the received signal * minus kW.sW according to the sampling time w, expressed as (1-々W).%]; • a buffer unit 'for storing the output of the second adding unit (1~~|&gt;]).%]; and a second adding unit having a first input terminal, a second input terminal and an output terminal 'for sampling time "between Z + 1S«幺Z + owed -1, its first input is received, its second input reads 0_ stored in the buffer unit, and its output is output 〜[«].Wind"] and (1 -~h -1)). - The framing device of claim 4, wherein the rolling edge amplitude generating unit is used for Receiving a channel length, and calculating a difference between the channel length and a guard interval, and determining a shrink sampling factor according to the difference. 7 9. The framing device of claim 8, wherein the downsampling factor is expressed as = 2*', and a roll length table is defined, and a sampling time is defined as "," , ", ruler is from 1 number ^ The length of the piping &amp; indicates that there is a unit time, and in the "Time Boxing Coefficient is expressed as %[«], and the edge amplitude generating unit includes: 36 1379561 1〇1 On May 18th, the first addition unit of the replacement page 2 is modified, coupled to the framed coefficient storage unit, and the first half-framed coefficients stored in the second C memory list are stored, and the frame-breaking coefficient is stored according to the sampling. Time? 2 is a specific boxing coefficient, expressed as ΧΚ!〇[η\ wKWn], n = 1,..., A:/D/2 ^K[K~Dn], n = K/D /2 + \y..., K/D~\ The first addition unit takes the specific boxing coefficient ~(9) 盥一^ to obtain, W=cW^t' the table of the table is the real number; And the hexagram unit is coupled to the first adding unit and the framing coefficient storage unit 'and at each sampling time", and the output core (7) Μ or %//) [«] is selected as the first sampling time Amplitude coefficients, wherein, ΖΚ! Square [η ^ x kid ln], n = 1, ..., ^ / 2 ^ / 2 χκι〇 [γι], n = KIDIl + \, .. "KID- \. 10. The framing device of claim 9, wherein the framing function is adjusted after the downsampling factor, and an adjustment boxing function 'is expressed as follows: ^αγΡ«], 1&lt;η&lt; Κ/2 C~RK[K-Dn], K/2 + \&lt;n&lt;K-\ Λ1; KID [n] = &quot;lC, K&lt;n&lt;L wkid[^ID + L~n], L + \&lt;n&lt;L + Kl 0, the other above L is a symbolized length of the symbol length ' representing the "sampling time". U. The framing device according to the first aspect of the patent application, wherein the 37 1379561 The boxed arithmetic unit includes: ιοί年月18 曰Fixed replacement page first multiplication unit• / X nj ^hii / \ ir — »· ».« The first round of the second input end _ to the end Guard unit, f coffee], multiply the received signal by the edge amplitude coefficient between "' is 'W·ah (four) ά ^ a second addition unit, having a first input, - spear input wall 與一輸出端,其第一輸入端接收該接收信號&lt;s[«],其第% 輸入端接收^dIX^o],用以依據取樣時間《,將兮接^ 信號減去Z《/£&gt;W,SW後輸出’表示為. 一緩衝單元,用以儲存該第二加法單元所輸出之 (1-〜/〇1^]).取];以及 一第三加法單元,具有一第一輸入端、一第二輸 入端與一輸出端,用以當取樣時間”介於 z + 厶+尤/Z) — 1,其第一輸入端接收2&amp;/£)[&quot;].^1”],其And an output terminal, the first input end receives the received signal &lt;s[«], and the first input end thereof receives ^dIX^o], according to the sampling time ", the splicing ^ signal is subtracted from Z" / £&gt;W, after SW output 'represented as. A buffer unit for storing (1-~/〇1^)).] and a third addition unit having a The first input end, the second input end and an output end are used when the sampling time is "between z + 厶 + especially / Z" - 1, and the first input receives 2 &amp; / £) [&quot;]. ^1"], its 具有一第一輸入端、 表示 第二輸入端讀取該緩衝單元内之(1_zw卜一邱_咖-1], 而其輸出端輸出々/士]·*^]與 (1 一 2請|&gt; - L]). - L] Wf/jD〇 -1] · - []之總和。 38Having a first input end, indicating that the second input end reads the buffer unit (1_zw Bu Yi Qiu_Caf-1), and its output terminal outputs 々/士]·*^] and (1~2 please | &gt; - L]). - L] Wf/jD〇-1] · - [] sum. 38
TW96146574A 2007-12-06 2007-12-06 A method to adjust ofdm-symbol window functions and a windowing apparatus to use this method TWI379561B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW96146574A TWI379561B (en) 2007-12-06 2007-12-06 A method to adjust ofdm-symbol window functions and a windowing apparatus to use this method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW96146574A TWI379561B (en) 2007-12-06 2007-12-06 A method to adjust ofdm-symbol window functions and a windowing apparatus to use this method

Publications (2)

Publication Number Publication Date
TW200926706A TW200926706A (en) 2009-06-16
TWI379561B true TWI379561B (en) 2012-12-11

Family

ID=44729773

Family Applications (1)

Application Number Title Priority Date Filing Date
TW96146574A TWI379561B (en) 2007-12-06 2007-12-06 A method to adjust ofdm-symbol window functions and a windowing apparatus to use this method

Country Status (1)

Country Link
TW (1) TWI379561B (en)

Also Published As

Publication number Publication date
TW200926706A (en) 2009-06-16

Similar Documents

Publication Publication Date Title
CN102111369B (en) Inter-subcarrier interference cancellation device and method
JP5281078B2 (en) Reception device, integrated circuit, digital television receiver, reception method, and reception program
CN100574303C (en) Method for synchronizing signal timing and synchronization circuit
EP0903898A2 (en) Equalizing method and equalizer for OFDM receiver
TW201041347A (en) Channel estimation apparatus and method thereof
CN101467412A (en) Signal detection in multicarrier communication system
TW200934189A (en) Methods and apparatus for estimating the channel impulse response
WO2010139234A1 (en) Method and device for estimating maximum doppler frequency offset
JP5896795B2 (en) Equalizer, receiver, and equalization method
JP2005102169A (en) OFDM receiving apparatus and OFDM receiving method
JP4659840B2 (en) Communication apparatus and channel estimation method
WO2009056044A1 (en) Peak value cancellation method, peak value cancellation device and reference peak value cancellation signal producing device
JP5347203B2 (en) Method and apparatus for estimating delay spread of multipath channel
US8385438B1 (en) System and method for adaptive synchronization
JP2010535452A (en) Method and apparatus for receiver synchronization
TWI379561B (en) A method to adjust ofdm-symbol window functions and a windowing apparatus to use this method
JP5349206B2 (en) Intercarrier interference canceling apparatus and intercarrier interference canceling method
JP4610401B2 (en) Channel estimation device
US9369329B1 (en) Low-complexity non-data-aided estimation of symbol time offset in OFDM systems
JP4125729B2 (en) Time synchronization method for communication system based on orthogonal frequency division multiplexing
JP7681533B2 (en) Wireless communication device and wireless communication method
EP1387544A2 (en) Synchronisation in multicarrier receivers
JP2007151097A (en) Delay profile analysis circuit and apparatus using the same
CN105450573B (en) Frequency deviation detection method and device
JP5473751B2 (en) OFDM signal combining receiver

Legal Events

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