TWI360974B - A synchronization method for ofdm systems - Google Patents

A synchronization method for ofdm systems Download PDF

Info

Publication number
TWI360974B
TWI360974B TW97105531A TW97105531A TWI360974B TW I360974 B TWI360974 B TW I360974B TW 97105531 A TW97105531 A TW 97105531A TW 97105531 A TW97105531 A TW 97105531A TW I360974 B TWI360974 B TW I360974B
Authority
TW
Taiwan
Prior art keywords
signal
interference
communication system
channel
frequency division
Prior art date
Application number
TW97105531A
Other languages
Chinese (zh)
Other versions
TW200937899A (en
Inventor
Wen Long Chin
Sau Gee Chen
Original Assignee
Univ Nat Chiao Tung
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 Univ Nat Chiao Tung filed Critical Univ Nat Chiao Tung
Priority to TW97105531A priority Critical patent/TWI360974B/en
Publication of TW200937899A publication Critical patent/TW200937899A/en
Application granted granted Critical
Publication of TWI360974B publication Critical patent/TWI360974B/en

Links

Landscapes

  • Synchronisation In Digital Transmission Systems (AREA)
  • Noise Elimination (AREA)

Description

1360974 九、發明說明: 【發明所屬之技術領域】 本發明係有關一種應用在通訊系統中的訊號同步方法,特別 是一種應用在正交分頻多工(OFDM)系統中之訊號同步方法。 【先前技術】 正乂分頻多工(Orthogonal Frequency Division Multiplexing, OFDM)是一種高效率的多通道調變解調變技術,對於多通道衰減具 有很強的抑制能力。正交分頻多工(〇FDM)目前被許多通訊標準所 採用,如DVB-T,DAB, xDSL,以802.11X為標準之無線區域網路 (WLAN)、以及以802_16x為標準之固定或移動式man系統,最 近之趨勢為3G以上之行動通訊系統幾乎都採用正交分頻多工 (OFDM)方式傳輸資料’其將可使用的頻寬被劃分為多個狹窄的頻 帶,資料就可以被平行的在這些頻帶上傳輸,然而其缺點為容易產生 同步的錯誤。 在 J. J. van de Beek ' M. Sandell 與 P. 〇_ Borjesson 所發表之“ML estimation of time and frequency offset in OFDM systems;5 (IEEE Trans. Signal Process·,vol.45, no.7)’提出符元及頻率偏移同時用延遲相關 (delayed-correlation )演算法估測,它是最大近似(maximum likelihood)估測,但只在可加性白色高斯雜訊通道(AWGN)的環境 下有較好的效能。 就像其他通訊系統一樣,在正交分頻多工系統亦有一些同步處 理議題需要考量。首先,未知之訊號延遲會造成符元時間偏移(symb〇1 time offset, STO )而需要粗略符元時間(coarse Symb〇i time, CST )及 細微符元時間(fine symbol time,FST)同步化’其在發射機與接收機 也存有載波頻偏的問通’以致於分數載波頻率偏移(fracti〇nalcarrjer 5 1360974 frequency offset, FCFO)、整數載波頻率偏移carrjer g>eqUenCy offset, ICFO)及剩餘載波頻率偏移(resjduai carrier fj*equenCy 〇ffset, - RCF0)必需消除;另外’類比-數位轉換器與數位·類比轉換器的取 樣時脈不一致也會造成取樣時脈頻率偏移(sampling cl〇ck fyeqUency offset, SCFO )。 在 Τ· M. Schmidl 與 D. C_ Cox 所發表之“Robust frequency and timing synchronization for OFDM,”(IEEE Trans. Commun” vol.45, no.12),提出一個利用時域的訓練符元,可用在靜態多重路徑的方法, 但還是存在一些不確定的區域。 在 H. Minn、V_ K. Bhargava 與 K. B. Letaief 所提之“A robust timing and frequency synchronization for OFDM systems;5 (IEEE Trans. Wireless Commun., vol_2, no.4),對上述的缺點提出了解決方案,但多 出的訓練符元會浪費系統資源,而且這些方法會找到最強的多重路 徑’而非第一個多重路徑’故不適用在細微符元時間估測(fme symb〇1 time estimation)。1360974 IX. Description of the Invention: [Technical Field] The present invention relates to a signal synchronization method applied in a communication system, and more particularly to a signal synchronization method applied in an orthogonal frequency division multiplexing (OFDM) system. [Prior Art] Orthogonal Frequency Division Multiplexing (OFDM) is a high-efficiency multi-channel modulation and demodulation technology with strong suppression capability for multi-channel attenuation. Orthogonal Frequency Division Multiplexing (〇FDM) is currently used by many communication standards such as DVB-T, DAB, xDSL, 802.11X-based wireless local area network (WLAN), and 802_16x-based fixed or mobile The man system, the most recent mobile communication system with a trend of more than 3G, uses Orthogonal Frequency Division Multiplexing (OFDM) to transmit data. It divides the usable bandwidth into multiple narrow frequency bands, and the data can be Parallel transmission over these frequency bands, however, has the disadvantage of being prone to synchronization errors. "ML estimation of time and frequency offset in OFDM systems; 5 (IEEE Trans. Signal Process·, vol.45, no.7)' proposed by JJ van de Beek ' M. Sandell and P. 〇 _ Borjesson The element and frequency offset are simultaneously estimated by the delayed-correlation algorithm, which is the maximum likelihood estimate, but is better only in the environment of additive white Gaussian noise channel (AWGN). As with other communication systems, there are some synchronization issues to be considered in orthogonal frequency division multiplexing systems. First, the unknown signal delay will cause symbol time offset (STO). Synchronic symbol time (CST) and fine symbol time (FST) synchronization are required. 'There is also a carrier frequency offset between the transmitter and the receiver' so that the fractional carrier Frequency offset (fracti〇nalcarrjer 5 1360974 frequency offset, FCFO), integer carrier frequency offset carrjer g>eqUenCy offset, ICFO) and remaining carrier frequency offset (resjduai carrier fj*equenCy 〇ffset, - RCF 0) Must be eliminated; in addition, the sampling clock inconsistency between the analog-to-digital converter and the digital-to-analog converter will also cause sampling cl〇ck fyeqUency offset (SCFO). In Τ·M. Schmidl and D. C_ Cox's "Robust frequency and timing synchronization for OFDM," (IEEE Trans. Commun" vol. 45, no. 12), proposes a training symbol using time domain, which can be used in the static multipath method. But there are still some uncertain areas. "A robust timing and frequency synchronization for OFDM systems; 5 (IEEE Trans. Wireless Commun., vol_2, no. 4), proposed by H. Minn, V_K. Bhargava and KB Letaief, proposes a solution to the above disadvantages. However, the extra training symbols will waste system resources, and these methods will find the strongest multipath 'not the first multipath' and therefore not applicable to fme symb〇1 time estimation.

雖然 K. Ramasubramanian 與 K_ Baum,所提之“An OFDM timing recovery scheme with inherent delay-spread estimation,,5 (Proc. Φ IEEE GLOBECOM’Ol. vol.5, pp. 3111-3115, Nov. 2001 ),在多路徑衰 減通道中可以辨認無交互符號干擾(ISI)之區域,但是要求精確之 ST估測時所使用之符元過多。 在 M. Speth,S_ Fechtel, G_ Fock,及 η. Meyer 所提之 “Optimum receiver design for OFDM-based broadband transmission-part II: a case study,5, (ffiEE Trans. Commun., vol.49, no.4, pp. 571-578, Apr. 2001 )-Although K. Ramasubramanian and K_Baum, the "An OFDM timing recovery scheme with inherent delay-spread estimation,, 5 (Proc. Φ IEEE GLOBECOM'Ol. vol. 5, pp. 3111-3115, Nov. 2001), Areas without inter-symbol interference (ISI) can be identified in the multipath fading channel, but too many symbols are required for accurate ST estimation. In M. Speth, S_Fechtel, G_Fock, and η. Meyer "Optimum receiver design for OFDM-based broadband transmission-part II: a case study, 5, (ffiEE Trans. Commun., vol.49, no.4, pp. 571-578, Apr. 2001)-

- 通道頻率響應(channel frequency response, CFR)必須先預測,接著 以反快速傅利葉轉換(IFFT)求得通道時脈響應(chaimd impulse response, CIR),其再被用於調整符元邊界,處理相當複雜D 6 【實施方式】 符元時間(SymbolTime)估測是整體正交分頻多工(〇FDM) 同步處理的第-階段’其提供後序階段的預估正交分頻k符元邊界 值’請參閱第1圖為-正交分頻多X通訊系統之三個連續的符元時間 時域示意圖,三個連續的符元為(/_;)th Symb〇卜(/)th Symb〇1及(/+7)也- The channel frequency response (CFR) must be predicted first, followed by the inverse fast Fourier transform (IFFT) to obtain the chaimd impulse response (CIR), which is then used to adjust the symbol boundary, which is fairly Complex D 6 [Embodiment] Symbol Time estimation is the first stage of the overall orthogonal frequency division multiplexing (〇FDM) synchronization process. It provides the estimated orthogonal frequency division symbol boundary of the subsequent stage. Value 'Please refer to Figure 1 for three consecutive symbol time-domain diagrams of the orthogonal frequency division multi-X communication system. Three consecutive symbols are (/_;)th Symb〇(/)th Symb 〇1 and (/+7) also

Symbd ’每一符元開始為一保護區間(GI),預估的符元時間通常落 在二個定義區域的其中一個,即圖中的ST1、ST2及ST3,壞的符元 時間誤差區域ST1及好的符元時間誤差區域ST3是在正交分頻多工 符儿之保護區間(GI)内(ng),通道的最大延遲擴展為Τί/。在⑺th Symbol中,付元時間誤差區域ST3只有通道效應而不會有交互符號 干擾(isi)現象,然而壞的符元時間誤差區域ST1與壞的符元時間 誤差區域ST2會分別受到(/_/沖Symbol及(/+/)th Symbol的碼間干擾 (inter-symbol interference, ISI)。因此,在壞的符元時間做粗略符元 時間(coarse symbol time, CST)估測後,接續以細微符元時間(fine symbol time, FST)估測用於細微調校〇FDM符元邊界,可避免交互 符號干擾(ISI )並使訊號對干擾加雜訊比 (Signal-to-Interference-and-Noise Ratio, SINR)達到最佳化。 第2圖所示為本發明一實施例之正交分頻多工(ofdm ) 通訊系統架構示意圖。於本實施例中,基頻帶訊息位元流(速率 為F=l/T)從資料源20經過訊號映射器21進入N點反向快速 傅立葉轉換22 ’串並聯轉換後分別送到n個子通道上,對N 個子通道上之訊號進行N點反向快速傅立葉轉換(IFFT) 22 處理實現正交調變’經由數位類比轉換器24轉換成為連續波 形,載於射頻上(升頻)放大發送,射頻中心載波頻率為/c, 為消除符元間之干擾,所以在循環前置插入單元23置入循環前 置(cyclic prefix,CP)。 接收端經由通道26接收到 調’即將信號從射頻帶移到基頻讯號後先進行射頻解 對類比信號取樣量化,經由類比數位=頻„),然後以週期τ 訊號,在循環前置移除單 、咨36轉換成為數位 (PPT) 3^;/^ ^ 各個子通道訊息流_、糾^及訊號解映射器32對 據流信號,最後傳2資串聯變換為原先的串行數 轉換訊比估測器38,連接類比數位 (OFDM)系統t的符 精出在正父分頻多工 正值,碁餘^ ’…、政正值及载波頻率偏移的誤差校 降頻計i 校正值分別傳送至循環前置移除單元35及 〜於上述實施例中,^為發射端第Η固子頻帶的頻域資料, 々為接收端第灸個子頻帶的頻域資料,气”是第/個符元中第” 個被傳輸的時間域取樣,I是包含循環前置(⑻的接收端資料, ,#奸頻道的數目,印)是接收到之連續時間訊 號是快速傅立葉轉換(FFT)的大小,馬=#+%是〇17〇]^ 系統中的符元長度,其包含循環前置(CP);厂八是取樣頻率, Λ是載波頻率,tfy是載波頻偏(CFO )經由子載波空間正規化 所產生’ ^Δ,Λ^/Λ«是估測的最大訊號對干擾加雜訊比(MSINR)符 元時間偏移(STO)’ g/滅靡是估測的最大訊號對干擾加雜訊比 (MSINR)載波頻偏(CFO)。另外,在發射端中#個複雜資料符元 是由#點反向快速傅立葉轉換22調變至W個子載波,最後一個符 元的iVG所形成的反向快速傅立葉轉換取樣會被複製成為循環前置 (CP),其會插入每一個OFDM符元的開始’因為置入循環前置(cp) 而形成一保護區間(guardinterval)’可盡量避免交互符號干擾(ISI) 現象及保持子載波間的正交關係,於接收端可使用快速傅立葉轉 換解調所枚到之訊號。 1360974 在第1圖中符元時間(ST) «Λ可能落在OFDM符元的三個誤 差區域的其中之一,在壞的ST1區域範圍為 +1 '好的 ST3 區域範圍為 -A/。+〜+1<«△幺〇、另外一個壞的ST2區域範圍為S7V-1。 符元時間偏移(STO)視為與理想符元時間的偏移量,理想符元時間 作記時間座標為〇,位於圖中ΑΑ’虛線’在此三個符元誤差區域内任 一點所接收的頻域資料訊號如式(1)所不: X ,,k = ^!,k + ................................. (1) 其中,无A為正確的資料訊號如式(2)所示:Symbd 'each symbol begins as a guard interval (GI). The estimated symbol time usually falls in one of the two defined areas, namely ST1, ST2 and ST3 in the figure, and the bad symbol time error area ST1. The good symbol time error region ST3 is within the guard interval (GI) of the orthogonal frequency division multiplex symbol, and the maximum delay of the channel is extended to Τί/. In (7)th Symbol, the pay element time error region ST3 has only channel effect without interaction symbol interference (isi) phenomenon, but the bad symbol time error region ST1 and the bad symbol time error region ST2 are respectively received (/_ Inter-symbol interference (ISI) of Symbol and (/+/)th Symbol. Therefore, after the coarse symbol time (CST) estimation is performed in the bad symbol time, the connection is continued. Fine symbol time (FST) estimation is used to fine tune the FDM symbol boundary to avoid inter-symbol interference (ISI) and signal-to-interference-and-noise ratio (Signal-to-Interference-and- Noise Ratio (SINR) is optimized. Fig. 2 is a schematic diagram showing the architecture of an orthogonal frequency division multiplexing (OFDM) communication system according to an embodiment of the present invention. In this embodiment, a baseband information bit stream (rate) For F=l/T) from the data source 20 through the signal mapper 21 to the N-point inverse fast Fourier transform 22 'serial-parallel conversion is sent to n sub-channels respectively, N-point reversal of the signals on the N sub-channels Fast Fourier Transform (IFFT) 22 processing to achieve quadrature modulation 'Converted into a continuous waveform by the digital analog converter 24, carried on the radio frequency (up-amplitude) to amplify and transmit, the RF center carrier frequency is /c, in order to eliminate the interference between the symbols, the insertion unit 23 is placed in the loop before the loop. The cyclic prefix (CP) is received by the receiving end via the channel 26. After the signal is moved from the radio frequency band to the base frequency signal, the radio frequency is sampled and quantized by the analog signal, and the analog digital bit = frequency „), and then The period τ signal, in the loop pre-removal, the 36 conversion to digital (PPT) 3 ^; / ^ ^ each sub-channel message stream _, correction and signal demapper 32 on the stream signal, the last transmission 2 The series conversion is the original serial number conversion ratio estimator 38, and the symbol of the analog-to-digital (OFDM) system t is selected in the positive-father frequency division multiplexing positive value, and the remaining ^ '..., the positive value and the carrier frequency The offset error dynamometer i correction value is respectively transmitted to the cyclic pre-removal unit 35 and in the above embodiment, where ^ is the frequency domain data of the Η Η 子 sub-band of the transmitting end, and 々 is the receiving end moxibustion The frequency domain data of the frequency band, the gas is the first / symbol The first time is sampled in the transmitted time domain, I is the size of the fast Fourier transform (FFT) that is included in the loop preamble ((8) of the receiver data, the number of the channel, printed) =#+% is 符17〇]^ The length of the symbol in the system, which includes the cyclic preamble (CP); the factory 8 is the sampling frequency, Λ is the carrier frequency, and tfy is the carrier frequency offset (CFO) through the subcarrier space. The resulting ^ Δ, Λ ^ / Λ « is the estimated maximum signal-to-interference plus noise ratio (MSINR) symbol time offset (STO) ' g / 靡 靡 is the estimated maximum signal to interference plus Signal ratio (MSINR) carrier frequency offset (CFO). In addition, the # complex data symbols in the transmitting end are modulated by the #点 inverse fast Fourier transform 22 to W subcarriers, and the inverse fast Fourier transform samples formed by the iVG of the last symbol are copied into the loop. Set (CP), which inserts the beginning of each OFDM symbol 'because of placing a loop preamble (cp) to form a guard interval (guardinterval)' to avoid inter-symbol interference (ISI) phenomena and to maintain inter-subcarriers In the orthogonal relationship, the received signal can be demodulated by the fast Fourier transform at the receiving end. 1360974 In Figure 1, the symbol time (ST) «Λ may fall within one of the three error regions of the OFDM symbol, and the range of the bad ST1 region is +1 'good ST3 region is -A/. +~+1<«△幺〇, another bad ST2 area range is S7V-1. The symbol time offset (STO) is regarded as the offset from the ideal symbol time. The ideal symbol time is recorded as the time coordinate 〇. In the figure, the 虚线 'dashed line' is at any point within the three symbol error regions. The received frequency domain data signal is not as shown in equation (1): X ,, k = ^!, k + .......................... ....... (1) Among them, no A is the correct data signal as shown in formula (2):

x{=YTwr^kxl<kw^^\^2) 1W w=〇 另外,斤*為干擾訊號加上可加性白色高斯雜訊通道(AWGN)訊號 v*,如式(3)所示: 是6_;2"/",且义=# + 乂是包括循環前置(CP)的長度。在 式(3)中的載波間的干擾(ICI)訊號為式(4):x{=YTwr^kxl<kw^^\^2) 1W w=〇 In addition, the kg* is the interference signal plus the additive white Gaussian noise channel (AWGN) signal v*, as shown in equation (3): It is 6_;2"/", and meaning=# + 乂 is the length including the loop preposition (CP). The inter-carrier interference (ICI) signal in equation (3) is equation (4):

N m*k n'=N~n& (4) 在式(3) t的符元間的干擾(ISI)訊號為式(5): Σ Knim-k+e/)Hmxl+^ (5) m ”'=N~n& ' 其中K為第m個子通道的通道頻率響應(channel Frequency Response, CFR )。 根據上述’正確的資料訊號所形成的功率如式(6)所示: (S ) 10 m\xf Ν· Ίν7 £ Hk 2 - (6) -1 Ν nl =〇 η2 =〇 且合併的干擾訊號與可加性白色高斯雜訊通道(awgn)訊號之功率 為: σ1 ~ E ΝN m*k n'=N~n& (4) The interference (ISI) signal between the symbols of equation (3) t is equation (5): Σ Knim-k+e/)Hmxl+^ (5) m "'=N~n& ' where K is the channel frequency response (CFR) of the mth subchannel. The power formed by the above 'correct data signal' is as shown in equation (6): (S) 10 m\xf Ν· Ίν7 £ Hk 2 - (6) -1 Ν nl =〇η2 =〇 and the power of the combined interference signal and the additive white Gaussian noise channel (awgn) signal is: σ1 ~ E Ν

Nk 111 ly2,k m\Y\i __ N-l N-l ,、 =Σ Σ ίχϋ—ε/) κ2 iV lm*k w, =0 «2=0 N-\ AM ( + Σ Σ %[,h^h Hk2 (7)Nk 111 ly2,km\Y\i __ Nl Nl ,, =Σ Σ ίχϋ—ε/) κ2 iV lm*kw, =0 «2=0 N-\ AM ( + Σ Σ %[,h^h Hk2 ( 7)

w, =//-wA n2=N^nA M~l yv-1 ( Λ + 2Σ Σ Σ W^~n^m-k+^ Hm2+a2 > m^k Wj =Α^-«δ n2~N~nL V* 其中’ 'r=£[|xu| ]=♦—!]、為傳送資料功率,< 為可加 性白色尚斯雜訊通道(AWGN)功率,SINR △,〜)的理論值是 咼度依靠符元時間偏移(STO)與載波頻偏(CFO)。 請參閲第3圖為本發明一實施例在正交分頻多工(〇FDM)通 訊系統中之訊號同步方法流㈣,包括:步驟S1接收通道内的頻域 資料訊號,分析頻域資料訊號如丨⑴所表示,並分析頻 域資料功率、干擾功率與AWGM功率,如式(6)與式(?)。 步驟S2計算全部子載波的平均職對干擾加雜概(sinr); 使用複高斯雜訊内嵌之二階與四階動量的包絡,可得出在多重路徑衰 減中真值訊號的SINR,對式(1)作統計運算產生式(8)、式(9) 與式(10): > (9)1360974w, =//-wA n2=N^nA M~l yv-1 ( Λ + 2Σ Σ Σ W^~n^m-k+^ Hm2+a2 > m^k Wj =Α^-«δ n2~ N~nL V* where ' 'r=£[|xu| ]=♦—!], for transmitting data power, < for additivity white Shangsi noise channel (AWGN) power, SINR △, ~) The theoretical value is that the temperature depends on the symbol time offset (STO) and the carrier frequency offset (CFO). Please refer to FIG. 3, which illustrates a signal synchronization method flow (4) in an orthogonal frequency division multiplexing (〇FDM) communication system according to an embodiment of the present invention, including: step S1 receiving a frequency domain data signal in a channel, and analyzing frequency domain data. The signal is represented by 丨(1) and analyzes the frequency domain data power, interference power and AWGM power, as in equations (6) and (?). Step S2 calculates the average job-to-interference plus sinr of all subcarriers; using the envelopes of the second and fourth orders of motion embedded in the complex Gaussian noise, the SINR of the true value signal in the multipath attenuation can be obtained, (1) For the statistical operation, the equations (8), (9) and (10): > (9) 1360974

Q 厶Q 厶

R ΓΣ ^ ΙΛ Ι + ΙΛ (10) 如此在第k個子載波的訊號功率&與干擾功私可由式⑴)與式⑴) 表示: ' (11)R ΓΣ ^ ΙΛ Ι + ΙΛ (10) Thus, the signal power & and interference function in the kth subcarrier can be expressed by equations (1) and (1)): ' (11)

Sk = (^l+Rk~Qk) / -5,, k k k’.............. (12) 在第k個子載波的估測SINR為式(13): = Sk / Ik. .............. (13) 所以平均全軒紐K得到整_ SINR,魏q如式(⑷ 1 …… 所示: η Κ (14) 可二= = 次平均,以此求== 號功率“關二彬,與干擾訊 ^~?Ιά;Ι.................. S: /: (16)Sk = (^l+Rk~Qk) / -5,, kk k'.............. (12) The estimated SINR at the kth subcarrier is Equation (13): = Sk / Ik. .............. (13) So the average Quanxuan New K gets the whole _ SINR, Wei q as the formula ((4) 1 ...... shows: η Κ (14) Can be === times average, so as to find == power "Guan Erbin, and interference signal ^~?Ιά;Ι.................. S: /: (16 )

式(16)中之總功率為使用兩個蒋& a ^ t /個符元的估測SINR7/'如式〇7)付凡的功率之幾何平均,則I A ‘..........二⑴)所示: 再對所有Z符元取平均值得到式( )王硝之SINR的簡化式:The total power in equation (16) is the geometric mean of the power of the estimated SINR7/'s equation (7) using two Chiang & a ^ t / symbols, then IA '... ....two (1)): Then average all the Z symbols to get the simplified formula of the SINR of the formula ( ):

12 S (18)1360974 上述之估測式只需要接收頻率域f料而不需先知道通道 與傳輸之資料。藉由對式(18)求取最大值以得^ 符元時間偏移(STO)與載波頻偏(CF〇),如式(19): (^^,MS1NR 5 ^f.MSlNR ) ~ 式斗⑼的二維搜尋問題非常複雜,在無12 S (18) 1360974 The above estimation method only needs to receive the frequency domain f material without first knowing the channel and transmission data. The maximum value of equation (18) is obtained to obtain the symbol time offset (STO) and the carrier frequency offset (CF〇), as in equation (19): (^^, MS1NR 5 ^f.MSlNR ) ~ The two-dimensional search problem of Dou (9) is very complicated, no

下,將式⑼分開成以個別解決獨立的符元時間偏移(s = 波頻偏(CFO)問題,如式(2〇)所示: 。戟Next, separate equation (9) to solve the independent symbol time offset (s = wave frequency offset (CFO) problem, as shown in equation (2〇): 戟

nAMSINR 8 fMSINR argmax{^K5 argC少 f M SINR )} MSIhfR 5 )^ ε (20)nAMSINR 8 fMSINR argmax{^K5 argC less f M SINR )} MSIhfR 5 )^ ε (20)

請看第4A圖所示為都會區中典型的具有29個路徑的通道中 SINR與在取樣雜巾_補,其^况,5腿在無符元 間的干擾(isi)的區域(·π至〇th之間)有一平垣的區域出現最大值。 雜的式(18)對SINR估測的取樣顯示圖如第4B圖所示,其所實 行的系統與第4A圖是相同的。第4B圖中可以看見式(18)對§驗 估測不如式U4)精確’細式⑽所產生的s脆輪廓依然清楚 的被保存,且第4B B的SINR在壞的ST1區域比第4A圖掉下的幅 度更明顯。式(18)的估測對所有符元作過求取平均值的動作,使 SINR輪廓可以保存在時變通道(time_variantchannei),如第5B圖戶斤 示,且圖中之正歸化都普勒頻率(N_ahzed ϋ〇_Γ Frequency5 NDp) 為〇·ι。另外,若是由式(14)的5臟估測式所計算之結果如第5A 圖所示,可得知式(14)並不適合在時變通道中使用。 第6A圖與第6B圖所顯示為式(14)與式(18)在各種CF〇值 中的SINIU古測,CF0的範圍為_1/2<^<1/2,圖中5聰的最大值 C S ) 13 出現在CFO=〇時。式(14)與式(18)在時變的通道令的騎 測如第7A圖與第7B圖所顯示,其NDF條件為0.1,第7B _ =)的在時變的通道t的讓估測是_,優於第J圖= 3圖中之步驟S31取樣符元之兩點’並產生此兩點之 ^對干擾加雜訊tb(s祖);為了減少搜尋整個範㈣所產生的計 算複雜度’、對況號於_Γ△及元"厕+'兩時點取樣, 元邊界的時間偏移量。 疋中 步驟S4用此兩點的訊號對干擾加雜訊比(SINR)相減,並過 ^出:錯誤訊號’使誤差訊號趨近於零;f 8圖中所顯示為本發明之 符元同步估,嗎構不意圖,將式〇8)在上述兩點之5臟相減並過 濾形成—錯誤訊m如果取樣的兩點^議_Γδ與 、卿Λ + γλ是分別在最佳ST駄邊與右邊並與其相狀距離相 等’其i是預估之符元時間偏移(ST0)的時間偏移量,錯誤訊號心(/) 等,0最彳i取;ιυ㈣是位在兩點之巾間;反之,在其它情況下錯誤 錢心0不等於0 ’則以早/遲閘(eai1y/late _,ELG )方式使誤差 。凡5虎趨近鱗。錯誤訊號eA(/) *等於G時錯誤訊號的定時朗器 之S曲線如式(21)所示: eM) - Vi(nAMSm-r^sfMSINR)~^{n^Msm ^τΑ,ε/Μ5ΙΝΚ\... (21) 其近似線彳峨型如帛9目所示’制於早/遲閑雜恢復迴路將誤差 值調整為0,其符號表示義意為: F{z) 迴圈過濾器; KF 迴圈過濾器增益; Κ1 時脈偵測器本質之增益 κν 壓控振盪器之增益; ΝΛζ) 的ζ轉換; ”Δ,Λ/Μν/?(7)的 Ζ 轉換; 1360974 Εδ(ζ) 心(/)的ζ轉換; Κ⑻ 過濾的錯誤訊息4(0的ζ轉換。 在第9圖中’範圍ηΔ的訊號(/)經過ζ轉換(z-transf〇rm) 後產生\(z)輸入至時脈/[貞測器(Time Detector) 71,並輸出_錯誤 訊號的Z轉換五Δ(ζ)至迴圈過濾器(LoopFilter) 79,、' & 瘦 圈過濾器72接收五△0)訊號產生五1(2)訊號,其為過濾後之錯誤气 號4(7)的Z轉換。在壓控振盪器(VC0) 73中,符元時間(ST)會 由式(22)调整:Please see Figure 4A for the SINR of a typical channel with 29 paths in the metropolitan area and the area of the interference (isi) between the sampled _ _ complement, its condition, and the 5 legs in the unsigned (·π Between 〇th) There is a flat area with a maximum value. The sampling display of the SINR estimation of the equation (18) is as shown in Fig. 4B, and the system implemented is the same as that of Fig. 4A. In Fig. 4B, it can be seen that equation (18) is not as good as equation U4. Accurate 'fine (10) is still clearly preserved, and the SINR of 4B B is in the bad ST1 area compared to 4A. The magnitude of the graph drop is more obvious. The estimation of equation (18) performs an average of all symbols, so that the SINR contour can be saved in the time-varying channel (time_variantchannei), as shown in Figure 5B, and the normalization of the map is The frequency (N_ahzed ϋ〇_Γ Frequency5 NDp) is 〇·ι. Further, if the result calculated by the 5-dirty estimation formula of the equation (14) is as shown in Fig. 5A, it can be understood that the equation (14) is not suitable for use in the time-varying channel. Fig. 6A and Fig. 6B show the SINIU ancient test of equations (14) and (18) in various CF values. The range of CF0 is _1/2 <^<1/2, The maximum value of CS ) 13 appears when CFO = 〇. The riding of the equations (14) and (18) in the time-varying channel is shown in Figures 7A and 7B, and the NDF condition is 0.1, and the 7B _ =) is estimated in the time-varying channel t. The measurement is _, which is better than the two points of the sampling symbol S3 in step S31 of the figure J3 and generates the interference of the two points plus the noise tb (szu); in order to reduce the search for the entire range (four) The calculation complexity ', the condition number is _Γ △ and the yuan " toilet + ' two points sampling, the time offset of the element boundary. In step S4, the two signals are subtracted from the interference plus noise ratio (SINR), and the error signal 'has the error signal close to zero; the symbol shown in FIG. 8 is the symbol of the present invention. Synchronous estimation, the structure is not intended, the formula 〇 8) in the above two points of the 5 dirty subtraction and filtering formed - error message m if the sampling of two points ^ _ Γ δ, qing Λ + γ λ are in the best ST The edge and the right side are equal to each other's distance. The i is the estimated time offset of the symbol time offset (ST0), the error signal heart (/), etc., 0 is the most 彳i; the ιυ(4) is in two In the case of the other, the error money center 0 is not equal to 0 ', and the error is made by the early/late gate (eai1y/late _, ELG). Where the 5 tigers approach the scales. Error signal eA(/) * The S curve of the error horn of the error signal equal to G is as shown in equation (21): eM) - Vi(nAMSm-r^sfMSINR)~^{n^Msm ^τΑ, ε/Μ5ΙΝΚ \... (21) The approximate line 彳峨 type is as shown in 帛9 mesh. The system adjusts the error value to 0 in the early/late idle loop. The sign indicates that the meaning is: F{z) KF loop filter gain; Κ1 clock detector essence gain κν voltage controlled oscillator gain; ΝΛζ) ζ conversion; Δ, Λ/Μν/?(7) Ζ conversion; 1360974 Εδ( ζ) Heart (/) ζ conversion; Κ (8) Filtered error message 4 (0 ζ conversion. In Figure 9 'range ηΔ signal (/) after ζ conversion (z-transf〇rm) produces \(z ) Input to the clock / [Time Detector 71], and output the Z signal of the _ error signal five Δ (ζ) to the loop filter (LoopFilter) 79, ' & thin circle filter 72 receives five △0) The signal generates five 1(2) signals, which is the Z-conversion of the filtered error number 4 (7). In the voltage controlled oscillator (VC0) 73, the symbol time (ST) will be expressed by the equation (22). )Adjustment:

^AMSINR (0 - ^AMSINR U ~ ^) + STS ...... ( 22 ) 最後輸出次0)訊號,其為运(/)訊號的Z轉換。 上述第9圖的迴路系統可以由式(23)所示: L(z)=-_ 1 + ζ~ι(ΚΓ-2)+ζ~2(1-Κτα)............... (23) 其中’尤;是時脈偵測器71本身的增益; 尺厂是過濾器72的增益; A是壓控振盪器(VCO) 73的增益; KT=KFKfKvNsTs。^AMSINR (0 - ^AMSINR U ~ ^) + STS (22) Finally output the 0) signal, which is the Z conversion of the transport (/) signal. The loop system of Fig. 9 above can be expressed by the equation (23): L(z)=-_ 1 + ζ~ι(ΚΓ-2)+ζ~2(1-Κτα)........ (23) where 'espeech is the gain of the clock detector 71 itself; the ruler is the gain of the filter 72; A is the gain of the voltage controlled oscillator (VCO) 73; KT=KFKfKvNsTs .

此系統在式(24)的條件下是穩定的: 〇<a<\ and 0 < < 4^+ β. (24) 經過上述調整後,最後於步驟S5將符元時間調整在最佳位置。 〇 在另一實施例中,於第3圖中之步驟S32將收到的頻域資料气 號乘以每一通道内之較小及較大之頻偏產生兩個訊號對干擾加雜訊 比(SINR);並於步驟S4用此兩個訊號對干擾加雜訊比(sinr)相 減,過濾出一錯誤訊號,使誤差訊號趨近於零;如第1〇圖中所顯示 為本發明之栽波頻率(CF)同步估測架構示意圖,其與第8圖的不同 處為在早刀支中頻率偏移f/ws/iw ~ ,在遲分支中頻率偏移 15 1360974 已 f WS1NR 七 了 f ’ 丁 f 是預估之載波頻偏(CFO)的頻率偏移量,其 錯誤訊號的頻率毁別器之S曲線如式(25)所示: 9 (’) _ 6 (只ΔΛ®·,^/邮/姻-) - ^ «JWMW,+ ~ ). ( 25 ) • 最後於步驟S5修正載波頻偏。 請參閱第11圖至第14圖以及第15圖至第18圖中所顯示為本 發明與其它估測法在各種條件情況下的效能比較,其所設定之環境參 數為一般都會區通道相似,一 OFDM系統N=256子通道,保護區間 Wg=AV8=32且只考慮資料子通道,正交相位變換鍵控(qpsk)調變, • 信號頻寬2.5 MHz及射頻為2.4 GHz,子載波範圍是8.68 kHz,符元 期間為115.2 /^。評估最大訊號對干擾加雜訊比估測器的效能是 用估測的正規化平均方差(mean-Squared error,MSE )。時脈恢復 迴路的參數設定如下:= 0.9997、& = 〇.3、正規化迴路頻寬 為,其中5是迴路頻寬並設為0.05,每一次模擬為1〇〇〇〇 個符元。 於符元估測實施時,本發明與A· J. AI-Dweik所提之最小干擾符 元估測(minimum-interference ST estimation)及[5]所提之時域中無 資料輔助符元估測(time-domain NDA ST estimation)之比較分別顯 • 示於第11圖與第12圖,第13圖為本發明之符元估測在各種二雜比 (SNR)情況下之效能比較,第14圖為本發明之符元估測在各種正 歸化都普勒頻率(NDF)情況下之效能比較;另外,於載波頻偏估測 實施時,本發明與A_ J. AI-Dweik所提之最小干擾符元估測 (minimum-interference ST estimation)之比較分別顯示於第 15 圖與 第16圖,第17圖為本發明之載波頻偏估測在各種訊雜比(SNR)情 ' 況下之效能比較,第18圖為本發明之載波頻偏估測在各種正歸化都 普勒頻率(NDF)情況下之效能比較。 上述模擬比較與分析中,本發明與A· j. A!_Dweik所提出的最小 干擾符元估測(mininrnm-imerferenceSTestimation)的比較,並分別 16 否有!2波,偏(CF〇)的情況作—完整之比較,圖中顯示不論是 裁波ϋ 的情況,本翻都可以有較佳的效能,且對於 載波頻偏(CFO)的影響較不敏感。 再者本發明之叙同步會朝向無符元間的干擾(⑸)的區域 ^集’且载波頻率同步侧會勤載波賴(cf〇)聚集最重要的 ^御_賴相她__ (雖)在低猶與 ^ 〇 兄代雜小H使縣發财極嚴苛的情況下也適This system is stable under the condition of equation (24): 〇 <a<\ and 0 << 4^+ β. (24) After the above adjustment, finally adjust the symbol time to the most in step S5. Good location. In another embodiment, the received frequency domain data gas number is multiplied by the smaller and larger frequency offset in each channel in step S32 in FIG. 3 to generate two signal-to-interference plus noise ratios. (SINR); and in step S4, the two signals are subtracted from the interference plus noise ratio (sinr), and an error signal is filtered to make the error signal approach zero; as shown in FIG. Schematic diagram of the carrier frequency (CF) synchronization estimation architecture, which differs from the figure 8 in the frequency offset f/ws/iw ~ in the early knives, and the frequency offset in the late branch 15 1360974 has f WS1NR VII f ' D is the estimated frequency offset of the carrier frequency offset (CFO), and the S curve of the frequency destroyer of the error signal is as shown in equation (25): 9 (') _ 6 (ΔΛ® only) ·, ^/mail/marriage-) - ^ «JWMW,+ ~ ). ( 25 ) • Finally, the carrier frequency offset is corrected in step S5. Please refer to Figures 11 to 14 and Figures 15 to 18 for comparison of the effectiveness of the present invention with other estimation methods under various conditions. The environmental parameters set are similar to those of the general metropolitan area. An OFDM system N=256 subchannels, guard interval Wg=AV8=32 and only consider data subchannels, quadrature phase shift keying (qpsk) modulation, • signal bandwidth 2.5 MHz and radio frequency 2.4 GHz, subcarrier range It is 8.68 kHz and the symbol period is 115.2 /^. The estimated maximum signal-to-interference plus noise ratio estimator is estimated using the mean-squared error (MSE). The parameters of the clock recovery loop are set as follows: = 0.9997, & = 〇.3, the normalized loop bandwidth is, where 5 is the loop bandwidth and is set to 0.05, and each simulation is 1 符 symbols. In the implementation of Fu Yuan's estimation, the present invention and A·J. AI-Dweik proposed minimum-interference ST estimation and [5] mentioned in the time domain without data-assisted symbol estimation The comparison of time-domain NDA ST estimation is shown in Fig. 11 and Fig. 12, respectively. Fig. 13 is a comparison of the performance of the symbol estimation of the present invention under various two-to-two ratio (SNR) conditions. Figure 14 is a comparison of the performance of the symbol estimates of the present invention in various positive normalized Doppler frequencies (NDF); in addition, in the implementation of carrier frequency offset estimation, the present invention and A_J. AI-Dweik The comparison of minimum-interference ST estimation is shown in Fig. 15 and Fig. 16, respectively, and Fig. 17 is the carrier frequency offset estimation of the present invention in various signal-to-noise ratio (SNR) conditions. The comparison of the performance below, the 18th figure is the performance comparison of the carrier frequency offset estimation of the present invention under various normalized normalized Doppler frequencies (NDF). In the above simulation comparison and analysis, the present invention compares with the minimum interference symbol estimation (mininrnm-imerferenceSTestimation) proposed by A·j. A!_Dweik, and 16 no respectively! 2 wave, partial (CF〇) case - complete comparison, the figure shows that whether it is the case of cutting, this turn can have better performance, and is less sensitive to the influence of carrier frequency offset (CFO) . Furthermore, the synchronization of the present invention will be directed toward the region of the interference between unsigned symbols ((5)) and the carrier frequency synchronization side will gather the most important _ _ _ _ _ _ In the case of low yue and ^ 〇 brother generation miscellaneous H, the county is also very strict

綜合上述,本發明針對正交分頻多工(〇FDM)系統中的符 明步_ H最大職對谓域賊(maxi_ SINR)之 方法,其_ Μ符元咖位於符元邊界有最大的s腿值,於無 輪資料的資訊情況下即可以解決同步的問題,所以本發明亦為種非 數據輔助(_-data-aided,NDA)演算法。且本發明可用於時變 時,多路徑衰落通道’且在低SNR、高CF〇值與高卿值情況下 有高的準確性。 ’In summary, the present invention is directed to a method for arbitrarily step-by-step thief (maxi_SINR) in a quadrature frequency division multiplexing (〇FDM) system, where the _ 元 元 咖 位于 位于 位于 位于 位于 位于 位于 位于 位于 元 元 元 元 元 元 元 元 元 元 元 元The s leg value can solve the synchronization problem under the information of the wheelless data, so the invention is also a non-data-aided (NDA) algorithm. Moreover, the present invention can be used for multipath fading channels when time varying, and has high accuracy at low SNR, high CF threshold and high value. ’

以上所述之實施例僅係為說明本發明之技術思想及特 點,其目的在使熟習此項技藝之人士能夠瞭解本發明之内容 並據以實施,當不能以之限定本發明之專利範圍,即大凡= 本發明所揭示之精神所作之均等變化或修飾,仍應涵 、 發明之專利範圍内。 本 【圖式簡單說明】 域示 第1圖所示為正交分頻多工通訊系統之三個連續的符元時 意圖。 第2圖所示為根據本發明一實施例之正交分頻多工通訊系絲扭 意圖。 11 '現架構示 17 (S ) 1360974 第3圖所“«本發明-實補在正交分·卫(〇fdm)通訊 系統中之訊號同步方法流程圖。 第4A圖與第4B圖所不為根據本發明在各種通情況下各取樣索 引的標示圖。 第5Α圖與第5Β圖所示為根據本發明在時變通道令於各種情 況下各取樣索引的標示圖。 第6=與第6Β圖所示為根據本發明在各種cf〇值情況下各取樣 索引的4示示圖。 第7A _第7B圖所示為根據本發明在時魏道巾於 情況下各取樣索引的標示圖。 ί υ值 第8騎4根據本發明—實施例之符元同步估啦構示意圖。 = 狀早/遲間時脈恢復迴路運算架構 意 圖所示為根據本發明-實施例之載波頻率同步估測架構示 斷谢轉其它估測法在各種舰條件情況 她曝結種卿條件情 【主要元件符號說明】 20 資料源 21 訊號映射器 22 N點反向快速傅立葉轉換 < S ) 18 1360974The embodiments described above are merely illustrative of the technical spirit and the features of the present invention, and the objects of the present invention can be understood by those skilled in the art, and the scope of the present invention cannot be limited thereto. That is, the equivalent changes or modifications made by the spirit of the present invention are still within the scope of the invention. [Simplified description of the diagram] Field diagram Figure 1 shows the three consecutive symbol time intentions of the orthogonal frequency division multiplexing communication system. Fig. 2 is a diagram showing the twisting intention of the orthogonal frequency division multiplexing communication system according to an embodiment of the present invention. 11 'present architecture shows 17 (S) 1360974 Figure 3 "The invention - the method of signal compensation in the Orthogonal Guardian (〇fdm) communication system. Figure 4A and Figure 4B It is a map of each sampling index in various cases according to the present invention. Fig. 5 and Fig. 5 are diagrams showing the index of each sampling index in various cases in a time varying channel according to the present invention. Figure 6 is a diagram showing four samples of each sample index in the case of various cf values according to the present invention. Fig. 7A - Fig. 7B are diagrams showing the index of each sample index in the case of a Weidao towel in accordance with the present invention. ί 第 第 骑 骑 根据 根据 根据 第 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = The architecture shows off the other estimation methods in various ship conditions. She is exposed to the conditions of the species [main component symbol description] 20 data source 21 signal mapper 22 N-point inverse fast Fourier transform < S ) 18 1360974

23 循環前置插入單元 24 數位類比轉換器 26 通道 30 資料區 32 訊號解映射器 33 等化器 34 N點快速傅立葉轉換 35 循環前置移除單元 36 類比數位轉換器 38 最大訊號對干擾加雜訊比估測器 71 時脈偵測器 72 迴圈過濾器 73 壓控振盪器 ST1、ST2、 ST3符元時間誤差區域 S1-S5 流程步驟 1923 Cycle pre-insertion unit 24 Digital analog converter 26 Channel 30 Data area 32 Signal demapper 33 Equalizer 34 N-point fast Fourier transform 35 Cycle pre-removal unit 36 Analog-to-digital converter 38 Maximum signal-to-interference plus miscellaneous Analog Estimator 71 Clock Detector 72 Loop Filter 73 Voltage Controlled Oscillator ST1, ST2, ST3 Symbol Time Error Region S1-S5 Flow Step 19

Claims (1)

1360974 ,. 一 ----- . 100年9月23日修正替換頁 : 十、申請專利範圍: • L一種正交分頻多工通訊系統中之訊號同步方法,包含下列步驟: 接收數個通道内的頻域資料訊號,分析該些頻域資料訊號;以及 找出該些頻域資料訊號的最大訊號對干擾加雜訊比,其步驟包含: 計算該些頻域資料訊號產生平均訊號對干擾加雜訊比; 產生兩個訊號對干擾加雜訊比(SINR);以及 將該兩個訊號對該干擾加雜訊比(SINR)相減,並過渡出 一誤差訊號,並使該誤差訊號趨近於零。 2. 如凊求項1所述之正交分頻多工通訊系統中之訊號同步方法,其中 產生該兩個訊號對干擾加雜訊比是由取樣每一該通道内之符元之兩點所 產生。 3. 如請求項丨所述之正交分頻多工通訊系統中之訊號同步方法,其中 產生該兩個訊號對干擾加雜訊比是由該些賴資料訊號乘以每—該通道 内之較小及較大之頻偏所產生。 ^ 4.如請求項1所述之正交分頻多工通訊系統中之訊號同步方法,其中 該頻域資料訊號包含資料訊號、干擾訊號以及可加性白色高斯雜訊通道 訊號,該頻域資料訊號之表示公式為: 心=I ’其中z△為正確的資料訊號,t為干擾訊號 加上可加性白色高斯雜訊通道(AWGN)訊號。 5·如請求項4所述之正父分頻多工通訊系統中之訊號同步方法,其中 該貢料訊號之表^公式為· 20 1360974 2丨〜…為—Λ),其中阶泛 « =〇 w 為 β - / 2 ;r / " ’ 為第W個子通道的通道頻率響應,X,,,為發射端第女 個子頻帶的頻域資料。 6. 如請求項4所述之正交分頻多工通訊系統中之訊號同步方法,其中 該干擾Λ號加上該可加性白色高斯雜訊通道訊號之表示公式為: m^kn =〇 1 上,1360974,. 一----- . September 23, 100 revised replacement page: X. Patent application scope: • L A signal synchronization method in orthogonal frequency division multiplexing communication system, including the following steps: Receive several Frequency domain data signals in the channel, analyzing the frequency domain data signals; and finding the maximum signal-to-interference plus noise ratio of the frequency domain data signals, the steps comprising: calculating the frequency domain data signals to generate an average signal pair Interference plus noise ratio; generating two signal-to-interference plus noise ratio (SINR); and subtracting the interference plus noise ratio (SINR) between the two signals, and transitioning an error signal and making the error The signal approaches zero. 2. The method for synchronizing signals in an orthogonal frequency division multiplexing communication system according to claim 1, wherein generating the two signals to interfere with the noise ratio is by sampling two points of each symbol in the channel Produced. 3. The method for synchronizing signals in an orthogonal frequency division multiplexing communication system as described in claim 1, wherein generating the two signals to interfere with the noise ratio is multiplied by each of the data signals in each channel Smaller and larger frequency offsets are generated. 4. The signal synchronization method in the orthogonal frequency division multiplexing communication system according to claim 1, wherein the frequency domain data signal comprises a data signal, an interference signal, and an additivity white Gaussian noise channel signal, the frequency domain The representation of the data signal is: heart = I 'where z △ is the correct data signal, t is the interference signal plus the additive white Gaussian noise channel (AWGN) signal. 5. The signal synchronization method in the positive-family crossover multiplex communication system according to claim 4, wherein the formula of the tribute signal is · 20 1360974 2丨~... is -Λ), wherein the order is «= 〇w is β - / 2 ; r / " ' is the channel frequency response of the Wth subchannel, X,,, is the frequency domain data of the female subband of the transmitting end. 6. The signal synchronization method in the orthogonal frequency division multiplexing communication system according to claim 4, wherein the interference apostrophe plus the additive white Gaussian noise channel signal is expressed by: m^kn = 〇 1 on, 其中外為可加性白色高斯雜訊通道(AWGN)訊號,^為第所個子通 道的通道頻率響應,為發射端第w個子頻帶的頻域資料,尤^為符 元間的干擾(ISI)訊號,又巧為載波間的干擾訊號。 7. 如請求項6所述之正交分頻多工通訊系統中之訊號同步方法,其中 該載波間的干擾訊號之表示公式為:The external one is an add-on white Gaussian noise channel (AWGN) signal, and ^ is the channel frequency response of the first sub-channel, which is the frequency domain data of the w-th sub-band of the transmitting end, especially the inter-symbol interference (ISI) signal. It is also an interference signal between carriers. 7. The signal synchronization method in the orthogonal frequency division multiplexing communication system according to claim 6, wherein the expression of the interference signal between the carriers is: m φΙ( η '=Ν ~η^ 該符元間的干優訊號之表示公式為: :以及 m ny~N~^^ 8.如請求項4所述之正交分頻多工通訊系統中之訊號同步方法,其中 分析該些頻域資料訊號更包括將該資料訊號及該干擾訊號加上該可加性 白色高斯雜訊通道訊號轉換成功率形式,資料訊號之功率為:m φΙ( η '=Ν ~η^ The expression of the good signal between the symbols is: : and m ny~N~^^ 8. In the orthogonal frequency division multiplexing communication system described in claim 4 The signal synchronization method, wherein analyzing the frequency domain data signals further comprises converting the data signal and the interference signal plus the additivity white Gaussian noise channel signal into a power form, wherein the power of the data signal is: 21 1360974 σΗ = E 所 |V|2 △-1 "-Ή古Σ Σ軋:“)" Ν Η ;以及 «I =0 «2=0 «•玄干擾§fl號加上該可加性白色雜訊通道訊號之功率為 σ 2 Ε Ν i,k N N 2 m2 Ν-ϊ N. Σ Σ Σ w少 m^k /ij =0 w2 =〇 //, + Σ Σ w: ηλ =M ~n& n2^N-n^ (w> ~η2)ε 2Σ Σ Yj H. + σ H\=N — rt& tt2=hi —η&”中,為傳送資料功率,' 為可加性白色高斯雜訊通道(规⑽)功 率’心與/^分別為第m個與第&子通道的通道頻率響應。9·如請求項i所述之正交分頻多項訊系統中之訊號同步方法,其中 計算該些頻域資料訊號包含: 對每一子載波估測訊號對干擾加雜訊比;以及 平均該些子紐之訊賴干擾加雜賊,產钱平均峨對干擾知 比0 雜tfl 10.如請求項9所述之正交分頻多工通訊系統中之訊號同步方法, 中對每一子載波估測訊號對干擾加雜訊比之步驟為: SI = -ί— V jr γ * 7 Λ I ,k Λ 1+1^ 其 22 丄湖97421 1360974 σΗ = E where |V|2 △-1 "-Ή古Σ Rolling: ")" Ν Η ; and «I =0 «2=0 «• 玄 § §fl plus the plus The power of the white noise channel signal is σ 2 Ε Ν i,k NN 2 m2 Ν-ϊ N. Σ Σ Σ w less m^k /ij =0 w2 =〇//, + Σ Σ w: ηλ =M ~n& n2^Nn^ (w> ~η2)ε 2Σ Σ Yj H. + σ H\=N — rt& tt2=hi —η&", for transmitting data power, 'is additive white Gaussian The channel (regulation (10)) power 'heart and /^ are the channel frequency response of the mth and the & subchannel respectively. 9. The method of signal synchronization in an orthogonal frequency division multiple frequency system as claimed in claim i, wherein calculating the frequency domain data signals comprises: estimating a noise-to-interference plus noise ratio for each subcarrier; and averaging the The signal of the sub-news is the interference of the thief, the average cost of the production is less than the interference. 0. The signal synchronization method in the orthogonal frequency division multiplexing communication system described in claim 9, for each sub- The steps of the carrier estimation signal to the interference plus noise ratio are: SI = -ί— V jr γ * 7 Λ I , k Λ 1+1^ 22 丄湖974 :以及 U·如請求項9所述之正交分頻多卫通訊系統中之訊號同步方法其 中平均該些子紐之城對干擾加雜訊比之絲公式為: η,^Σ”; ·。 厶ιAnd U. The signal synchronization method in the orthogonal frequency division multi-wei communication system according to claim 9, wherein the average of the sub-news city-to-interference plus noise ratio formula is: η, ^Σ";厶ι 12·如4求項u所述之正交分頻多卫通訊系統中之訊號同步方法,其 中更包括求取最大值以制-符元_偏移與1波頻偏,其求取最大 值之表示公式為: ^amsinr = argmax {ηΊηΑ F 、1 J Ά、 Α 5 6 fMSlNR )/ ^/msinr = arg max {η\ηκ 〇、1. L n&,ef ^ V ^MSlNRi^f)> 其中G是載波頻偏(CF0)經由子栽波空間正規化所產纟,似是符 元時間。12. The signal synchronization method in the orthogonal frequency division multi-wei communication system according to the fourth item, wherein the method further comprises: obtaining a maximum value to perform a symbol-offset and a wave frequency offset, wherein the maximum value is obtained. The expression is: ^amsinr = argmax {ηΊηΑ F , 1 J Ά, Α 5 6 fMSlNR ) / ^/msinr = arg max {η\ηκ 〇, 1. L n&,ef ^ V ^MSlNRi^f)&gt Where G is the carrier frequency offset (CF0) produced by the normalization of the sub-carrier space, which seems to be the symbol time. 13.如請求項12所述之正交分頻多工通訊系統中之訊號同步方法,其 中該誤差訊號具有一定時甄別器之s曲線,其表示式為: 以’)-6(¾麵-;△,&顺)-伽△觸讀)其中。 是符元邊界的時間偏移量。 Η.如請求項!2所述之正交分頻多工通訊系統中之訊號同步方法,其 中該誤差訊號具有一頻率甄別器之S曲線,其表示式為: ’=伽一,Wr ")-伽一,心婦厕,)·其中 ^ /是載波頻偏的頻率偏移量。 (S ) 23 1360974 15. 如請求項1所述之正交分頻多工通訊系統中之訊號同步方法,其 中使該誤差訊號趨近於零是使用時脈恢復迴路。 16. 如請求項15所述之正交分頻多工通訊系統中之訊號同步方法,其 中該時脈恢復迴路為早/遲閘時脈恢復迴路,其表示公式為: Κτζ-\\-αζ~ι) L{Z)= 1 + Ζ^(Κτ-2)+ζ-\ΐ-Κτα),其中心為全部之增益。13. The signal synchronization method in the orthogonal frequency division multiplexing communication system according to claim 12, wherein the error signal has a s curve of a time discriminator, and the expression is: ')-6 (3⁄4 plane - ; △, & 顺) - 伽 △ touch)). Is the time offset of the symbol boundary. Η. As requested! 2 The signal synchronization method in the orthogonal frequency division multiplexing communication system, wherein the error signal has an S curve of a frequency discriminator, and the expression is: '=gamma, Wr ")-gamma, heart The toilet,)· where ^ / is the frequency offset of the carrier frequency offset. (S) 23 1360974 15. The signal synchronization method in the orthogonal frequency division multiplexing communication system according to claim 1, wherein the error signal is brought close to zero by using a clock recovery loop. 16. The signal synchronization method in the orthogonal frequency division multiplexing communication system according to claim 15, wherein the clock recovery loop is an early/late gate clock recovery loop, and the expression is: Κτζ-\\-αζ ~ι) L{Z)= 1 + Ζ^(Κτ-2)+ζ-\ΐ-Κτα), the center of which is the total gain. 24twenty four
TW97105531A 2008-02-18 2008-02-18 A synchronization method for ofdm systems TWI360974B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW97105531A TWI360974B (en) 2008-02-18 2008-02-18 A synchronization method for ofdm systems

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW97105531A TWI360974B (en) 2008-02-18 2008-02-18 A synchronization method for ofdm systems

Publications (2)

Publication Number Publication Date
TW200937899A TW200937899A (en) 2009-09-01
TWI360974B true TWI360974B (en) 2012-03-21

Family

ID=44867156

Family Applications (1)

Application Number Title Priority Date Filing Date
TW97105531A TWI360974B (en) 2008-02-18 2008-02-18 A synchronization method for ofdm systems

Country Status (1)

Country Link
TW (1) TWI360974B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8385387B2 (en) * 2010-05-20 2013-02-26 Harris Corporation Time dependent equalization of frequency domain spread orthogonal frequency division multiplexing using decision feedback equalization
US9860861B2 (en) 2016-04-20 2018-01-02 Hong Kong Applied Science And Technology Research Timing offset estimation in an OFDM-based system by SINR measurement
WO2017181352A1 (en) * 2016-04-20 2017-10-26 Hong Kong Applied Science and Technology Research Institute Company Limited Timing offset estimation in an ofdm-based system by sinr measurement
CN112285688A (en) * 2019-07-22 2021-01-29 财团法人工业技术研究院 Signal sensing system and method

Also Published As

Publication number Publication date
TW200937899A (en) 2009-09-01

Similar Documents

Publication Publication Date Title
JP4159030B2 (en) Timing synchronization method for wireless networks using OFDM
TWI407719B (en) A symbol time synchronization method for ofdm systems
TW201101760A (en) ID-cell index search algorithm and carrier frequency offset estimation
TWI360974B (en) A synchronization method for ofdm systems
Zhang et al. Clustered pilot tones for carrier frequency offset estimation in OFDM systems
US9369329B1 (en) Low-complexity non-data-aided estimation of symbol time offset in OFDM systems
Zhang et al. Window function and interpolation algorithm for OFDM frequency-offset correction
You et al. Pilot-assisted fine frequency synchronization for OFDM-based DVB receivers
CN107018112B (en) NC-OFDM cognitive radio narrow-band interference resistant cognitive user synchronization method
Zhou et al. Successive interference cancellation-based weighted least-squares estimation of carrier and sampling frequency offsets for WLANs
Kung et al. Frequency domain symbol synchronization for OFDM systems
Chin et al. A low-complexity minimum-interference symbol time estimation for OFDM systems
Li et al. A novel time and frequency synchronization algorithm for LTE system
Yin et al. Low-Complexity Fine Frequency Synchronization for MB-OFDM Based UWB Systems.
Kim et al. Residual frequency offset and phase compensation for OFDM systems
Dona et al. MMSE channel estimation using two-dimensional filtering in rapid time-variant environments
Sheng et al. Joint estimation of Doppler spread and carrier frequency offset for OFDM systems
Brada et al. Blind estimation of the carrier frequency offset in OFDM systems
Fang et al. A frequency offset estimation algorithm for OFDM systems
Li et al. Frequency offset Estimation for OFDM systems using ICI self-cancellation schemes
CN105743836B (en) Ofdm system frequency deviation estimating method based on multi-path separation
Prema et al. Joint Timing, Carrier Frequency and Sampling Clock Offset Estimation for MIMO OFDM WLAN Systems
KR101259164B1 (en) Method and apparatus of estimating frequency offset based on local optimum estimation in wireless communication system
KR101275692B1 (en) Method and apparatus of estimating frequency offset based od training symbols in wireless communication system
Mu et al. A new carrier frequency synchronization scheme for OFDMA downlink of WiMAX