577197577197
發明之背景 發明之界定 本發明係針對-些通訊系統。詳言之,本發明係針對 一些無線通訊系統内之接收機。更明確地說,本發明係針 對此等無線通訊系統内之接收機有關的自動增益控制系 統。 ' 相關技藝之說明 一些接收機在類似無線電和行動電話通訊系統等無 線電系統内之用途,在本技藝中係眾所周知。第1圖係顯示 一典型之超外差式接收機設計1 〇。在此,一射頻(RF)信號, 係在天線15上面被接收,以及將會提供給一 RF放大器2〇。 此RF信號,將會被其RF放大器20放大,以及會在一混頻器 2 5内’與一來自一本地振盡器3 〇之信號相混合。此將會產 生一中頻(IF)信號,其將會在一 iF放大器35内被放大,以 及會在一帶通濾波器40内被濾波。此濾波之if信號,將會 再次在一 IF放大器35内被放大,以及會在一帶通濾波器4〇 内,被一IF放大器45放大,以及會在一乘積檢波器50内, 與一來自一拍頻振盪器55之信號相混合。此結果係一被一 帶通放大器60放大及被一類比數位(A/D)轉換器65數位化 以供進一步處理之信號。 在此等接收機中,較強之信號係需要較小之放大器增 益,以及一極強信號,不被放大至當被放大時會使所接收 之資訊信號失真、會使系統組件超載、以及可能會損害到 此等組件的程度,係很重要的。基於此一理由,彼等接收 4 (請先閲讀背面之注意事項再填寫本頁) -訂— 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 577197 A7 B7 五、發明説明(2 ) "" — (請先閲讀背面之注意事項再填寫本頁) 機典型地係具有某種自動增益控制(AGC)系統,其可控制 或多之系統放大器20、35、45、和60,以使此等放大之 信號,維持在某一定之範圍内(此控制舉例而言,可能是經 由一靶加至该等放大器之偏壓)。在第i圖中,其AGC單元 70,可接收其IF放大器45所輸出之iF輸入,以及可使用它 來產生一些可控制該等尺卩放大器2〇和IF放大器35和45之 偏壓信號。 發明之概要 本發明之一貫施例,提供了一無線接收機有關之自動 增益控制系統,其可快速地區別所希望之頻帶内信號,與 一些重疊進其目標頻帶内之高功率頻帶外信號。此系統可 在一接收之信號,經過若干基本上可限制該信號之功率至 其頻帶内者的濾波器之前及之後,測量其功率。藉由比較 其接收之信號經濾波後的頻帶内能量,與其濾波前之總能 量,其將有可能決定是否有一新的頻帶内信號已來臨。此 新的頻帶内信號之出現,接著會以其常態化自關聯之多臨 界值來做確認,以確認一新的所希望之頻帶内信號的存在。 圖示之簡單説明 本發明之一實施例的此等和其他特徵,將可藉由閱讀 下文配合所附諸圖對其較佳實施例之詳細說明,而有更明 確之瞭解,其中: 第1圖係顯示一依據先存技藝之通訊接收機的結構; 第2圖係顯示一依據本發明之一實施例的通訊接收機 之結構; 本紙張尺度適用中國國豕標準(CNS) A4規格(210X297公爱) 、發明説明( 編Γ圖係、顯示—依據本發明之—實施例的自動增益控 制機構之結構; f 4和5圖係顯示—增純制之信號的希望特性; 圖係顯示此實施例中A/D轉接器飽和之特性;而 第7A 7B、8 A、和8B圖,則係顯示彼等頻帶内和頻 帶外信號之特性。 兒明 之 運 g 本發明之-實施例的接收機之基本結構,係顯示在第 中在此其一 I頻帶天線115,可接收一射頻(RF) RF L #u以及可將其提供給—RF放大器12〇,以及此頻帶内 之特疋頻道或信號,較佳地係藉由改變該等本地振盪器 130和180來做選擇。在此實施例中,該尺1?信號較佳地係遵 照IEEE 802.1 1 la標準,具有一在5 GHz頻帶内之頻率,以 及係做正交調變,藉以傳送6至54 Mbps。在此一實施例中, 该仏號可傳送高至54 Mbps之資料,以及係位於十二個2〇 MHz中之一内,其中有八個在5.15-5.3 5 GHz頻帶内,以及 有四個在5.75 - 5.85 GHz頻帶内。此一實施例中之信號, 係一使用52個間隔312.5 kHz之副載波做編碼的正交分頻 多工化信號(OFDM)。然而,理應瞭解的是,雖然以下之詳 細說明,係在一 IEEE 802.1 1 1a標準之背景中完成,本說明 書所說明之本發明,係可應用至許多不同類型之通訊系 統’以及並非受限於一些運作於此IEEE 802.1 1 a標準中 系統。舉例而言,誠如下文之說明,本發明在說明上係 作於一 IEEE 802.1 la系統中之短和長的教練碼元(trainin 6 (請先閲讀背面之注意事項再填寫本頁) 本紙張尺度適用中國國家標準(CNS) A4規格(210 X 297公釐) 577197 A7 -------PZ______ 五、發明説明(4 ) symbol)’但理應注意的是,其相關之教學,可被一般化為 任何由一或多正弦波所構成之教練序列。因此,舉例而言, 彼等功率之測量,可基於一存在於一包含多數正弦波之教 練碼元内的最慢頻率之正弦波的半個周期來完成,而每一 多數正弦波,係具有一為其最慢頻率之正弦波的整倍數之 頻率。在802.1 la系統中,此會轉譯成一短教練碼元序列之 一半。 上述放大之RF信號,會與一自一本&RF振盪器13〇供 應至一混頻器125之信號相混合,產生一中頻(IF)信號,而 被饋送至一 IF放大器135。該等本地RF振盪器13〇和本地IF 振盪器180之頻率和值,最好是在515-5 35和5.75_5 85 (5沿 之範圍内,而使其RF振盪器頻率對IF振盪器頻率之比率為 4:1。在此一實施例中,該等本地振盪器13〇和18〇,最好是 在一浮接IF之組態中,彼等在其中均為可變,而非為一固 定之IF組態,其中舉例而言,僅有本地RF振盪器13〇為可 變。 该等放大之IF信號,係分別供應至一同相混頻器 175-IP和一正交混頻器175_Q,此等同相混頻器175_ιρ和一 正交混頻器175-Q中的一個,係直接受到一本地汀振盪器 180之驅動,以及同相混頻器175-Ip和一正交混頻器i75_q t的另一個,則係在一相移器185中做9〇。之相移後,受到 上述本地IF振盪器之驅動,在此一方式中,上述接收之 信號的同相(IP)和正交(Q)成分,將會分別在該等同相混頻 器175-IP和正交混頻器175_q之輸出處得到。 本紙張尺度適用巾g g家標準(cns) A4規格(21〇\297公^〇 " ---------BACKGROUND OF THE INVENTION Definition of the Invention The present invention is directed to some communication systems. Specifically, the present invention is directed to receivers in some wireless communication systems. More specifically, the present invention is directed to an automatic gain control system for receivers in such wireless communication systems. '' Description of Related Techniques The use of some receivers in radio systems such as radio and mobile telephone communication systems is well known in the art. Figure 1 shows a typical superheterodyne receiver design. Here, a radio frequency (RF) signal is received on the antenna 15 and will be provided to an RF amplifier 20. This RF signal will be amplified by its RF amplifier 20, and will be mixed in a mixer 25 'with a signal from a local shaker 30. This will produce an intermediate frequency (IF) signal that will be amplified in an iF amplifier 35 and filtered in a band-pass filter 40. The filtered if signal will be amplified again in an IF amplifier 35, and will be amplified in an band pass filter 40, by an IF amplifier 45, and in a product detector 50, and one from one. The signals of the beat frequency oscillator 55 are mixed. The result is a signal amplified by a band-pass amplifier 60 and digitized by an analog-to-digital (A / D) converter 65 for further processing. In these receivers, a stronger signal requires a smaller amplifier gain, and an extremely strong signal, which is not amplified so that when it is amplified, it will distort the received information signal, overload system components, and possibly It is important that these components are damaged. For this reason, they received 4 (please read the notes on the back before filling out this page)-order — this paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 577197 A7 B7 V. Description of the invention ( 2) " " — (Please read the notes on the back before filling this page) The machine typically has some kind of automatic gain control (AGC) system, which can control or more system amplifiers 20, 35, 45, and 60, so that these amplified signals are maintained within a certain range (for example, this control may be a bias voltage applied to the amplifiers via a target). In Fig. I, its AGC unit 70 can receive the iF input output from its IF amplifier 45, and can use it to generate some bias signals that can control these size amplifiers 20 and IF amplifiers 35 and 45. SUMMARY OF THE INVENTION One embodiment of the present invention provides an automatic gain control system related to a wireless receiver, which can quickly distinguish signals in a desired frequency band from some high-power out-of-band signals that overlap into its target frequency band. This system can measure the power of a received signal before and after passing through a number of filters that substantially limit the power of the signal to those in its frequency band. By comparing the energy in the filtered frequency band of the received signal with the total energy before filtering, it will be possible to determine whether a new signal in the frequency band has arrived. The appearance of the signal in this new frequency band will then be confirmed with its multiple thresholds of normalized self-correlation to confirm the existence of a new signal in the desired frequency band. These and other features of one embodiment of the present invention are briefly illustrated in the drawings, and will be more clearly understood by reading the following detailed description of the preferred embodiment with the accompanying drawings, wherein: The figure shows the structure of a communication receiver according to the prior art. The second figure shows the structure of a communication receiver according to an embodiment of the present invention. The paper size is in accordance with China National Standard (CNS) A4 (210X297). (Public love), description of the invention (editing the graph system, showing the structure of the automatic gain control mechanism according to the embodiment of the present invention; f 4 and 5 graphs showing the desired characteristics of the signal of increasing pureness; the graph system shows this The characteristics of the saturation of the A / D adapter in the embodiment; and Figures 7A, 7B, 8 A, and 8B show the characteristics of their in-band and out-of-band signals. The basic structure of the receiver is shown in the first one of its I-band antenna 115, which can receive a radio frequency (RF) RF L #u and can be provided to-RF amplifier 12o, and special features in this band Channel or signal, preferably by The local oscillators 130 and 180 are changed to make a choice. In this embodiment, the ruler signal preferably conforms to the IEEE 802.1 1 la standard, has a frequency in the 5 GHz band, and is orthogonal. Modulation to transmit 6 to 54 Mbps. In this embodiment, the 仏 number can transmit data up to 54 Mbps, and is located in one of twelve 20 MHz, eight of which are at 5.15- 5.3 In the 5 GHz band, and four in the 5.75-5.85 GHz band. The signal in this embodiment is an orthogonal frequency division multiplexed signal (OFDM using 52 subcarriers spaced 312.5 kHz apart). However, it should be understood that although the following detailed description is completed in the context of an IEEE 802.1 1 1a standard, the invention described in this specification is applicable to many different types of communication systems' and is not subject to It is limited to some systems operating in this IEEE 802.1 1 a standard. For example, as described below, the present invention is described as a short and long coaching symbol (trainin 6 ( Please read the notes on the back first (Write this page) This paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) 577197 A7 ------- PZ______ 5. Description of the invention (4) symbol) 'But it should be noted that its Relevant teaching can be generalized to any coaching sequence composed of one or more sine waves. Therefore, for example, the measurement of their power can be based on a training symbol that exists in a coaching symbol that contains most sine waves The slowest frequency sine wave is completed in half a cycle, and each majority sine wave has a frequency that is an integral multiple of its slowest frequency. In 802.1la systems, this translates into half of a short coaching symbol sequence. The amplified RF signal is mixed with a signal supplied from a & RF oscillator 13 to a mixer 125 to generate an intermediate frequency (IF) signal, which is then fed to an IF amplifier 135. The frequency and value of the local RF oscillator 13 and the local IF oscillator 180 are preferably in the range of 515-5 35 and 5.75_5 85 (5 edges, so that the RF oscillator frequency is equal to the IF oscillator frequency. The ratio is 4: 1. In this embodiment, the local oscillators 13 and 18 are preferably in a floating IF configuration, where they are all variable instead of A fixed IF configuration where, for example, only the local RF oscillator 13 is variable. The amplified IF signals are supplied to a common-phase mixer 175-IP and a quadrature mixer, respectively. 175_Q, one of the equivalent phase mixer 175_ιρ and a quadrature mixer 175-Q, is directly driven by a local clock oscillator 180, and the in-phase mixer 175-Ip and a quadrature mixer The other of i75_q t is 90% in a phase shifter 185. After the phase shift, it is driven by the local IF oscillator. In this way, the received signal is in phase (IP) and positive The cross (Q) component will be obtained at the outputs of the equivalent phase mixer 175-IP and the quadrature mixer 175_q. Associate (cns) A4 size (21〇 \ 297 billion public ^ " ---------
:…卜-磬…: (請先閲讀背面之注意事項再填寫本頁) -訂, 五、發明説明(5 ) 該等混頻之IF信號,將會行經一些低通濾波器14〇_Ip 和140-Q,藉以選擇其所希望之頻道,以及移除彼等不相 關而在頻邊上退隔之成分,以及將會被兩個串聯之基帶放 大器145-IP和145-Q放大。雖然在每一支路中係顯示兩個基 帶放大器,不同數目之放大器係可加以使用。在一特定之 順序中,使用一些具有適當選定之可程式化增益的基帶放 大的成乎可開發出任何希望之基帶增益步驟的配置。 該等低通濾波器14(MP和140_Q,最好係一些具有一在 28 MHz下之3 dB交角的二極點之橢圓濾波器。自類比移至 數位域,該等基帶放大器輸出,係饋至彼等A/D轉換器 190-IP和190-q,彼等可數位化該等同相和正交成分之信 號,較佳地以一 80 MHz之頻率,而至一九位元之解析度: 和一-500 mV至500 mV之輸入動態範圍。 該等A/D轉換器,最好係一些管線A/D轉換器;然而, 本發明係不受限於此。舉例而言,△或其他轉換器,可 用以取代彼等。 類比頻道滤。皮為,矛口 /或防混疊渡;皮器,可有利地 放置在該等A/D轉換器190-IP*190_Q之前。在此較佳之實 施例中,料類比渡波器之組合,可執行4犯之相鄰阻斷 信號拒斥和20 dB之交替阻斷信號拒斥。在一 16犯較大之 相姊阻斷k號和32 dB較大之交替阻斷信號的最糟情況 下’ -在其A/D轉換器輸人處之接收阻斷信號,可較其頻 帶内信號高U dB。誠如本技藝中所熟知,—相鄰阻斷信 號’係-與其所關注之頻帶相鄰或相重疊的干擾信號,而 577197 A7 -------丨丨___B7 _ 五、發明説明(ό ) 一交替阻斷信號,則係一遠離其所關注之頻帶的干擾信號。 該等數位化之I/Q成分信號,係供應至一自動增益控制 (AGC)單兀170 ’其有關本發明之運作,將在下文有更詳細 之說明。此AGC 170,可如下文更詳細之說明,分析該等 ι/Q成分h號,以及可基於其產生一些增益控制信號。此等 增益控制信號,如第2圖中之虛線所顯示,係供應至該等放 大器 120、135、145-IP、和 145-Q。 更明確地說,誠如第3圖中所示,該等來自A/D轉換器 190-IP和190-Q之數位化巧信號,係行經一些泄漏桶式濾波 器245-IP和245-Q與有限脈沖嚮應(FIR)濾波器2〇5_Ip、 21CMP、和205-Q、和 210-Q。該等第一FIR2〇5_Ip 和 2〇5-Q, 係一些抽取濾波器,彼等可自彼等對應之資料流,消除每 隔一個之樣本,藉以就一正常8·5 MHz單邊帶頻寬數據 包’使其資料取樣率,自80 MHz減少至40 MHz。該等第 一尸11121〇-1?和21〇-(^,係一些標準低通濾波器,彼等可在 將資料傳送至其自關聯器225和一功率偵測器220之前,移 除任何殘餘之相鄰或交替之阻斷信號。在其Agc 170内, 將會做二次功率測量,一次自其第二FIR濾波器21〇之輸 出’由其功率偵測器220完成,另一次自其A/D轉換器190-IP 和190-Q之輸出,由另一功率偵測器215完成。此等測量如 下文較詳細之說明,係供應至其AGC控制邏輯電路230。 雖然此一實施例係使用數位FIR,其他類型包括類比 濾、波器等濾波器,可用以取代彼等。若此系統不會被過度 取樣’該等濾波器係較佳地為類比型。 本紙張尺度顧巾關家標準(⑽)M規格⑵⑽撕公爱) 9 (請先閲讀背面之注意事項再填寫本頁) 訂— 577197 A7 _____B7_ 五、發明説明(7 ) 其AGC控制邏輯電路230,可接收一些來自功率偵測 器215和220之功率測量,以及使用彼等來控制一增益控制 產生器235,藉以產生每一RF放大器120、IF放大器135、 和個別基帶放大器145-IP和145-Q有關之類比增益控制信 號。在此實施例中,該AGC控制邏輯電路230,可提供一 控制字組,在此較佳之實施例中為十位元長,給其增益控 制產生器235,以及此增益控制產生器235,可產生該等放 大器有關之適當控制信號。此等增益控制信號,將會回授 至該等RF放大器120、IF放大器135、和基帶放大器145-IP 和145-Q,藉以如上文所說明控制每一之增益。 此實施例係使用一專用積體電路,來具現該AGC控制 邏輯電路230 ;然而,一適當程式規劃之處理器,内欲式或 分立式’或其他適當之裝置,同樣可被使用。理應注意的 是,雖然第3圖係顯示該AGC 170内之各個組件,為彼此各 自分離’其兩個或多個單元,有可能被整合進一單元内。 舉例而a ’该AGC控制邏輯電路230,係顯示與該等fir 205、210、功率偵測器215、220、和自關聯單元225相分離; 然而,此等數個可使結合進一單一處理器内,而做適當程 式規劃,以執行此等功能。此外,一程式規劃之處理器, 並不需要被使用,以及一或多之此等組件,可被具現在專 用之硬體中。 該AGCn〇可控制一 DC抵補控制單元24〇,藉以提供一 些類比抵補控制信號,給一或多之基帶放大器ΐ45_ιρ和 145-Q。其DC抵補控制在完成上,可確保該等提供給放大 本紙張尺度適用中國國家標準(CNS) Α4規格(210X297公釐) (請先閲讀背面之注意事項再填寫本頁) 、可| 10 577197 A7 B7 五、發明説明(8 夯和八/D轉換為19〇-IP*19〇_q之類比信號,係適當地被集 中及量化。 (請先閲讀背面之注意事項再填寫本頁) AGC運作— 在此實施例中,該控制邏輯電路230,首先會檢查看 看,邊^號是否會充分地使任一 A/D轉換器19〇-IP*19〇_q 飽和。右然,一急速下降之增益控制程序,將會被執行; 若否,亦將在下文做說明,一基本增益控制程序,便會被 執行。 其次’其基本增益控制邏輯電路23〇,將會決定其接 收之信號,是否在一如下文所說明之較佳範圍内。若然, 便不需要增益控制;否則,便執行下文將做更詳細說明之 增益控制程序。 接著,該AGC系統170,將會嘗試使用一些如下文將 做更詳細說明之強信號和弱信號偵測技術,來識別一頻帶 内信號。若有一信號被發現到,此偵測程序便算完成;若 否,此偵測程序便會就該信號之次一部分被重複。彼等弱 信號偵測和強信號偵測,係一些獨立和互補之特徵。誠如 本說明書之進一步說明,就強信號偵測而言,一使接收增 |中之下降為必要的較強信號的到來,將可決定一信號可 月b存在,而就弱信號偵測而言,其將可決定一信號之可能 存在,係由於所測量之頻帶内功率,至少正比於其Agc處 之功率(當不需要一增益改變時),緊接是一超過臨界值之 自關聯。理應注意的是,若做出一增益改變,其最好使弱 仏號偵測禁能’通常係長達數微秒,因為直至其整個自關 本紙張尺度適用中國國家標準(哪)A4規格(21〇><297公釐) 11 A7 ^" ~____B7 五、發明説明(9 ) —-- 、有關之視窗填滿彼等後增益改變值前,自關聯將不會是 有放的因此,弱信號偵測,在使用上係針對一些相對於 阻斷信號或雜訊不夠大之到達信號,以及強信號债測,係 針對一些較大之到達信號。以及就強信號偵測而言,一新 的4 “虎之已到達,如下文所說明,在決定上係基於一粗增 ι下降或急速下降增益是否產生。 基本增益控制 在運作中,其AGC 170必須調整彼等接收機增益,以 使其接收之^號,可被其A/D轉換器i 9〇適當地量化。若此 一仏號在其A/D轉換器輸入處過大,此信號將會因飽和而 失真。若此一信號在其A/D轉換器輸入處過小,其a/D轉換 器190之量化雜訊,將使得其接收之信號雜訊(S/N)比,對 板準備測而言係過低。基於此一目地,其Agc控制邏輯電 路230 ’係如上文所述,使用其增益控制單元235,以數位 方式控制該等類比可變增益級段。較佳地,此實施例之增 盈控制,在該等結合之RF和IF級段120和135内,係具有93 dB - 51 dB之動態範圍,以及在其基帶級段145内為42 dB。:… Bu- 磬…: (Please read the notes on the back before filling this page)-Order, V. Description of the invention (5) These mixed IF signals will pass through some low-pass filters 14〇_Ip And 140-Q, so as to select their desired channels, and remove their unrelated components that are separated on the frequency side, and will be amplified by two serially connected baseband amplifiers 145-IP and 145-Q. Although two baseband amplifiers are shown in each branch, a different number of amplifiers can be used. In a particular sequence, using some baseband amplification with a properly selected programmable gain, a configuration that can develop any desired baseband gain step. The low-pass filters 14 (MP and 140_Q are preferably some two-pole elliptical filters with a 3 dB cross angle at 28 MHz. Moving from analog to the digital domain, the output of these baseband amplifiers is fed to Their A / D converters 190-IP and 190-q, they can digitize the signals of the equivalent phase and quadrature components, preferably at a frequency of 80 MHz to a resolution of nineteen bits: And an input dynamic range of -500 mV to 500 mV. These A / D converters are preferably some pipeline A / D converters; however, the present invention is not limited to this. For example, Δ or other Converter, can be used to replace them. Analog channel filter. The skin is, spear and / or anti-aliasing crossing; the leather device can be advantageously placed before these A / D converters 190-IP * 190_Q. In the preferred embodiment, a combination of analog and analogue wave filters can perform adjacent block signal rejection of 4 offenses and 20 dB of alternate block signal rejection. On the larger side of 16 offenders, block k and 32 Worst case of alternating blocking signals with larger dB '-the receiving blocking signal at the input of its A / D converter can be U dB higher than the signal in its frequency band. As is well known in the art, “adjacent blocking signals” are interfering signals adjacent to or overlapping with the frequency band of interest, and 577197 A7 ------- 丨 丨 B7 _ V. Description of the invention (Ό) An alternate blocking signal is an interference signal far from the frequency band of interest. The digitized I / Q component signals are supplied to an automatic gain control (AGC) unit 170 ' The operation of the invention will be described in more detail below. This AGC 170 can be described in more detail below, analyzing the ι / Q component h number, and can generate some gain control signals based on it. These gain control signals As shown by the dashed lines in Figure 2, these are supplied to these amplifiers 120, 135, 145-IP, and 145-Q. More specifically, as shown in Figure 3, these are from A / D The digital signals of the converters 190-IP and 190-Q pass through some leaky barrel filters 245-IP and 245-Q and FIR filters 205_Ip, 21CMP, and 205-Q , And 210-Q. The first FIR205_Ip and 205-Q are some decimation filters, which can be obtained from their corresponding data. Stream, eliminating every other sample, thereby reducing the data sampling rate from a normal 8.5 MHz single-sideband bandwidth packet 'from 80 MHz to 40 MHz. 2o-(^) are standard low-pass filters that can remove any remaining adjacent or alternating blocking signals before transmitting data to their autocorrelator 225 and a power detector 220. Within its Agc 170, secondary power measurements will be made, once from the output of its second FIR filter 21, by its power detector 220, and once from its A / D converters 190-IP and 190. The output of -Q is completed by another power detector 215. These measurements are supplied to its AGC control logic 230 as explained in more detail below. Although this embodiment uses digital FIR, other types including filters such as analog filters and wave filters can be used instead. If the system is not to be oversampled, the filters are preferably of analog type. This paper scales the family standard (⑽) M specification ⑵⑽ tears and love) 9 (Please read the precautions on the back before filling this page) Order — 577197 A7 _____B7_ V. Description of the invention (7) Its AGC control logic circuit 230 Can receive some power measurements from the power detectors 215 and 220 and use them to control a gain control generator 235 to generate each RF amplifier 120, IF amplifier 135, and individual baseband amplifiers 145-IP and 145 -Q related analog gain control signal. In this embodiment, the AGC control logic circuit 230 can provide a control block, which is ten bits long in this preferred embodiment. The gain control generator 235 and the gain control generator 235 can provide Generate appropriate control signals for these amplifiers. These gain control signals will be fed back to the RF amplifier 120, IF amplifier 135, and baseband amplifiers 145-IP and 145-Q, so as to control the gain of each as described above. This embodiment uses a dedicated integrated circuit to embody the AGC control logic circuit 230; however, a properly programmed processor, internal or discrete 'or other suitable device, can also be used. It should be noted that although Figure 3 shows the individual components within the AGC 170, two or more units may be integrated into one unit to separate them from each other. For example, a 'the AGC control logic circuit 230 is shown to be separated from the firs 205, 210, power detectors 215, 220, and the auto-correlation unit 225; however, these several can be combined into a single processor And do proper programming to perform these functions. In addition, a programmed processor does not need to be used, and one or more of these components can be implemented in dedicated hardware. The AGCn0 can control a DC offset control unit 24, thereby providing some analog offset control signals to one or more of the baseband amplifiers ΐ45_ιρ and 145-Q. The DC offset control is completed, which can ensure that these paper sizes are applicable to the Chinese National Standard (CNS) Α4 specification (210X297 mm) (please read the precautions on the back before filling this page). May | 10 577197 A7 B7 V. Explanation of the invention (8 rams and 8 / D conversions to analog signals of 19-IP * 19〇_q are appropriately concentrated and quantified. (Please read the notes on the back before filling this page) AGC Operation — In this embodiment, the control logic circuit 230 first checks to see if the edge number will sufficiently saturate any A / D converter 19〇-IP * 19〇_q. Right, a The rapidly decreasing gain control program will be executed; if not, it will be explained below, a basic gain control program will be executed. Secondly, its basic gain control logic circuit 23 will decide to receive it. Is the signal within a better range as described below. If so, gain control is not required; otherwise, the gain control procedure described in more detail below is performed. Next, the AGC system 170 will attempt to use Some will do as follows Detailed description of strong and weak signal detection technology to identify a signal in a frequency band. If a signal is found, the detection process will be completed; if not, the detection process will be part of the next part of the signal Repeat. Their weak signal detection and strong signal detection are independent and complementary features. As further explained in this specification, as far as strong signal detection is concerned, once The arrival of a strong signal will determine the existence of a signal b, and in terms of weak signal detection, it will determine the possible existence of a signal because the power in the measured frequency band is at least proportional to its Agc Power (when a gain change is not required), it is an auto-correlation that exceeds a critical value. It should be noted that if a gain change is made, it is best to disable the weak sign detection. Microseconds, because the paper size of this paper is up to the Chinese national standard (Which) A4 specification (21〇 > < 297 mm) 11 A7 ^ " ~ ____ B7 V. Invention description (9) ----, Relevant windows fill them Before the gain changes, the auto-correlation will not be released. Therefore, the weak signal detection is used for some arriving signals that are not large enough compared to the blocking signal or noise, and the strong signal debt measurement is for some Larger arrival signals. As far as strong signal detection is concerned, a new 4 "Tiger has arrived, as explained below, the decision is based on whether a coarse increase or rapid decrease gain is generated. Basic gain control In operation, its AGC 170 must adjust the gain of their receivers so that the ^ number they receive can be properly quantified by their A / D converter i 90. If this number is in their A / D converter The input is too large and this signal will be distorted by saturation. If this signal is too small at the input of its A / D converter, the quantization noise of its a / D converter 190 will cause the signal-to-noise (S / N) ratio it receives to be too small for board preparation. low. For this purpose, the Agc control logic circuit 230 'uses the gain control unit 235 as described above to digitally control the analog variable gain stages. Preferably, the gain control of this embodiment has a dynamic range of 93 dB to 51 dB in the combined RF and IF stages 120 and 135, and 42 dB in its baseband stage 145.
其功率偵測器215,可藉由就總數為16個樣本,就一 802.1 1a信號(400 ns)内之一半前同步短碼元幅窗,加總一 瞬時功率計算之幅窗,來估計該等A/D轉換器190-IP和 190-Q之總數位化功率。舉例而言,考慮一來自9_位元a/d 轉換器190而範圍在[_256,255]之信號,和此信號在半前同 步短碼元幅窗内之16-位元樣本幅窗有關之功率的測量 值。為完成此,其AGC控制邏輯電路230,將會計算其A/D 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 12 (請先閲讀背面之注意事項再填窝本頁)Its power detector 215 can estimate the value by adding a window of instantaneous power calculation for a total of 16 samples and a half-preamble short symbol window within an 802.1 1a signal (400 ns). Wait for the total power of the A / D converters 190-IP and 190-Q. For example, consider a signal from the 9-bit a / d converter 190 in the range [_256,255], which is related to the 16-bit sample window of this signal in the semi-preamble short symbol window. The measured value of the power. To accomplish this, its AGC control logic circuit 230 will calculate its A / D. This paper size is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm). 12 (Please read the precautions on the back before filling this page)
577197 A7 _B7_ _ 五、發明説明(10 ) 轉換器輸出流adcoutput上面之瞬時功率adcpwrl為 15 adcpwrl = ^(real(adcoutput[k]))2 Λ-(imag(adcoutput[k]))2 (1) k=0 此功率測量值,接著會被放進一對數表内,其中之最大值 為零。因此,就每一值在-256處之一完全導軌(railed)的輸 出而言,其對數表輸出將為零。一完全導軌正弦波之功率, 將為-3 dB ;若每一樣本為128,其功率將為-6 dB,等等。 其AGC控制邏輯電路230,係使用此總功率估計值, 來維持其信號在該等A/D轉換器190-IP和190-Q處之範圍 内。若其信號功率被決定係在範圍外(但並未使該等A/D轉 換器190-IP和190-Q飽和),一粗增益改變將會被完成,以 使該信號重回其範圍内。更明確地說,若其AGC控制邏輯 電路230偵測到,此總測量功率adcpwrl(此實施例中,在-63 -OdB之範圍内),係大於其最大希望之A/D轉換器信號尺 度,其所希望之增益值gaintarget,此信號尺度在設定上’ 係大至足以使其量化雜訊夠小,但亦小至足以使其ADC飽 和不是問題,而係包括該信號和任何潛在阻斷信號之尺 度,以及係在一粗增益下降中,被其AGC控制邏輯電路 230,以下列之方程式降低 gaintarget=gaintarget+(coarsepwr_const-adcpwrl) (2) 其中,coarsepwr_const,係其粗增益下降有關之額外增盈 (第4圖),例如,-17dB。此額夕卜增益損失之被使用,係由 於其進入信號可能對量化嫌過大,但如下文更詳細之說 明,未大至足以觸發一急速下降一舉例而言,若該信號偶 13 (請先閲讀背面之注意事項再填寫本頁) 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 577197 A7 B7 五、發明説明(11 ) 而飽和,但不足以觸發一急速下降。在此等情況中,使其 增益下降多於其基於一飽和波形之功率測量之gaintarget 所指示的值一一極強烈之值,將會是有用的。因此,加進 上述實驗決定之coarsepwr_const值,旨在增加其增益下 降,使更快速地收斂至其所希望之信號尺度。此結果係被 用來經由其增益控制產生器235,產生該等放大器有關之適 當控制信號。 若其總測量功率adcpwrl,係小於其最小希望之A/D轉 換器信號尺度,此所希望之增益值gaintarget,將會被其 AGC控制邏輯電路230,以下列之方程式增加 gaintarget=gaintarget+(totalsizede sired-adcpwrl) (3) 其中,totalsizedesired,係彼等粗增益改變期間之目標A/D 轉換器信號尺度,亦即,其A/D轉換器輸出在一希望之信 號(第5圖)不存在下所希望之尺度,在此較佳實施例中,為 大約-17 dB。 AGC急速下降增益控制 若其接收之信號,會經常使該等A/D轉換器190-IP和 190-Q飽和,一精確功率測量便可能無法得到;然而,該 信號確係部份地在範圍外。此一資訊可被用來迅速降低其 增益。更明確地說,一飽和計數器adcsat,係由其AGC控 制邏輯電路230來建立,以計數該等I或Q A/D轉換器輸出 樣本之飽和數目。一對可藉由下載一不同之臨界值來做改 變的飽和臨界值adcsat_thrh和adcsat_thrl,係被用來計數任 何A/D轉換器輸出範圍之可能缺乏。因此,一飽和將會被 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 14 (請先閲讀背面之注意事項再填寫本頁) -、可| 577197 A7 B7 五、發明説明(Π ) 偵測到,若 adcoutput^ (adcsat_thrh+192) (4) 或若 adcoutput^ (adcsat_thrh+192) (5) 其中,adcsat_thrh,係一小於其最大A/D轉換器輸出值之高 臨界值,其係指明該A/D轉換器輸出之高側上面的飽和, adcsat-thrl係一大於其最小A/D轉換器輸出值之低臨界 值,其係指明該A/D轉換器輸出之低側上面的飽和(第6 圖),以及該等固定值係屬具現相依性。adcsat_thrh在設定 上,係使略低於其最大A/D轉換器輸出,而adcsat_thrl在設 定上,係使略高於其最小A/D轉換器輸出。此之有用是因 為其可允許一些接近其A/D轉換器之飽和但未實際飽和的 信號,被分類為飽和信號,以便有更多之靈活性。在一 adcsat_icount周期(較佳地小於或等於八,四分之一-碼元 802.1 1a測量幅窗内之周期數)之樣本幅窗期間,若該等A/D 轉換器輸出樣本之飽和數目,超過一飽和臨界值量 adcsat_thresh,其AGC控制邏輯電路230,將會指示一急速 增益下降,以及gaintarget將會減少一預定之量 quick_drop,舉例而言,一 -30dB之增益改變。在此較佳之 實施例中,該adcsat_thresh,係就一 8·周期幅窗中之12個飽 和的臨界值被設定(使彼等飽和可能獨立地基於彼等I和Q ADC) 〇 此技術可有利地以下列之方式來加以具現。在校準或 任何增益改變之後,一 AGC安定時間將會發生。在adcpwrl 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 15 (請先閲讀背面之注意事項再填寫本頁) 、可| 表— 577197 A7 B7 五、發明説明(l3 ) 之後,其對應於所測量功率之量的變數,將會被重置,以 及一 acc_count計數器,較佳地為一在AGC運作期間連續循 環之8-位元增量的計數器,亦會被重置。 該計數器acc_count,將會發生下列事件: mod(acc_count,16)=0 :重置adcpwrl累加器 mod(acc_count,16)=l :清除adcpwrl累加器上面之重置 mod(acc_count,16)=2 :儲存adcpwrl mod(acc_count,16) ^ adcsat icount 和 adcsat=l(當 set_thresh飽和,舉例而言八個飽和已被計數出時被宣 稱),其飽和計數器係已超過adcsat_thresh,以及一急速增 益下降應被執行。誠如所顯示,在此較佳之實施例中,該 等adcpwrl值,係每16個周期被計算一次,以及此系統記錄 到一大於adcsat_icount之值(較佳地為8)前,將會期待adcsat 被宣稱。 AGC數據包偵測 一旦其接收機之信號係在範圍内,其AGC控制邏輯電 路230,便會偵測到一希望之數據包的出現。基於此一目 地,其AGC控制邏輯電路230,可決定一頻帶内功率估計 值,可使用該等FIR濾波器205-IP、210-IP和205-Q、210-Q, 來降低所有相鄰和交替之阻斷信號,使較802.1 la所指定之 最大位準下的頻帶内信號功率低20 dB,以及可如下文之說 明,比較adcpwrl與Hrpwrl。此在完成上是為得到有關其A/D 轉換器190-IP或190-Q處之量化信號能量究在頻帶内亦或 頻帶外的資訊-一有助於尋找該等所希望之數據包的資訊。 16 (請先閲讀背面之注意事項再填寫本頁) 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 577197 A7 B7 五、發明説明(14 、更明確地說,考慮第7A圖中所顯示之信號,如上文所 述,计异一 A/D轉換器輸出之功率,將可決定其係具有一 ㈣率’比方說_12撕,其中,撕係一些低於上文參考 方私式(1)所說明之全導轨信號尺度的信號之rms尺度的 測夏值。然而,如第7B圖所示,該信號通過其第二FIR 210 IP或210-Q ’冑會喪失其大部份之功率,&及會被降低 至25 dBr之位準“一大約85%之減少量。由於該信號大 部份之功率,係被其FIR21〇-Ip*21〇-Q阻斷,其被假定係 一頻帶外信號。 參照第8A圖中所顯示之信號,此信號亦係具有一如同 在其A/D轉換器輸出處所測得大約-12 dBr之全功率。然 而,如第8B圖所示,通過其第二FIR 21(Mp或21〇_q,僅使 降低其降低至大約-15 dBr——大約僅10%之減少量。由於 该信號大部份之功率,係被其FIR21〇-Ip或21〇_Q傳遞,其 被假定係一頻帶内信號。 藉由此一瞭解,其頻帶内功率,係計算為彼等瞬時功 率測量值之總和,較佳地在一 32個樣本中,類似上文所述 全功率計算adcpwrl之0.8//S幅窗。firpwrl係其基於某些小 於所得總樣本數目之樣本數目的最低者,諸如出自其偵測 為220中之32樣本幅窗内的32個樣本中之28個樣本。理應注 意的是,firpwrl有關之樣本數目,係大於adcpwri有關之樣 本數目,因為firpwd係用於細增益控制,其中之精密度係 重要的,而adcpwrl係用於粗增益控制,其中略帶雜訊之功 率估計值係可行。亦應注意的是,雖然此小於總樣本數目 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公爱)577197 A7 _B7_ _ V. Description of the invention (10) The instantaneous power adcpwrl on the converter output stream adcoutput is 15 adcpwrl = ^ (real (adcoutput [k])) 2 Λ- (imag (adcoutput [k])) 2 (1 ) k = 0 This power measurement is then placed in a logarithmic table, where the maximum value is zero. Thus, for every fully railed output at -256 for each value, its logarithmic table output will be zero. The power of a complete rail sine wave will be -3 dB; if each sample is 128, its power will be -6 dB, and so on. Its AGC control logic 230 uses this total power estimate to maintain its signal within the range of these A / D converters 190-IP and 190-Q. If its signal power is determined to be out of range (but does not saturate these A / D converters 190-IP and 190-Q), a coarse gain change will be made to bring the signal back into its range . More specifically, if its AGC control logic 230 detects that the total measured power adcpwrl (in this embodiment, in the range of -63 -OdB), it is greater than its maximum desired A / D converter signal scale , Its desired gain value gaintarget, this signal scale is set to be large enough to make its quantization noise small enough, but also small enough to saturate the ADC is not a problem, but includes the signal and any potential blocking The scale of the signal is reduced by its AGC control logic circuit 230 during a coarse gain drop, and is reduced by the following equation: gaintarget = gaintarget + (coarsepwr_const-adcpwrl) (Figure 4), for example, -17dB. This amount of gain loss is used because the incoming signal may be too large for quantization, but as explained in more detail below, it is not large enough to trigger a rapid decline. For example, if the signal is even 13 (please first Read the notes on the reverse side and fill in this page) This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 577197 A7 B7 V. Description of the invention (11) It is saturated, but not enough to trigger a rapid decline. In these cases, it would be useful to reduce its gain more than the value indicated by its gaintarget based on a saturated waveform power measurement—a very strong value. Therefore, the coarsepwr_const value determined by the above experiment is added to increase the gain and decrease so that it can converge to its desired signal scale more quickly. This result is used to generate appropriate control signals for these amplifiers via its gain control generator 235. If its total measurement power adcpwrl is smaller than its minimum desired A / D converter signal scale, the desired gain value gaintarget will be increased by its AGC control logic circuit 230, with the following equation: gaintarget = gaintarget + (totalsizede sired -adcpwrl) (3) where totalsizedesired is the target A / D converter signal scale during their coarse gain change, that is, their A / D converter output is in the absence of a desired signal (Figure 5) The desired scale is about -17 dB in this preferred embodiment. If the AGC's rapidly-decreasing gain control signal often saturates these A / D converters 190-IP and 190-Q, an accurate power measurement may not be available; however, the signal is partially in the range outer. This information can be used to quickly reduce its gain. More specifically, a saturation counter adcsat is established by its AGC control logic circuit 230 to count the number of saturations of the I or Q A / D converter output samples. A pair of saturation thresholds, adcsat_thrh and adcsat_thrl, which can be changed by downloading a different threshold, are used to count the possible lack of any A / D converter output range. Therefore, a saturation will be applied to Chinese paper standard (CNS) A4 specification (210X297 mm) for this paper size 14 (Please read the notes on the back before filling this page)-、 可 | 577197 A7 B7 V. Description of the invention ( Π) It is detected that if adcoutput ^ (adcsat_thrh + 192) (4) or if adcoutput ^ (adcsat_thrh + 192) (5) where adcsat_thrh is a high critical value that is less than its maximum A / D converter output value, It indicates saturation on the high side of the A / D converter output. Adcsat-thrl is a low critical value that is greater than its minimum A / D converter output value. It indicates the low side of the A / D converter output. The saturation above (Figure 6) and the fixed values are now dependent. adcsat_thrh is set to be slightly lower than its maximum A / D converter output, while adcsat_thrl is set to be slightly higher than its minimum A / D converter output. This is useful because it allows some signals that are close to the saturation of the A / D converter but not actually saturated to be classified as saturated signals for more flexibility. During a sample window of an adcsat_icount cycle (preferably less than or equal to eight, a quarter-the number of cycles in the measurement window of symbol 802.1 1a), if the A / D converter outputs a saturated number of samples, Above a saturation threshold value adcsat_thresh, its AGC control logic 230 will indicate a rapid gain drop and gaintarget will decrease by a predetermined amount of quick_drop, for example, a gain change of -30dB. In this preferred embodiment, the adcsat_thresh is set for a threshold value of 12 saturations in an 8-period window (so that their saturation may be independently based on their I and Q ADCs). This technique may be advantageous Land is realized in the following way. After calibration or any gain change, an AGC settling time will occur. In adcpwrl, this paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 15 (Please read the precautions on the back before filling this page), OK | Table — 577197 A7 B7 V. After the description of the invention (l3), Its variable corresponding to the amount of measured power will be reset, and an acc_count counter, preferably an 8-bit increment counter that continuously cycles during AGC operation, will also be reset. The counter acc_count will have the following events: mod (acc_count, 16) = 0: reset the adcpwrl accumulator mod (acc_count, 16) = l: clear the reset above the adcpwrl accumulator mod (acc_count, 16) = 2: Store adcpwrl mod (acc_count, 16) ^ adcsat icount and adcsat = l (declared when set_thresh is saturated, for example, eight saturations have been counted), its saturation counter has exceeded adcsat_thresh, and a rapid gain drop should be carried out. As shown, in this preferred embodiment, the adcpwrl values are calculated every 16 cycles, and before the system records a value greater than adcsat_icount (preferably 8), adcsat will be expected Be declared. AGC packet detection Once the signal from its receiver is within range, its AGC control logic 230 will detect the presence of a desired packet. Based on this, its AGC control logic circuit 230 can determine the power estimation value in a frequency band. The FIR filters 205-IP, 210-IP, and 205-Q, 210-Q can be used to reduce all adjacent sums. Alternate blocking signals make the signal power in the frequency band below the maximum level specified by 802.1la 20 dB lower, and adcpwrl and Hrpwrl can be compared as described below. This is done to get information about the quantized signal energy at its A / D converter 190-IP or 190-Q in or out of the band-a help to find these desired packets News. 16 (Please read the precautions on the back before filling this page) This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) 577197 A7 B7 V. Description of the invention (14. More specifically, consider Figure 7A The signal shown in the figure, as mentioned above, will be determined by the power output of the A / D converter. The summer measurement of the rms scale of the signal of the full rail signal scale described in the private formula (1). However, as shown in Figure 7B, the signal will lose its second FIR 210 IP or 210-Q '胄Most of the power, & will be reduced to a level of 25 dBr "a reduction of about 85%. Because most of the power of this signal is blocked by its FIR21〇-Ip * 21〇-Q It is assumed to be an out-of-band signal. Referring to the signal shown in Figure 8A, this signal also has a full power of approximately -12 dBr as measured at the output of its A / D converter. However, as shown in As shown in Figure 8B, with its second FIR 21 (Mp or 21〇_q, only reducing it to about -15 dBr- The reduction is only about 10%. Because most of the power of this signal is passed by its FIR21o-Ip or 21〇_Q, it is assumed to be a signal in a frequency band. From this understanding, the power in the frequency band , Calculated as the sum of their instantaneous power measurements, preferably in a 32 sample, similar to the full power calculation of the 0.8 // S window of adcpwrl described above. Firpwrl is based on some less than the total sample obtained The lowest number of samples, such as 28 out of 32 samples within the 32 sample window it detected as 220. It should be noted that the number of samples related to firpwrl is greater than the number of samples related to adcpwri Because firpwd is used for fine gain control, the precision of which is important, and adcpwrl is used for coarse gain control, where the power estimate with a little noise is feasible. It should also be noted that although this is smaller than the total sample The paper size applies to China National Standard (CNS) A4 (210X297 public love)
(請先閲讀背面之注意事項再填寫本頁) -、盯_ MV/ 五 、發明説明(15 ^頻V内功率計算之目地,係較佳地被使用,其他使用此 等樣本所發生之後類比數位轉換器處理,通常將使用所 得到之樣本。 頁 此小於總樣本數目之被使用,係由於在干擾期數,舉 ^而。纟㈣干擾之碼元交界處,由於其碼元遷移而變 乂為其所希望之頻帶内的頻帶内功率成分處的干擾之高頻成 /刀所致,一暫時性頻帶内功率尖波可能會發生。此將會不 自然地顯現為其頻帶内工力率中之步階。纟該干擾之發射器 處加幅窗,舉例而言,在其碼元交界處,使用一為其一半 之先前值加至其一半之後繼值的值,將可如同低通濾波地 使此略微降低,以致其餘賴,可傳遞其必要之頻譜遮 罩。此等瞬時高頻峰值雖被降低,仍然會存在。當一相鄰 之干擾出現時,此在干擾信號内之暫時性高頻成分,實際 上係在其所希望之信號有關之頻帶内,以致當無所希望之 信號出現時,該頻帶内功率測量值,可得到一有關數樣本 之急速尖波,像是其頻帶内功率之增加。為防止它,其係 使用32個樣本中之最低28個,以便此暫時性尖波,藉由不 計數彼等值而歸零,以及彼等臨界值做相應之調整,以抵 補因漏失四個樣本所致降低之功率測量值。然而,一曰有 一關注之信號出現,在建立其功率測量值中,最好使用所 有之樣本,來做一詳細之測量。此第二功率測量值係稱作 firpwr一all。使用以上所述之功率資訊,彼等所希望之信 號,便可在兩種方式中被發現:強信號偵測和弱信號偵測。 強信號偵測將首先加以說明。 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 18 577197 A7 B7 五、發明説明(l6 ) 強信號偵測 一旦有一如上文所說明之粗增益下降或急速增益下 降發生,其AGC控制邏輯電路230,便會設定一旗標 strongsignal。此一旗標將會保持為高邏輯位準,直至該信 號在其A/D轉換器190-IP或190-Q,被決定為在範圍内為 止,以及其演算法將會繼續進行,藉以如上文所說明,完 成一 firpwrl測量。就此點而論,flag_relpwr係計算如下 flag_relpwr=set if(firpwrl>relpwr+adcpwr) (6) (其中,relpwr係其頻帶内信號之絕對數位尺度相對於其 A/D轉換器190-IP或190-Q處之絕對總數位信號的一個實驗 臨界值變數),因而嘗試注意大部份計算之功率係在頻帶 内。若flag_relpwr係高邏輯位準,以及strongsignal係高邏 輯位準,則一新的、極強之信號係已出現。在此一方式中, 此實施例可允許檢查一過度取樣之進入信號,其會因過度 取樣而具有一所希望頻率範圍外之數位化頻率,以及可在 決定一所希望信號業已被發現到前,先決定其大部份功率 是否在頻帶内。 因此,當flag_relpwr係高邏輯位準,以及strongsignal 係高邏輯位準時,其signal_found旗標會被宣告,一細增益 改變便會如下文所述地完成,以及一旦彼等聯貫之增益改 變之數目,係等於或大於其被認為構成一成功之AGC運作 的最小增益改變數目,其AGC程序使算完成,亦即,當其 中已有足夠之增益改變,而於其系統啟通時,確保有一全 幅之可程式放大器線性上昇。 19 (請先閲讀背面之注意事項再填寫本頁) 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 五 發明說明(Π ) 在弱h號偵測中,其如下文所界定之短序列的常態化 哥耳外在測里上係為尋找此較佳之實施例中任何具有〇 . 8 # S之周期的頻帶内者。此係一與以上說明之強信號偵測程 序同時執行之兩步驟程序。首先,此系統會等候其自關聯 處理器225所測量之常態化自關聯超過一第一常態化自關 聯幅度臨界值m 1 thres。 其關聯處理器225,最好在測量802.1 la之自關聯數據 包上,係藉由在一數據包之開始,取得一短教練碼元中之 32個樣本,以及比較其每一樣本與一來自其前行短教練碼 兀之對應樣本。更明確地說,一 A/D轉換器流adc〇utput之 自關聯,係表示為 self corr = ΕΣα^οι//Ρ^[^] · conj{adcoutput[n - 32])]2 ^adcoutput[nf (?) 其中,其分母係一常態化因數。其分子可見於χ[ηρσ χ[η_32] 相同時,將會相當地高,以及舉例而言,於彼等不相關聯 時,將會相當地低。因此,此測量可被用作自關聯之一良 好之指示器。 偵測何時其自關聯輸出超過m 1 thres,可因而偵測出一 進入數據包之存在;然而,其亦將偵測出彼等干擾,因為 彼等可具有一些亦可自關聯之結構。基於此一理由,此實 施例係有利地採用另一試驗。一旦其常態化自關聯超過 mlthres,此系統便進入一迴路,以及就❿卜⑽加―max周期, 在一變數mltally中,計數一其常態化自關聯超過一第二常 20 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公爱) 577197 五、發明説明(is ) 態化自關聯幅度臨界值m2thres的次數,其中,m2thres* 小於或等於mlthres。若mltally > m2count—thr,其常態化 自關聯之計數的臨界值>1112化代3,在1111(:01^—11^\(—有 關其自關聯計數之幅窗長度)周期過去前,弱信號偵測可被 偵測。 誠如上文所注意到的,此基於mlcount—max之幅窗化 技術的被使用,是因為彼等干擾和雜訊兩者,可具有一短 暫地超過一臨界值之自關聯,但當橫跨連續周期期間所得 之樣本幅窗被使用時,此發生之機會將會減少。舉例而言, 一後繼之幅窗,將會包含許多與其先前幅窗相同之樣本, 但此先前幅窗將不會包含其出自後繼幅窗之大部份新近樣 本,以及其後繼之幅窗,將不會包含其出自先前幅窗之最 早樣本。因此,舉例而言,若相鄰頻道中有兩個8〇2Ua碼 元被傳送,以致彼等在頻率上係相隔2〇 MHz,其第一碼元 之最後0·8 // s,將會完全與其次一碼元之次一 〇 8 # $保護 周期相匹配,而建立成自關聯,但此尖波相較於一前同步 碼,將會迅速消退,其中,就一前同步碼期間而言,係預 期一平直常態化之自關聯結果。 因此,此實施例提供了一種針對一自關聯測量之執行 -二臨界值幅t化程序的方法。其_臨界值係被用來決定 L唬可能出現在頻帶内,在彼等抵補樣本之不同幅窗 内,一第二臨界值被超過之次數將會被計數,藉以進一步 决疋该頻帶内信號,是否為一所希望之信號。此一作為是 為防止熱雜Λ之暫時性關聯,加上防止一干擾信號之資料 本紙張尺目繼(Please read the notes on the back before filling in this page)-、 _ MV / V. Description of the invention (15 ^ frequency V power calculation purpose is better used, other analogies after the use of these samples occurred Digital converter processing will usually use the obtained samples. Pages smaller than the total number of samples are used because of the number of interference periods, for example. 纟 ㈣ Interference at the symbol boundary changes due to its symbol migration.乂 Caused by the high-frequency component / knife of interference at the power component in the band within the desired frequency band, a power spike in a temporary band may occur. This will appear unnaturally in its power rate A stepped window at the transmitter of the interference, for example, at its symbol boundary, using a value that is half of its previous value and added to its subsequent value will be like low-pass filtering This makes the ground slightly lower, so that the rest can pass its necessary spectrum mask. Although these instantaneous high-frequency peaks are reduced, they will still exist. When an adjacent interference occurs, this is temporary in the interference signal. High frequency components, real It is in the frequency band related to its desired signal, so that when no desired signal appears, the power measurement value in this frequency band can get a sharp spike of related data samples, such as the increase of power in its frequency band. To prevent it, it uses a minimum of 28 out of 32 samples so that this temporary spike is reset to zero by not counting their values, and their critical values are adjusted accordingly to compensate for missing four. The power measurement value caused by each sample is reduced. However, a signal of interest appears in the first place. In establishing its power measurement value, it is best to use all the samples to make a detailed measurement. This second power measurement value is called Make firrpr all. Using the power information described above, their desired signals can be found in two ways: strong signal detection and weak signal detection. Strong signal detection will be explained first. Paper size applies Chinese National Standard (CNS) A4 specification (210X297 mm) 18 577197 A7 B7 V. Description of the invention (l6) Once a strong signal is detected, the coarse gain decreases or rapid gain as explained above When a drop occurs, its AGC control logic circuit 230 will set a strongsignal flag. This flag will remain at a high logic level until the signal is detected by its A / D converter 190-IP or 190-Q. Decided to be within the range, and its algorithm will continue, so as to complete a firpwrl measurement as described above. In this regard, flag_relpwr is calculated as follows flag_relpwr = set if (firpwrl > relpwr + adcpwr) (6) (Among them, relpwr is an experimental threshold value of the absolute digital scale of the signal in its frequency band relative to the absolute total digital signal at its A / D converter 190-IP or 190-Q), so try to pay attention to most of the calculations. Power is in the frequency band. If flag_relpwr is a high logic level and strongsignal is a high logic level, then a new, extremely strong signal has appeared. In this manner, this embodiment may allow an oversampled incoming signal to be checked, which may have a digitized frequency outside a desired frequency range due to oversampling, and may be determined before a desired signal has been discovered. , First determine whether most of its power is in the frequency band. Therefore, when flag_relpwr is at a high logic level and strongsignal is at a high logic level, its signal_found flag will be announced, a small gain change will be completed as described below, and once the number of their consecutive gain changes, Is equal to or greater than the minimum number of gain changes that it considers to constitute a successful AGC operation, and its AGC procedure is completed, that is, when there are sufficient gain changes in it, and when its system is turned on, a Programmable amplifier rises linearly. 19 (Please read the precautions on the back before filling out this page) This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm). 5 Description of the invention (Π) In the weak h detection, it is defined as follows The normalization of the short sequence of the short sequence is to find in the preferred embodiment any of the frequency bands with a period of 0.8 #S. This is a two-step procedure performed simultaneously with the strong signal detection procedure described above. First, the system waits for the normalized autocorrelation measured by its autocorrelation processor 225 to exceed a first normalized autocorrelation amplitude threshold m 1 thres. Its association processor 225, preferably on the self-associated data packet measuring 802.1la, obtains 32 samples in a short coaching code symbol by the beginning of a data packet, and compares each sample with a sample from Corresponding samples of his short trainer. More specifically, the self-correlation of an A / D converter stream adc〇utput is expressed as self corr = ΕΣα ^ οι // Ρ ^ [^] · conj {adcoutput [n-32])] 2 ^ adcoutput [ nf (?) where the denominator is a normalization factor. Its molecules can be seen to be quite high when χ [ηρσ χ [η_32] is the same, and, for example, to be relatively low when they are not related. Therefore, this measurement can be used as a good indicator of autocorrelation. Detecting when its auto-correlation output exceeds m 1 thres can thus detect the presence of an incoming packet; however, it will also detect their interference because they may have some structure that can also auto-correlate. For this reason, this embodiment advantageously uses another test. Once its normalized auto-correlation exceeds mlthres, the system enters a loop, and the “max period” is added. In a variable mltally, a count of its normalized auto-correlation exceeds a second constant. This paper scale applies China National Standard (CNS) A4 specification (210X297 public love) 577197 5. Description of the invention (is) The number of times the threshold value m2thres of the auto-correlation amplitude is normalized, where m2thres * is less than or equal to mlthres. If mltally > m2count_thr, the threshold value of normalized self-correlation count > 1112 generation 3, before the period of 1111 (: 01 ^ —11 ^ \ (— the length of the window of its auto-correlation count)) period elapse Weak signal detection can be detected. As noted above, this windowing technique based on mlcount-max is used because both their interference and noise can have a transient over one Threshold autocorrelation, but when sample windows obtained across consecutive periods are used, the chances of this happening will decrease. For example, a subsequent window will contain many of the same as its previous window Samples, but this previous window will not include most recent samples from its successor, and its subsequent windows will not include its earliest samples from its previous window. So, for example, if Two 802Ua symbols are transmitted in adjacent channels, so that they are separated by 20MHz in frequency. The last 0 · 8 // s of the first symbol will be completely different from the next symbol. Once a 108 # $ protection period matches, and is established into self-association However, compared with a preamble, this sharp wave will fade away quickly, and in terms of a preamble period, it is expected that a straight normalized auto-correlation result is obtained. Therefore, this embodiment provides a method for The implementation of self-correlation measurement-a method of two threshold value amplitude t program. Its _ threshold value is used to determine that Lbl may appear in the frequency band. Within the different windows of their offset samples, a second threshold value is The number of times will be counted, so as to further determine whether the signal in the frequency band is a desired signal. This action is to prevent the temporary association of thermal noise Λ, plus the information to prevent an interference signal. Maki
…:— if, (請先閲讀背面之注意事項再填寫本頁) 訂丨 A7 '_______B7 五、發明説明(19 1 ~--- 段期間的自關聯。 此外’為進一步強化對抗雜訊和干擾信號之能力,此 實施例最好要求致能-弱信號偵測結果,一潛在之摘測數 據包,必須增加其頻帶内信號功率,使多達至少某一定量, 以及其增加係至少正比於其總信號功率之任何增加,其信 说功率至少為某-定最小尺度。此可於一新的頻帶内信號 低於一干擾^號或接近雜訊基值時,提供額外之靈敏度, 因而不會觸發強信號债測,但有尋找弱信號債測之價值。 一旦mltally〉mlthresh,至少有三件事可能會阻止弱 信號偵測發生。首先,若ycOK=0,弱信號偵測將不會發生。 ycOK係一漸減計數器,其若被決定某一定尺度 (flag—firstep)之頻帶内信號中有增加,以及至少正比於其 總功率(flag—relstep)中之任何增加,而使所測量之行卬〜^ 有某一疋最小尺度(flag—flrpwr),便會被重置至 yc〇Kmax(在此實施例中為四個),藉以致能弱信號偵測, 則一到來之新的頻帶内信號,係有可能低於一干擾信號或 接近雜訊基值,因而不會觸發強信號偵測,但有尋找弱信 號偵測之價值。為確保此種辨識作用在一有限時段内發 生,正當ycOK > 0畢竟發生之際,以上所述便勢必會發生。 為執行此荨步驟之計异’ firpwrl* acjCpWr之舊值,係儲存 為firpwr(2-4)和adcpwr(2-4)。彼等足夠值會被儲存,以致 若該信號在一可程式放大器斜進期間被偵測到,足夠之差 異將會存在於其第一與最後之測量值間,而超過其所定之 臨界值。 22 (請先閲讀背面之注意事項再填寫本頁) 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公爱) 577197 A7 B7 五、發明説明(20 另一何以弱信號偵測可能不會發生之理由是,因為 gc_count係大於零。gc_count係測量其自短碼元增量内之最 後增盈改變起,在一增益改變後,就自其3之起始值的每一 有效之firpwrl測量值,因其AGC控制邏輯電路230而減少之 時間。其觀念是,在一增益改變後,其中直至一自關聯有 效前,係有一最小之時間量。 最後’若該信號業已以其他方法偵測到,弱信號偵測 將不會發生,因為那樣將不需要使用弱信號偵測來找尋它。 AGC數據包偵測-DC抿補洁昤 上述雙臨界值配置,係成功於在弱信號偵測期間針對 干擾信號降低虛假之數據包偵測;然而,其並非特別成功 於防止相對於彼等總能自關聯之Dc信號的虛假偵測。在其 A/D轉換器190之輸出處,通常會有一小量之DC成分,故 此實施例係使用一二分接頭DC陷波濾波器,作為一泄漏桶 式濾波器一更明確地說,一二分接頭IIR濾波器具有之一轉 移曲線的形式為 (請先閱讀背面之注意事項再填寫本頁) •、η— y[n\ a-l 其中,x係其輸入信號,y係其輸出信號,以及“係一濾波 裔參數(在此一情況中為32)—其係使用AGC控制邏輯電路 230所提供之DC位準的估計值,藉以將此DC成分消除掉。 其AGC控制邏輯電路230,可基於一些當前之增益設定, 自一查表得到此一位準。 AGC完成簇庠…: — If, (please read the precautions on the back before filling out this page) Order 丨 A7 '_______B7 V. Description of the invention (19 1 ~ --- self-correlation during the period. In addition,' for further strengthening against noise and interference Signal capability, this embodiment preferably requires enable-weak signal detection results, a potential test packet, must increase its signal power in the band to at least a certain amount, and its increase is at least proportional to Any increase in its total signal power means that its power is at least a certain minimum scale. This can provide additional sensitivity when the signal in a new frequency band is lower than an interference signal or close to the noise base value, so it does not Will trigger a strong signal debt test, but there is value in finding a weak signal debt test. Once mltally> mlthresh, at least three things may prevent weak signal detection from happening. First, if ycOK = 0, weak signal detection will not happen YcOK is a decreasing counter. If it is determined that there is an increase in the signal in a certain scale (flag-firstep), and at least proportional to any increase in its total power (flag-relstep), the measured The trip 卬 ~ ^ has a certain minimum scale (flag_flrpwr), it will be reset to ycoKmax (four in this embodiment), so that weak signal detection is enabled, and the new one The signal in the frequency band may be lower than an interference signal or close to the noise base value, so it will not trigger strong signal detection, but it has the value of finding weak signal detection. In order to ensure that such identification occurs within a limited period of time When ycOK > 0 occurs after all, the above will certainly happen. To implement the different steps of this step, the old value of firpwrl * acjCpWr is stored as firpwr (2-4) and adcpwr (2-4 ). Their sufficient values will be stored so that if the signal is detected during a ramp of the programmable amplifier, sufficient differences will exist between its first and last measured values, exceeding its defined threshold. 22 (Please read the precautions on the back before filling this page) This paper size is applicable to Chinese National Standard (CNS) A4 (210X297 public love) 577197 A7 B7 V. Description of the invention (20 Another reason why weak signal detection is possible The reason it doesn't happen is because gc_coun t is greater than zero. gc_count is measured from the last gain change in short symbol increments. After a gain change, each valid firpwrl measurement value from its starting value of 3 is due to its AGC control. The logic circuit 230 reduces the time. The concept is that after a gain change, there is a minimum amount of time until an auto-correlation is valid. Finally, if the signal has been detected by other methods, weak signal detection It will not happen, because then it will not need to use weak signal detection to find it. AGC packet detection-DC 抿 fill clean 昤 The above dual threshold configuration is successful in reducing falsehoods for interference signals during weak signal detection Packet detection; however, it is not particularly successful in preventing false detections with respect to DC signals that they can always auto-correlate. At the output of its A / D converter 190, there is usually a small amount of DC component, so this embodiment uses a two-tap DC notch filter as a leaky barrel filter. More specifically, a The two-junction IIR filter has a transfer curve of the form (please read the precautions on the back before filling this page) •, η— y [n \ al where x is its input signal and y is its output signal, And "It is a filter parameter (32 in this case)-it uses the estimated value of the DC level provided by the AGC control logic circuit 230 to eliminate this DC component. Its AGC control logic circuit 230, This level can be obtained from a look-up table based on some current gain settings.
577197 A7 B7 五、發明説明(21 ) 一旦該信號經由強信號偵測或弱信號偵測被發現 到,細增益改變將會被完成,在此較佳之實施例中,若 consec gainchanges < min gainchanges。以及在此較佳之 實施例中,每一細增益改變在完成上,將基於方程式 gain_change=adc—desired—size-firpwrl—all (9) conSec_gainchangeS,就強信號偵測而言,係自零處開始, 以及就弱信號偵測而言,係自2處開始,因為其係意謂一在 AGC中所度過之時間的粗測量值,以及其大約要用兩個增 益改變時間,來執行一成幅窗之自關聯。其係就每一粗和 細增益改變而遞增。當無增益改變完成時,其將會被重置, 以及強#號偵測’並不會決定有一信號出現。此一特徵係 打异就更精密之增益,或為確定其增益會在PA完成斜進後 被設定,而確保其AGC内係已度過一最小量之時間。 上文所說明之較佳實施例,在表示上係僅基於解釋之 目地,以及本發明不應詮釋為受到如此之限制。本發明之 變更形式,係為本技藝之專業人員在閱讀過此說明書後所 輕易理解,以及本發明和所附申請專利範圍,同樣係涵蓋 此等變更形式。 (請先閲讀背面之注意事項再填寫本頁) •訂· 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 24 577197 A7 B7 五、發明説明(22 ) 元件編號對照 10…超外差接收機設計 140-IP,140-Q…低通濾波器 15…天線 145-IP,145-Q…基帶放大器 20…RF(射頻)放大器 170…AGC(自動增益控制)單元 25…混頻器 175-IP…同相混頻器 30…本地振盪器 175-Q…正交混頻器 35-"IF(中頻)放大器 180…本地IF振盪器 40…帶通濾波器 185…相移器 45…IF放大器 190-IP,190-Q …A/D 轉換器 50…乘積檢波器 245-IP,245-Q···泄漏桶式濾波器 55…拍頻振盪器 205-IP,210-IP,205_Q,210-Q … 60…帶通放大器 有限脈沖嚮應(FIR)濾波器 65…類比數位(A/D)轉換器 215…功率偵測器 70…AGC(自動增益控制)單元 220…功率偵測器 115···寬頻帶天線 225…自關聯器 120—RF放大器 225…自關聯處理器 125···混頻器 230…AGC控制邏輯電路 130···本地RF振盪器 235…增益控制產生器 135*"IF放大器 (請先閲讀背面之注意事項再填窝本頁) -訂· 本紙張尺度適用中國國家標準(CNS) A4規格(210X297公釐) 25577197 A7 B7 V. Description of the invention (21) Once the signal is found through strong signal detection or weak signal detection, the fine gain change will be completed. In this preferred embodiment, if consec gainchanges < min gainchanges . And in this preferred embodiment, each fine gain change is completed based on the equation gain_change = adc_desired_size-firpwrl_all (9) conSec_gainchangeS, in terms of strong signal detection, starts at zero As far as weak signal detection is concerned, it starts at two places because it means a rough measurement of the time spent in the AGC, and it takes about two gains to change the time to perform 10% Self-association of windows. It is incremented for each coarse and fine gain change. When no gain change is completed, it will be reset, and strong # detection will not determine that a signal appears. This feature is a more precise gain, or to determine that the gain will be set after the PA has completed the ramp, and to ensure that its AGC has passed a minimum amount of time. The preferred embodiments described above are shown for explanation purposes only, and the invention should not be construed as being so limited. Modifications of the present invention are easily understood by those skilled in the art after reading this specification, and the scope of the present invention and the appended patents also covers such modifications. (Please read the precautions on the back before filling this page) • Order · This paper size applies Chinese National Standard (CNS) A4 specification (210X297 mm) 24 577197 A7 B7 V. Description of the invention (22) Component number comparison 10 ... Heterodyne receiver design 140-IP, 140-Q ... low-pass filter 15 ... antenna 145-IP, 145-Q ... baseband amplifier 20 ... RF (radio frequency) amplifier 170 ... AGC (Automatic Gain Control) unit 25 ... mixing 175-IP ... in-phase mixer 30 ... local oscillator 175-Q ... quadrature mixer 35- " IF (intermediate frequency) amplifier 180 ... local IF oscillator 40 ... band pass filter 185 ... phase shifter 45 ... IF amplifier 190-IP, 190-Q ... A / D converter 50 ... product detector 245-IP, 245-Q ... Leaky barrel filter 55 ... beat frequency oscillator 205-IP, 210-IP , 205_Q, 210-Q… 60… Band-pass amplifier limited pulse response (FIR) filter 65… Analog digital (A / D) converter 215… Power detector 70… AGC (Automatic Gain Control) unit 220… Power Detector 115 ... Broadband antenna 225 ... Autocorrelator 120—RF amplifier 225 ... Autocorrelation processor 125 ... Mixer 230 AGC Control Logic Circuit 130 ... Local RF Oscillator 235 ... Gain Control Generator 135 * " IF Amplifier (Please read the precautions on the back before filling in this page)-Order · This paper size applies to Chinese National Standards (CNS ) A4 size (210X297 mm) 25