533692 A7 ___ B7__ 五、發明說明（/ ) 與相關的專利交互參考 本專利申請案主張在2〇〇〇年6月2臼申請之美國專 利申請案優先權第60/209,057號。 · 本發明的背景 本發明一般與通訊網路有關，特別是與測試用作傳輸 資料的電話線路品質的系統有關。 如同習知的技術：公共交換電話網路，例如所謂的舊 式電話服務(POTS)線路，原本是設計用來傳輸語音，其涵 蓋有限的頻寬（例如大約4KHz);今天則希望使用相同的 POTS線路來傳輸資料。然而，資料訊號通常較聲音訊號 有不同的頻率特性，包括有較大的頻寬。結果，一條能正 常傳輸聲音訊號的POTS線路，對於傳輸資料訊號可能無 法正常工作、甚至根本不能傳輸。電話公司必須要知道那 些線路是合適的，例如合格的，而那些線路不適合用來傳 輸資料；電話公司也必須知道爲什麼特定的線路無法支援 資料傳輸，及這些錯誤會發生在何處，如此他們才能決定 該傳輸線路是否能被校正。 線路合格認證是針對描述用戶迴路品質的整體能力， 因爲與線路傳輸聲音(例如POTS)，或是資料服務的能力有 關。在此特別針對支援高速資料傳輸線路的合格認證，諸 如ISDN，SDSL，ADSL或是HDSL。事先的淘汰是對不經矯 正動作即無法支援資料服務的用戶迴路，作出具有高信任 度的描述能力；事先的合格認證是對能支援資料服務的用 戶迴路，作出具有高信任度的描述能力。 4 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） — — — — — — — — — — — — — I I 1 I I 1 ( — — — — — — II I I (請先閱讀背面之注意事項再填寫本頁) 533692 A7 ___B7___ 五、發明說明（2 ) 電信公司（TEL中央控制室，s)對解決傳輸資料的用戶迴 路之合格3忍δ登具有兩個問題，第一^是策略性的：電信公司 不願意佈署用來傳輸資料的新興科技（例如ISD&或是 ADSL)，因爲在他們的認知裡，足夠的用戶迴路具有足夠 的高品質，能使佈署具有經濟上的成功，是不確定的。這 使得早期採用者卻步，因爲作爲第一個傳送一種在其存取 網路可能不會成功的科技，是具有重大風險的。如果能給 予電信公司一種科技，使在初次佈署時大幅減少風險，則 他們成保有巾場占有率，並在面對競爭時保持領先。 第二個問題是戰術性的，並且是在電信公司決定佈署 一種特定科技之後，不管是主動或是被動地，當服務可能 由用戶提出要求，或是由電信公司爲傳送而鎖定，或是電 信公司迫於法令必須提供足量的線路給CLEC(中央控制室 mpetitive Local Exchange Carrier，競爭性區域交換載具）時 ’都有需要對提供服務的線路作品質檢驗。舉例而言，如 果一家電信公司要進入市場並提供新的服務，他們會想要 鎖定那些在盒子之外最有可能支援服務的和（或）以最少工 作的用戶迴路；另一個例子是，一家電信公司在從用戶接 到新的服務要求時，需要基於他們線路的狀況，接受或拒 絕該新服務要求的資訊。 線路品質驗證通常以單端或是雙端測試。若以雙端進 行測試’會派遣一位技工到受測線路的兩端，受測的線路 自網路中離線，然後測試儀器裝在線的兩端。測試儀器共 同合作’以決定該線路是否支援所需型式的資料傳輸。雙 5 尺度適用中國國家標準（CNS)A4規格（210 X 297公釐) " --------------------訂--------- (請先閱讀背面之注意事項再填寫本頁) 533692 A7 -____Β7 __________ 五、發明說明（；） 端的測試因爲在線路的兩端裝設儀器所費的時間和成本， 故並不値得推荐；此外，雙端測試通常提供的是針對資料 服務的特定型式之結果，所以如果資料服務的型式改變時 ，測試就要重做。 以xDSL線路品質驗證而言，這兩種方法現在都有在 使用。一種方法是派遣一位技術人員到受測線路的最上端 ，諸如機房（中央控制室)或是數位迴路載具(DLC)。技術人 員從受測線路離線，然後接上一系列有特殊用途的儀器， 如用來偵測橋狀分接頭的負載線圏或時域反射儀，或是用 來找出阻抗錯誤的數位多用電表。這樣的方法只能給予在 線路上的訊號衰減(我們發現是線路表現中非常重要的預警 )一個不完全的指示，而且在大量佈署特定型式的高速資料 服務時，太耗時且昂貴。 其它則嘗試使用自動量測儀器，其發展用來測試影響 語音服務錯誤的線路，進行單端的預先品質檢驗，並且依 賴電信公司爲他們線路所保有的紀錄，以提供額外的資訊 。纜線資訊能指出線路是否是以已知的不適合特定資料服 務的方式規劃，舉例而言，線路的管線儀的紀錄和是否已 裝設負載線圏都可能指示該線路無法支援一特定型的服務 。然而，纜線紀錄的檢閱通常要透過人工方式進行，結果 造成了無法令人接受的長期的測試時間；此外，電信公司 通常發現需要的紀錄是不正確的，因爲需要的資訊輸入不 正確或是從未更新。更有甚者，影響數位資料服務速度的 線路特性並不必然造成任何傳統語音服務的衰減。許多年 6 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） --------------#-------訂---------線丨1· (請先閱讀背面之注意事項再填寫本頁) 533692 A7 __B7___ 五、發明說明（> ) 來，對電信公司來說，沒有任何理由去保存他們現在需要 用來爲高速資料服務的線路作品質檢驗的資訊紀錄，因此 ，纜線紀錄通常不包含所需的資訊。 > 目前現有的預先品質檢驗的技術限制造成電信公司無 法以高信任度預知他們的線路是否能支援特定型式的資料 服務，這些限制已迫使服務供應商限制他們提供高速資料 服務給特定的區域，在這些區域他們相信這些線路有能力 支援他們，或只能承諾客戶其服務只能以小於他們可能的 全速下運作。 如果電信公司能快速而不需大筆花費地決定那些線路 能支援高速資料服務，以及其線路運作的速度，對公司是 有極大益處的。 本發明的摘要 以前述的背景，本發明的一個目標是提供一種方法和 裝置，用來執行線路品質檢驗測試，並且對這樣的測試結 果加以分類。線路被測試來決定或估計不同的線路特性。 線路的物理特性被估計(例如線長、線徑、***損失）；在 線路上存在的儀器，諸如負載線圏、橋狀分接頭、終端和 類似的也會被決定。線路能支援的資料傳輸率的預估値會 藉由線路測量和預估的狀況而定，因此，在低頻（少於 100KHZ)下線路的行爲會被用作預估在高頻(超過100KHZ) 線路行爲的替代物。測試和量測比較喜歡在在低頻下執行 ，這樣他們可以透過電信公司的語音交換來執行。結果會 依據一定的標準而分類，並提供一個易於分辨的線路狀況 7 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公爱） ---------------------訂---------線—AVI (請先閱讀背面之注意事項再填寫本頁) 533692 A7 ---- ---B7___—_ 五、發明說明（^ ) 。分類可以藉由使用特別爲此目的設計之軟體的電腦而加 以執行。 在一較佳的實施例中，分類的結果包含一個第^一類別 ，其中標出無法支援一特定階層的高速存取的線路；分類 的結果也包含一第二類別，其中標出能支援一特定階層的 高速存取的線路；分類的結果也包含一第三類別，其中標 出目前不行，但是在移除障礙後能夠支援特定的高速存取 的線路；一第四類別，其標出被選擇的線路特性落在測試 系統涵蓋範圍之外，也會被包含在內。每一類別被賦予一 各別的顏色，爲使線路的狀況能被輕易地被識別。測試和 分類可能會爲各種不同的高速存取階層而執行。客戶以不 同的費率被收費，取決於在其線路可以使用的服務等級。 附圖的簡要說明 本發明藉由以下更詳細的說明和相對應的附圖能更容 易地被了解，其中： 圖一是本發明裝置的方塊圖； 圖二是本發明方法的流程圖； 圖三是本發明方法的流程圖；以及 •圖四是用來表示一數據機模型的資料結構草圖。 詳細的說明 現今對提供高速存取，諸如整體服務數位網路(ISDN) 或是不同型式的數位用戶迴路(xDSL)服務給上網際網路的 用戶，存在著大量的需求。這些用戶大多數希望透過他們 現有的舊式電話服務（POTS)線路來接收這類存取。pots 8 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） --------------------訂---------丨 (請先閱讀背面之注意事項再填寫本頁) 533692 A7 _ —_B7_____ 五、發明說明（6) (請先閲讀背面之注意事項再填寫本頁) 線路原本是建設用來攜帶語音，而並不盡然是爲與ISDN 或xDSL有關的更高的頻率。POTS線路的某些比率能提供 用戶所需的高速存取。預知現有POTS線路支援高速存取 的能力通常是勞力密級的、需要大量的時間，且需要有技 術的人來解釋結果。 本發明提供對現有POTS線路自動化的測試，用來估 計線路提供高速服務的能力，並更進一步地以一種易於了 解的格式，呈現測試的結果。在線路上執行的測試，作爲 線路特徵的一部份，會採取各式各樣不同的量測和估計。 線路的特徵，其會藉由測試而被預估，包含了線路的*** 損失、線路的相位不平衡、線路的長度和線徑。額外的由 線路測試所決定的特徵可能包含線徑的改變的決定和估計 、橋狀分接頭、負載線圈，和其它路徑原件，諸如分岐器 、濾波器、終端，和相類似的東西。此外，速度干擾器， 其爲纜線圍束中其它的高速服務，也會被列入考慮。線路 測試的第二部份包含使用一個基於測得和估計而得的特徵 的程序，來預測線路會支援的資料傳輸率。此外’預測的 可靠度和涵蓋範圍也會被包含在內，作爲測試結果的一部 份。透過測試程序所決定或估計的線路特徵的描述如下所 示。 假設一條電話線沒有其它附加的情況，在線路上服務 的運作會明顯地受加入訊號的線路整體衰減或***損失的 影響。目前電信公司藉由佈署人員到線的兩端，透過手持 式儀器量測在不同頻率(例如ΙΟΟΚΗζ，300ΚΗΖ等)下的*** 9 本紙張尺度適用中國國家標準（CNS)A4規格（210 x 297公釐〉 533692 A7 _______B7_____ 五、發明說明（1 ) 損失，來量測***損失。現在的測試系統包含一從線路單 端量測來估計一線路的***損失的方法。 線路的相位不平衡也是預估線路處理高速存取能力的 一個重要考量。一條電話線是由兩條金屬線所組成，稱爲 尖端(Tip)和環(Ring)。通常尖端線和環線應該有相同的電 氣特性，兩條線最好能平衡。在一條平衡的線中，每條線 的阻抗、電流和電感會相同。如果一條線的阻抗、電流和 電感和另外一條不同，則不平衡便會存在。不平衡會導致 迴路中兩條線間差異的相位轉移。較高速的數據機在載體 頻率上編碼許多位兀到相位轉移，因此甚至是不平衡微小 的不穩定都會導致減少的資料流量、錯誤和再訓練。在 ISDN中，相位的轉移會造成從一脈衝來的能量會重疊到同 步訊號中或到由另一脈衝所占用的時間，結果造成內部的 訊號扭曲和（或）同步的損失。類比和數位的傳輸方法都會 在相位不平衡中，對些微的不穩定有所感應。偵測不平衡 是重要的，因爲大的不平衡數値會藉由降低迴路的電流値 而影響POTS。阻抗不平衡可能會大到(2 kilo-ohms或更多) 連捉住一個撥號音都變爲不可能，或是響鈴電流可能不足 以啓動在電話或是數據機中的響鈴電路。同樣重要的是當 資料傳輸列入考慮時，要偵測數値低於2 kilo-ohm的不平 衡。任何系列的阻抗和造成相位轉移的聲音會對資料傳輸 量有負面的影響，它可能會在該用戶迴路上達成。 線長、線徑和延著線路之線徑的改變能透過線路的測 試和量測或是透過電信公司的紀錄而被預估。線路的長度 10 Ϊ紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） " ' -------------·-------訂---------線丨♦ (請先閱讀背面之注意事項再填寫本頁) 533692 A7 _ __..._ B7 _ 五、發明說明（S ) 在高速存取中扮演一個角色，因爲線路可能會太長而無法 支援高速傳輸存取。線路線徑和線路上任何線徑的改變也 會影響線路支援高速存取的能力。線徑的改變被視^阻抗 ，並會影響高速訊號在線路上傳輸。 電話網路通常是爲POTS訊號最佳化。一個將網路最 佳化的方法是過在電話線路上將負載線圈包含在內，負載 線圈改善類比訊號在線路上的頻率反應，在線路上的負載 線圏會產生一個回應，而影響線路處理高速服務的能力。 在美國，許多用戶線路有橋狀分接頭，這是由於電話 線路佈設在房屋分部中的方式，電話線路預先佈設在外， 以決定分部中房屋確實的定位。線路通過許.多房屋的計畫 中位置的附近，當房屋稍後建造起來，建商會將電話單元 與先前佈設的電話線路中的最近的點相連接。先前的線路 未連接的部份會產生橋狀分接頭，橋狀分接頭會從終端反 射訊號，反射的訊號接著會傳回用戶線路，並與用戶線路 上的訊號互相干擾。最有害的干擾會發生在當反射的訊號 與進入的訊號是在相位之外，在這種情況下，反射的訊號 會對在用戶線路上進入的訊號產生嚴重的干擾。 _橋狀分接頭的長度指出了在原始和反射訊號之間的相 位差。對高速數位訊號來說，其頻率會延伸到大約1MHZ ，例如ADSL訊號，如果橋狀分接頭的長度介於約200到 700吋，則會產生重大的取消。在美國，建造許多房屋分 部所留下的橋狀分接頭的長度是落在這個範圍內。因此， 偵測並移除橋狀分接頭之能力，對想提供高速數位資料服 11 ^---- _ 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） (請先閱讀背面之注意事項再填寫本頁) ·· 訂i 線— φ· 533692 A7 _______B7_ _ 五、發明說明（7 ) 務給他們用戶的電信公司是有用的。 像ADSL的新服務的產生，如果電信公司的中央控制 室（中央控制室）裝備好了去處理提供這些新服務給用戶， 不可能有任何預知。舉例而言，如果受測的線路裝有分歧 器和一半的響鈴器但並沒有電話時，標準的測試系統會被 誤導認爲在用戶的線路上有響鈴。ADSL在線路上加入一 中央控制室分歧器和假定分歧器。中央控制室分歧器使用 二極體，提供極性敏感度和落在ADSL訊號領域範圍外的 傳導飽和値。 中央控制室分歧器的電流只能在環邊比尖端邊更正向 時通過。沒有已知的測試系統所執行的量測允許這樣的情 況，因此這些現在所使用的測試系統無法偵測一正常安裝 的中央控制室分歧器。爲了要正確地將線路的特徵找出， 決定或是預估線路上殘存的元件是重要的。 迴路的線路模型從在線路上的物理量測而產生，最好 是用裝設在電話網路中的自動量測系統，在較佳的實施例 中，線路模型包含迴路資訊，其包括長度、線徑、橋狀分 接頭的指示，以及負載線圏的指示及其衰減。取得該迴路 資訊的技術在泰瑞丹公司的不同的專利申請中有描述，或 是包含在商業可取得的產品，如Celerity，此爲在美國位 於伊利諾州Deerfield的泰瑞丹公司通訊部門所售。較佳的 技術可在美國專利申請中找到：由Groessl等人於1999年 4月24日所申請的第〇9/294,563號；由Rudinski等人於 1999年9月30日申請的第〇9/410,236號；由Faulkner等 12 ______________--------^---------^ — L^wl Γ 清先閱讀背面之泛意事項再填寫本頁) 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） 533692 A7 ___B7_____ 五、發明說明（，。） 人於1999年9月30日申請的第09/410,222號；由Holeva 等人於1999年9月30日申請的第〇9/409，141號；由 McGill等人於1999年9月30日申請的第〇9/410,2$7號， 前述的申請在此納入作爲參考。 進一步來說，該模型也可以包含有時可能會被認爲是 DC錯誤資訊的資訊。這樣的錯誤能在線路上用直流量測 偵測到。此資訊包含尖端接地錯誤、環接地錯誤和交叉錯 誤。線路的模型也包含低頻噪音的量測和在線路上以何種 終端的評估，諸如機械式的或是電子式的電話或是網路介 面設備。這樣的錯誤和狀況能以商業可取得的產品來偵測 ，諸如由位於美國伊利諾州Deerfield的泰瑞丹公司通訊部 門所出售的4TEL II，但是其它的產品也是商業可取得進f了 量測。更進一步如何取得這樣的訊息的資訊能在美國專利 5,699,402中找到，其在此被納入作參考。 這些量測的技術是較佳的，因爲他們允許使用相對低 頻量測來決定線路的狀況，使用低頻量測允許測試系統藉 由現有的交換機轉換到一特定的線路’現有的交換機通常 有低頻回應。舉例而言，進行直接對橋狀分接頭的量測時 ，需要至少300KHZ的頻寬。如同以下將描述的，線路的 表現能藉由使用低頻量測和預測來預估。 在線路模型中那些項目要包含在內並不是本發明重要 的限制，通常更多的資訊會導致線路更精確的分級，但是 也會花更多的時間，並需要更複雜的量測設備或軟體。更 進一步來說，那些品質在預測表現時最有幫助將會隨不同 13 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） (請先閱讀背面之注意事項再填寫本頁) 訂·· 線· 533692 A7 _—____B7 _ 五、發明說明（I丨） 網路而異，舉例而言，橋狀分接頭在美國比在許多歐洲國 家更加普遍，所以線路模型用於美國包含橋狀分接頭是否 存在的指示可能是重要的，但是這樣的質當用在歐>州的網 路時可能變得不重要。一組較佳的線路量測和特性將會透 過使用實證而被部份地選擇是所期望的。 爲分類特定的線路至適當的類別，兩個額外的模型被 建立：一個模型描述線路上的雜訊的特性，另一則定義在 高速數位服務所使用的數據機的能力。 與從該線路取得的量測，有關特定線路的資訊相比， 雜訊模型最好是從有關一特定網路由區域電信公司運作的 方式的資訊而取得。雖然量測線路上真實的雜訊能被用作 可能的替代性實行，以現有的自動化量測設備量測高頻或 寬域的雜訊仍是困難的。這樣的設備通常透過一交換機連 接至線路，該交換機有低通頻率反應，使得量測在高頻或 寬域雜訊中的高頻變得困難。舉例而言，寬域或脈衝雜訊 的直接量測需要1·1ΜΗζ的頻寬，使用現存的交換設備， 自動連接一測試系統到一線路不太可能提供足夠的頻寬， 用來可靠地執行所需的雜訊量測。 雜訊的模型能選擇性地從有關一家電信公司如何運作 其網路的資訊而取得。在電話線路上一個主要的雜訊來源 是串音。電話線路通常在一纜線圍束從中央控制室繞徑到 用戶的附近，到所有附近用戶的線路也包含在纜線圍束內 。雖然在線設計中已努力要降低從一條線路到另一線路伴 隨而來的訊號量，但仍會有一些伴隨的訊號，而此伴隨的 14 本紙張尺度適用^國國家標準（CNS)A4規格（210 X 297公釐）一' ' 一 n n n n ϋ n n n n n n ϋ I >ϋ ϋ ϋ I I n n 一 4OJI I n n n I n ϋ I I (請先閱讀背面之注意事項再填寫本頁) 533692 A7 __B7__ 五、發明說明（A) 訊號稱作串音雜訊。從一線路到另一線路的結合量取決於 傳輸訊號的頻率，和串音雜訊的量’因此取決於在纜線圍 0 束中所傳送的何種其它型式的訊號。 服務供應商遵循有關在纜線圍束中傳送的複合訊號的 設計規則，舉例而言，他們可能指定用在ADSL或是T1 或是其它傳送高頻訊號的訊號最大的線路數量，以確保沒 有太多會造成串音雜訊的高頻訊號。基於服務供應商的設 計規則，在一特定線路上期望的預估最大雜訊能藉由計算 或是透過實驗而得。 爲了要預測一指定線路在高速資料服務下的表現’量 化在一頻率範圍中衝擊服務型式的雜訊影響是重要的。舉 例而言，在100-200KHZ範圍的雜訊影響ISDN，SDSL和 HDSL ,但是ADSL對高達1·1ΜΗζ和T1/E1的雜訊敏感。 因此在較佳的實施例中，雜訊模型會包含給每個受測線路 可能會傳送的資料服務型式的雜訊衰減係數。這些係數是 藉由估計而決定，不論是透過計算，或是基於在代表性的 纜線圍束所進行的量測，出現在由每個資料服務使用的頻 率範圍內的雜訊。每一在纜線圍束中其它干擾訊號的組合 ，每一服務(ISDN，xDSL)用一些由服務供應商通常視爲典 型的組合，都有不同的衰減係數。雜訊模型可能會以每一 中央控制室或DLC交換機單元爲基礎被定義。這使得雜訊 模型能爲典型的資料服務和在任何特定埸所發現的迴路外 觀客製化。 預估的雜訊衰減係數可用分貝來表示，並可類推至線 15 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐〉 (請先閱讀背面之注意事項再填寫本頁) -------訂 i 線 533692 A7 _____B7_________ 五、發明說明（〇 ) 路上額外的損失。適當的雜訊係數可被加至特定線路的衰 減，造成校正過雜訊的衰減。舉例而言，假設雜訊模型指 # 出在ΙΟΟΚΗζ —些干擾訊號的組合雜訊係數爲6dB的ISDN ，且線路量測指出在ΙΟΟΚΗζ特定線路25db的衰減’作爲 ISDN服務絢路速度分級流程的一部份，衰減會被結合’所 以線路模型會指出在ΙΟΟΚΗζ 31dB的已校正雜訊的衰減。 無論如何，在此描述的實施例，雜訊模型是獨立在線路模 型之外的，並用以下描述的方法使用。 一特定雜訊模型的衰減係數，當在決定線路所支援的 速度時，會與線路的預測衰減一同使用，然而’這是在以 下所述，作爲速度分類流程一部份的分析背景下所達成的 〇 在較佳的實施例中，每一個雜訊模型以表格方式組織 ，每個服務形式(ISDN，ADSL，SDSL等)有一項目列在—雜 訊模型表格中，其中有以分貝表示的衰減係數的項目’給 不同的干擾訊號的組合(4 Tl，8 ADSL等），在表格中的每一 行與一不同的干擾訊號的組合相符合。每個中央控制室或 DLC可能有各別定義的表格。 •除了線路特性和雜訊特性外，資料傳輸率會被使用在 線路上的特定的數據機影響。在線路的用戶末端需要一台 數據機，在機房或在其它線路末端的數位迴路集中器需要 第二數據機。雜訊和線路特性會不一樣地影響不同的數據 機，因此，必須提供一個數據機對的模式。 對一特定的雜訊模型和線路模型，能將一特定的中央 16 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） (請先閱讀背面之注意事項再填寫本頁) Φ 訂ί 533692 A7 ___ __ B7__ 五、發明說明（#) 控制室或DLC數據機和客戶點的數據對能提供的上行和下 行的資料傳輸率的特徵找出，這些傳輸率會隨長度而改變 ’通常當迴路的長度增加時會減少。特定的長度與資料傳 輸率的關係會隨不同供應商的數據機之間改變，也會隨相 同長度的線路模型和不同構造(橋狀分接頭、線徑混合的結 合等）之間而改變。一個數據機模型爲每一組標準的線路模 型定義一指定的中央控制室/DLC上行和下行的速度，和爲 一指定的雜訊模型定義客戶點數據機組（干擾訊號的規劃） 〇 在較佳的實施例中，標準的線路模型是下列的其中一 種： 單線徑迴路(.32mm，.4mm，.5mm，.6mm，.8mm，26AWG， 24AWG，26AWG)在一對雙絞線或四絞線中。每個從500ft 到以500ft爲增加單位的各別長度存在著不同的模型。 線徑比例從整個迴路長度的10°/。到80%相異的混合線 徑迴路。每一各別的線徑組合，和每一各別的長度從 500ft.到以500ft.爲增加單位的各別長度，都存在一各別的 模型。 •一組工業和電信公司標準的線路模型，包括那些定義 爲 CSA，MIDCSA 和 ANSI 的。 測試單元2是以包含有關可能存在於執行速度預測的 線路上的數據機組表現的資訊的多數的數據機模型來程式 化。較佳的是，每一可能存在於線路上的數據機組都有一 數據機模型，每一數據機模型能以表格型式規劃。對每一 17 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） -------------Φ------ 丨訂---------線-- (請先閱讀背面之注意事項再填寫本頁) 533692 B7 五、發明說明（π ) 標準的線路模型，在數據機模型表中都有一列項目，其中 有不同的雜訊模型（干擾訊號的組合）上行和下行的速度的 項目。在一簡單的實施例中，線路模型可能單用一總長度( 包括橋狀分接頭長度)和橋狀分接頭計數來代表。更精確可 以替代地藉由使用更多變數來代表一線路模型而達成’舉 例而言，一線路可用一線段列表成模型，那些線段的長度 、線徑和位移，包含任何橋狀分接頭線段，以及線路是否 包含負載線圏或其它特色也可以紀錄在模型中。此外’爲 模型以分貝形式、在從ΙΟΟΚΗζ到1.1MHz的多數的頻率 量測的衰減也可以被包含在內。使用更多的變數來描述一 線路模型，當評估受測線路時，會導致更精確的分類’但 是一數據機模型也需要更大量的資料要被儲存。 在纜線圍束的雜訊會影響資料傳輸率，因此，數據機 模型表格應在表格中的每一項包含不同的資料傳輸率，其 與不同的雜訊等級相關連，如同由雜訊模型所辨識的。在 表格中的每一列與一雜訊模型相符，亦即，一些干擾訊號 的組合(例如4條T1線路，8條ADSL線路）。如前所述， 任何特定線路的干擾訊號是從有關何種訊號能在纜線圍束 中的線路上傳輸的運作規則所得到的，以及雜訊模型是用 來衰減其表現。如圖4所示，干擾訊號的數量當移到表格 右邊時會逐漸增加，行414A代表沒有干擾訊號的表現， 行414B代表當纜線圍束包含4條T1和12條ADSL訊號 在一 25對纜線圍束中的表現；同樣地，當產生的雜訊增加 時，預測的速度會減少，減少的速度是由從雜訊所產生的 18 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） " _____________Φ________^_________$ — #1 (請先閱讀背面之注意事項再填寫本頁) 533692 A7 _B7 _______- 五、發明說明（ 衰減所造成的，因此，使用圖4的資料結構去預測一資料 傳輸率，代表一替代性方法，即雜訊模型能用來衰減在線 0 路上的表現。 藉尋找與受測線路構造上最接近的列來索引表格內容 ，然後在符合一些特定的雜訊模型之列中的不同行間進行 索引，一特定的表格項目被選出。該選擇提供預期的資料 傳輸率給由標準的線路模型所呈現的建造的數據機組’ $ 選定的干擾訊號的組合。 在較佳的實施例中，在數據機表格中的資料項目是從 實際的量測而得。實際的量測能從實地的代表性線路上進 行，或是更佳地，它們可以在實驗室的設定下進行’在其 中線路長度和線路規劃設定都可以受控制。每一數據機組 實際上有兩個與每一作業狀況相關連的資料傳輸率’一爲 上行速度，一爲下行速度。上行速度代表資料從用戶數據 機至機房數據機的傳輸率；下行速度代表資料從機房數據 機到用戶數據機的傳輸率。許多資料服務，諸如ADSL ’ 是設計爲下行資料傳輸率比上行資料傳輸率高，結果上行 和下行傳輸需要不同的速度預測，因此，上行和下行的資 料傳輸率有各別的表格。 圖4說明一資料結構，其可用於一特定數據機組的一 數據機模型。需要的數據機模型的數量是由電信公司的作 業政策所定，其與電信公司使用何種數據機和其用戶被允 許使用何種數據機有關。資料結構400是存放在測試單元 2的電腦記億體中。描繪的模型是爲下行資料傳輸率，相 19 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） (請先閱讀背面之注意事項再填寫本頁) ·· 訂— 533692 A7 五、發明說明（β ) 似的表格也爲上fT貝料傳輸率存在著。行401代表標準的 線路模型’定義表格中列的索引。爲簡化說明，模以 圖型方式呈現’每一模型被顯示來代表一有不同線徑和有 或沒有橋狀分接頭線段的線路。如同上所述，線路模型能 包含許多變數。在一較佳的實施例中，每一線路模型藉由 使用代表一特疋値或相關變數範圍的數位資料來表示。藉 由將使用由測試單元2所作之量測的受測線路的模型，與 行402中的一項作配對，選出其中一個資料列，諸如 412A...4120 列402是表格行的索引。每~跨列4〇2的項目都與一 特定的雜訊模型（干擾訊號的組合）相關連。藉由選擇一特 定的雜訊模型，資料欄其中之一，諸如414A...414B會被 選擇。選擇的列與行定義了在數據機模型中的一個項目。 可能存在的雜訊狀況的數量是由電信公司的作業流程所指 定，因爲在說明的實施例中，雜訊模型是藉由在纜線圍束 中線路的使用所決定，這樣的選擇提供了在一使用特定數 據機的特定線路上的預測的下行資料傳輸率，該流程會重 覆以獲得上行部份的資料傳輸率。 •爲一特定標準線路、爲一些雜訊模型所選擇的上行和 下行的速度是作爲指定給受測線路的速度上限。更進一步 的評估，包含將與選擇的標準線路模型相關的衰減，和受 測線路的預估衰減作比較(de-rated給特定的雜訊模型），導 致一速度分級的決定。此評估也在任何選擇的標準線路模 型和受測線路之間的差異被納入考慮。 20 (請先閱讀背面之注意事項再填寫本頁) ΙΦ------ 丨訂---------線！ 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） 533692 A7 _____B7 —_ 五、發明說明（β ) 在較佳的實施例中’線路速度的預測是以一種易於理 解的型式來呈現，線路測試的結果可能會以類別分開或是 分類，其中每一類別代表一個不同的線路狀況。分^是藉 由使用爲此目的而設計的軟體的電腦來執行’其可被包含 成爲測試單元的一部份’或是成爲一從測試單元接收資料 的獨立的系統。分類可以更進~步地以顏色編碼，以使測 試結果容易辨識。當一特定的顏色組被描述，任何顏色或 是它指示的方法也可以被用來辨識線路測試的測試結果的 不同類別。舉例而言，紅色可以代表該線路無法支援高速 存取，綠色代表該線路可以支援高速存取，黃色代表該線 路目前無法支援高速存取’但是若阻礙移除時就可以，而 灰色代表線路的特性落在系統的專業範圍外。 灰色的特性描述提供兩種好處，第一，它使一個使用 者輕易地了解特定線路的狀態是不可知的；第二，它藉由 減少測試的涵蓋範圍而增加了分類預測的可靠度。舉例而 言，如果分類是灰色的線路不列入系統專業領域(及包含整 體受測線路的20%)的一部份，則全部受測線路的80%的一 群線路會被賦予預估達95%準確度的特性。如同在此所用 的，·「專業領域」代表系統的預測需依賴使用者的狀況。 如果無論如何分類是灰色的線路被納入成爲系統的專業領 域的一部份，則所有測試線路的100%的線路組可以賦予有 85%的準確度。因此，當預測有高度的不確定性時，線路 被描述成灰色，或是不確定的，而不是預測爲紅色、黃色 或是綠色。這讓使用者更能信賴紅色、黃色或是綠色的預 21 本紙張尺度適用中國國家標準（CNS)A4規格（21〇 X 297公釐） <請先閱讀背面之注意事項再填寫本頁) #------ —訂---------線丨· 533692 A7 ------B7__ 五、發明說明（/?) 測°當線路被分類成灰色時，電信公司可能會選擇進行更 完整的測試，或只是單純地選擇在纜線圍束中的其它線路 $ Ο 線路的測試和分類最好是在不同的高速存取速率下執 行，諸如 ADSL 於 2Mb/sec，ADSL 在 4Mb/sec，ADSL 在 8Mb/sec等，並依線路所支援的高速傳輸率來收取不同的 費用。舉例而言，接收2Mb/sec的存取的客戶以第一級費 率收費，接收4Mb/sec存取的另一客戶以不同（更高）的費 率收費，因此，線路可由電信公司所提供的不同服務類別 來分類。 對應線路模型和速度指定至一服務類別的指定可以被 視爲分類作業的第二步。電信公司較喜歡提供允許此對應 的商業規則，電信公司提供其欲提供的服務類別，舉例而 言，如果一電信公司已決定它將提供512Kbps的ADSL服 務，則已分配速度大於512Kbps的線路將會爲此服務被分 類爲綠色。 如果線路被指定速度低於512K，它可能被指定爲紅色 或黃色類別，舉例而言，如果線路模型顯示在線路上影響 狀況的服務，諸如一分歧對、一橋狀分接頭或是一負載線 圏，可能線路需要修理，然後以更高的速度運作。電信公 司會提供商業模型，其顯示是否電信公司要在線路上嘗試 修復。電信公司只在預測的線路速度在希望等級的50%之 內時會修復一條線路，或是電信公司會在商業模型中顯示 其會以一橋狀分接頭或是一負載線圈來修復一線路，但是 22 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） --------------------訂---------線—AWI (請先閱讀背面之注意事項再填寫本頁) 533692 A7 _ B7___ 五、發明說明（) 不會嘗試修復一分歧對。如果線路模型顯示一電信公司要 修復的狀況，線路會被分類爲黃色，否則，線路會被分類 參 爲紅色。 當在線路上執行的量測無法立即對應到任何設計至測 試系統的線路模型，則線路會被指定爲灰色，或是不確定 的類別。替代地，上述的許多預測技術提供與預測有關的 信心因數。當預測的信心在一些門檻之下，線路也會被分 類爲灰色。精確的信心門檻也是電信公司的商業模型的一 項功能，舉例而言，一家有許多可用線路的電信公司可能 會被定一高的信心門檻，儘管高的信心門檻會造成許多線 路被列爲不合格，對電信公司而言乾脆使用其它線路會比 較便宜。另一方面，一家線路有短缺線路的電話公司會設 定一個較低的門檻，即使這樣的設定增加了建置來服務的 線路稍後需要修復或是更換的風險，電信公司可能會藉由 使用更多的線路用作高利潤資料服務而賺得Μ乡。 電信公司特定的商業模型最好是在測試單元中編碼於 軟體中。 參照圖一所顯不的現在的裝置：測試單元2包含一電 壓來源和量測單元3，以及一預測單元4，測試單元2位在 機房5的父換器(或DLC)6之後或是櫃，交換機6連 接測試單元2到電話線路9的tiP7和ring8，交換機6使得 測試單元2很容易地測試多條線路，測試可以視需要或是 定期地來進行。 測試單兀2透過父換機6被連接至尖端7和環8線， 23 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） ----- (請先閱讀背面之注意事項再填寫本頁) #---- 訂i 533692 A7 ____B7 ____ 五、發明說明（w ) 電壓來源和量測單元3提供一來源訊號到尖端7和環8線 ，並量測有關特定受測線路的不同的特性’量測可以包含 電壓値（真實的和想像的）、電流、電容、阻抗、電阻和相 類似的，量測可用多種不同的頻率重覆’測試單元2使用 由電壓來源和量測單元3所構成的量測’來預估受測線路 的特性，此特性是指線路的性質和在線路上元件的存在。 一旦特性被預估’預估的單元使用該特性來預測線路 能支援的資料傳輸率’此預估的過程會由一所謂的專家系 統來進行，專家系統會利用神經網路、統計模型、決策樹 模型以及其它模型，或是這些模型型式的組合。 預測單元4會與電壓來源和量測單元3被共同放置在 機房或是DLC中，或是，預測單元會被放置在遠端，並藉 由一線路或是無線傳輸介面與量測單元溝通。 一旦一線路被量測，該線路會透過電換機6從測試單 元2離線，而其它線路可以連接至測試單元，然後執行測 試於下一條線路。一旦一組線的預估已估計出來，線路測 試的結果會被分類爲詳述於下的。 參考圖二流程1所示：流程10的第一步包含測試線路 ，當量測不同訊號的參數於線路內和線路本身時，測試單 元連接至一選定的線路並***訊號到線路中，***的訊號 可能以多種的頻率和等級提供。 在下一步驟20，線路的特性從量測而被估計，作爲一 個例子，特性包含線路的***損失、線路的相位不平衡、 線路的長度和線徑。額外的特性包含線徑改變、橋狀分接 24 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公爱)" ""~ -------------#-------訂---------線-♦ (請先閱讀背面之注意事項再填寫本頁) 533692 A7 _____B7_____ 五、發明說明（A) 頭、負載線圏和其它路徑參數，諸如分歧器、過濾器、終 端和在線路中或是延線路類似的東西。 接下來的兩個步驟建立數據機和雜訊環境以供使用， 當評估線路能支援何種資料傳輸率。步驟21選擇下一(第 一）個數據機模型來使用，然後步驟21選擇下一（第一)雜 訊模型。 下一步驟25使用一流程來預估基於量測和預估的特性 ，線路能支援的資料傳輸率，此外，預測的可靠度和涵蓋 範圍會是測試結果的一部份’該流程會被實現成爲一專家 系統。 步驟30包含決定線路測試結果是否落在系統的應用範 圍之外，如果線路的特性真的落在系統的應用範圍之外， 則步驟35被執行。在步驟35有這樣的條件的結果會以顏 色編碼。儘管任何顏色都可被用於任何類別，在一特定的 實施例中，灰色被用於以顏色編碼這個類別，因此使得有 這樣實驗結果的線路能輕易地被辨識。如果線路特性沒有 落在系統的應用範圍之外，則步驟40被執行。 步驟40決定是否從測試系統的線路測試結果顯示該線 路能支援高速存取，如果情況是如此，則步驟45被執行。 步驟45以顏色將此結果編碼爲綠色。對任何顯示爲無法支 援高速存取，步驟50被執行。 步驟50決定是否從測試系統示該線路目前無法支援高 速存取的線路測試結果，但是該線路在障礙移除後能支援 高速存取。障礙可能包含舉例而言，一負載線圏或是一橋 25 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公爱) '—- -------------#------- —訂--------- 線*丨黌 (請先閱讀背面之注意事項再填寫本頁) 533692 A7 __B7____ 五、發明說明（/Μ 狀分接頭。如果線路在移除障礙後能支援高速存取’此線 路會被顏色編碼爲黃色，如同步驟55所示，如果線路在移 除障礙後仍無法支援高速存取，則步驟60被執行。 步驟60爲剩餘線路的結果顏色編碼爲紅色，表示該線 路無法支援高速存取，這有許多的原因(例如線路太長）。 下一步驟，步驟65，爲選定的數據機和雜訊模型提供 資料傳輸率和顏色編碼給使用者。步驟70決定一選定的數 據機模型是否有更多的雜訊模型可供考慮。 步驟80決定是否有更多的數據機模型可供考慮，當再 也沒有數據機模型可供使用，流程1結束。在此時，該線 路被指定一上行和下行的資料傳輸率，和爲每一可能的雜 訊組合以及重要的數據機模型，給服務供應商。 藉由顏色編碼的結果，符合品質驗證標準能支援高速 存取的線路使用者，可能會被接觸以提供高速存取給使用 者。分類可以在以每一服務階級的基礎上來進行，一線路 可能在G.lite(小於1.5Mb/sec的ADSL)分類爲綠色，但是 在G.dmt(大於1.5Mb/sec但是小於或等於8Mb/sec的 ADSL)被分類爲紅色。在這個例子，指定爲紅色可能是因 爲線路太長，無法支援G.dmt，此爲電信公司無法解決的 問題類型。雖然上述的說明討論從多條線路的測試結果， 此方法當測試結果是從只能提供單一線路的情況，也是有 用的。 參考圖3流程2所示，第一步驟85是要取得在流程1 所量測和（或）預估的路特性，以及每一數據機模型和雜訊 26 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公爱） --------------------訂---------線— (請先閱讀背面之注意事項再填寫本頁) 533692 A7 ___B7_ 五、發明說明（W) 模型組合的上行和下行資料傳輸率預測。下一步驟90，選 擇定義服務類別的下一（第一）服務供商，與此服務類別有 關的資訊可能包含最小和最大的保證上行和下行資料傳輸 率、是否包含voice-over-xDSL服務、供應商網路頻寬的 可取得性等等。 下兩個步驟建立數據機和雜訊環境以供使用，步驟 1〇〇選擇下一(第一)數據機模型來使用，接著步驟11〇選擇 下一(第一)雜訊模型。下一步驟120使用服務類別資訊， 連同線路特性和速度資訊，來執行服務類別的指定，亦即 ，預測線路支援特定數據機和雜訊模型服務的能力，此外 ，預測的可靠度和涵蓋範圍可能是指定結果的一部份，該 流程可被實現爲一專家系統。 步驟130包含決定線路可取得的資料(量測、速度預測 、服務能力預估等）否落在系統的應用範圍外，如果該資訊 確實落在系統的應用範圍之外，則步驟135被執行。在步 驟155有這標準的結果會以顏色編碼。雖然任何顏色都可 以用於這個類別，在一特定的實施例中，灰色被用來爲這 個類別指定作顏色編碼，因此使得結果容易被辨識。如果 可取得的資訊沒有落在系統的應用範圍之外，則步驟140 被執行。 步驟140決定是否可取得的資訊顯示線路爲特定的數 據機和雜訊模型支援特定的服務，但此情況，則步驟145 被執行。步驟145以顏色將此結果編碼爲綠色。對任何顯 示爲無法支援特定服務，步驟150被執行。 27 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） (請先閱讀背面之注意事項再填寫本頁) I --I 1 I I I ^ -----I I I I · 533692 A7 ___ B7_ _ 五、發明說明（〇 步驟150決定是否可取得的資訊顯示線路現在無法支 援特定的服務，但是該線路在障礙移除後能支援該服務。 障礙可能包含速度障礙（負載線圏、橋狀分接頭等或可 能是供應商的限制（頻寬、路由能力、商業政策問題等）。 如果線路在移除（或是定址）障礙後能支援特定服務類別， 此線路會被顏色編碼爲黃色，如同步驟155所示，如果，線 路在移除障礙後仍無法支援特定服務，則步驟160被執行 〇 步驟160爲剩餘線路的結果顏色編碼爲紅色，表示該 線路無法支援特定服務類別，這有許多的原因(例有如沒有 可用的DSLAM能力），線路被顏色編碼爲黃色或紅色會與 解釋原因爲何相關聯。 下一步驟165，爲選定的數據機和雜訊模型的特定服 務類別提供顏色編碼給使用者。步驟170決定一選定的數 據機模型是否有更多的雜訊模型可供考慮。下一步驟180 決定是否有更多的數據機模型可供考慮，當再也沒有數據 機模型可供使用，則下一步驟190，決定是否有更多的服 務類別可供考慮，當再也沒有服務類別可供考慮，流程2 結束。在此時，該線路被指定一顏色編碼給每一可能的服 務類別的組合以及重要的雜訊和數據機模型，給服務供應 商。 在說明了本發明的較佳實施例後，對那些其它使用這 些觀念的實施例的那些一般習知的技術，現在變得明顯。 雖然描述的是一特定線路組的特性和線路元件，但是其它 28 本纸張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） (請先閱讀背面之注意事項再填寫本頁) 訂----- 線！ 533692 A7 _____ —__B7_ 五、發明說明（ >心） 的線路特性和線路元件也可以被包含在內。此外，包含作 爲工具的一部份的軟體可以一電腦程式產品而實例化，其 包含一電腦可用的媒體，舉例而言，這樣的電腦可_用的媒 體能包含一可讀取的記憶裝置，諸如硬碟機、光碟機、數 位影音光碟、或是電腦磁碟機，都有電腦可讀取的程式碼 片段儲存於其上。電腦可讀取的媒體也可包含通訊連結， 不論是光學、有線或是無線，其有程式碼片段在其上，以 數位或類比訊號傳輸。相同地，提出說明的是本發明不應 侷限於已說明的實施例，而只應受限於附加的申請專利範 圍的精神和領域。 另外，必須要知道的是一個兩維的表格結構會爲數據 機模型來描述，其中表格中的欄被由線路的長度所定義。 必須要知道的是線路的長度能被視爲只是另一定義線路的 變數，依此方式，兩維表格可以視覺化視爲一維淸單。 其它在電腦記憶體中的資料結構也可能儲存數據機模 型而不背離發明的觀念，舉例而言，描述一各別的上行和 下行表格被用來識別資料傳輸率。必須了解的是當相同的 雜訊模型被用作上行和下行傳輸，上行和下行値可以藉由 單純地在圖4的每一數據機模型表格的子記錄包含一上行 和一下行資料傳輸率而獲得。 同樣也描述表格查詢被用來決定與一特定雜訊模型有 關的資料傳輸率，同樣地，資料傳輸率可藉由計算而得。 假設雜訊不存在，雜訊因素可被用來衰減在線路上的速度 預測。 29 尺度適用中國國家標準（CNS)A4規格（210 X 297公f ~ -------------#------ 1訂---------線--AWI (請先閱讀背面之注意事項再填寫本頁)533692 A7 ___ B7__ V. Description of the invention (/) Cross-reference with related patents This patent application claims the priority of US Patent Application No. 60 / 209,057 filed on June 2, 2000. Background of the invention The present invention relates generally to communication networks, and more particularly to systems for testing the quality of telephone lines used for transmitting data. As is known in the art: public switched telephone networks, such as the so-called legacy telephone service (POTS) lines, were originally designed to carry voice and cover a limited bandwidth (for example, approximately 4KHz); today they want to use the same POTS Line to transmit data. However, data signals usually have different frequency characteristics than audio signals, including larger bandwidths. As a result, a POTS line capable of transmitting audio signals normally may not work properly or may not transmit at all. The telephone company must know which lines are suitable, such as qualified, and those lines are not suitable for transmitting data; the telephone company must also know why specific lines cannot support data transmission, and where these errors occur, so that they can Determines whether the transmission line can be corrected. Line conformity certification refers to the overall ability to describe the quality of the user's circuit, as it relates to the ability of the line to transmit sound (such as POTS) or data services. This is especially for certifications that support high-speed data transmission lines, such as ISDN, SDSL, ADSL or HDSL. The prior elimination is to describe the user circuits that cannot support data services without corrective action, and to provide a high-confidence description capability. The pre-qualified certification is to describe the user circuits that can support data services, and have a high-confidence description capability. 4 This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) — — — — — — — — — — — — II 1 II 1 (— — — — — — II II (Please read first Note on the back, please fill out this page again) 533692 A7 ___B7___ V. Description of the invention (2) The telecommunications company (TEL central control room, s) has two problems in solving the problem of the user circuit of transmitting data. The first problem is ^ Is strategic: telecommunications companies are reluctant to deploy emerging technologies (such as ISD & or ADSL) used to transmit data, because in their perception, enough user circuits have high enough quality to make deployment economical Success is uncertain. This discourages early adopters, because being the first to deliver a technology that may not be successful in their access to the network is a major risk. If telecommunications companies can be given a technology, In order to reduce the risk significantly during the initial deployment, they will retain market share and stay ahead of the competition. The second problem is tactical and determined by the telecommunications company. After the deployment of a specific technology, whether it is active or passive, when the service may be requested by the user, or locked by the telecommunications company for transmission, or the telecommunications company is forced by law to provide a sufficient number of lines to CLEC ( The central control room, mpetitive Local Exchange Carrier, often requires quality inspections of the lines that provide the service. For example, if a telecommunications company wants to enter the market and provide new services, they will want to Lock out those customer loops that are most likely to support service and / or work the least outside of the box; another example is when a telecommunications company receives a new service request from a customer, it needs to accept based on the condition of their line Or reject the information requested by the new service. Line quality verification is usually performed by single-ended or double-ended testing. If testing with double-ended testing 'will send a technician to both ends of the tested line, the tested line comes from the network Offline, then test instruments are installed at both ends of the line. The test instruments work together to determine if the line supports the required type of Material transmission. Double 5 scales are applicable to China National Standard (CNS) A4 specifications (210 X 297 mm) " -------------------- Order ----- ---- (Please read the precautions on the back before filling this page) 533692 A7 -____ Β7 __________ V. Description of the invention (;) The test of the terminal is not because of the time and cost of installing the instrument at both ends of the line, so it is not Not recommended; in addition, double-ended testing usually provides results for a specific type of data service, so if the type of data service changes, the test needs to be redone. In terms of xDSL line quality verification, both methods are now in use. One method is to send a technician to the upper end of the circuit under test, such as a computer room (central control room) or a digital loop carrier (DLC). The technician goes offline from the line under test, and then connects a series of special-purpose instruments, such as a load line or time domain reflectometer to detect bridge taps, or a digital multi-purpose to find impedance errors. Electricity meter. This method can only give an incomplete indication of signal attenuation on the line (we found it to be a very important early warning in line performance), and it is too time-consuming and expensive when a large number of high-speed data services of a specific type are deployed. Others have attempted to use automatic measuring instruments, which have been developed to test lines that affect voice service errors, perform single-ended pre-quality checks, and rely on telecommunications companies to maintain records for their lines to provide additional information. Cable information can indicate whether the line is planned in a way that is known to be unsuitable for a particular data service. For example, the line's plumbing record and whether a load line has been installed may both indicate that the line cannot support a particular type of service . However, the review of cable records is usually performed manually, resulting in unacceptable long-term testing times. In addition, telecommunications companies often find that the required records are incorrect because the required information is entered incorrectly or Never updated. What's more, the line characteristics that affect the speed of digital data services do not necessarily cause any attenuation of traditional voice services. For many years, 6 paper sizes are applicable to China National Standard (CNS) A4 (210 X 297 mm) -------------- # ------- Order ----- ---- Line 丨 1 · (Please read the notes on the back before filling out this page) 533692 A7 __B7___ V. Description of Invention (>) There is no reason for telecommunications companies to save what they need now Information records for quality inspection of line works for high-speed data services. Therefore, cable records usually do not contain the required information. > The current technical limitations of pre-quality inspection prevent telecommunications companies from predicting with high confidence whether their lines can support certain types of data services. These restrictions have forced service providers to restrict their provision of high-speed data services to specific regions. In these areas they believe that these lines are capable of supporting them, or can only promise customers that their services can only operate at less than their full speed possible. It would be of great benefit to the telecommunications company to determine quickly and without a great deal of expense which lines can support high-speed data services and the speed at which their lines can operate. SUMMARY OF THE INVENTION In the foregoing context, it is an object of the present invention to provide a method and apparatus for performing a line quality inspection test and classifying the results of such tests. The lines are tested to determine or estimate different line characteristics. The physical characteristics of the line are estimated (such as wire length, wire diameter, insertion loss); instruments present on the line, such as load coils, bridge taps, terminations, and the like are also determined. The estimation of the data transmission rate that the line can support will depend on the measurement and estimation of the line. Therefore, the behavior of the line at low frequencies (less than 100KHZ) will be used to estimate the high frequency (over 100KHZ). Alternative to line behavior. Testing and measurement are preferred to be performed at low frequencies so that they can be performed through telecommunication company voice exchange. The results will be classified according to certain standards and provide an easily distinguishable line condition. 7 This paper size applies to China National Standard (CNS) A4 specifications (210 X 297 public love) ------------- -------- Order --------- Line—AVI (Please read the precautions on the back before filling this page) 533692 A7 ---- --- B7 ___——_ V. Description of the invention (^). Classification can be performed by using a computer with software specially designed for this purpose. In a preferred embodiment, the classification result includes a first category, where the lines that cannot support high-speed access of a specific level are marked; the classification result also includes a second category, where it is marked that can support a high-speed access. High-speed access lines of a specific level; the classification results also include a third category, which indicates that the current is not possible, but can support a specific high-speed access line after removing the obstacle; a fourth category, which is marked by The selected line characteristics fall outside the scope of the test system and are also included. Each category is assigned a separate color so that the condition of the line can be easily identified. Testing and classification may be performed for various high-speed access levels. Customers are charged at different rates, depending on the level of service available on their line. Brief description of the drawings The present invention can be more easily understood through the following more detailed description and the corresponding drawings, in which: Figure 1 is a block diagram of the device of the invention; Figure 2 is a flowchart of the method of the invention; Third is a flowchart of the method of the present invention; and Figure 4 is a sketch of the data structure used to represent a data model. Detailed description Today there is a great demand for providing high-speed access, such as the Integrated Services Digital Network (ISDN) or different types of Digital Subscriber Loop (xDSL) services to Internet users. Most of these users want to receive such access through their existing legacy telephone service (POTS) lines. pots 8 This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) -------------------- Order -------- -丨 (Please read the notes on the back before filling this page) 533692 A7 _ —_B7 _____ 5. Description of the invention (6) (Please read the notes on the back before filling this page) The line was originally built to carry voice, and Not necessarily for higher frequencies related to ISDN or xDSL. Certain ratios of POTS lines can provide the high-speed access required by users. Predicting the ability of existing POTS lines to support high-speed access is usually labor-intensive, requires a lot of time, and requires skilled people to explain the results. The present invention provides automated testing of existing POTS lines to estimate the ability of the line to provide high-speed services, and further presents the results of the tests in an easy-to-understand format. The tests performed on the line, as part of the characteristics of the line, take a variety of different measurements and estimates. The characteristics of the line, which can be estimated by testing, include the insertion loss of the line, the phase imbalance of the line, the length and diameter of the line. Additional features determined by line testing may include decisions and estimates of changes in wire diameter, bridge taps, load coils, and other path elements such as dividers, filters, terminations, and the like. In addition, speed jammers, which serve other high-speed services in cable enclosures, will also be considered. The second part of the line test involves using a procedure based on measured and estimated characteristics to predict the data transfer rates that the line will support. In addition, the reliability and coverage of the prediction are also included as part of the test results. A description of the line characteristics determined or estimated through the test procedure is shown below. Assuming that there is no additional condition for a telephone line, the operation of the service on the line will be significantly affected by the overall attenuation or insertion loss of the line to which the signal is added. At present, telecommunications companies measure the insertion at different frequencies (such as 100KΗζ, 300KΗZ, etc.) through hand-held instruments by deploying personnel to both ends of the line. Mm> 533692 A7 _______B7_____ 5. Description of the invention (1) Loss to measure insertion loss. The current test system includes a method for estimating the insertion loss of a line from a single-ended measurement of the line. The phase imbalance of the line is also a pre-measurement. An important consideration in assessing the ability of a line to handle high-speed access. A telephone line is composed of two metal wires, called a tip and a ring. Usually the tip and the ring should have the same electrical characteristics. Two The wires should be balanced. In a balanced wire, the impedance, current, and inductance of each wire will be the same. If the impedance, current, and inductance of one wire are different from the other, an imbalance will exist. Imbalance will cause Phase shift of the difference between two lines in the loop. Higher-speed modems encode many bits on the carrier frequency to phase shift, so it is even unbalanced. Will cause reduced data flow, errors, and retraining. In ISDN, the phase shift will cause the energy from one pulse to overlap into the synchronization signal or to the time taken by another pulse, resulting in internal Signal distortion and / or loss of synchronization. Both analog and digital transmission methods are sensitive to slight instabilities in phase imbalances. It is important to detect imbalances because large imbalances can borrow The POTS is affected by reducing the loop current. Impedance imbalance may be so large (2 kilo-ohms or more) that it becomes impossible to catch even a dial tone, or the ringing current may not be sufficient to start the call or It is the ring circuit in the modem. It is also important to detect the imbalance of data below 2 kilo-ohm when data transmission is taken into account. Any series of impedance and the sound that causes phase shift will affect the data transmission The quantity has a negative impact, and it may be achieved on the customer's circuit. Changes in wire length, wire diameter, and wire diameter along the wire can be tested and measured by the wire or through the telecommunications company's The record is estimated. The length of the line is 10 Ϊ The paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) " '------------- · ---- --- Order --------- Line 丨 ♦ (Please read the precautions on the back before filling this page) 533692 A7 _ __. . . _ B7 _ 5. The invention description (S) plays a role in high-speed access, because the line may be too long to support high-speed transmission access. Changes in line diameter and any line diameter on the line will also affect the line's ability to support high-speed access. Changes in wire diameter are considered impedance and will affect high-speed signal transmission on the line. The telephone network is usually optimized for POTS signals. One way to optimize the network is to include the load coil on the telephone line. The load coil improves the frequency response of the analog signal line. The load line on the line will generate a response, which affects the line processing high-speed service. Ability. In the United States, many subscriber lines have bridge taps. This is due to the way telephone lines are placed in the house branch, and the phone lines are laid out in advance to determine the exact location of the house in the branch. The line passed Xu. Near the location in the multi-house plan, when the house is built later, the builder will connect the telephone unit to the nearest point on the previously deployed telephone line. The unconnected part of the previous line will generate a bridge tap. The bridge tap will reflect the signal from the terminal. The reflected signal will then return to the user line and interfere with the signal on the user line. The most harmful interference occurs when the reflected signal and the incoming signal are out of phase. In this case, the reflected signal will cause severe interference to the incoming signal on the subscriber line. The length of the bridge tap indicates the phase difference between the original and reflected signals. For high-speed digital signals, the frequency will extend to about 1MHZ, such as ADSL signals. If the length of the bridge tap is between about 200 to 700 inches, there will be a significant cancellation. In the United States, the length of bridge taps left by the construction of many housing divisions falls within this range. Therefore, the ability to detect and remove the bridge taps, and to provide high-speed digital data services 11 ^ ---- _ This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) (please first Read the notes on the back and fill in this page) ·· Order i-line — φ · 533692 A7 _______B7_ _ V. Description of the invention (7) It is useful for telecommunications companies that provide services to their customers. The creation of new services like ADSL, if the telecommunications company's central control room (Central Control Room) is equipped to handle the provision of these new services to users, there can be no prediction. For example, if the line under test is equipped with a splitter and half of a ringer but there is no telephone, the standard test system may be misled into thinking that there is a ring on the user's line. ADSL adds a central control room splitter and hypothetical splitter on the line. The central control room splitter uses a diode to provide polarity sensitivity and conduction saturation outside the range of the ADSL signal field. The current in the central control room splitter can only flow when the ring side is more forward than the tip side. The measurements performed by known test systems do not allow this, so these test systems currently in use cannot detect a properly installed central control room diverter. In order to correctly find the characteristics of the line, it is important to determine or estimate the remaining components on the line. The circuit model of the loop is generated from physical measurements on the line. It is best to use an automatic measurement system installed in the telephone network. In a preferred embodiment, the circuit model contains loop information, which includes length, line Indication of diameter, bridge tap, and indication of load line and its attenuation. The technology for obtaining this loop information is described in Teradyne's different patent applications, or it is included in commercially available products such as Celerity, which is the telecommunications division of Teradyne Corporation in Deerfield, Illinois, USA For sale. Better techniques can be found in U.S. patent applications: No. 09 / 294,563, filed by Groessl et al. On April 24, 1999; No. 09 /, filed by Rudinski et al. On September 30, 1999 No. 410,236; Faulkner et al. 12 ______________-------- ^ --------- ^ — L ^ wl Γ Please read the general meanings on the back before filling in this page) This paper size applies to China National Standard (CNS) A4 specification (210 X 297 mm) 533692 A7 ___B7_____ V. Description of the invention (,.) No. 09 / 410,222 filed by a person on September 30, 1999; by Holeva et al. In September 1999 No. 09 / 409,141, filed on the 30th; No. 09 / 410,2 $ 7, filed by McGill et al. On September 30, 1999. The aforementioned application is incorporated herein by reference. Furthermore, the model can include information that may sometimes be considered DC error information. Such errors can be detected on the line using DC measurements. This information includes tip grounding errors, ring grounding errors, and crossover errors. The line model also includes measurement of low frequency noise and evaluation of which terminals are on the line, such as mechanical or electronic telephones or internet interface equipment. Such errors and conditions can be detected with commercially available products, such as the 4TEL II sold by Teradyne's Communications Division in Deerfield, Illinois, but other products are also commercially available and measured. . Further information on how to obtain such information can be found in US Patent 5,699,402, which is incorporated herein by reference. These measurement techniques are better because they allow the use of relatively low frequency measurements to determine the condition of the line. The use of low frequency measurements allows the test system to switch to a specific line from an existing switch. Existing switches usually have low frequency response. . For example, when performing direct measurement on bridge taps, a bandwidth of at least 300KHZ is required. As will be described below, the performance of the line can be estimated by using low frequency measurements and predictions. It is not an important limitation of the present invention to include those items in the line model. Usually more information will lead to more accurate classification of the line, but it will also take more time and require more complicated measurement equipment or software. . Furthermore, those qualities that are most helpful in predicting performance will vary with 13 paper sizes that apply the Chinese National Standard (CNS) A4 (210 X 297 mm) (Please read the notes on the back before filling out this page ) Order line 533692 A7 ____B7 _ 5. Description of invention (I 丨) The network varies, for example, bridge taps are more common in the United States than in many European countries, so the line model is used in the United States to include The indication of the presence of a bridge tap may be important, but such quality may become unimportant when used in a European network. It is expected that a better set of circuit measurements and characteristics will be partially selected through the use of empirical evidence. To classify a particular line into the appropriate category, two additional models were established: one model describes the characteristics of the noise on the line, and the other defines the capabilities of the modem used in high-speed digital services. Compared with the measurement obtained from this line and the information about a specific line, the noise model is preferably obtained from the information about the way a specific network routing area operates a telecommunications company. Although real noise on the measurement line can be used as a possible alternative, it is still difficult to measure high frequency or wide area noise with existing automated measurement equipment. Such equipment is usually connected to the line through a switch, which has a low-pass frequency response, making it difficult to measure high frequencies in high-frequency or wide-range noise. For example, direct measurement of wide-area or pulse noise requires a bandwidth of 1.1 MHz. Using existing switching equipment, it is unlikely to automatically connect a test system to a line to provide sufficient bandwidth for reliable execution. Required noise measurement. Noise models can be selectively obtained from information about how a telecommunications company operates its network. A major source of noise on telephone lines is crosstalk. Telephone lines are usually routed in a cable bundle from the central control room to the vicinity of the user, and lines to all nearby users are also included in the cable bundle. Although efforts have been made in the online design to reduce the amount of signals accompanying from one line to another, there will still be some accompanying signals, and the accompanying 14 paper standards are applicable to the national standard (CNS) A4 specification ( 210 X 297 mm) a '' one nnnn ϋ nnnnnn ϋ I > ϋ ϋ ϋ II nn one 4OJI I nnn I n ϋ II (please read the precautions on the back before filling this page) 533692 A7 __B7__ 5. Description of the invention (A) The signal is called crosstalk noise. The amount of combination from one line to another depends on the frequency of the transmitted signal, and the amount of crosstalk noise 'therefore depends on what other type of signal is transmitted in the cable bundle. Service providers follow the design rules for composite signals transmitted in cable bundles. For example, they may specify the maximum number of lines used for ADSL or T1 or other signals that transmit high-frequency signals to ensure that Many high-frequency signals that cause crosstalk noise. Based on the design rules of the service provider, the estimated maximum noise expected on a particular line can be calculated or experimentally obtained. In order to predict the performance of a given line under high-speed data services, it is important to quantify the noise impact on the service type in a frequency range. For example, noise in the 100-200KHZ range affects ISDN, SDSL, and HDSL, but ADSL is sensitive to noise up to 1.1MHz and T1 / E1. Therefore, in a preferred embodiment, the noise model will include a noise attenuation coefficient for the data service type that may be transmitted for each circuit under test. These coefficients are determined by estimation, whether by calculation or based on measurements made on representative cable envelopes, and noise that appears in the frequency range used by each data service. Each combination of other interference signals in the cable bundle, each service (ISDN, xDSL) uses a combination that is generally considered typical by the service provider and has a different attenuation coefficient. Noise models may be defined on a central control room or DLC switch unit basis. This allows the noise model to be customized for typical data services and the appearance of loops found in any particular frame. The estimated noise attenuation coefficient can be expressed in decibels, and can be analogized to line 15. This paper size is applicable to China National Standard (CNS) A4 (210 X 297 mm) (Please read the precautions on the back before filling this page) ------- Order i-line 533692 A7 _____B7_________ 5. Invention description (〇) Additional losses on the road. Appropriate noise coefficients can be added to the attenuation of specific lines, resulting in attenuation of corrected noise. For example Assume that the noise model refers to ISDN with the combined noise coefficient of 6dB for some interference signals, and the line measurement indicates the attenuation of 25db on a specific line at 100KΗζ as part of the ISDN service speed classification process. The attenuation will be combined, so the line model will indicate the attenuation of the corrected noise at 100KΗζ 31dB. However, in the embodiment described here, the noise model is independent of the line model and is used in the method described below. The attenuation coefficient of a specific noise model is used with the predicted attenuation of the line when determining the speed supported by the line. However, 'this is the following In the preferred embodiment, each noise model is organized in a table, and each service form (ISDN, ADSL, SDSL, etc.) has a list of items. In the —noise model table, there are items of attenuation coefficient expressed in decibels' for different combinations of interference signals (4 Tl, 8 ADSL, etc.). Each row in the table is associated with a different combination of interference signals. Yes. Each central control room or DLC may have its own defined form. • In addition to the line characteristics and noise characteristics, the data transmission rate will be affected by the use of specific modems on the line. A user terminal at the line requires one Modem, a digital circuit concentrator in the computer room or at the end of other lines requires a second mode. Noise and line characteristics will affect different modems differently. Therefore, a mode of a modem pair must be provided. For a specific The noise model and line model can adapt a specific central 16 paper size to the Chinese National Standard (CNS) A4 specification (210 X 297 mm) (Please read first Please fill in this page again before filling in this page) Φ Order 533692 A7 ___ __ B7__ V. Description of the invention (#) Find out the characteristics of the uplink and downlink data transmission rates that can be provided by the data in the control room or DLC modem and customer point. These transmission rates will change with the length 'usually decreases when the length of the loop increases. The relationship between the specific length and the data transmission rate will vary between modems from different vendors, and will also vary with the same length of the line model and It varies between different structures (bridge taps, combined wire diameters, etc.). A modem model defines a specified central control room / DLC uplink and downlink speed for each set of standard line models, and The specified noise model defines the customer point data unit (interference signal planning). In a preferred embodiment, the standard line model is one of the following: Single wire loop (. 32mm ,. 4mm ,. 5mm ,. 6mm ,. 8mm, 26AWG, 24AWG, 26AWG) in a twisted pair or quad twisted pair. There are different models for each length from 500ft to 500ft increments. The wire diameter ratio is 10 ° / from the entire loop length. Up to 80% different hybrid wire loops. Each individual wire diameter combination, and each individual length from 500ft. To 500ft. To increase the individual lengths of the units, a separate model exists. • A set of standard line models for industry and telecommunications companies, including those defined as CSA, MIDCSA, and ANSI. Test unit 2 is programmed with a majority of modem models containing information about the performance of the modems that may be present on the line performing the speed prediction. Preferably, each data unit that may exist on the line has a modem model, and each modem model can be planned in a tabular format. For each 17 paper sizes, the Chinese National Standard (CNS) A4 specification (210 X 297 mm) is applied ------------- Φ ------ 丨 Order ----- ---- Line-(Please read the notes on the back before filling this page) 533692 B7 V. Description of the invention (π) The standard line model has a list of items in the model table of the modem, which have different miscellaneous items. A signal model (combination of interference signals) is an item of the speed of the uplink and downlink. In a simple embodiment, the line model may be represented solely by a total length (including bridge tap length) and bridge tap counts. More precise can instead be achieved by using more variables to represent a line model. For example, a line can be modeled with a list of line segments whose length, diameter, and displacement include any bridge-like tap segments, And whether the line contains a load line or other features can also be recorded in the model. In addition, the model is in decibels, ranging from 100KΗζ to 1. Attenuations for most frequencies measured at 1MHz can also be included. Using more variables to describe a circuit model will lead to more accurate classification when evaluating the circuit under test ', but a modem model also requires a larger amount of data to be stored. Noise around the cable will affect the data transmission rate. Therefore, the modem model table should include different data transmission rates in each entry in the table, which are related to different noise levels, as if by the noise model Recognized. Each column in the table corresponds to a noise model, that is, a combination of some interference signals (for example, 4 T1 lines, 8 ADSL lines). As mentioned earlier, the interference signal of any particular line is derived from the operating rules regarding which signals can be transmitted on the lines in the cable bundle, and the noise model is used to attenuate its performance. As shown in Figure 4, the number of interference signals will gradually increase when moved to the right of the table. Line 414A represents the performance of no interference signals, and line 414B represents when the cable bundle contains 4 T1 and 12 ADSL signals at 25 pairs. The performance in the cable bundle; similarly, when the noise generated increases, the predicted speed will decrease. The reduced speed is caused by the 18 paper sizes generated from the noise. The Chinese paper standard (CNS) A4 specification ( 210 X 297 mm) " _____________ Φ ________ ^ _________ $ — # 1 (Please read the precautions on the back before filling out this page) 533692 A7 _B7 _______- 5. Description of the invention (caused by attenuation, therefore, use the information in Figure 4 The structure to predict a data transmission rate represents an alternative method, that is, the noise model can be used to attenuate the performance on line 0. The table content is indexed by finding the column closest to the structure of the line under test, and then matches some specific Indexing between different rows in the Noise Model column, a specific form item is selected. This selection provides the expected data transfer rate to that presented by the standard line model The built data unit '$ selected combination of interference signals. In a preferred embodiment, the data items in the data machine table are obtained from actual measurements. The actual measurements can be taken from representative lines in the field Or better, they can be performed under laboratory settings, where line length and line planning settings can be controlled. Each data unit actually has two data transmissions associated with each operating condition Rate 'one is the uplink speed and the other is the downlink speed. The uplink speed represents the transmission rate of data from the user modem to the computer room modem; the downlink speed represents the transmission rate of data from the computer room modem to the user modem. Many data services such as ADSL It is designed that the downlink data transmission rate is higher than the uplink data transmission rate. As a result, the uplink and downlink transmissions require different speed predictions. Therefore, there are separate tables for the uplink and downlink data transmission rates. Figure 4 illustrates a data structure that is available A modem model for a particular modem. The number of modem models required is determined by the telecommunications company's operating policy The data structure 400 is stored in the computer memory of the test unit 2. The model depicted is for the downlink data transmission rate, which is related to the data modem used by the telecommunications company. 19 This paper size is in accordance with Chinese National Standard (CNS) A4 (210 X 297 mm) (Please read the precautions on the back before filling this page) ·· Order — 533692 A7 V. Description of Invention (β) There is a transmission rate of fT material. Row 401 represents the standard line model's index of the columns in the definition table. To simplify the description, the model is presented in a graphical manner. Or a line without bridged tap segments. As mentioned above, the route model can contain many variables. In a preferred embodiment, each line model is represented by using digital data representing a specific range or range of related variables. By matching the model of the circuit under test measured by the test unit 2 with one of the rows 402, one of the data rows is selected, such as 412A. . . 4120 column 402 is the index of the table row. Each item spanning 402 is associated with a specific noise model (combination of interference signals). By choosing a specific noise model, one of the data columns, such as 414A. . . 414B will be selected. The selected columns and rows define an item in the modem model. The number of possible noise conditions is specified by the telecommunications company's operational process, because in the illustrated embodiment, the noise model is determined by the use of the lines in the cable bundle, and this choice provides A predicted downlink data transmission rate on a specific line using a specific modem, the process is repeated to obtain the data transmission rate of the upstream part. • The uplink and downlink speeds selected for a specific standard line and some noise models are used as the upper speed limit assigned to the line under test. Further evaluations include comparing the attenuation associated with the selected standard line model with the estimated attenuation of the line under test (de-rated to a specific noise model), resulting in a speed classification decision. This assessment also takes into account the differences between any selected standard line model and the line under test. 20 (Please read the precautions on the back before filling this page) ΙΦ ------ 丨 Order --------- line! This paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) 533692 A7 _____B7 —_ V. Description of the invention (β) In the preferred embodiment, the 'line speed prediction is an easy-to-understand type To show, the results of the line test may be separated or classified by category, where each category represents a different line condition. The point is performed by using a computer designed with software for this purpose, 'which may be included as part of the test unit' or become a stand-alone system that receives data from the test unit. Classification can be further color-coded to make test results easier to identify. When a particular color group is described, any color or method of its indication can also be used to identify different categories of test results for line testing. For example, red can indicate that the line cannot support high-speed access, green can indicate that the line can support high-speed access, yellow indicates that the line cannot currently support high-speed access', but it can be used when it is removed, and gray represents the line's Features fall outside the professional scope of the system. The gray characterization provides two benefits. First, it makes it easy for a user to know the status of a particular line. Second, it increases the reliability of classification prediction by reducing the coverage of the test. For example, if the line that is classified as gray is not included in the system specialty area (and includes 20% of the total tested lines), then a group of 80% of all tested lines will be given an estimated 95 % Accuracy characteristics. As used herein, "professional area" means that the prediction of the system depends on the situation of the user. If in any case the gray line is classified as part of the professional area of the system, then 100% of the line groups of all test lines can be given 85% accuracy. Therefore, when the forecast has a high degree of uncertainty, the line is described as gray or uncertain, instead of being predicted as red, yellow, or green. This allows users to rely more on the red, yellow, or green colors. 21 This paper is sized to the Chinese National Standard (CNS) A4 (21 × 297 mm). < Please read the notes on the back before filling this page) # ------ —Order --------- line 丨 · 533692 A7 ------ B7__ 5. Description of the invention (/ ?) Measurement When the line is classified as gray, the telecommunications company may choose to perform a more complete test, or simply select other lines in the cable bundle. $ 〇 The test and classification of the line is preferably different High-speed access rate, such as ADSL at 2Mb / sec, ADSL at 4Mb / sec, ADSL at 8Mb / sec, etc., and charge different fees depending on the high-speed transmission rate supported by the line. For example, a customer receiving 2Mb / sec access is charged at the first rate, and another customer receiving 4Mb / sec access is charged at a different (higher) rate, so the line can be provided by the telecommunications company To the different service categories. The assignment of the corresponding line model and speed to a service class can be considered as the second step of the classification operation. Telecom companies prefer to provide business rules that allow this correspondence. Telecom companies provide the types of services they want to provide. For example, if a telecommunications company has decided that it will provide 512Kbps of ADSL services, the lines that have been allocated speeds greater than 512Kbps will be This service is classified as green. If a line is assigned a speed below 512K, it may be assigned a red or yellow category. For example, if the line model shows services that affect conditions on the line, such as a branch pair, a bridge tap, or a load line, The line may need repairs and then run at a higher speed. The telecommunications company provides a business model that shows whether the telecommunications company is trying to repair online. The telecommunications company will repair a line only when the predicted line speed is within 50% of the desired level, or the telecommunications company will show in the business model that it will repair a line with a bridge tap or a load coil, but 22 This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) -------------------- Order --------- Line—AWI (Please read the precautions on the back before filling out this page) 533692 A7 _ B7___ V. Description of Invention () will not attempt to repair a different pair. If the line model shows a condition to be repaired by a telecommunications company, the line will be classified as yellow, otherwise, the line will be classified as red. When the measurement performed on the line cannot immediately correspond to any line model designed to the test system, the line will be designated as gray or an indeterminate category. Alternatively, many of the forecasting techniques described above provide a predictive confidence factor. When the predicted confidence is below some threshold, the line will also be classified as gray. A precise threshold of confidence is also a function of the business model of a telecommunications company. For example, a telecommunications company with many available lines may be set a high confidence threshold, although a high confidence threshold may cause many lines to be classified as not. Qualified, it would be cheaper for telecommunications companies to simply use other lines. On the other hand, a telephone company with a short line will set a lower threshold. Even if such a setting increases the risk that a line built to serve needs to be repaired or replaced later, the telecommunications company may Many lines are used as high-margin data services to earn M Township. Telco-specific business models are best coded in software in test units. Referring to the present device shown in Fig. 1, the test unit 2 includes a voltage source and measurement unit 3, and a prediction unit 4. The test unit 2 is located behind the parent switch (or DLC) 6 in the computer room 5 or a cabinet Switch 6 connects test unit 2 to tiP7 and ring8 of telephone line 9. Switch 6 enables test unit 2 to easily test multiple lines. The test can be performed as needed or periodically. The test unit 2 is connected to the tip 7 and the ring 8 line through the parent switch 6. The paper size applies to the Chinese National Standard (CNS) A4 (210 X 297 mm) ----- (Please read the Note this page, please fill in this page) # ---- Order i 533692 A7 ____B7 ____ V. Description of the invention (w) The voltage source and measurement unit 3 provides a source signal to the tip 7 and ring 8 wire, and measures the specific Measure different characteristics of the line 'Measurement can include voltage (real and imagined), current, capacitance, impedance, resistance and similar, the measurement can be repeated with a variety of different frequencies' Test unit 2 uses a voltage source The measurement 'composed by the measurement unit 3 is used to estimate the characteristics of the circuit under test. This characteristic refers to the nature of the circuit and the presence of components on the line. Once the characteristics are estimated, the 'estimated unit uses this characteristic to predict the data transmission rate that the line can support'. This estimation process will be performed by a so-called expert system, which uses neural networks, statistical models, and decision making. Tree models and other models, or a combination of these model types. The prediction unit 4 and the voltage source and the measurement unit 3 are placed together in a computer room or a DLC, or the prediction unit is placed at a remote end and communicates with the measurement unit through a line or a wireless transmission interface. Once a line is measured, the line will be taken offline from the test unit 2 through the electric switch 6, and other lines can be connected to the test unit and then the test is performed on the next line. Once the estimates for a set of lines have been estimated, the results of the line tests will be classified as detailed below. Refer to Figure 2 as shown in flow 1: the first step of flow 10 includes testing the circuit. When measuring the parameters of different signals in the line and the line itself, the test unit is connected to a selected line and inserts the signal into the line. Signals may be provided at multiple frequencies and levels. In the next step 20, the characteristics of the line are estimated from the measurement. As an example, the characteristics include the insertion loss of the line, the phase imbalance of the line, the length and the diameter of the line. Additional features include change of wire diameter, bridge-like tapping. 24 This paper size is applicable to China National Standard (CNS) A4 (210 X 297 public love) " " " ~ ----------- -# ------- Order --------- Line- ♦ (Please read the precautions on the back before filling this page) 533692 A7 _____B7_____ V. Description of the invention (A) Head and load line圏 and other path parameters, such as splitters, filters, terminations, and the like in the line or extension line. The next two steps establish the modem and noise environment for use when evaluating what data transfer rate the line can support. Step 21 selects the next (first) modem model to use, and then step 21 selects the next (first) noise model. The next step 25 uses a process to estimate the data transmission rate that the line can support based on the measured and estimated characteristics. In addition, the reliability and coverage of the prediction will be part of the test results. 'The process will be implemented Become an expert system. Step 30 includes determining whether the line test result falls outside the application range of the system. If the characteristics of the line really fall outside the application range of the system, step 35 is performed. Results with such conditions in step 35 will be color coded. Although any color can be used for any category, in a particular embodiment, gray is used to color-code this category, so that lines with such experimental results can be easily identified. If the line characteristics do not fall outside the application range of the system, step 40 is performed. Step 40 determines whether the line test results from the test system show that the line can support high-speed access. If this is the case, step 45 is performed. Step 45 encodes this result as green in color. For any high-speed access that appears to be unsupportable, step 50 is performed. Step 50 determines whether the test result indicates that the line cannot currently support high-speed access from the test system, but the line can support high-speed access after the obstacle is removed. Obstacles may include, for example, a load line or a bridge. This paper size is applicable to the Chinese National Standard (CNS) A4 specification (210 X 297 public love). '--------------- # ------- —Order --------- Line * 丨 黉 (Please read the precautions on the back before filling this page) 533692 A7 __B7____ 5. Description of the invention (/ M-shaped tap. If the line can support high-speed access after removing the barrier, this line will be color-coded yellow as shown in step 55. If the line cannot support high-speed access after removing the barrier, step 60 is performed. Step 60 The results of the remaining lines are color-coded red, indicating that the line cannot support high-speed access, for many reasons (for example, the line is too long). The next step, step 65, provides data transmission for the selected modem and noise model. Rate and color coding to the user. Step 70 determines whether a selected modem model has more noise models to consider. Step 80 determines whether there are more modem models to consider, when there are no more modems. The model is ready for use and process 1 ends. At this point, the line Specify an uplink and downlink data transmission rate, and for each possible noise combination and important modem model, to the service provider. With color-coded results, the quality verification standards are met to support the use of high-speed access lines Or, it may be contacted to provide high-speed access to users. Classification can be done on a service-level basis. A line may be classified as green in G.lite (ADSL less than 1.5Mb / sec), but G.dmt (ADSL greater than 1.5Mb / sec but less than or equal to 8Mb / sec) is classified as red. In this example, red may be specified because the line is too long to support G.dmt, which cannot be resolved by the telecommunications company Although the above description discusses the test results from multiple lines, this method is also useful when the test results are from a single line only. Refer to the flow 2 of Figure 3, the first step 85 is to Obtain the road characteristics measured and / or estimated in process 1, as well as each modem model and noise. 26 This paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 cm). ) -------------------- Order --------- line — (Please read the notes on the back before filling this page) 533692 A7 ___B7_ Five 2. Description of the invention (W) Model prediction of uplink and downlink data transmission rate. The next step 90 is to select the next (first) service provider that defines the service category. The information related to this service category may include the minimum and maximum Ensure the uplink and downlink data transmission rates, whether to include voice-over-xDSL services, the availability of the provider's network bandwidth, etc. The next two steps establish a modem and noise environment for use. Step 100 select The next (first) modem model is used, and then the next (first) noise model is selected in step 110. The next step 120 uses the service category information, together with the line characteristics and speed information, to perform the service category designation, that is, predict the ability of the line to support specific modem and noise model services. In addition, the reliability and coverage of the prediction may be As part of the specified result, the process can be implemented as an expert system. Step 130 includes determining whether the available data (measurement, speed prediction, service capability estimation, etc.) of the line falls outside the application range of the system. If the information does fall outside the application range of the system, step 135 is performed. The results of this criterion in step 155 are color-coded. Although any color can be used for this category, in a particular embodiment, gray is used to specify color coding for this category, thus making the results easy to recognize. If the available information does not fall outside the scope of application of the system, step 140 is performed. Step 140 determines whether the available information display line supports a specific service for a specific data machine and noise model, but in this case, step 145 is performed. Step 145 encodes this result as green in color. Step 150 is performed for anything that appears to be unable to support a particular service. 27 This paper size is in accordance with China National Standard (CNS) A4 (210 X 297 mm) (Please read the precautions on the back before filling this page) I --I 1 III ^ ----- IIII · 533692 A7 ___ B7_ _ 5. Description of the invention (0. Step 150 determines whether the available information indicates that the line cannot support a specific service now, but the line can support the service after the obstacle is removed. The obstacle may include a speed obstacle (load line, bridge-like Taps, etc. may be a supplier limitation (bandwidth, routing capabilities, business policy issues, etc.) If a line can support a specific service category after removing (or addressing) obstacles, the line will be color coded yellow, As shown in step 155, if the line cannot support a specific service after removing the obstacle, step 160 is performed. Step 160 color-codes the result of the remaining line to red, indicating that the line cannot support a specific service category. There are many (For example, if no DSLAM capability is available), the color coding of the line as yellow or red will be associated with explaining why. Next step 165, Provide color coding for the selected modem and specific service category of the noise model to the user. Step 170 determines whether a selected modem model has more noise models to consider. The next step 180 determines whether there are more Modem model is available for consideration. When there is no modem model available, the next step 190 is to determine whether there are more service categories to consider. When there are no more service categories to consider, process 2 ends. At this time, the line is assigned a color code to each possible combination of service categories and important noise and modem models to service providers. After describing the preferred embodiment of the invention, Other generally known techniques using embodiments of these concepts are now becoming apparent. Although the characteristics and circuit components of a particular circuit group are described, the other 28 paper standards are applicable to the Chinese National Standard (CNS) A4 specification (210 X 297 mm) (Please read the notes on the back before filling this page) Order ----- Line! 533692 A7 _____ —__ B7_ V. Description of the invention (> Heart The circuit characteristics and circuit components can also be included. In addition, the software included as part of the tool can be instantiated as a computer program product, which contains a computer-usable medium. For example, such a computer can The media used can contain a readable memory device, such as a hard drive, optical drive, digital video disc, or computer drive, with computer-readable code segments stored on it. Computer-readable The media taken can also include communication links, whether optical, wired or wireless, with code segments on them, transmitted as digital or analog signals. Similarly, it is proposed that the invention should not be limited to what has been described Examples should be limited only by the spirit and scope of the scope of additional patent applications. In addition, it must be known that a two-dimensional table structure will be described by a modem model, where the columns in the table are defined by the length of the line. It must be known that the length of a line can be regarded as just another variable that defines the line. In this way, a two-dimensional table can be visualized as a one-dimensional bill. Other data structures in computer memory may also store the modem model without departing from the concept of the invention. For example, describing separate uplink and downlink tables is used to identify the data transfer rate. It must be understood that when the same noise model is used for uplink and downlink transmission, uplink and downlink can be achieved by simply including an uplink and a downlink data transmission rate in the sub-records of each modem model table in Figure 4. obtain. It also describes that the table query is used to determine the data transmission rate related to a specific noise model. Similarly, the data transmission rate can be calculated. Assuming noise is not present, the noise factor can be used to attenuate speed predictions on the line. 29 scales are applicable to China National Standard (CNS) A4 specifications (210 X 297 male f ~ ------------- # ------ 1 order --------- line --AWI (Please read the notes on the back before filling this page)