TW399380B - System and method for optimal virtual path capacity dimensioning with broadband traffic - Google Patents

System and method for optimal virtual path capacity dimensioning with broadband traffic Download PDF

Info

Publication number
TW399380B
TW399380B TW85108504A TW85108504A TW399380B TW 399380 B TW399380 B TW 399380B TW 85108504 A TW85108504 A TW 85108504A TW 85108504 A TW85108504 A TW 85108504A TW 399380 B TW399380 B TW 399380B
Authority
TW
Taiwan
Prior art keywords
virtual
virtual path
block
ratio function
capacity
Prior art date
Application number
TW85108504A
Other languages
Chinese (zh)
Inventor
Szabolcs Malomsoky
Wlodek Holender
Original Assignee
Ericsson Telefon Ab L M
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ericsson Telefon Ab L M filed Critical Ericsson Telefon Ab L M
Application granted granted Critical
Publication of TW399380B publication Critical patent/TW399380B/en

Links

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)
  • Telephonic Communication Services (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A general dimensioning method and system for allocating limited transmission resources to various virtual paths defined on top of a physical network. A two-level hierarchical structure is defined with a layer of one or more virtual paths on top of a layer of physical network elements. Traffic demand is specified for each virtual path and the Entropy Rate Function is used as a blocking measure. The loads on the various links are balanced by equalizing blocking probabilities and the optimal allocation of network physical resources is determined.

Description

A7 B7 五、發明説明(’) 在 35 U. S. C. § 119(e) & 37C. F· R § 1.78(a)(1)之下的罄在壤 此非臨時申請專利範圍優先權係基於一前美國臨時專利 申請,標題為”用於一般形式流量之寬頻ATM網路之有效 的大小 1 測方法"(Efficient Dimensioning Methods p〇r Broadband ATM Networks With General Type of Traffic),声號. 60/001,169 (代理人檔案編號27946-00094),於1995年7月 14 日以 Wlodek Holender 與 Szabolcs Malomsoky 之名義中 請,其亦授權給本發明之相同申請人。 相關發明參考 此非臨.時之美國專利申請案所包含之標的係相關於第 08/5 14,235賦之共同申請之非臨時美國專利申請案,標題 為''利用寬頻流量最佳化邏輯網路容量大小之系統與方法 ” (System and Method For Oniitnal Logical CapacityA7 B7 V. Description of the invention (') Exhaustion under 35 USC § 119 (e) & 37C. F · R § 1.78 (a) (1) US Provisional Patent Application, entitled "Efficient Dimensioning Methods p Broad Broadband ATM Networks With General Type of Traffic", Sound Signal. 60 / 001,169 (Attorney Docket No. 27946-00094), on July 14, 1995, in the name of Wlodek Holender and Szabolcs Malomsoky, and was also licensed to the same applicant of the present invention. For related inventions, refer to this non-temporal US patent The subject matter included in the application is a non-provisional U.S. patent application related to the co-application of 08/5 14,235, entitled "System and Method for Optimizing the Size of Logical Network Capacity Using Broadband Traffic" For Oniitnal Logical Capacity

Dimensioning With Broadband Traffic),於 1995 年 8 月 1 1 日 申請(代理人檔案編號27946-00093 ),及第08/513,723號非 臨時美國專利申請案,標題為"用以調整在一虛擬路徑寬 頻網路路梭之系統與方法"(System and Method For Adaptive Routing On A Virtual Path Broadband Network) j 於1995年8月1 1日申請(代理人檔案編號27946-00095)。這 些非臨時美國專利申請案及所揭示於此合併作為參考。 發明背景 1.技術領域 本發明係相關有效量測雷信綑路女小之系統與方法,更 確切地說係關於一種技術,其用於使用熵比率函數作為一 -4 - 本纸張尺戾適W中國國家標準(CMS ) A4規格(210X297公釐) (請先閱讀背面之注意事項再填寫本頁) 裝--- -丁 A7 B7 五、發明説明(2 種區塊刻音之有限管體網路上,量測已定義之虛擬路蔣的 大小0 2·相關技藝敘述 電仏設備與其他位於一地方區域之通訊裝置,傳統上是 經由稱作「地區交換站(l〇cal exchanges)」之交墙設備互相 連接。地區交換站接著與中繼線交換(trunk exchanges)互 連。位於一地方區域,彼此之間分隔並連接不同地區交換 站之電話/數據設備,經由一於電信網路中鏈接—起之地區 及中繼站夂換複合群組而相互通訊,一電信網路因此包括 多個互連網路元件,例如地區交換站、中繼站交換站、行 動無線交換站、長途交換站及其組合。在每個網路階層 =,由了網路元件(例如一交換站)至另一網路元件之流 量可採取通過相異交換站之不同路後。 於—網路中,有效能的通訊設施網路流言管理蒙軲暑 ,夠的可用電路,以處理每個㈣之流量需求,而不會在 取後選擇之流量路徑上過合的姑塞。亦要求網路擁塞儘可 說的在所有最後選擇路徑上平均,且在實際提供之路徑中 未吏用的谷1旎最+仆以確保資源的甩玟利用。此外,運 一罔路〈包忐公司預算有限,所以必須儘可能有效地利用 母一網路中現存的資源。 曰f過/,、一通訊網路中之流量管理已包括有週期性地測 罔中*量模式的程序,並改變電路處踗滅之组態,以 :放地%進流量管理。而且,對於一特定區域或特定範 ’及^該範圍中之地方性事件,預測會有高的通話密 -5 (請先閲讀背面之注意事項再填寫本頁) ^----Dimensioning With Broadband Traffic), filed on August 11, 1995 (Attorney Docket No. 27946-00093), and Non-Provisional U.S. Patent Application No. 08 / 513,723, entitled " for adjusting the bandwidth of a virtual path System and Method For Adaptive Routing On A Virtual Path Broadband Network (j) was applied on August 11, 1995 (agent file number 27946-00095). These non-provisional US patent applications and their disclosures are incorporated herein by reference. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and method for effectively measuring thunderbolt bundles, and more specifically to a technique for using an entropy ratio function as a -4-the paper size Applicable to China National Standard (CMS) A4 specification (210X297 mm) (Please read the precautions on the back before filling out this page) Installation ----Ding A7 B7 V. Invention Description (Limited tube with 2 block engravings) On a physical network, the size of a defined virtual circuit is measured. 2 · Relevant technical description Electric equipment and other communication devices located in a local area are traditionally referred to as "local exchange stations (10cal exchanges)" The cross-wall equipment is connected to each other. The regional exchanges are then interconnected with trunk exchanges. They are located in a local area, separated from each other and connected to the telephone / data equipment of different regional exchanges, and linked via a telecommunications network The local area and relay stations exchange communication with each other in a composite group. A telecommunications network therefore includes multiple interconnection network elements, such as regional switching stations, relay switching stations, mobile wireless switching stations, Long-distance switching stations and their combinations. At each network level =, traffic from a network element (such as a switching station) to another network element can take a different path through a different switching station. In order to effectively manage the network rumors of communication facilities, there are enough circuits available to handle the traffic requirements of each network, without excessive congestion in the selected traffic path after the retrieval. The network is also required Congestion can be said to be averaged over all the last selected paths, and the unused valleys in the actual provided paths are used to ensure the utilization of resources. In addition, Yunyi Road (including the company's limited budget) Therefore, it is necessary to make effective use of the existing resources in the parent-network as far as possible. The flow management in a communication network has included a procedure for periodically measuring the medium-quantity mode and changing the circuit location. The configuration of the destruction is managed by: inflow of% of land. Moreover, for local events in a specific area or a specific range, and ^, a high call density -5 (please read the note on the back first) Please fill in this page for more details) ^ ----

、1T 本纸張尺度適用中國囤家標率(CNS ) Λ4規格(210x 297公釐) 五、發明説明( Α7 Β7 度時,更多的路徑及電路就可加入一網路中。習知網路管 理系統亦可改變網路所撰路徑之間相關的流晉畲載分布, 以¥效地充分利用網路而不會使服務品質(Q〇S)大幅7 降。然而’習知流量網路管理系統及程序通常汲汲於増加 網路中電路及路徑之可用性,以便控制個別的流量需求, 而非藉在較抽象階層由重新量測其路徑與電路大小以脅紐 構一網路,使得整個網路之效能發揮極致。 除了要完全發揮一電信系統中管理實體網路資源之效能 的固有需求外,近幾年在資訊及通訊技術的成長已建立新 2濟機會的財富與管理上的挑戰。提供電信服務之賣方繼 、..Λ地面對新的顧客需求。透過電信網路提供一般的聲音通 訊已二不夠了。今日之用戶需要的是,可以即時以及透過 封包切換網路二者上,不僅是聲音訊號的傳送,而且還需 要資料、聲頻、视頻及多媒體訊號的傳送上的能力。因為 其在提供寬頻電信設施上所提升的能力,非同步傳送模式 (ATM)技術的重要性漸增。 ATM技術主要的特徵為在網路資调利用上的彈性。一種 發展此彈性之可成方式為整體或部分地將f體絪路資源劃 分為邏輯定義的資源。例如,藉由將—個實體㈣劃分I ^個虛擬過路’此—實體網路之操作、維護及管理可相令 私詹地W化。以此程序,對於整個實體網路之每—通話之 t析與解量路㈣題之工作,可純mm虚擬網 路上相當簡單之路押門1 — 仫問畸β母一個別的虛擬網路比一 整體的整個實體網路的複雜 , 、,θ '又季义λ],這使仔流I路徑問題、 1T This paper size is applicable to China's standard of household standard (CNS) Λ4 specification (210x 297 mm) 5. Description of invention (Α7 Β7 degrees, more paths and circuits can be added to a network. Know network The road management system can also change the distribution of related traffic between the routes written by the network, in order to make full use of the network effectively without degrading the quality of service (QOS). However, 'Learn the traffic network Route management systems and procedures usually draw on the availability of circuits and paths in the augmented network in order to control individual traffic requirements, rather than constructing a network by re-measuring its path and circuit size at a more abstract level, making The performance of the entire network is extreme. In addition to fully utilizing the inherent needs of managing the performance of physical network resources in a telecommunications system, the growth of information and communications technology in recent years has established new wealth opportunities and management of economic opportunities. Challenges. Vendors who provide telecommunications services are facing new customer needs. Providing general voice communications through telecommunications networks is no longer enough. Today's users need to be available instantly and through On both packet switching networks, not only the transmission of sound signals, but also the ability to transmit data, audio, video and multimedia signals. Because of its enhanced ability to provide broadband telecommunications facilities, the asynchronous transmission mode ( The importance of ATM technology is increasing. The main feature of ATM technology is the flexibility in the use of network resources. A possible way to develop this flexibility is to divide the f-system resources into logically defined resources in whole or in part. For example, by dividing one entity into one virtual passway, the operation, maintenance, and management of the physical network can be implemented in a private way. With this procedure, for every call in the entire physical network, The t analysis and the work of solving the road problems can be quite simple on the pure mm virtual network. Gate 1 — Asking about a different virtual network is more complicated than the entire physical network as a whole. θ 'also quarter meaning λ], which makes the flow path problem

(請先閱讀背面之注意事項再填寫本頁} 哀--- -ip > ti -1 mm A7 B7 4 五、發明説明( 之解決較為容易。 a實體資源之劃分亦可由租用網路中存有不同的流量形 .占不同服務級別或變化的流量需求而成為必要。不是提 供一分離的實體網路給每個顧客申請,取而代之的是一服 矛力供應者可設足一些虛擬網路’定義在一單一共同實 體下部結構(infrastructure)之頂端。 此虛擬網路架構组態上新的彈性,要求有效的大小測量 =具,万法以及演繹法。因為在未來所提供之電信服務本 貝很難加以預測,任何用以管理虛擬網路資源之组態的大 小量測方法必須要能夠處理所有形態之寬頻流量。假如一 虚擬,·祠路之组態必須經常的修改,以適應所提供之流量形 恐上的改變,那麼網路的大小再量測及组態控制系統之運 算放flb亦也需要很高。一種選定以執杆網路大小再量測及 重新組態一網路之演繹法,應該在一段遠比每個虛擬網路 眭間要短之期間内執行其計算。 "千多現存 < 大小量測技術就被大小量測技術要能模擬— 般流量分伟之需求所排除。因為使用Erlang區塊測量所致 的限制’大部分通常所使用之大小量測方法不能操控—般 的流·量模式。本發明之方法及系統可克服這些缺點。 發明概迷 本發明之目的之一即是可為使用者量測一使用一般流量 分忡筷式之網路大小。本發明之另一目的是要能使用一最 小的汁算資源來執行大小量測的計算。本發明之再—目的 是要能執杆—種具有低計算禎雜麼而因此速度較快之大+ 7- 本纸浪尺度遣财丨侧家縣率(CNS )从规格(2iGx297公潑) (請先閱讀背面之注意事項再填寫本頁) 裝--- 訂-- 經浐部中^"'^^,-^-"消於合竹^印4·^ A7 ---_________B7 五、發明説明(5 ) 夏測演繹法。本發明之又一目的异处私+压 ^ Η旳疋要此夠在儘可能短的時 7内,執行大小重新量測的計算。本發明尚有另一目的, 是要提升虛擬網路大小量測及虚擬路徑大小量測處理 *6匕 月b ° 給定一包括多個實體鏈路之實體網路,其中每個實體鏈 路具有一預定之傳輸容量,本發明之系統及方法說明一種 提供-般流量模式之大小量測技術,此大小量測工作被視 f於各樣實體鏈路上之負載平衡問題。對於虚擬路徑大小 f測 < 最佳解法係相當於在各種虛擬路徑上選擇配置容 量,其中在每個虛擬路徑上之區塊是要儘可能於各種不同 鏈路上一致。 某一方面而言’本發明包括一種可有效能地量測一具有 ^個貫體鏈路之電信網路大小的方法,而該實體鏈路係與 多個父換站或節點互連。多個實體鏈路與一或多個虛擬路 徑相關。於電信網路中,每個虛擬路徑於一對交換站或節 點之間提供一個單獨可切換的連接。所提供之流量是針對 每個虚擬路徑予以特定,而傳輸容量的限制則是由電信網 路的每個貫體鏈路加以設定。所提供之流量與其他計算參 數之間的關係,是於電信網路上使用一種熵_區塊測量加以 模擬,而且容量係配置給受到各種實體鏈路之傳輸容量限 制之多個虛擬路徑’這使得不同虚擬路徑上之區塊機率在 一可預測之錯誤範圍中儘可能的一致。 圈式簡單敘迷 由^考之後較佳貫施例之詳細敘述以及所伴隨之圖 -8- 本紙張尺度適;1丨中國國家標丰(CNS ) A4規格(210X297公釐 (請先閲讀背面之注意事項再填寫本萸) I - ·. '訂 A7 B7 五、發明説明(6 式’可以更加了解本發明之方法與系統,其中: 圖1 ,-員v⑨t罔路之方塊圖,於其中虛擬路徑大小量调 可被完成; 圖2顯示一範例ATM單元結構之方塊圖; 圖3顯示一 ATM網路中,龙加^、土上 t 、, 、 多個互連虛擬路徑及處擬通道 疋方塊圖; 圖4 員π ATM網路中,貞擬路徑及虚擬通道之交錯連 接與交換動作之方塊圖; =5顯不-不同服務級別供應與標準層之π打丁 b_isdn 參考模式之方燒圖.; _圖6顯示-提供虛擬出租線路(VLL)服務atm網路之方塊 圖^示-包括ΑΤΜ交錯連接之多層.咖.基 <万塊圖; 圖圖請示-虛擬路徑及實體鏈路之間的枯撲學關係之方塊 圖9顯示一流程圖,說明用於實體網路中定 虛擬路徑之圖解下推式演繹法之各種步騾; ( 圖10顯示一相關本發明之熵比率函數特徵々_表. 徵所示之熵比率函數’其位移參數特 圖1 2顯示—流程圖,說明使用相關於本發明 數之虛擬路徑大小量測的各種步驟。 …、 較隹實施合丨躱破 經"部中JA^^'^h-T消於合竹心卬4·水 A7 --_ _ B7 ~ " -- - -- . , III — 五、發明説明(7 ) 曹知網路之擁塞控制 首先翱閱圖1,其中顯示一傳統公共電信網路之圖解示意 圖’ ί亥電#網路包括多個地區交換站2 1至2 6,而每個地區 叉換站有多個地區用戶與其連接並以電話設備27表示。兩 個地區交換站21及24是表示為具有遠端用戶多工級 (remote subscriber multiplex stages) 28 輿29 邀 l相德之, 並接著具有地區用戶27連接其上。圖1之網路亦包括多個 中繼交換站3 1至3 4,其主要是服務將不同的地區交換站與 另一個相互連接’以提供網路不同部分之間的路徑。中繼 交換站3 1顯示連接一行動交換站3 5,該行動交換站3 5包 括一對服務多個以3 8表示之行動無線電話用戶的圖式基地 台3 6與3 7。此外,其他電信服務,例如資料庫及智慧型網 路,亦可與所示之不同的交換站相連接。在每個網路之交 換站2 1至3 5之間,顯示的是多個通訊路徑3 〇,每個通訊 路徑包括多個通訊電路’包括電纜,光學鏈路或無線鏈 路’以於網路中之不同交換站之間傳送聲音及/或數據通 訊。 圖1之網路也包括一網跋抟在,丨玄統4 〇,其藉由诵訊娃路 4 1 (以虚線表示)與網路φ之矣個交換站7 1 $ 3 5相速拉, 以將控制訊號傳輸至每個交換姑,並從每個交換站接收流 重貞ft。網路控制系統4 0發出命令’以動態地重組網路中 个Μ流量路徑之通訊路徑,並控制網路交換站之警報系 統,以便來微調和缓網路中的擁塞狀況。 ATM乐税概眾 -10- (諳先閱讀背面之注意事項再填寫本頁) ·11— ·1 ·-- - «1 —II · 幕-- ^-- 本纸乐尺度適用中國國家標準(CNS ) A4規格(210X297公釐) A7 B7 五、發明説明( 於將几成之公共電訊傳送網路中银产找 -, 、 于現在發生了一此改變。 么共電信網路之操作員之主要 二 ,,,.,, . 的^ —為配置一種單一形 式的技術’以處理其共同下部結 之傳送與交換動作一個這㈣有電信服務形式 (ATM)技術。 卩為相步傳送模式 目前已實施的ATM是想要藉由邊上 ^ „ + , u 胥措由建儿—载運(bearer)通信網 路來滿足這些需求,該通作铜政 ,,,k , u 、彳….‘罔路實質上有"頻寬粒狀 (bandwidth aranillaritv),且可香丄 一 ^ ) 」見服非當鬲的頻寬連接。此 石司"頻克粒狀"係指一網路之转徵,甘+ 』崎又狩徵,其申一诵話所雲之頻 見可於整個通話期間自由變化。 公共電信網路中ATM技街之使用對於相關服務,增加頻 寬粒狀U位元率服務之提供,與多媒體服務之提供, 提供共同叉換及傳送之能力。由於這些特性,atm由國祭 電報電話諮詢委員會(CCITT),&isnd(b_isdn)服務之核 心技術而選定。這是不管ATM缺點,包括低速等時服務 (low speed isochronous service)之傳送延遲,於一網路中加 入複雜度,及新的執行參數之導入(例如單元_損失與擁 塞 ’ cell-loss and congestion),將說明如下。 ATM網路可使用plesiochronous數位層級(pdhh或因此數 位層級(SDH),或兩者来完成。再者,不論該等起自atm 與S Τ Μ (同步傳送模式)間之多重轉換的限制何時產生,純 粹ATM岣可作為一網路之載運,而所導致之執行表現下降 則可被應付處理。 於圖2中顯示之ATM單元結構為ATM技術之核心。一 -11 - 本紙張尺度这《 t國國家標準(CNS ) A4規格(210X 297公釐) (讀先閱讀背面之注意事項再填寫本頁) -11 A7 ______________B7 五、發明説明(9 ) atm單元有53位元组之固定長度,其化分為一個5位元组 之標頭及一個48位元組之資訊欄(也稱之為酬载, 1 0 a d)。ATM單元之標頭以一些欄位欄架構,且其主要功 能之一為透過一或多個切換節點,以支援從起點至終點之 ATM單元之路徑。每個保持之資訊係相對小的保持,以減 少交換節點之内部緩衝器之大小,並限制在那些緩衝器中 的詢問延遲=ATM於一連接導向模式中操作。從—模擬觀 點而言是重要的,因為利用良好建立之電路_切換數學模式 之結果以最佳化網路資源之配置及控制是可能的。 ATM單元標頭之主要功能為虛擬連接之辨識。路徑atm 單元之資訊係包含於兩櫚中··一虛擬路徑辨識元(νρι),其 決定ATM單元屬於哪個虛擬路徑,及—虚擬通道辨識元 (VCI),其決定ATM單元屬於虛擬路徑中的哪個虚擬通 道。 —虛擬通道為一動態地可配置的終端-對-終端(endqo-end) 連接 。光傳 輸鍵路 每秒可 通過數 百個百 萬位元 ,然而 虛擬通道一鏈路每秒只可填入幾千位元。因此,大量的同 時之虛擬通道可於一單一傳輸鏈路上提供。 反之,一虚擬路徑於端點之間為半永久(sem卜permanent) 連接。每個虛擬路徑可通過大量同時連接之虛擬通道。由 於一大群的虚擬通道以一單一單元而一起操控及切換,一 虛擬路徑之總處理需求小於一虛裡電路之總處理需求,且 結果每個(虛擬)電路有較快的處理,因而導致更有效能的 使用網路資源。虛擬路徑之網路管理相對地為簡單而且有 -12- 尺度適用肀囤國家栋準(CNS ) Λ4現格Γ210Χ297公釐) ' (請先閲讀背面之注意事項再填寫本頁(Please read the notes on the back before filling in this page} Sorry ----ip > ti -1 mm A7 B7 4 V. Explanation of the invention (The solution is easier. A) The division of physical resources can also be stored in the leased network There are different traffic patterns. It becomes necessary to account for different service levels or changing traffic requirements. Instead of providing a separate physical network for each customer to apply, instead a service provider can set up some virtual networks. At the top of a single common entity infrastructure. The new flexibility in the configuration of this virtual network architecture requires effective size measurement, measurement, and deductive methods. Because the telecommunications services provided in the future are very important. It is difficult to predict. Any size measurement method used to manage the configuration of virtual network resources must be able to handle all forms of broadband traffic. If it is virtual, the configuration of the temple road must be frequently modified to adapt to the provided If the traffic flow changes, then the network size measurement and configuration control system operation flb also need to be very high. One type is selected to perform the network size measurement. The deductive method of reconfiguring a network should perform its calculation in a period much shorter than that of each virtual network. &Quot; Thousands of existing " Size measurement technology must be capable of Simulation-the need for general traffic sharing is ruled out. Because of the limitation caused by the use of Erlang block measurement, most of the commonly used size measurement methods cannot be controlled-general flow and volume mode. The method and system of the present invention can overcome These are the shortcomings of the invention. One of the objectives of the present invention is to measure the size of a network using a general traffic distribution chopsticks for users. Another object of the present invention is to be able to use a minimum of computing resources to Perform calculation of size measurement. The purpose of the present invention is to be able to execute the pole-a type with low computational complexity and therefore a high speed + 7- This paper waves the scale of wealth 丨 side home county rate (CNS) From the specifications (2iGx297 male splash) (Please read the precautions on the back before filling in this page) Install --- Order-In the economic section ^ " '^^,-^-" 消 于 合 竹 ^ 印 4 · ^ A7 ---_________ B7 V. Description of the invention (5) Summer test deduction method. Another purpose is to be private and to press ^ Η 旳 疋 It is necessary to perform the calculation of the re-measurement of the size in the shortest possible time 7. The present invention also has another purpose is to improve the virtual network size measurement And virtual path size measurement processing * 6 d. B ° Given a physical network including multiple physical links, where each physical link has a predetermined transmission capacity, the system and method of the present invention provide a method for providing- The size measurement technology of the general traffic mode. This size measurement work is considered as the load balancing problem on various physical links. For the virtual path size measurement < the best solution is equivalent to selecting the configuration capacity on various virtual paths. Among them, the blocks on each virtual path are as consistent as possible on various different links. In one aspect, the present invention includes a method for efficiently measuring the size of a telecommunications network having a plurality of continuous links, and the physical link is interconnected with a plurality of parent stations or nodes. Multiple physical links are associated with one or more virtual paths. In a telecommunications network, each virtual path provides an individually switchable connection between a pair of switching stations or nodes. The traffic provided is specific for each virtual path, and the transmission capacity limit is set by each continuous link of the telecommunications network. The relationship between the provided traffic and other calculation parameters is simulated on the telecommunication network using an entropy_block measurement, and the capacity is allocated to multiple virtual paths that are limited by the transmission capacity of various physical links. This makes The probability of blocks on different virtual paths is as consistent as possible within a predictable range of errors. The circle-style simple narrative consists of a detailed description of the preferred embodiment after the examination and the accompanying figures. 8- This paper is of appropriate size; 1 丨 China National Standards (CNS) A4 specifications (210X297 mm (please read the back first) For the matters needing attention, fill in this 萸) I-·. 'Order A7 B7 V. Description of the invention (Formula 6') You can better understand the method and system of the present invention, where: Figure 1,-Block diagram of the member v⑨t 罔 Road, in which Virtual path size adjustment can be completed; Figure 2 shows a block diagram of an example ATM unit structure; Figure 3 shows an ATM network, Longjia ^, Tu t,,, and multiple interconnected virtual paths and processing channels疋 Block diagram; Figure 4 Block diagram of the interleaved connection and exchange action of virtual path and virtual channel in a member π ATM network; = 5 Display-Different service level supply and standard layer π hitting b_isdn reference mode Burning diagrams; _Figure 6 shows-block diagram of atm network providing virtual leased line (VLL) service ^-Multi-layer including ATM interleaved connection. Coffee. Basic < 10,000 block diagrams; diagrams and diagrams-virtual paths and entities Block diagram of the relationship between links in the map. Figure 9 shows a flow chart that says Various steps of a graphical push-down deductive method for determining a virtual path in a physical network; (Figure 10 shows a characteristic of the entropy ratio function of the present invention. Figure 12 shows a flow chart illustrating the various steps of using the virtual path size measurement related to the number of the present invention.....卬 4 · 水 A7 --_ _ B7 ~ "---., III — V. Description of Invention (7) Cao Zhi Network Congestion Control First look at Figure 1, which shows a traditional public telecommunication network Schematic diagram 'ίHAI 电 # The network includes multiple regional exchange stations 2 1 to 26, and each regional interchange station has multiple regional users connected to it and represented by telephone equipment 27. Two regional exchange stations 21 and 24 is represented as having remote subscriber multiplex stages 28, 29 and 29 are invited, and then there are regional users 27 connected to it. The network in Figure 1 also includes multiple relay exchange stations 3 1 to 3 4, which mainly serve to connect different regional exchanges with one another Connect 'to provide paths between different parts of the network. The relay switching station 31 shows a connection to a mobile switching station 35, which includes a pair of mobile radio subscribers serving a plurality of mobile radiotelephone users, indicated by 38. Schematic base stations 36 and 37. In addition, other telecommunications services, such as databases and smart networks, can also be connected to the different switching stations shown. Switching stations 2 1 to 3 in each network 5 shows multiple communication paths 3 0, each communication path includes multiple communication circuits 'including cables, optical links or wireless links' to transmit sound between different switching stations in the network and / Or data communication. The network in Fig. 1 also includes a network post in Xuantong 4 〇, which uses the chanting message path 4 1 (indicated by a dotted line) and an exchange station in the network φ 7 1 $ 3 5 phase speed Pull to transmit the control signal to each switching hub and receive the stream from each switching station. The network control system 40 issues a command ′ to dynamically reorganize the communication paths of the M traffic paths in the network and control the alarm system of the network switching station to fine-tune and alleviate the congestion in the network. ATM Music Tax Summary-10- (谙 Please read the notes on the back before filling in this page) · 11— · 1 ·--«1 —II · Curtain-^-This paper's music standards are applicable to Chinese national standards ( CNS) A4 specification (210X297 mm) A7 B7 V. Description of the invention (finding the percentage of silver products in the public telecommunications transmission network--, this has now changed. What is the main operator of the telecommunications network? Two ,,,,,, ^ — to configure a single form of technology 'to handle the transmission and exchange actions of its common nodes. There is a form of telecommunication service (ATM) technology. 卩 The phased transmission mode is currently available. The implementation of the ATM is to meet these needs by the side ^ „+, u 胥 measures by Jianer-bearer communication network, which is used for copper management ,,, k, u, 彳 ... ' Kushiro is essentially "bandwidth aranillaritv" and can be used for a wide range of bandwidth connections. This "Shi Ke Grain" refers to the signing of a network, Gan + "Qi Qi", whose frequency can be freely changed throughout the call. The use of ATM technology streets in public telecommunications networks provides the ability to jointly switch and transmit the related services, increase the provision of bandwidth granular U-bit rate services, and provide multimedia services. Because of these characteristics, atm was selected by the core technology of CCITT, & isnd (b_isdn) service. This is regardless of the shortcomings of ATM, including the transmission delay of low speed isochronous services, adding complexity to a network, and the introduction of new execution parameters (such as cell_loss and congestion) ), As explained below. ATM networks can be done using the plesiochronous digital hierarchy (pdhh or consequently the digital hierarchy (SDH), or both. Furthermore, whenever these restrictions arise from multiple conversions between atm and STM (Synchronous Transfer Mode)) Pure ATM 岣 can be used as a network carrier, and the resulting decrease in execution performance can be dealt with. The ATM unit structure shown in Figure 2 is the core of ATM technology. I-11-This paper scales this "t National Standard (CNS) A4 specification (210X 297 mm) (Read the precautions on the back before filling this page) -11 A7 ______________B7 V. Description of the invention (9) The atm unit has a fixed length of 53 bytes. It is divided into a 5-byte header and a 48-byte information column (also known as payload, 10 ad). The header of the ATM unit is structured with some field columns, and its main function is One is to support the path of the ATM unit from the start point to the end point through one or more switching nodes. Each of the held information is a relatively small hold to reduce the size of the internal buffers of the switching node and is limited to those buffers Inquiry Late = ATM operates in a connection-oriented mode. It is important from a simulation point of view, because it is possible to optimize the configuration and control of network resources using the results of well-established circuits_switching mathematical modes. ATM units The main function of the header is the identification of virtual connections. The information of the path atm unit is contained in two palms. A virtual path identifier (νρι) determines which virtual path the ATM unit belongs to, and the virtual channel identifier (VCI) ), Which determines which virtual channel in the virtual path the ATM unit belongs to. — A virtual channel is a dynamically configurable end-to-end connection. Optical transmission keys can pass hundreds of millions per second However, a virtual channel can only fill a few thousand bits per second. Therefore, a large number of simultaneous virtual channels can be provided on a single transmission link. On the contrary, a virtual path is semi-permanent between endpoints ( sem (permanent) connection. Each virtual path can pass a large number of simultaneously connected virtual channels. Because a large group of virtual channels are controlled and switched together in a single unit, The total processing requirement of the virtual path is less than the total processing requirement of a virtual circuit, and as a result, each (virtual) circuit has faster processing, resulting in more efficient use of network resources. The network management of the virtual path is relatively Simple and has -12- scales applicable to the National Standards of Construction (CNS) Λ4 is now Γ210 × 297 mm) '(Please read the precautions on the back before filling this page

gft., —^^1 11 - I —^ϋ In Hi -» \ .、 、T A7 — B7 五、發明説明(10) 效的。 如圖2所示,ATM單元標頭在比較使用者網路介面(UNI, user-network interface)與網路節點介面(NNI,network-node interface)上有些微的不同。UNI包括四位元的一般流程控 制(GFC,generic flow control)且用以確保一終端與網路之 間的有效容量之公平及有效使用。一酬載形態指示器(ΡΤΙ, payload type indicator)欄係用以指示ATM單元是否包含使 用者資訊或特定的網路資訊.例如用於維護之目的。一單 兀損失優先權(CLP, cell loss priority)欄編碼二階優先權 (two-level priorty),且用以當其由於網路條件而變得必須 抛棄该等單元。標頭資訊由包含於標頭錯誤控制(HEC, header error control)欄之核對和(check sum)而保護 u ATM單几之使用允許資訊傳送率以適應實際的服務需 求。取決於容量需求,每時間單位之單元數目增加至用以 貫行貝料之貫際媒介之傳送位元率限制。除了資料單元, 尚具有訊號化及維護之單元輿閒置單元。訊號化單元於網 路之終端使用者之間使用,或於網路之節點之間使用,且 其功能為設定一服務,例如一連接。當閉置單元用來填入 馇輪宝奮以高读傳送媒介率時,維護單元提供atm層^監 督。 參考圖3,其顯示一 ΑΤΜ網路中.多個互連接虛擬路徑 及虚擬通道之方塊圖。從一交換設計者之觀點,"νρ切換 (VP switching)係指一只使用辨識欄之上部分之單元 之交诲,意即,較#欄(VPI)。相反的,在,,vp/vc Ζ換 ___ -13- 本紙張尺度適(2i〇x297^i^----- 1- 1! . l:i _i I— -I --1 ^^二- I i— 11. -I -* - 1 ► ,¾ ,-° (請先閲讀背面之注意事項再填寫本頁) A7 ------------------B7 五、發明説明(H ) (VC switching)整個辨識櫚被使用及。一 VP/VC路徑包括多個互連接vp/vc程度。交換及縱橫-連接 可在V P或v c P白層上執行。虛擬路徑辨識元(νρι)及虚擬通 道辨識元(να)定4 一 ATM電路中的兩連線_•㈣操控 及各&冑從網路架構的觀點,—虛擬路徑(VP)為一個 別連接群,在一ATM網路之路徑地圖中的一”高速公路 "(Mghway)形態。在網路營理中的—個重要工作為配晉德 輸容量之JP硪量給每個如此之高速公路(即—虛擬路徑), 以最佳化網路執行。此最佳化工作為頻寬管理或虛擬路徑 大小量測技術之方針,且為本發明一觀點之標題,將於下 討論。 表者圖4,顯示一 ATM網路中,虛擬路徑及虛擬通道之 縱橫-連接與切換之觀念。虛擬路徑辨識元(νρι)及虚擬通 道辨識元(VCI)只在特定的鏈中有效。在每個縱橫-連接/交 換,新的VPI/VCI值以實際埠與提供ATM單元辨識之 vPI/VCI值之組合而指派給該單元。然後執行一範例ATM 單元之路徑,以轉換表格之幫助,如表格〗所示如 連接-從埠 VPI VCI 連接-至埠 VPI VCI A 1 - C 10 A 2 一. D 6 B 5 3. C 7 4 B 5 2 D 3 5 -14- 本紙張尺度適用中國國家標车(CMS ) A4規格(210x297公釐〉 (請先閱讀背面之注意事項再填寫本頁)gft., — ^^ 1 11-I — ^ ϋ In Hi-»\.,, T A7 — B7 V. Description of the invention (10) Effective. As shown in Figure 2, the ATM unit header is slightly different in comparing a user-network interface (UNI) with a network-node interface (NNI). The UNI includes four-bit general flow control (GFC) and is used to ensure the fair and efficient use of effective capacity between a terminal and the network. A payload type indicator (PTI) column is used to indicate whether the ATM unit contains user information or specific network information, such as for maintenance purposes. A cell loss priority (CLP) column encodes two-level priorty and is used to discard those cells when it becomes necessary due to network conditions. The header information is protected by a check sum included in the header error control (HEC) column. U The use of ATM documents allows the information transfer rate to meet actual service requirements. Depending on capacity requirements, the number of units per time unit is increased to the transmission bit rate limit of the intermediary medium used to carry the shellfish. In addition to the data unit, it also has a signalization and maintenance unit and an idle unit. The signaling unit is used between end users of the network or between nodes of the network, and its function is to set a service, such as a connection. When the closed unit is used to fill in the high-read transmission media rate, the maintenance unit provides atm level supervision. Referring to FIG. 3, a block diagram of multiple interconnected virtual paths and virtual channels in an ATM network is shown. From the point of view of a switching designer, " νρ switching (VP switching) refers to the intersection of a unit using the upper part of the identification column, that is, compared to the # column (VPI). In contrast, in the case of, vp / vc Zn ___ -13- The paper size is appropriate (2i〇x297 ^ i ^ ----- 1- 1!. L: i _i I- -I --1 ^^二-I i— 11. -I-*-1 ►, ¾,-° (Please read the notes on the back before filling this page) A7 ----------------- -B7 V. Description of the Invention (H) (VC switching) The entire identification tree is used. A VP / VC path includes multiple interconnected vp / vc levels. Switching and aspect-connection can be performed on the VP or vc P white layer. The virtual path identifier (νρι) and the virtual channel identifier (να) determine the two connections in an ATM circuit. • Control and each & 胄 From the perspective of network architecture, the virtual path (VP) is Individual connection groups, a "highway" (Mghway) form in the path map of an ATM network. One of the important tasks in network management is to allocate the JP volume of Jinde's transmission capacity to each such Highway (ie, virtual path), with optimized network implementation. This optimal chemical is used as the guideline for bandwidth management or virtual path size measurement technology, and is the title of an aspect of the present invention, which will be discussed below. Figure 4 shows an ATM network , The concept of cross-link-connection and switching of virtual paths and virtual channels. The virtual path identifier (νρι) and virtual channel identifier (VCI) are only valid in a specific chain. In each aspect-connection / exchange, a new VPI The / VCI value is assigned to the unit by a combination of the actual port and the vPI / VCI value that provides ATM unit identification. Then execute the path of an example ATM unit to help with the conversion table, as shown in the table, such as the connection-slave port VPI VCI connection-to port VPI VCI A 1-C 10 A 2 I. D 6 B 5 3. C 7 4 B 5 2 D 3 5 -14- This paper size applies to China National Standard Vehicle (CMS) A4 specification (210x297 male) Li> (Please read the notes on the back before filling in this page)

A7 B7 五、發明説明(12 表格1 ATM單元為一 ATM傳送系統之某本的客工單元,每個般 元或資訊單元包括其本身的連接與路徑資訊。此特性致使 服務頻道之直接的多工或解多工’其中每個頻道可實行不 同的位元率。每個ATM單元由包含於虛擬路徑辨識元(vpI) 及虚擬通道辨識元(VCI)搁之標頭的資訊辨識及路徑。如上 所述,一务取敗姆為一兩终鸪點,如切換系統、區域 網路(LAN)閘路、或釔人網路閘路,之間的豸工啻炊群。 一vp撂银一南擗踗鸦線诚+明的亩拉邏輯_,wvpT值辨 識特定的虚擬路徑。 如上所述,ATM祜俶十所使用的虚捧玖蔣允許多個虚擬 通道fVCd,以一單一單元操控。具有共同特性之虛擬通 迢’如相同品質服務(Q〇S),可群集在以一單元被傳送、 處理、及答撕士 一赶士。此蓰性辞集簡化了八丁以系統之操 作及維撙_。 虛擬路徑及虚擬通道皆可用以提供ATM網路中的半永久 (semi-permanent)路徑。路徑以縱橫_連接元件或沿著一路 徑之多工器之”路徑連接表格”(path c〇nnect taMes)設定, 從一操作提供系統中建立與釋放。虚擬通道亦可由訊號化 一使用者與網路之間或於網路所建立之連接’為一要龙即 X 換(on-demimd swuching)使用。 atm技術之重要特徵之一為其協定架構’且係以稱之為 ,核心及邊緣”(core-and-edge)之定則建立。該協定功能說 明傳送之資訊形態,例如重新傳送、流程控制、及延遲均 -15- (請先閱讀背面之注意事項再填寫本1) 袈—— B7 五、發明説明(13 ) 等化,係於ATM網路之邊緣之終端中執行。此流下一有效 的、服務獨立的,,核心"網路,其只包括簡單的單元傳送及 交換功能。在此核心之ATM節點,既没有資訊櫚的錯誤核 對亦沒有任何的流程控制。單元資訊被簡單的讀取,然後 使用HEC校正可能影響位址之單一位元錯誤,且單元然後 切換至其目的地。 ATM tw凋階層(aaL,adaptation layer)在網路之邊緣上 使用以増進服潞的提供。如圖5所示,is於B_LSDN服務 I CCITT參考模式想像AAL包括服務仿附功能。如圖$折 不,在ATM標準中|有三層。第一層為定義膏際|面及框 協f之會搭辱。第二ATM層與所選的實際媒介無關,並定 # ^疋結構,提供多工與解多工及VpL/Vci轉槔以摊制邏 輯網路中的單元流贫。I 3售今AAL,其提供服務與ΑΤΜ 層之間的重要協定,藉此允許服務獨立ατμ傳送。執 行原始服務格式與一 ATM單元之資訊欄之間的映對。由 AAL所提供的範例功能包括不同長度的封包描述,序列排 號,時序恢復,及執行監控。 在電信網路之ATM發展 ATM技術之一種使用可用於用戶前端,以於用戶區域網 路之中或之間提供高速資料通訊。此外,ATM可用作下部 結構資源,該下部結構資源對於所有在—用戶前端網路之 服務而言為共同的,包括聲音及視頻通訊,資料傳送及多 媒體應用。 A T Μ節點導入一公共電信網路之範例服務係提供虛擬出 -16- 本紙張尺度適州中_家標準(CMS ) Α4規格(2丨Οχ 297公釐) A7 _____ B7___ 五、發明説明(14) 租線(VLL)服務。VXL服務係基於一虚擬路徑觀念,並允 許線容量直接地適合顧客需求,江容易的變更而不用修改 界面結構。大量邏輯連接可透過使用者網路界面(UNIs)提 供給一使用者。而且,一顧客適合的服務品質亦可提供給 一顧客,符合使用者之需求。因此,多種服務級別,服務 及執行參數之品質可被選定。例如,聲音服務要求低的傳 送延遲但可容忍高位元錯誤,而相反的,資料通訊可容忍 較大的網路延遲,但對於位元錯誤卻是敏感的。因此,一 特定應用之服務階層之品質可於一服務提供者與一顧客之 間合約上的同意’並手動地或自動地審核以確保依從 (compliance) 〇 如圖6所示,据供一 ATM網路所完成之虚擬出租線(VLL) 服務之例子。網路終端A至E,每個分別透過流程實施節點 60 1至605而連接至ATM縱橫-連接節點6丨丨至6 14。ATM網 路包括多個ATM縱橫-連接節點6 11至6 14,每個縱橫-連接 節點可提供在虚擬路徑及在虛擬通道階層之路徑。流程實 施節點601至605係位於ATM網路之邊緣,以保護網路避免 可说的超載。此功能確保當連接設定後不會受到干擾。額 外的服務可由加入服務至一或多個縱橫-連接節點61丨至 6 14。而π成。在圖6之網路中,以波浪線62丨敘明一終端 C與D之間的範例虛擬路徑。—終端六與6之間的第一虚擬 連接由虛線63 1敘明,而終端C與Ε之間的第二虛擬連接由 點線6 3 2敛明。 除了圖6所示虛擬出租線以外,其他的服務,例如 ___ -17- 本纸張尺度適f丨’關家標準(CNS ) Α4規格(210x797公楚)----- (請先閱讀背面之注意事項再填寫本頁 袈—— 經浐部中夾打^-/:J:iJT,消合竹"卬;^ A7 ________________B7 五、發明説明(15 ) SMDS/C刪與訊財繼,可以地取決於㈣接服務至網 路中的ATM即點之要求而加入。在常駐區域,ΑτΜ技術可 用以提供新的改良的娛樂服務,例如一要求即顯示之終端 使用者。一ATM網路之彈性使得提供大眾服務,例如長距 離教育,在家採購,及遊戲是可能的。 圖7顯示一 ATM絪路,其太傳送網路上配罾。 該層包括一用戶前竑絪玖層7〇1,—區域傳送網路層7〇2, —地域性傳送網路層703,及一國際性傳送網路層7〇4。多 個ATM商業網路節點711至714控制來自用戶前端715與 LANs 7 16之具料流程,為在區域傳送網路7〇5之sdh縱橫_ 連接節點722 ,服務各別的多個add_dr〇p多工器 (ADM)721。區域縱橫_連接節點722係依次透過地域性傳 运網路層I地域性縱橫_連接節點73丨連接,其中兩個係由 add-drop多工器(ADM)732連接。在區域傳送網路層7〇2, —對ATM存取節點乃3,與SDH環之中,包括add_dr〇p多 工器721,服務縱橫·連接722,並甩於具有達到一*stm_ 1(每秒155百禺位元)容量之用戶存取,B_ISDN服務之標準 化存取率。 既存之流量’例如普通舊式的電話服務(p〇TS,plain 〇ld Telephone SerVlce)亦可用遠端多工器及其他提供最後區域 迴路連接义存取節點,實行此環網路。ATM存取節點723 被分旱以存取來自—位置之不同的服務,且可包括使用不 同的VP/vr,夕整音與資料。在atm存取節點723中,atm 流量係集中以更有效的利用傳輸容量。 —-- 18 - 本紙張尺度適削,國國家辟(CNTn^T^x 297/A$ J ^ - ....... ! 11 - I ....... - In I. - -- I *tin ·-! - ....... .......... In--eJ (請先鬩讀背面之注意事項再填寫本頁) A7 A7 籽淡部中而 VP T,>vifr 合竹=(ί卬ί'::. _________________Β7 五、發明説明(16 ) ATM存取節點之大小可取決於所需的容量而改變。在地 域性傳送網路層703中,ATM縱橫_連接乃3用以路徑地區 區域之間的流量。在國際性傳送網路層7〇4中,如圖7所 示,ATM示不可見的。在一 ATM重疊網路中,適當的如圖 7所示,例如訊框中繼SMDS/CNDS之該等服務可簡單的加 入。B-ISDN之功能亦可由加人適#的軟體與硬體,而加入 存取及地域性節點。如圖7所示,一網路管理系統,例 如根據CCITT之TMN標準之操作,可被完成以提供必須的 網路管理功能給網路之Sdh與ATM元件。 子系統750之ATM網路之管理可根據網路管理系統之電 訊管理及操作提供(TM0S)家族而完成,而訪絪路管理系統 係由本發明之授權者Telef〇naktieb〇laget LM Encss〇n公司 提供。如此之網路可包括不同的功能,例如根據本發明所 教示而兀成之路控演繹法及擁塞控制,將於下敘明。 虚擬路程容量大_小量测 在大小量測一電信網路中之可用模式係以包含二階層結 構之大小量測問題視之,其中該二階層結構包括一第一實 際網路層,具分離的拓撲學與特定的鏈容量,及一第二虚 擬路徑層,具虛擬路繼與其特定的路徑。一流景要求係提 供給此模式之虛擬路徑。只處理大小量測網路容蚤之工 作,虚擬路徑,屮π介"0,以路徑化。每個虛擬路徑可透 過不同的實體鏈路通過,但將模擬一只包括一單一路径之 高速公路。每個虛擬路徑將具有一特徵區塊值,且一個特 徵配置答量值只具有和虛擬路徑一樣多模式的變數。 -19- 本纸張尺度適用中國國家標孪(CMS ) Μ規格(210X297公楚) (請先閱讀背面之注意事項再填寫本頁) 袈--- 訂----- A7 _________________B7 五、發明説明(17 ) 名岡”供應流量'1 (offered traffic)用以參者沿菩每個虚擬 路徑之傳輪容量之時間變化要求。名詞"流量需求 demand)用以指示每個鏈之供應流署的睡閽革均值。當在 Μ網路上的流量特徵可由—單_參數㈣模擬時,:流 量為同次的單級流量。當供應流量為非同次時,览通常使 用一多級之Poissonian分佈模擬。 曰供應流量亦可用一常態分佈來模擬。此參考正常的流 量。最後,網路大小量測亦可基於測量所決定之實際流 量° 多數使用者之傳送需求可結合成_聚集的流量流。例 如,=些使用者可同時地從DaIIas傳送訊息至St〇ckh〇lm。 替代官理廷些各別的多重傳送,以一通過一寬頻中繼站線 群,結合並傳送這些各別的多重傳送是更有效的。如上所 述,一虛擬通道為一動態可配置之終端-對-終端連接。一 虛梃路徑為一邏輯组態,其允許多重虛擬通道以一單一單 兀 起被官理並切換。此平均切換減低了全面的處理需 求’亚加速傳送。因為一虚擬路徑之管理比起虛擬通道或 各別X際電路之管理’是簡單的而且更有效的,由此技術 可獲得網路資源之顯著地改良運用。 虚提路徑大小量測模式 所考慮的基本模式為一種在一固定路徑下操作之連接導 向網路。假如我們定義實體網路為包括一組任意的連接鏈 J,則每個虎癡路徑(vp)或路徑r為一順序的列夷,允开件 為J的子隹。虛擬路徑與實體鏈路之間的關係可用一路 本紙張尺度述--- (請先閱讀背面之注意事項再填寫本页)A7 B7 5. Description of the invention (12 Form 1 ATM unit is a guest unit of an ATM transmission system. Each element or information unit includes its own connection and path information. This feature makes the service channel more direct. Each channel can implement different bit rates. Each ATM unit is identified by the information contained in the header of the virtual path identifier (vpI) and the virtual channel identifier (VCI) and the path. As mentioned above, the failure of a task is one or two end points, such as a switching system, a local area network (LAN) gate, or a yttrium network gate. A south 擗 踗 row line Cheng + Ming Mu pull logic _, wvpT value to identify a specific virtual path. As mentioned above, the virtual holding channel used by ATM 祜 俶 ten allows multiple virtual channels fVCd to be controlled in a single unit . Virtual communications with common characteristics, such as the same quality service (Q0S), can be clustered in a unit to be transmitted, processed, and answered in a hurry. This vocabulary simplifies the operation of the system And dimension_. Virtual paths and virtual channels are available Provides semi-permanent paths in ATM networks. Paths are set in the aspect_connection element or the "path connection table" (path c〇nnect taMes) of a multiplexer along a path to provide the system from one operation Creation and release. The virtual channel can also be signaled by the connection between a user and the network or 'on-demimd swuching'. One of the important characteristics of atm technology is Its protocol framework is also established with the rules called core-and-edge. The protocol function describes the form of information transmitted, such as retransmission, flow control, and delay. First read the notes on the back before filling in this 1) 袈 —— B7 V. Invention Description (13) The equalization is performed in the terminal of the edge of the ATM network. This stream is effective, service independent, core " Network, which only includes simple unit transmission and exchange functions. The ATM node in this core has neither error checking of the information tree nor any process control. The unit information is simply read, and then HEC is used to calibrate it. A single bit error that may affect the address, and the unit then switches to its destination. The ATM twitter layer (aaL, adaptation layer) is used on the edge of the network to provide advanced services. As shown in Figure 5, is In the B_LSDN service I CCITT reference mode imagine that AAL includes service imitation function. As shown in Figure A, there are three layers in the ATM standard. The first layer is the definition of the gypsum | face and the frame association f. The second ATM The layer is irrelevant to the actual medium selected, and has a # ^ 疋 structure, which provides multiplexing and demultiplexing and VpL / Vci conversion to spread the unit flow in the logical network. I3 sells today's AAL, which provides an important agreement between the service and the ATM layer, thereby allowing service independent ατμ transmission. The mapping between the original service format and the information column of an ATM unit is performed. Example functions provided by AAL include packet descriptions of different lengths, sequence numbering, timing recovery, and execution monitoring. Development of ATM in telecommunication networks One use of ATM technology can be used in the user's front end to provide high-speed data communication in or between user area networks. In addition, ATM can be used as a substructure resource, which is common to all services in the user front-end network, including audio and video communications, data transmission and multimedia applications. The example service that AT Μ node introduces into a public telecommunication network is to provide virtual output -16- This paper size is suitable for _ home standard (CMS) A4 specification (2 丨 〇χ 297 mm) A7 _____ B7___ V. Description of the invention (14 ) Leased Line (VLL) service. The VXL service is based on a virtual path concept and allows line capacity to be directly adapted to customer needs. Jiang easily changes it without modifying the interface structure. A large number of logical connections can be provided to a user through the user network interface (UNIs). Moreover, the quality of service suitable for a customer can also be provided to a customer, meeting the needs of the user. Therefore, multiple service levels, quality of service and execution parameters can be selected. For example, sound services require low transmission delays but can tolerate high bit errors. Conversely, data communications can tolerate large network delays but are sensitive to bit errors. Therefore, the quality of the service level of a particular application can be agreed on in the contract between a service provider and a customer, and manually or automatically audited to ensure compliance. As shown in Figure 6, according to an ATM An example of a virtual leased line (VLL) service performed by the network. The network terminals A to E are each connected to the ATM cross-connect nodes 6 丨 丨 to 6 14 through process implementation nodes 60 1 to 605, respectively. The ATM network includes multiple ATM cross-connect nodes 6 11 to 6 14 and each cross-connect node can provide paths at the virtual path and at the virtual channel level. The process implementation nodes 601 to 605 are located at the edge of the ATM network to protect the network from arguably overload. This function ensures no interference when the connection is set. Additional services may be provided by joining services to one or more cross-connect nodes 61 丨 to 6 14. And π Cheng. In the network of FIG. 6, an exemplary virtual path between terminals C and D is described by a wavy line 62. -The first virtual connection between terminals six and 6 is illustrated by the dotted line 63 1, and the second virtual connection between terminals C and Ε is illustrated by the dotted line 6 3 2. In addition to the virtual rental line shown in Figure 6, other services, such as ___ -17- This paper is suitable for the standard (CNS) Α4 size (210x797), please read first Note on the back, please fill in this page again 袈 ——Tap in the Ministry ^-/: J: iJT, Digestive Bamboo "卬; ^ A7 ________________B7 V. Description of Invention (15) SMDS / C delete and Xun Cai Ji It can be added depending on the requirements of the ATM point-to-point connection service to the network. In the resident area, the ATm technology can be used to provide new and improved entertainment services, such as end-user display upon request. An ATM network The flexibility of the road makes it possible to provide mass services, such as long-distance education, home purchasing, and games. Figure 7 shows an ATM road, which is configured on the transmission network. This layer includes a user front layer 70. 1, —regional transmission network layer 702, —regional transmission network layer 703, and an international transmission network layer 704. Multiple ATM business network nodes 711 to 714 control from the user front end 715 and LANs The material flow of 7 16 is the sdh aspect in the regional transmission network 705. At point 722, each of the multiple add_droop multiplexers (ADM) 721 is served. The regional aspect_connection node 722 is connected in turn through the regional transport network layer I regional aspect_connection node 73, two of which are connected It is connected by the add-drop multiplexer (ADM) 732. In the area transport network layer 702,-to the ATM access node 3, and the SDH ring, including the add_droop multiplexer 721, service aspect · Connect 722 and drop it to users with a capacity of * stm_1 (155 Hbits per second), standardized access rate of B_ISDN service. Existing traffic, such as ordinary old-style telephone service (p〇TS (Plain 〇ld Telephone SerVlce) can also use a remote multiplexer and other access nodes that provide the final area loop to implement this ring network. ATM access node 723 is divided to access different services from the location , And can include the use of different VP / vr, evening tone and data. In the ATM access node 723, the ATM traffic is concentrated to use the transmission capacity more efficiently. --- 18-This paper is suitable for standard, country and country (CNTn ^ T ^ x 297 / A $ J ^-.......! 11-I .......-In I.- -I * tin ·-!-....... .......... In--eJ (Please read the notes on the back before filling this page) A7 A7 VP T, > vifr 合 竹 = (ί 卬 ί '::. _________________B7 V. Description of the invention (16) The size of the ATM access node can be changed depending on the required capacity. In the regional transport network layer 703, ATM Aspect_Connection is used to route traffic between regions. In the international transport network layer 704, as shown in FIG. 7, the ATM is not visible. In an ATM overlay network, suitable as shown in Figure 7, such services as frame relay SMDS / CNDS can simply be added. The functions of B-ISDN can also include access and regional nodes by adding software and hardware. As shown in FIG. 7, a network management system, such as the operation according to the TMN standard of CCITT, can be completed to provide the necessary network management functions to the Sdh and ATM components of the network. The management of the ATM network of the subsystem 750 can be completed according to the telecommunications management and operation provision (TM0S) family of the network management system, and the visit to the Kushiro management system is made by Telef〇naktieb〇laget LM Encss〇n company provide. Such a network may include different functions, such as a road control deduction method and a congestion control method based on the teachings of the present invention, which will be described below. Large virtual path capacity_Small size measurement available in a telecommunications network is considered as a size measurement problem that includes a two-level structure, where the two-level structure includes a first actual network layer with separation Topology and specific chain capacity, and a second virtual path layer, with virtual paths following its specific path. The first-class view requirement is to provide a virtual path for this model. It only deals with the work of measuring the size of network flea. The virtual path, 屮 π 介 " 0, is used for pathization. Each virtual path can pass through a different physical link, but will simulate a highway that includes a single path. Each virtual path will have a characteristic block value, and a characteristic configuration response value has only as many pattern variables as the virtual path. -19- This paper size applies to China National Standards (CMS) M specifications (210X297). (Please read the precautions on the back before filling out this page) 袈 --- Order ----- A7 _________________B7 V. Invention Explanation (17) Minggang's "offered traffic" 1 is used to participate in the time change requirements of the capacity of the wheel along each virtual path. The noun "quote demand" is used to indicate the supply flow of each chain The average value of the sleeping sleeper. When the flow characteristics on the M network can be simulated by -single_parameter, the flow is a single-stage flow of the same time. When the supply flow is not the same time, it usually uses one or more levels. Poissonian distribution simulation. That is, the supply flow can also be simulated with a normal distribution. This refers to normal flow. Finally, the network size measurement can also be based on the actual flow determined by the measurement ° The transmission needs of most users can be combined into _ aggregated Traffic flow. For example, some users can send messages from DaIIas to StOckhlm at the same time. Instead of some separate multiple transmissions, they can be combined and transmitted through a broadband relay station line group. many Transmission is more efficient. As mentioned above, a virtual channel is a dynamically configurable terminal-to-terminal connection. A virtual path is a logical configuration that allows multiple virtual channels to be managed in a single unit. And switching. This average switching reduces the overall processing requirements of 'sub-accelerated transmission. Because the management of a virtual path is simpler and more effective than the management of virtual channels or individual X-circuits', this technology can obtain Significantly improved use of road resources. The basic mode considered for the virtual path size measurement mode is a connection-oriented network operating under a fixed path. If we define a physical network as including a set of arbitrary connection chains J, Then each tiger path (vp) or path r is a sequence of columns, and the opening is allowed to be a child of J. The relationship between the virtual path and the physical link can be described in this paper. (Read the notes on the back and fill out this page)

五、 發明説明( 18 徑矩陣/定義,其元件為: A7 B7 (式子1 圖圖Γ圖 及,體鏈路之間的拓撲學關係之方塊 · 虛挺路徑KPl包含實體鏈路戶I及尸2 ,而虛擬 路包含實體鏈路尸2及尸3_。 而虛挺 ” θ、]勺走量關係配置給不同的VPs,且相應在實I* 鍵路上所配置的容量係以矩陣格式給定如下: (請先閲讀背面之注意事項再填荇本頁) 裝. X〇VP =r Cpftys (式予2 ) fχ為上面(式子1)所定義形態之路徑矩陣,CVp為虚# 各亡容量向量,而為實體鏈路容量向量。 衣不配置給不同的VPs容量之實體鏈路容量向量汗 能超過任意實體鏈路上的可用實體容量。此限制可用簡等 的限制關係來表示: XC-VP ~ Cpf^s < Cphys (式子3V. Description of the invention (18-diameter matrix / definition, its components are: A7 B7 (Equation 1 diagram, figure Γ diagram, and the block of the topological relationship between body links. The virtual path KPl contains the physical link household I and But the virtual path contains the physical link corpse 2 and corpse 3_. The virtual relationship "θ,] is allocated to different VPs, and the corresponding capacity allocated on the real I * key path is given in a matrix format. Set it as follows: (Please read the precautions on the back before filling this page). X.VP = r Cpftys (Eq. 2) fχ is the path matrix of the form defined above (Eq. 1), CVp is virtual #each The dead capacity vector is the physical link capacity vector. The physical link capacity vector that is not allocated to different VPs capacity can exceed the available physical capacity on any physical link. This limitation can be expressed by a simple restriction relationship: XC -VP ~ Cpf ^ s < Cphys (Equation 3

1T 部 屮 A il 1\ _τ 消 fr A 印 其中’ Cmv,為一敘述在每個炙實體鏈路上的可用實體容巧 I向量。由於等式(式子3 )為一向量不等式,必須注意ΐ 是相應兩向量之元件必須滿足此不等式。如圖8所示的ί 單例子,路徑矩陣%為: (式子4) 本紙張尺度適ϋΓΡ -21 - 囤國家榡準(CMS ) Λ4規格(210X297公釐) A7 A7 (式子5 f請先閲讀背面之注意事項再填符本頁} 五、發明説明(19) 其中,虚擬路徑客量與相應配置給VPs之實體鏈路容量之 間的關係為: —10' 'cr Cl ' 1 1 * Cl 二 Cl+C2 _0 1 VP C7 phys 其中 "cr -C. β〇_ 為虛擬路技容量何量,且 iyr> *** C1+C2 ΥΓ C2 _ 為 只體缝路之配罾向量。 假如該程序具有一資源,該資源有無限的容量可接收所 有的呼叫並避免所有的堵塞,一給定路徑,的呼叫要求程 序可為任一固定的程序·以此我們知道假的佔據分佈^尤, 表示此假的無限容量資源之佔據程度,並共同地參考此技 藝之”供應流量,1。 虛擬路徑大小量測的問顥*太發明的系统輿方法中定 義,具具有兩個要點。首先,配置給每個虛擬路徑之傳輸 容量必須最佳化,以最小化一傳送費用函數。第二,對於 每個實體鏈路’配置給在此鏈中流通之不同虛擬路徑之容 量不得超過限制於實體鏈路之會際的傳輸容量。 配給不同虛擬路徑之實體容量可由—連續函數趨近> 該函數假設為範圍中之任一實數值。結果資源最佳 化工作不要求分離的最佳化及其所有的伴隨複雜度。在此 發明之大小量測問題,在不同VPS之間分享的負載並未考 -22- 本紙张尺度適川屮园國家標準(CNS ) A4規格(210X297公釐)The 1T part 屮 A il 1 \ _τ eliminates fr A, where ‘Cmv, is a vector describing the available physical capacity of each physical link. Since the equation (Equation 3) is a vector inequality, it must be noted that ΐ is that the elements of the corresponding two vectors must satisfy this inequality. As shown in the single example shown in Figure 8, the path matrix% is: (Equation 4) The paper size is ΓΡ -21-Country Standard (CMS) Λ4 specification (210X297 mm) A7 A7 (Equation 5 f Please Read the notes on the back before filling in this page} 5. Description of the invention (19) Among them, the relationship between the virtual path passenger volume and the physical link capacity allocated to the VPs is: —10 '' cr Cl '1 1 * Cl two Cl + C2 _0 1 VP C7 phys where " cr -C. Β〇_ is the amount of virtual road technology capacity, and iyr > *** C1 + C2 ΥΓ C2 _ is the distribution vector of the body suture road only If the program has a resource, the resource has unlimited capacity to receive all calls and avoid all congestion. A given path, the call requires the program to be any fixed program. From this we know the false occupation distribution ^ You, express the occupation degree of this false unlimited capacity resource, and collectively refer to the "supply flow" of this technique, 1. The question of virtual path size measurement is defined in the system method of the invention, which has two points First, the transmission capacity allocated to each virtual path must be Optimize to minimize a transmission cost function. Second, for each physical link, the capacity allocated to the different virtual paths circulating in this chain must not exceed the transmission capacity that is limited to the intervening physical links. The physical capacity of different virtual paths can be approximated by a continuous function. The function is assumed to be any real value in the range. As a result, resource optimization does not require optimization of separation and all its accompanying complexity. Invented here The size of the measurement problem, the load shared between different VPS is not considered -22- This paper size fits the Sichuan Gion National Standard (CNS) A4 specification (210X297 mm)

A7 五 '發明説明(2〇 ) 慮。供應流量假設以每個虚擬路徑定義。再者,因為網路 有固定路徑,供應流量之路徑係由虛擬路徑之選擇來固 定。 下推式”大小量測技街 虛擬路徑大小量測工作在本發明中视為一負載平衡問 ^其中負載為一適當選擇區塊測量之值,最佳化解法 係相應配置VP容量之選擇,纟中在每個虛擬路徑上的區塊 係儘可能以平均的方式分佈。使得區塊分佈平均的方法之 測量不同虛擬路徑上區塊值的發散(divergence) ’並然 後取小化此發散。此趨近可利用任意標準最小化演繹法來 完成,例如,眾所皆知的模擬韌化(annealmg)技術。 、相關的方法是首先確認具有最高區塊值之虚擬路徑, f後由重新配置其他vps容量,直到虛擬路徑不在為具有 取尚區塊之VP,以最小化此虛擬路徑之區塊。此公式相應 最小-最大最佳化之問題,且可由下列所述分析地形成。 。假如我們在0虛擬路徑以5(κρί)表示區塊;則具有最大 區塊之VP為,其中最大值是從所有VPs中取出 的。透過虛擬路徑集合之區塊測量的最大值對於vp大小量 "、J問題足義目標函數(亦為費用函數)。因此,最佳化程序 之目的為找出目標函數之最小值,如下: min maxiB^VP^) (式子6) 其中最小值透過所有的適當的组態而定義。 由於此技術包括下所有考慮之v p s之間的最高區塊值, _____________- 23 - 本紙張尺纽 ^^财辟(CNS) -—-- (請先閱讀背面之注意事項再填寫本頁} 装—A7 Five 'Invention Description (20). Supply flow is assumed to be defined per virtual path. Furthermore, because the network has a fixed path, the path for supplying traffic is fixed by the choice of a virtual path. The “push-down” size measurement technology street virtual path size measurement work is considered as a load balancing problem in the present invention, where the load is a value of a properly selected block measurement, and the optimal solution is the choice of correspondingly configured VP capacity. The blocks on each virtual path are distributed as evenly as possible. The method of making the block distribution average measures the divergence of block values on different virtual paths and then minimizes this divergence. This Approach can be accomplished using any standard minimization deduction method, for example, the well-known analog toughening (annealmg) technology. The related method is to first confirm the virtual path with the highest block value, and then reconfigure other vps capacity until the virtual path is no longer a VP with a block to accept, in order to minimize the block of this virtual path. This formula corresponds to the problem of minimum-maximum optimization, and can be formed analytically as described below. If we At 0 the virtual path is represented by 5 (κρί); then the VP with the largest block is, where the maximum value is taken from all VPs. Measured through the blocks of the virtual path set The maximum value of V is the objective function (also the cost function) for the vp size and J problem. Therefore, the purpose of the optimization procedure is to find the minimum value of the objective function, as follows: min maxiB ^ VP ^) (Equation Sub 6) The minimum value is defined by all appropriate configurations. Since this technology includes the highest block value between all vps considered, _____________- 23-This paper ruler ^^ 财 闯 (CNS)-- -(Please read the notes on the back before filling this page}

、1T A7 ------------------- B7 五、發明説明(21) ~種使用此技術以解決一最佳化問題之演繹法稱為"下推 式"演繹法。此演繹法遵循此事實,平均區塊分伟係相應 未限定VP大小量測問題之最佳解法。結果,最佳解法配置 容量給每個VP,κ得在每個VPs之區塊在一錯誤範圍中為 2等的。然而,如此之解法並不總是可信賴的,因為不同 實體鏈路之容量限制,需要於實體鏈路中傳送之所有VPS 之間分旱。 圖9頭7F — Λ fa網路所定義之大小量測虛擬路徑下推式演 繹法之說明實施例之不同步驟。大小量測處理從9〇2開 始,為不同的VPs之連接拓撲學之定義。不同的vps亦组合 為-vP大小量測集合。然後在9〇3中,州群集為了每個 傳送之實體鏈路。在904中,然後產生傳輸容量之初始配 置給每個VP。在905中,選定—降低區塊之目標圖形。為 了設定一目標,必須首先選擇—邊组測量〇在本發明之較 佳貫施例中,熵比率函數,如下一段所述,作為—區塊測 置。該目標值用以設定大小量測演繹法之中斷條件。 在906中,每個VPs之區塊係透過每個實體鏈路決定。在 907中,假如在一單一實體鏈路中通過之不同vps没有面對 相同或相似的區塊階程度,則目前配置給每個vps之容量 被修訂,以在一錯誤範圍中均等化vps之區塊值。容量= 由未配置的實體容量,或由一較少生產之vp至一較多生產 之VP重新配置已配置之容量,而加至VPs。此容量重新調 整之執行並不違背任一實體鏈路之容量限制。 在908中,此處理導致一或多個實體鏈路之辨識,其為此 _____-24- 本紙張尺度適州中 1¾國家標準(CNS ) A4&格(210>7^7^5"7 (請先閲讀背面之注意事項再填离本ί ) 袈--- 丁 — 、-° 經浐部中^"'^/;Jy T.消抡"竹"卬 y -.. I ... I..... .. ·1··· _ A7 B7 五、發明説明(22) 最佳化程序之瓶頸。杜VP堵塞之—實體鏈路為最高的,且 該不可由容量重新配置減低之堵塞係以一臨界鏈路(critical link)被參考。每個臨界鏈路決定可在實體鏈路通過之vps 上元成之最低的區塊。下推式演繹法之主要工作之一為在 最佳化程序之每級,辨識一給定虛擬路徑之臨界鏈路集 合。 —在9〇8中辨識一臨界鍵路,實禮容量可用如此之方 法’在此臨界鏈路通過之不同虛擬路徑之間重新配置。必 須注意的是,當一實體鏈路被發現為一臨界鏈路々印 時’其不具有未配置的容量„結果,只有通過一臨界鏈路 之VPs之間的容量重新配置’在演繹法達到大小量測程序 此級之後為可能的。 接著’在909中,具配置容量之VPs從所有需要被大小量 測之VPs之集合中去除。然後在910中,可用的實體鏈路容 量以前步驟所去除之VPs配置容量來降低。 該大小量測的問題因此降低為最小化剩餘VPs集合之最 南區塊機率之最佳化的問題。此允許使用遞迴重複進入演 繹法’而達到完成此程序。 兩步驟之區塊值現在作為剩餘大小量測問題之初始值。 在9 Π中’最佳化程序遞迴遞重複直到所有實體鏈路之容 量被配置為止《總而言之,此貪心式(greedy_type)之演繹 法從大小量測整個的Vps集合開始,並當剩餘之大小量測 虛擬路徑集合變成一空集合時,在9丨2中結束。 必頊強碉的是此所處理形態之每個大小量測演繹法之執 __ - 25 - 本纸張尺度適準(CNS ) A4規格(210X 297公釐) '一" (請先閱讀背面之注意事項再填寫本頁) 訂 A7 五 '發明説明(23 ) -?":^屮^扔^而:^-7消於合竹;印; 行,不需要完全的遵照圖9所敘述之步驟。大小量測演鋒 法之某些步驟之執行順序在細部執行上可能與圖9所救述 <步騾不同。 在足VPs木θ之分析地辨識該臨界鏈路之問題被證 明是-個困難的工作。習知技術中並未揭示直接從供應流 I,及實體鏈路容量限制決定臨界鏈路的方法。因此,下 推式演繹法利用-種反覆的趨近來辨識臨界料。廣擇法 由使用一甩於所有VPs之平均的大的區塊值來初始所有的 VPs。所選擇的初始區塊值必須夠大,以使得vp容量之初 始配置值之總和不超過不同實體鏈路之可用的實體容量。 '藉由餘留在最佳化程序之所有虛擬路徑集合上的區塊程 度 < 緩慢及平均的減低,當該鏈路首先達反一橫越實體鏈 路疋貫際的容量限制時,則臨界鏈路在每一階層均被辨識 , 出。 I 丨 利用一燏區塊測量之大小量谢 在大小量測處理之每一級用於辨識臨界鏈路之上述程序 <速度及效能係取決於模擬中區塊測量之複雜度。一般而 Erlang區塊測量(亦稱為時間擁塞區塊公式)係用以夹 定網路中V P容量之最佳的配置。 ; 本技術合併作為區塊測量之熵比率函數,以得到比那些 | 藉使用Erlang區塊測量所蒋得的結果更好。熵比率函數之 I 使用允許任意流量分佈之模橱,且大部分的情況,此計| <芫成要比某於其他區塊測量之計算而言快很多。亦可發 現’臨界鏈路之反覆搜尋可實質地被改良,主要遵循此事 -—___ - 26 - 、 本紙張尺度適( CNS ) A4^# ( 210X 297^1 j ' ' c请先閱讀背面之注意事項再填寫本頁) 裝----1T A7 ------------------- B7 V. Invention Description (21) ~ A deduction method using this technology to solve an optimization problem is called " Push " Deductive Method. This deduction method follows the fact that the average block score is the best solution to the corresponding undefined VP size measurement problem. As a result, the optimal solution allocates capacity to each VP, and κ must be equal to 2 in an error range for each VPs block. However, such a solution is not always reliable, because the capacity constraints of different physical links need to be divided between all VPSs transmitted in the physical links. Fig. 9 head 7F-different steps of the illustrated embodiment of the push-down deductive method of size measurement virtual path defined by the Δ fa network. The size measurement process starts from 902, which is the definition of the connection topology of different VPs. Different vps are also combined into a -vP size measurement set. Then in 903, the state cluster for each transmitted physical link. In 904, an initial configuration of transmission capacity is then generated for each VP. In 905, the target graphic of the selected-lower block is selected. In order to set a target, the edge group measurement must first be selected. In the preferred embodiment of the present invention, the entropy ratio function is described in the following paragraph as a block measurement. The target value is used to set the interruption condition of the size measurement deduction method. In 906, the blocks of each VPs are determined through each physical link. In 907, if different vps passing through a single physical link do not face the same or similar block level, the capacity currently allocated to each vps is revised to equalize the vps in an error range. Block value. Capacity = Reconfigure the configured capacity from unconfigured physical capacity, or from a less-produced vp to a more-produced VP, and add to the VPs. The implementation of this capacity readjustment does not violate the capacity limit of any physical link. In 908, this process results in the identification of one or more physical links, which is _____- 24- this paper size is 1¾ National Standard (CNS) A4 & Grid (210 > 7 ^ 7 ^ 5 " 7 (Please read the notes on the back before filling in this book) 袈 --- 丁 —,-° In the Ministry of Economic Affairs ^ " '^ /; Jy T. Elimination " 竹 " 卬 y-.. I ... I ..... .. · 1 ··· _ A7 B7 V. Explanation of the invention (22) Bottleneck of the optimization procedure. Du VP blocked-the physical link is the highest, and the unavailable capacity Reconfiguration to reduce the congestion is referred to a critical link. Each critical link determines the lowest block that can be passed on the vps of the physical link. The main work of the push-down deduction method One is to identify a set of critical links for a given virtual path at each stage of the optimization procedure.-Identify a critical bond in 908. The actual capacity can be used in such a way that the critical link passes through it. Reconfiguration between different virtual paths. It must be noted that when a physical link is found to be a critical link seal, 'it does not have unconfigured capacity.' As a result, Only the reconfiguration of the capacity between VPs through a critical link is possible after the deductive method reaches the level measurement procedure level. Then, in 909, the VPs with configured capacity are from all VPs that need to be measured by size Then, in 910, the available physical link capacity is reduced by the VPs allocation capacity that was removed in the previous step. This size measurement problem is therefore reduced to the best that minimizes the southernmost block probability of the remaining VPs set. This problem allows the use of recursive recursive deduction to complete this process. The two-step block value is now used as the initial value for the remaining size measurement problem. In 9 Π 'Optimize recursive recursive repetition Until the capacity of all physical links is configured, "In short, this greedy_type deduction method starts with the size of the entire Vps set, and when the remaining size measurement virtual path set becomes an empty set, in 9 丨The end of the 2nd. It is necessary to insist on the implementation of each size measurement deduction method __-25-The paper size is correct (CNS) A4 size (210X 297 Li) '一 " (Please read the notes on the back before filling out this page) Order A7 Five' Invention Note (23)-? &Quot;: ^ 屮 ^ Throw ^^: ^-7 It is not necessary to completely follow the steps described in Fig. 9. The execution order of some steps of the size measurement and deduction method may differ in detail from the steps shown in Fig. 9 < Analyzing the problem of the critical link analytically proved to be a difficult task. The method of determining the critical link directly from the supply flow I and the physical link capacity limitation has not been disclosed in the conventional technology. Therefore, the push-down deduction method uses a kind of iterative approach to identify the critical material. The wide selection method starts by using a large block value averaged over all VPs to initialize all VPs. The selected initial block value must be large enough so that the sum of the initial configuration values of the vp capacity does not exceed the available physical capacity of the different physical links. 'With the degree of block remaining on all virtual path sets of the optimization process < slow and average reduction, when the link first reaches the capacity limit across the physical link across the physical link, then Critical links are identified at each level. I 丨 Size measurement using a block measurement Thank you The above procedures for identifying critical links at each stage of the size measurement process < Speed and performance depend on the complexity of the block measurement in the simulation. Generally, Erlang block measurement (also known as the time congestion block formula) is used to determine the optimal configuration of V P capacity in the network. This technique is combined as a function of entropy ratio for block measurements to get better results than those obtained by using Erlang block measurements. The I of the entropy ratio function uses a mold cabinet that allows any flow distribution, and in most cases, this calculation is much faster than the calculation of some other block measurement. It can also be found that the iterative search of the critical link can be substantially improved, mainly following this matter --- ___-26-, the paper size is appropriate (CNS) A4 ^ # (210X 297 ^ 1 j '' c Please read the back (Notes to fill out this page)

、1T tnn i · A7 B7 五、發明説明(24 實之結果為熵比率函數為凸函數(c〇nvex functi⑽、。在, 摘比率函數敘述-大小量測演繹法之前,揭示橘 ^ 之特徵是有用的。 数块比率函數作為一區魂測詈 如早先所提,區瑰測量之選擇係臨界下推式演繹法。基 於熵比率函數之區塊測量之一般式子將接著被推衍並應用 於範例情沉,其Φ供應流量係替代地由單級(sin幻e_ciasd 及由多級(multi-class') Pm.ssonian 分怖模擬。 摘比率函數在此技藝中為習知的’並用以模擬實體鏈路 1¾層的擁塞’爹見們如,J γ Hui,數位與類比電境系統 之國際雜誌之”用於多層流量之呼叫許可與頻寬指派之擁 塞測量(A Congestion Measure for Call Admission and Bandwidth Assignment for Multi-Layer Traffic,International Journal 〇f Digital & Analog Cabled System)(1990),但其尚未用來解決 虛擬路徑階層或網路階層上大小量測或計劃之問題。此 外’熵比率函數已經用來定義一實體鏈路之”有效容量”之 概念。須特別注意的是,使用熵比率函數之大小量測技術 並未限定為遵循一 P 〇 i s s ο η分佈之供應流量,且該系統與方 法可適用於任意形態的流量分佈,包括由測量所決定之流 量分饰。 飽和的區塊機率可被定義成為流量要求超過一特定傳輸 容量之機率值。飽和機率亦可稱為"尾標機率”(tail probabiHty),因為其表示供應流量分佈之尾標之機率量。 對與此尾標機率之習知趨近稱為Chernoffs Bound ’於下推 -27- _ (請先閱讀背面之注意事項再填寫本頁) HH 11 nn g - tl^i · 裝丨 訂 « n^lf m· .^m --1- - 本紙張尺度適用中囤國家標準(CNS ) A4規格(210X297公釐) A7 B7 五、 發明说明(25 衍。 X為一任意分佈隨機變數,而c為給定值。對於所有的正 值ί ’可導出下列邊界值: X>C) = P(e^>e^)< ^-Ιη(£·(β^ (式子7) 其中尸(X>C)為隨機變數Ζ取大於C之值。 此邊界值之推析係基於Markov不等式。項目/η表 不log moment的產生函數,亦稱為cmnuland函數/z〇)。最 緊的邊界值(tightest bound)(亦稱為chernoff's邊界值)係以 相關5,最大化指數sC- //(功而獲得。最大值在時達 到’其中’為方程式之唯一正解,〆⑷。 可客易的看出⑷係以s而增加,界此確保其根之單— 性。〆⑴以s而增加因為μ"(5)$0 /此係由於第二導函數 係等於(移位)分饰之變化之事實。&最大的指數亦由7 表示且稱之為场比率函數。'瑪比率④數為μ—函數 凸擁塞轉換,且可以下列式子表示: (式子8 由於障峰1而且岭X卜 -分佈之左尾標之相似式子可由下列不等式推衍: -lnP{X < 〇 > {-sC-lnE(e^)) . / 燒比率函數分佈之右與左位標之間^ > 〇~sC> (式子9 ) 的關係可用參數S表 (請先鬩讀背面之注意事項4"衣頁」 m^i I* nn i k .n -28- 本纸張尺度適州中围國家標準(CNS ) A4纟見格(210X297公 A7 五、發明説明(26) 示, 如下 ^-X (~^(.S)) - Ιχ (C(—-S)) (式予ΙΟ) Q此,由改變參數的符號s,可從熵比率函八、 標切換至左位標,反之亦然。 刀 <右4 用於同質Poisson流董之熵比率函數 當供應流量為同質時太小量測虚擬路徑之熵比率承 用被認為式第一的。同質P〇iss〇n流量可由—寬頻=數的名 P中間呼叫到達率(mean ca!l arrival rate>,及、々 '' =凄 之中間參數乃(亦以中間保持時間作為參考)特徵:個^ 要求β因此之由中間呼叫到達率及中間保持時間流ί Π ’屋生。同質流量之cumuiand函數以下面關係說明:丨 <請先閲讀背面之注意事項再填寫本頁j 裴 p{esP-\) (式子1 訂--- 用於同質流量之結果配置容量C及掹士玄.^ & 分里L夂堝比率函數I可由下列 子給定: •^n I -- i C= ppesp (式子12 好浐部中-欠|1準而.-i!T.消於合w.^.卬-9.C人. 且 /(〇=〆(〇(:-〆/((:)) (式子13 (式子14) 方程式之解法(式子 -29 本紙張尺度適/f]中國國家標準(CNS ) Α4規格(210X297公整 A7 B7 五、發明説明( 271T tnn i · A7 B7 V. Explanation of the invention (24 The result is that the entropy ratio function is a convex function (c0nvex functi⑽.) Before the abstract ratio function description-size measurement deduction method is revealed, the characteristics of tangerine ^ are Useful. Several block ratio functions as a region soul test. As mentioned earlier, the selection of the area rose measurement is a critical push-down deductive method. The general formula for block measurement based on the entropy ratio function will then be derived and applied. In the case of the example, the Φ supply flow is instead simulated by a single-stage (sin magic e_ciasd and multi-class') Pm.ssonian. The ratio function is known in this art and is used to Simulate congestion at layer 1¾ of the physical link. Daddy, for example, J γ Hui, International Journal of Digital and Analog Electrical Systems, "A Congestion Measure for Call Admission for Multi-layer Traffic" and Bandwidth Assignment for Multi-Layer Traffic, International Journal 〇f Digital & Analog Cabled System) (1990), but it has not been used to solve the size measurement or calculation at the virtual path level or the network level. In addition, the 'entropy ratio function has been used to define the concept of "effective capacity" of a physical link. It must be noted that the size measurement technique using the entropy ratio function is not limited to follow a P oiss η-distributed supply flow, and the system and method can be applied to any type of flow distribution, including flow ornamentation determined by measurement. Saturated block probability can be defined as the probability that the flow requirement exceeds a specific transmission capacity. The saturation probability can also be called " tail probiHty " (tail probabiHty), because it represents the probability amount of the tail of the distribution of the supply flow. The approach to this tail probability is called Chernoffs Bound 'Under Push- 27- _ (Please read the precautions on the back before filling in this page) HH 11 nn g-tl ^ i · Binding «n ^ lf m ·. ^ M --1--This paper standard is applicable to national standards (CNS) A4 specification (210X297 mm) A7 B7 V. Invention description (25). X is a random variable with arbitrary distribution, and c is a given value. For all positive values, the following boundary values can be derived: X > C) = P (e ^ > e ^) < ^ -Ιη (£ · (β ^ (Equation 7) where cadmium (X > C) is a random variable Z taking a value greater than C. The inference of this boundary value is based on Markov's inequality. The item / η table is not a generating function of the log moment, also known as cmnuland function / z〇). The tightest bound (also called chernoff's boundary value) is obtained by maximizing the exponent sC- // (with the correlation 5. The maximum value reaches 'where' is the only positive solution of the equation, 时It is easy to see that ⑷ increases with s, and this ensures the unity of its roots. 〆⑴ increases with s because μ " (5) $ 0 / This is because the second derivative function is equal to (shift ) The fact that the decoration is changed. The largest index is also represented by 7 and is called the field ratio function. The 'ma ratio ④ number is μ—a function of convex congestion conversion, and can be expressed by the following formula: (Equation 8 due to Barrier peak 1 and the similarity of the left tail of the ridge X b-distribution can be derived from the following inequality: -lnP {X < 〇 > {-sC-lnE (e ^)). / Right of the burn rate function distribution ^ ≫ 〇 ~ sC > (Equation 9) The parameter S table can be used (please read the note on the back 4 " clothing page first) m ^ i I * nn ik .n -28- The size of this paper is in accordance with the National Standard (CNS) A4 of Shizhou (210X297 male A7) 5. The description of the invention (26) is as follows ^ -X (~ ^ (. S))-Ιχ (C (-S )) (Formula I ) Q Here, by changing the sign s of the parameter, you can switch from the entropy ratio function to the left symbol, and vice versa. Knife < Right 4 The entropy ratio function for homogeneous Poisson flow director when the supply flow is homogeneous Too small to measure the entropy ratio inheritance of the virtual path is considered to be the first. The homogeneous P0iss〇n traffic can be-broadband = number of name P intermediate call arrival rate (mean ca! L arrival rate>, and 々 ' The intermediate parameter of '= 凄 is (also using the intermediate hold time as a reference) characteristics: a ^ request β is therefore caused by the intermediate call arrival rate and the intermediate hold time flow ί Π' house. The cumuiand function of homogeneous traffic is explained by the following relationship:丨 < Please read the precautions on the back before filling in this page j 裴 p (esP- \) (Formula 1 order --- for the result of homogeneous flow allocation capacity C and Shi Xuan. ^ &Amp; Fen Li L The cauldron ratio function I can be given by: • ^ n I-i C = ppesp (Equation 12 in the good part-owing | 1 quasi. -I! T. 合 合 w. ^. 卬- 9.C 人. And / (〇 = 〆 (〇 (:-〆 / ((:)) (Equation 13 (Equation 14) Solution of the equation (Equation-29 This paper is suitable for the standard / f) Chinese national standard CNS) Α4 size (210X297 well integer A7 B7 V. invention is described in (27

CC

C_ 乂 PP -l, (式予1 5 圖10揭繪用於單位頻寬要求^中不同供應流量值之熵比 率函數特徵之圖表。圖u顯示配置容量C之函數表示、 移參數$。 、 ' 如圖10及1.1所示,熵比率函數具有三個顯著的特性。首 先,熵比率函數為一凸函數,其在分布之中間值,即當C —P時,達到〇之最小值。第二,位移參數5從負值變化為 正值,以在分布之中間值,即當c= P時,增加c的值。從 圖11可見,當C<p時,位移參數5為負值;而當 時’位移參數S為正值。第三,位移參數S為單一的 (monotonic),且以配置給一虚擬路徑之容量函數而增加。 轉換參數*s可因此而解譯為一機率分伟位移參數。當位移 參數為負值時,機率分佈位移至比較相應於位移參數〇值 之機率分侔的左邊。位移參數為正值時,機率分佈位移至 右邊。 用於多鈒Poisson流量之熵比率函數 流量模式可擴展至由一多級Poisson分佈而特徵化供應流 量’且相應如此供應流量模式之摘比率函數於下面推衍。 以多級分佈之熵測量取代單級分佈之熵測量推導出熵比 率函數不再以配置容量C顯性地表示。避免此問題熵比率 函數以位移參數s表示,圪移參數s亦用為控制參述。以增 30 本紙張尺度適用中國國家標準(CMS ) A4規格(210X297公釐) υ^— 1: ft (請先閱讀背面之注意事項再填寫本頁} 訂 < n^— In —i-i mu un Α7 Β7 28 五、發明説明( 口此參數疋絕對值,配置容量可隱性地被改變’因此 熵測量被正的增加。 U 5 表示級K其中/的值從1至幻之隨機流量。定義峰波 頻見要求L,中間呼叫到達率r,.,及每級保持時間h... 严 一 1 i 9 1¾. 于—r!. !·且隨機變數JTAcumulative值之預測值為: ΚΙ斗— (式子16) 用於多級流量之熵比率函數Λ (C1)為飽和機率之負對 Σχ* 、 <*1 數之估計值: (讀先閱讀背面之注意事項再填寫本頁) > ZXt (Ο (式子1 7) IT-' 其中&為流量級之數目,而匚為配置給執行此合計多級流量 之VP之容量。 下列關係式為用於所有分侔之橘比率函數之—般特性: hx,C_ 乂 PP -l, (Eq. 15) Figure 10 reveals a graph of the characteristics of the entropy ratio function for different supply flow values in the unit bandwidth requirement ^. Figure u shows the functional representation of the capacity C and the shift parameter $. As shown in Figures 10 and 1.1, the entropy ratio function has three significant characteristics. First, the entropy ratio function is a convex function, which is the middle value of the distribution, that is, when C-P, it reaches a minimum of 0. Second, the displacement parameter 5 changes from a negative value to a positive value, so as to increase the value of c when c = P. As can be seen from Figure 11, when C < p, the displacement parameter 5 is negative; At that time, the displacement parameter S was positive. Third, the displacement parameter S was monotonic and increased by a capacity function allocated to a virtual path. The conversion parameter * s can be interpreted as a probability split Displacement parameter. When the displacement parameter is negative, the probability distribution is shifted to the left of the probability score corresponding to the value of the displacement parameter 0. When the displacement parameter is positive, the probability distribution is shifted to the right. Entropy ratio function flow model can be extended to a multi-stage Poisson The distribution rate is characterized by the supply flow 'and the corresponding ratio function of the supply flow pattern is derived below. The entropy measurement of the multi-level distribution is replaced by the entropy measurement of the single-level distribution. To avoid this problem, the entropy ratio function is expressed by the displacement parameter s, and the displacement parameter s is also used as a control reference. In order to increase 30 paper sizes, the Chinese National Standard (CMS) A4 specification (210X297 mm) is applied. ^^ 1: ft (Please read the precautions on the back before filling in this page} Order < n ^ — In —ii mu un Α7 Β7 28 V. Description of the invention (This parameter 疋 absolute value, the configuration capacity can be changed implicitly '' therefore The entropy measurement is positively increased. U 5 represents the random flow of level K where / has a value from 1 to magic. For the definition of the peak wave frequency, see the requirement L, the intermediate call arrival rate r,., And the hold time of each level h ... -1 i 9 1¾. In —r !.! · And the predicted value of the random variable JTAcumulative value is: ΚΙ 斗 — (Equation 16) The entropy ratio function Λ (C1) for multistage flow is the negative pair of saturation probability. Σχ *, < * 1 number estimates: Note this page, please fill in this page) > ZXt (〇 (Formula 1 7) IT- 'where & is the number of traffic levels, and 匚 is the capacity allocated to the VP that performs this total multi-level traffic. The following relationship is General characteristics of the tangerine ratio function for all tillers: hx,

E 1% (式子1 8 ^浐·部中-ΛΊ η 消抡合^:.^卬4-;?;. 相同地可知Vc)為C之凸函數 用於夕級Poisson流里之對數轉率(i〇g m〇ment)產生函數 係由下面關係式給定: ^(s) = YJPi(ssp, -1) -31 本紙張尺度適州中國國家標準(CNS ) Α4規格(210Χ297公釐) (式子19 鉍浐部中""'^;0公^消於"竹71卬絮 A7 p~_________B7 _ 五、發明説明(29) 而摘比率函數具有一般的格式: ^x(C(s))= ^ (sC-^^(s)) = s—Jux(s)~Jux(s) (式子 2 〇 ) 用於多級流量之熵比率函數可以位移參數s表示,如卞: TU =Σ {sPiPfispi - p{QSPi - 0) (式子 21) ί=Ι 1=1 ν Ν 其中’配置容量c尚可功能地相關於位移參數J,如下: C⑻=丢"㈡=Σ A·< (式子2 2 ) z=l \ 以多級分佈之熵測量取代單級分佈之熵使得此問題分析 性更複雜’因為熵測量不再以配置容量C顯性地表杀。此 複雜度係由於位移參數s不能從方程式(式子2 2 )中消杳。 然而’由於方程式(式子2 1 )以位移參數s表示编比率函 數,我們可由改變ί以取代C。因此容量值可由方程式(式 子2 2 ),在演繹法之每個反覆步驟計算。必須注意的是,j I 0值係相應熵測量之〇值。大小量測演繹法係由設定所有 的VPs之s為〇而初始。 用於常態分佈流量之熵比率函數 烟比率函數亦可用於不同於那些基於p〇iss〇nian供應流量 分佈之其他的流量模式。兩種重要的流量模式於下討論。 第一流里模式係基於一正常供應流量分佈之编比率函數。 熵比率函數之相應表示由此流量模式推衍出。第二流量模 式係基於實際流量流程測量’而非供應流量分佈之顯性假 -32- 準(CNS ) A4規格(210X297公釐) ' (讀先閲讀背面之注意事項真填寫本育〕 ΐτ ,( /// 五、發明説明( 30 A7 B7 設所推衍之熵比率函數。用於正常的分佈流量之摘比率函數可參見驗之 触心Large Deviat祕_加此㈣π (SPnnger_Verlag,1985),其由下列式子定義: (C-m)2 w): (式子23 其中所為中間值,而σ為常態分佈σ )之變化。再 者, C-m σ2 (式子24 k兩個關係式可直接以控制參數S產生下列熵比率函數之 簡單式子:Jx(C(s))= -γ- s2 (式子25) 因此在正常的分佈流量之情沉下,熵比率函數以一簡單的 一次函數(且為凸函數)表示。 測量流量之熵比率函數 由於未來網路中所提供服務之多樣性將比今日所使用的 要來得多’非常可能的是未來寬頻網路在網路供應流量的 型式上將歷經極大的變化。所以,所有基於流量分佈之特 定理想化表示之假設的模式,均可能因其固有的不變性而 變得不夠。利用由流量測量所推演之熵比率函數可提供一 解法給此不同流量模式之估計問題。 -33- 本纸張尺度適用中囚國家標準(CNS ) A4規格(210X 297公釐) (請先閲讀背面之注意事項再填寫本頁j 丨,衣 丁 1 --5 i, ! A7 B7 五、發明説明(31 上所討論之流量模式已由呼叫-階層時間標度(calMevei time scale)定義。相反的,流量測量平均性於一標準Atm 單元之時間標度上定義。此呼叫-階層時間標度可視為趨近 單位-階層時間標度。在一呼叫期間隨機地改變流量要求可 因此由一或多個敘述呼叫-階層時間標度上定值寬頻要求之 參數而摘要敘述。 最近被建議的是熵比率函數可從呼叫-階層上的流量測量 來估計。例如參考N. G. Duffield等人,ATM流量流之熵: 用以估 Q〇S 参數之工具 Entropy of ATM Traffic Streams : A Tool for Estimating QoS Parameters){OMm Institute for Advanced studies, 1994)。 该比率函數上超載之流量實施 作為一區塊測量之熵比率函數之解譯,假如在每個實體 鏈路 < 平均供應流量低於該鏈上相應之可用的實體容量, 即假如,其可良好的工作。然而,此條件可被某 些理論的超載條件上達反。考量下面基於同質p〇iss〇n流量 及時間擁塞區塊測量之例子(即區塊公式)。 表格2列出二個配置容量值及由一固定區塊值〇 μ計算之 相應流量要求。注意在最末—項時,流量係大於配置狀 沉,雖然區塊相對地是較低的值。 34E 1% (Equation 1 8 ^ 浐 · 中 中 -ΛΊ η elimination coupling ^:. ^ 卬 4-;?;. Similarly, Vc) is a convex function of C for logarithmic conversion in Poisson flow The rate (i〇gm〇ment) generation function is given by the following relationship: ^ (s) = YJPi (ssp, -1) -31 The paper size is the state standard of China (CNS) A4 (210 × 297 mm) (Equation 19 in the Bismuth tritium " "'^; 0 公 ^ 消 于 " Bamboo 71 卬 7A7 p ~ _________ B7 _ 5. Explanation of the invention (29) and the ratio function has a general format: ^ x ( C (s)) = ^ (sC-^^ (s)) = s-Jux (s) ~ Jux (s) (Equation 2 〇) The entropy ratio function for multi-stage flow can be expressed by the displacement parameter s, such as卞: TU = Σ {sPiPfispi-p {QSPi-0) (Equation 21) ί = Ι 1 = 1 ν Ν Where 'the configuration capacity c can still be functionally related to the displacement parameter J, as follows: C⑻ = drop " ㈡ = Σ A · < (Equation 2 2) z = l \ Replacing the entropy of the single-level distribution with the entropy measurement of the multi-level distribution makes this problem more analytically 'because the entropy measurement is no longer apparently killed by the configuration capacity C . This complexity is because the displacement parameter s cannot be eliminated from the equation (Equation 2 2). However, since the equation (Equation 2 1) represents the ratio function with the displacement parameter s, we can change C to replace C. Therefore, the capacity value can be calculated by the equation (Equation 2 2) at each iterative step of the deductive method. It must be noted that the value of j I 0 is the value of the corresponding entropy measurement. The size measurement deduction method is initiated by setting s of all VPs to zero. Entropy Ratio Function for Normally Distributed Flow The smoke ratio function can also be used for other flow patterns than those based on the p0issoonian supply flow distribution. Two important traffic patterns are discussed below. The first-rate model is an edited ratio function based on a normal supply flow distribution. The corresponding representation of the entropy ratio function is derived from this flow pattern. The second flow mode is based on the actual flow flow measurement 'not the dominant false-32-standard (CNS) A4 specification (210X297 mm)' (read the precautions on the back and fill in this education first) ΐτ, (/// V. Description of the invention (30 A7 B7 Let's deduct the entropy ratio function. For the ratio function used for normal distribution flow, please refer to My Heart's Heart Large Deviat Secret _ plus this ㈣π (SPnnger_Verlag, 1985), It is defined by the following formula: (Cm) 2 w): (where Equation 23 is the intermediate value, and σ is the normal distribution σ). Furthermore, Cm σ2 (the two relational expressions of 24 k can be directly expressed as The control parameter S produces the following simple expression of the entropy ratio function: Jx (C (s)) = -γ- s2 (Equation 25) Therefore, under the condition of normal distribution flow, the entropy ratio function is a simple linear function (And it is a convex function). The entropy ratio function for measuring the traffic will be much more diverse than the services provided in the future due to the diversity of services provided in the future network. It is very likely that the future broadband network will supply network traffic. Admiral has undergone tremendous changes. So, all flow-based The hypothetical models of the specific idealized representation of the quantity distribution may be insufficient due to their inherent invariance. Using the entropy ratio function derived from flow measurement can provide a solution to the estimation of this different flow model. -33 -This paper size applies to the National Prisoner's National Standard (CNS) A4 specification (210X 297 mm) (Please read the precautions on the back before filling in this page j 丨, Yiding 1 --5 i,! A7 B7 V. Invention Note (31) The traffic pattern discussed above has been defined by the call-level time scale (calMevei time scale). In contrast, the flow measurement average is defined on the time scale of a standard ATM unit. This call-level time scale Can be considered as approaching unit-hierarchical time scale. Randomly changing traffic requirements during a call can therefore be summarized by one or more parameters describing the broadband requirements on the call-hierarchical time scale. Recently suggested is The entropy ratio function can be estimated from call-level traffic measurements. For example, refer to NG Duffield et al., Entropy of ATM Traffic Flow: Entropy of ATM Traffic St reams: A Tool for Estimating QoS Parameters) {OMm Institute for Advanced studies, 1994). The overloaded traffic on this ratio function is interpreted as an entropy ratio function measured as a block, provided that at each physical link < average The supply flow is lower than the corresponding available physical capacity on the chain, ie if it works well. However, this condition can be reversed by some theoretical overload conditions. Consider the following example based on the measurement of homogeneous poisson traffic and time congestion blocks (ie, the block formula). Table 2 lists the two configured capacity values and the corresponding flow requirements calculated from a fixed block value of 0 μ. Note that in the last term, the flow rate is greater than the configuration state, although the block is relatively low. 34

木纸伕尺度適用中國國家標準(CNS )Α4規格(210X297公釐) (讀先閱讀背面之注意事項再填济本育)The size of the wooden paper rafters applies to the Chinese National Standard (CNS) A4 specification (210X297 mm) (read the precautions on the back before filling in the education)

A7 B7 五、發明説明(32A7 B7 V. Description of Invention (32

此範—例中顯示熵區塊測量需要擴展,以當達反條件 五時覆蓋超載情沉。在數學上,如此之擴展可輕 易地完成。如前所述!熵比率函數為—在五上具有一 θ 小〇值之凸函數。摘比率函數之左分支定義超載範圍(Ζ 圖10及11)。在此範圍,一熵比率函數之遞增係相應—配 置容量遞減及控制參數之負值。由改變控制參數之符號, 基於熵比率函數之下推式演繹法可簡單地擴展,以$ =超 載範圍。如此之擴展只須要對原始的大小量測演繹法作二 量的修改。' 乂 剩餘的問題本質上主要屬觀念性問題,即如付 |」奸洋此擴 展。熵比率函數之左分支係相應機率量之左尾標之趨近 如下: ^P(X^IAC(s)) (式子26) 且可以一配置容量C之使用之熵測量解譯。 因為在初始狀態,資源之配置超過實際的資源,即 -35- 本纸張尺度適/Π中闼囤家標準(CNS ) A4規格(210X297公釐) (請先閲讀背面之注意事項再填寫本頁)This example—the example shows that the entropy block measurement needs to be extended to cover the overload situation when the anti-condition 5 is reached. Mathematically, such an extension can be done easily. As mentioned before! The entropy ratio function is a convex function with a value of θ small 0 on five. The left branch of the ratio function defines the overload range (Z Figures 10 and 11). In this range, an increase in the entropy ratio function is the corresponding decrease in the configuration capacity and the negative value of the control parameter. By changing the sign of the control parameter, the push-down deduction method based on the entropy ratio function can be simply extended to $ = overload range. Such an extension would require only two modifications to the original size measurement deduction method. '乂 The remaining problems are essentially conceptual problems, such as paying | The left branch of the entropy ratio function is approximated by the left tail of the corresponding probability quantity as follows: ^ P (X ^ IAC (s)) (Equation 26) and can be interpreted by using an entropy measurement to configure the capacity C. Because in the initial state, the allocation of resources exceeds the actual resources, that is -35- this paper is suitable for the standard / Π 闼 闼 standard (CNS) A4 specifications (210X297 mm) (Please read the precautions on the back before filling in this page)

-----' V 訂 A7 B7 五、發明説明(33 ) £(;)<已如,結果當相應一.熵比率函數遞增時,此應用必 須為遞減的。 在超載範圍中最佳化目的解譯之一可為如下所述。為增 進此範圍之熵應用測量之分侔的平均性,最大的資源用戶 (即VP具有最低的熵)被辨識,且此外項的表示(extreme representative)為遞減的。最大用戶遞減應用係相應一超載 範圍遞增搞比率函數。因此,此趨近係相應於一最隹化問 通之最大最小公式f〇rmulati〇n)。必須注意的 是,限制條件項在由初始地落入超載範圍之值開始應用。 於此,我們可使用相應熵應用測量之平均分佈為可被使 用之最佳育源應用(即使最佳應用為不可實行的)。對於右 分支範圍亦遵循相同的解釋,以滿足限制條件,熵應用測 量在每個臨界鏈路係平均的產生。再者,所有通過臨界鏈 路4 VPs I客量以一種均等產生這些vps之熵應用值之方式 被配置。 如先前所述’在摘曲線之右邊範圍之最佳化目的為增加 具有最南區塊V P之配置容量(即v p有最小的熵區塊測 量)。此相應於一最佳化問題之最大最小公式。必須注意的 是’左邊範圍.之最佳化目的可解譯為一其右邊範圍之最佳 化問題,其在最早的左邊範圍公式化最佳目的,以項目Μ 區塊n (blocking)取代項目”應用,,(utilizati〇n),且以項目„ 具有最大區塊之VP" (VP With the largest blocking)取代項 目”取大的貝源用卢(largest consumer of resources) p 兩種不同之用於璃比率函數之左邊及右邊分支之最佳化 -36- 本紙張尺度適州中因國家標隼(CNS ) A4規格(210X29"?公疫) (請先閱讀背面之注意事項再填寫本頁)----- 'V order A7 B7 V. Description of the invention (33) £ (;) < As already said, when the corresponding one. Entropy ratio function increases, this application must be decreasing. One of the interpretations for optimization purposes in the overload range may be as follows. In order to increase the averaging of the measured entropy in this range, the largest resource user (that is, the VP has the lowest entropy) is identified, and the representation of the other term (extreme representative) is decreasing. The maximum user decrement application is a ratio function that increases the overload range accordingly. Therefore, this approximation corresponds to a maximum-minimum formula for minimization (formulati). It must be noted that the restriction term is applied at a value that initially falls within the overload range. Here, we can use the average distribution measured by the corresponding entropy application to be the best breeder application that can be used (even if the best application is not feasible). The same interpretation is followed for the right branch range to meet the constraints. The entropy application measures the average generation of each critical link system. Furthermore, all the VPs I passengers passing the critical link 4 are configured in such a way that the entropy application values of these vps are evenly generated. As mentioned previously, the optimization of the range on the right side of the abstraction curve is to increase the allocation capacity of the southernmost block V P (that is, v p has the smallest entropy block measurement). This corresponds to a maximum-minimum formula for an optimization problem. It must be noted that the optimization purpose of 'the left range. Can be interpreted as an optimization problem of the right range, which formulates the best purpose in the earliest left range, replacing the project with the project M block n' Application, (utilizati〇n), and replace the project with the project "VP with the largest blocking" (VP With the largest blocking) to take the largest consumer of resources p (two different uses for Optimization of the left and right branches of the glass ratio function -36- The paper size is suitable for the national standard (CNS) A4 specification (210X29 "? Public Plague) in Shizhou (Please read the precautions on the back before filling this page)

^^""'中决打^^,,;!^消於"竹"印;^ A7 ---------------_B7_ 五、發明説明(" ~ ' 目的孓公式導致相同的最佳化程序。在這兩種情沉下,熵 比率函數必須為遞增的。此可由增加控制參數s之絕對值 而元成。假如負載不超過可用的資源,則位移參數將為正 值’且配置給不同虚擬路徑之容量可連續地上升,直到所 有可用的實際資源被配置為止。反之假如負載超過可用杳 源、’、則位移參數為負的。在此情況下配置容量必須逐漸地 遞減直到到達實際資源限制為止。 使用埃比率函數之V p大小量測演繹法 我們可使用上述之熵比率函數特徵,以有效地解決V p大 小量測問題。如早先所解釋,vp大小量測問題目的是在給 疋—供應流量分佈之多個預定義VPS之間配置限制的實際 網路資源。一個使用熵比率函數為一區塊測量之vp大小量 測演繹法之實施例顯示於圖1 2中。 此程序由—系列初始步騾1202至1206開始。所有大小量 測處理之VPs在12〇2組合為一 VP大小量測集合。在網路中 每個實體鏈路之傳輸容量限制於12〇3中說明。於12〇4中, 一熵比率函數之上限集合/…尤每個選擇性地以一虛擬路徑 說明。 二 必須注意的是,步驟12〇2至1204可用取決於完成考量之 任意順序來完成。再者,只有當具有一截去右尾標之供應 流量分佈之真實希望存在時,即只有當尸(z>c)為〇,其中尤 .係太於某個有限值C >仏γ時,才於丨204中說明。假如供 應流量分佈具有一截去的右尾標,則理論上量取網路資源 以完成0區塊是可能的。然而,如此之情況是很少發生 _ -37- "^浪尺度適/丨]中囚國家標準(CNS)A4規格(公趁) ~~----- (請先閲讀背面之注意事項再填寫本頁) n^i— : «m · 訂 A7 B7 35 五、發明説明( 的 於1205中,其他初始步驟包括每個虛擬路徑之大於等於 區塊值之選擇。如其他所述,在一熵比率函數及相應νρ< 區塊之該等值之間具有一反向關係。结果,大的區塊值值 係相應小的熵比率函數值。使用較早發展之關係,不同 VPs之%始容量配置亦於1205中計算。 於12〇6中,這些初始配置透過每個實體鏈路而累積並比 較該實體鏈路之預確切訂定之傳輸容量。假如初始配置因 此而超载了一或多個實體鏈路,則遵循之計算步騾修改成 使用基於位移參數s之負值之公式。此結果遵循超載情況 之演繹法相應大小量測問題為位移參數s之負值反映的事 實。 於初始程序中隱含的另一步騾為選擇一供應的流量模 式’且假如此模式不是基於測量’則推衍相應關係之熵比 率函數ί,配置容量C,及位移參數j。此步驟未於圖12中 顯示。' 初始步驟1202至1206之後’大小量測技術執行遞迴步驟 1207至1215。圖12之遞迴技術顯示一種二階層遞迴,其中 V Ρ遞迴演繹法首先反覆地配置容量給ν ρ大小量測遞迴集 合之VPs,如1207至1210所顯示,直到一或多個實體鏈路 達到全滿(即1 〇〇% )應用為止。 實體鏈路容量完全配置時稱為臨界鏈路。因此,步驟 12 0 7至12 10之網效應係遞迴地辨識臨界鏈路。當臨界鏈路 辨識程序適當地在每個反覆的程序級,辨識一實體鏈路以 -38- +、纸張尺度適川中國國家榡率(CNS ) Λ4規格(210X 297公釐) (請先閱讀背面之注意事項再填寫本頁) 訂 好泸部中央i?4'-XJ”;!T.消於合竹^卬^':: A 7 B7 ----—____________________________ 五、發明説明(%) 作為一臨界鏈'路,如實行的演繹法在一給定的時間,恰妤 可辨識並處理多於一個臨界鏈路。 在本發明之一實施例中,臨界鏈路之辨識由遞增完成, 如1207所示,目前的熵比率函數由使用一熵比率函數之函 數表示式之固定量而估計,而該熵比率函數係取決於供應 的流量模式。如此表示式之例子可在用於同質p〇iss〇nian 流量之方程式(式子1 5 ),用於多級P〇issonian流量之方程 式(式子2 1 ),及用於常態分佈流量之方程式(式子2 3與式 子2 5 )中發現。必須注意的是,一熵比率函數估計之增量 通系為貝值。可能發生的是大小量測演澤法越過最佳化值 並超過配置容量。 位移參數之值^在1208中對於每個大小量測集合之VPS計 算。必須注意的是,位移參數對於相應之V p,表示圖1 〇 之熵容量圖之斜率。大小量測集合中配置給Vps之增加容 量在1209中,利用熵比率函數之增量值計算。步驟12〇7至 1209可在不同於圖12所示之基於完成考量之序列中執行。 配置給不同VPs之容量在1210中對於每個實體鏈路而累 積’並在12 1 1中比較該實體鏈路之總容量。假如一鏈之未 配置實體容量掉到一目前極限值之下,則該鏈被判定為臨 界鏈路。 如果於實體鏈路之辨識比較結果為一臨界鏈路,則計算 進行至12 12。假如沒有任何實體鏈路被發現為一臨界鏈 路,則步驟1207至1210或同等者反覆地進行直到一臨界鏈 路被發現為止。在包括具有一去除右尾標之供應流量模式 -39- 本紙張尺度適州中S國家標準(CNS ) Λ4规格(2^0^797公楚) 一 '~~~~~ (#先閲讀背面之注意事項再填寫本瓦) -Λ'木 訂 A7 B7 部 中 il il 消 11· 合 li 卬 f: 五、發明说明(37 之極少情況下’反覆程序有時候不能再辨識任何臨界鏈路 是可能的。在此情況下:當熵比率函數到達其最大值 ,如1204所述時,該計算自動地終止。 辨識一臨界鏈路之後,VP大小量測演繹法輸出結果,並 重新陳述該問題,如1212至1213所示。每一次在1211中辨 識一或多個實體鏈路是否為臨界鏈路,在1212中,VP大 小量測演繹法.繼續進行,處理目前在一臨界鏈路通過之每 個VPs所配置的容量以產生輸出。在12丨3中,在一臨界鏈 路通過之VPs從大小量測集合中去除。假如沒有任何vps剩 下來以被大小f測,則在1216中終止大小量測演繹化。 假如一或多個VPs剩餘以被大小量測,則大小量測集合 在1215中重新疋義’以只包含如此之vps。由於在一臨界 鍵路通過之VPs從大小量測集合中去除,且由於這些去除 的VPs用掉了部分的實體鏈路容量,大小量測工作減為仍 然於大小量測集合中剩餘的VPs之未配置實體鏈路容量之 分佈。此在12 15中由配置給去除vps之量,經由減少不同 時際鏈之可用容量而芫成,該VP s係相應最後在1211中發 現之臨界鏈路。在本發明之另一實施例中,相同的效應由 包持VPs熵比率函數之值而完成,該VPs係從V P大小量測 集合而估計得到。因為配置給去除VPs之容量係由1212產 生’此計算可被輕易地執行。在12 15重新陳述該問題之 後,在演繹法返回步騾1207之前,熵比率函數對於仍然在 大小量測集合剩餘之所有VPs,以一固定量增加。 雖然本發明之方法及裝置之較佳實施例已伴隨圖式加以 -40 本紙张尺度適川中囡國家標準(CNS〉A4规格(210X297公釐) (請先閲讀背面之注意事項再填寫本頁} 1T- 五、發明説明( 38 A7 B7 詳細說明,必須強調的是,本發明並不侷限於所揭示之實 施例,且其可在不悖離本發明精神之範圍下做不同的變 換,修改,及取代。 4 (請先閱讀背面之注意事項再填寫本頁 -1. v^-- 1 . 訂 -41 - 本纸張尺度適用中囤國家標準(CNS ) Α4規格(210Χ 297公釐)^^ " " 'Zhong Jue ^^ ,,;! ^ Suppressed by " Bamboo "India; ^ A7 ---------------_ B7_ 5. Explanation of the invention (" ~ 'Purpose 孓 formula leads to the same optimization procedure. In both cases, the entropy ratio function must be increasing. This can be achieved by increasing the absolute value of the control parameter s. If the load does not exceed the available resources , The displacement parameter will be positive 'and the capacity allocated to different virtual paths can be continuously increased until all available actual resources are configured. Otherwise, if the load exceeds the available source,', the displacement parameter is negative. In this case, the allocation capacity must be gradually reduced until the actual resource limit is reached. Using the V p size measurement deduction method of the Angstrom ratio function, we can use the above entropy ratio function characteristics to effectively solve the V p size measurement problem. As explained earlier, the purpose of the vp size measurement problem is to allocate limited actual network resources between multiple predefined VPSs in the supply-distribution flow distribution. A deduction of the vp size measurement using an entropy ratio function for a block measurement Example of the method In Figure 12, this procedure starts from the initial steps of a series of 1202 to 1206. The VPs of all size measurement processing are combined into a VP size measurement set at 1202. The transmission capacity of each physical link in the network It is limited to the description in 1203. In 1204, the upper set of an entropy ratio function / ... each is optionally described with a virtual path. Two It must be noted that steps 1220 to 1204 are available depending on Complete in any order of consideration. Furthermore, only when there is a real hope for a supply flow distribution with a truncated right tail, that is, only when the corpse (z > c) is 0, where especially. Only when there is a limit C > 仏 γ, it will be explained in 204. If the supply flow distribution has a truncated right tail, it is theoretically possible to measure network resources to complete block 0. However, this is the case The situation is rare _ -37- " ^ 浪 度 尺度 / 丨] National Prisoner's National Standard (CNS) A4 specification (public time) ~~ ----- (Please read the precautions on the back before filling this page ) n ^ i—: «m · order A7 B7 35 V. Description of the invention (in 1205, other initial steps package The selection of each virtual path is greater than or equal to the block value. As mentioned elsewhere, there is an inverse relationship between an entropy ratio function and the values of the corresponding νρ < blocks. As a result, large block value values are Correspondingly small entropy ratio function value. Using the earlier development relationship, the% starting capacity allocation of different VPs is also calculated in 1205. In 1206, these initial allocations are accumulated and compared through each physical link. The pre-defined transmission capacity of the channel. If the initial configuration therefore overloads one or more physical links, the calculation steps to be followed are modified to use a formula based on the negative value of the displacement parameter s. This result follows the fact that the corresponding size measurement problem of the deductive method of the overload situation is reflected by the negative value of the displacement parameter s. Another step implied in the initial procedure is to select a supplied flow mode 'and if this mode is not based on measurement', the entropy ratio function ί of the corresponding relationship is derived, the allocation capacity C, and the displacement parameter j. This step is not shown in Figure 12. 'After initial steps 1202 to 1206' the size measurement technique performs recursive steps 1207 to 1215. The recursive technique in Figure 12 shows a two-level recursion, where the V P recursive deduction method first repeatedly allocates capacity to ν ρ and measures the VPs of the recursive set, as shown in 1207 to 1210, until one or more entities The link reaches full application (that is, 100%). When the physical link capacity is fully configured, it is called a critical link. Therefore, the net effect of steps 12 07 to 12 10 recursively identifies the critical link. When the critical link identification procedure is appropriately at each iterative procedure level, identify a physical link with -38- +, paper size suitable for China National Standard (CNS) Λ4 specification (210X 297 mm) (please first Read the notes on the back and fill in this page) Order the central i? 4'-XJ ";! T. Eliminate the ^^^^ ':: A 7 B7 ----——____________________________ 5. Description of the invention ( %) As a critical chain, if a deductive method is implemented, more than one critical link can be identified and processed at a given time. In one embodiment of the present invention, the identification of critical links is incremented by Done, as shown in 1207, the current entropy ratio function is estimated from a fixed amount using a function expression of the entropy ratio function, and the entropy ratio function depends on the flow pattern of the supply. An example of such an expression can be used in Equations for homogeneous p0iss〇nian flow (Equation 15), equations for multi-stage Poissonian flow (Equation 2 1), and equations for normal distributed flow (Equation 2 3 and Equation 2) 5). It must be noted that the increment of an entropy ratio function estimate The value is a shell value. What may happen is that the size measurement algorithm exceeds the optimized value and exceeds the configured capacity. The value of the displacement parameter ^ In 1208, the VPS calculation for each size measurement set. It must be noted that the displacement For the corresponding V p, the parameter represents the slope of the entropy capacity graph in Figure 10. The increase capacity allocated to Vps in the size measurement set is calculated in 1209 using the incremental value of the entropy ratio function. Steps 1207 to 1209 may be Executed in a sequence based on completion considerations different from that shown in Figure 12. Capacity allocated to different VPs is accumulated for each physical link in 1210 'and the total capacity of the physical link is compared in 12 1 1. If The unconfigured physical capacity of a chain falls below a current limit value, the chain is determined as a critical link. If the identification and comparison result of the physical link is a critical link, the calculation proceeds to 12 12. If there is no If any physical link is found to be a critical link, steps 1207 to 1210 or equivalent are repeated iteratively until a critical link is found. Include a supply flow pattern with a right tail removed Zhang scale Shizhou National S National Standard (CNS) Λ4 specification (2 ^ 0 ^ 797 Gongchu) a '~~~~~ (#Read the precautions on the back before filling in this tile)-Λ' wood order A7 B7中 il il elimination 11. · li 卬 卬 f: 5. Explanation of the invention (in the rare case of 37, it is sometimes possible that the iterative procedure can no longer identify any critical link. In this case: when the entropy ratio function reaches its maximum value As described in 1204, the calculation is automatically terminated. After identifying a critical link, the VP size measurement deduction method outputs the result and restates the problem, as shown in 1212 to 1213. Every time in 1211, it is identified whether one or more physical links are critical links. In 1212, the VP size measurement deduction method is continued. The current capacity of each VPs passed by a critical link is determined by Produces output. In 12 and 3, the VPs passing through a critical link are removed from the size measurement set. If no vps is left to be measured by the size f, the size measurement deduction is terminated in 1216. If one or more VPs remain to be measured by size, the size measurement set is redefined in 1215 to include only such vps. Because the VPs passing through a critical link are removed from the size measurement set, and because these removed VPs use part of the physical link capacity, the size measurement work is reduced to the remaining VPs in the size measurement set. Distribution of physical link capacity is not configured. This is formed in 12 15 by removing the amount of vps, by reducing the available capacity of the intertemporal chain. The VP s is the corresponding critical link finally found in 1211. In another embodiment of the present invention, the same effect is achieved by enclosing the value of the entropy ratio function of VPs, which is estimated from the V P size measurement set. Because the capacity allocated to remove VPs is generated by 1212 ', this calculation can be easily performed. After restatement of the problem at 12 15 and before deductive method returns to step 1207, the entropy ratio function increases by a fixed amount for all VPs remaining in the size measurement set. Although the preferred embodiment of the method and device of the present invention has been added to the drawing along with the figure -40 The paper size is suitable for the national standard of Sichuan (CNS> A4 specification (210X297 mm) (Please read the precautions on the back before filling this page} 1T- 5. Description of the invention (38 A7 B7 detailed description, it must be emphasized that the invention is not limited to the disclosed embodiments, and it can be changed and modified differently without departing from the spirit of the invention, 4 (Please read the notes on the back before filling in this page-1. V ^-1. Order -41-This paper size applies to the national standard (CNS) Α4 specification (210 × 297 mm)

Claims (1)

第851〇85〇4號專利申請案 申清專利範圍修正本(88年4月) 六、申請專利範圍 1.二以電腦執行之量測虛擬路徑大小的方法,該虛擬路 徑定義於—承載一般流量之電信網路上,該網路具有受 限傳輪容量之多個互連鏈路,該㈣大小方 步驟: J 挑選一適用的熵比率函數,以模型化該電信網路之 每個虛擬路;上的負冑’其中該摘比率函數是藉由理想 化一電信網路上所提供流量之特性來決定; 、定一化解演繹法,使用熵比率函數以作為一區塊 測I,運算該區塊測量以解出用於該一般流量 衡問題;及 於—計算系統中執行計算,此系統係利用併行該熵 比率函數之負載平衡演繹法,以在該虛擬路徑上產生— 儘可旎平均之負載分佈,其中該用於模型化所提供流量 勺區塊測量乃是熵比率函數,A(c) ’該熵比率函數可以 機率二4繹法的近似計算,而任意分佈之隨機變數( 大於等於預選定值(q,並且該摘比率函數另外也是_ 凸函數,以獲得在其分佈平均上之最小零值。 2·:種以電腦執行之量測虛擬路徑大小的方法,該虛擬路 fe定義於_承载一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小方法包括 步騾: 挑選一適用的熵比率函數,以模型化該電信網路之 每個虛擬路徑上的負載; 選疋使用熵比率函數之化解演繹法,其中該使用 --------L k-----^-I訂------}線 (請先閱讀背面之注$項再填寫本頁) 經濟部中央標準局負工消費合作社印製Amended Patent Scope of Patent Application No. 85108504 (April 88) 6. Application for Patent Scope 1.2 The method of measuring the size of a virtual path executed by a computer, the virtual path is defined in-bearing general On a telecommunication network with a large amount of traffic, the network has multiple interconnected links with limited wheel capacity. The steps are as follows: J Pick an applicable entropy ratio function to model each virtual path of the telecommunication network. ; On the negative 上 'where the extraction ratio function is determined by idealizing the characteristics of the traffic provided on a telecommunications network; and, a deductive deduction method, using the entropy ratio function as a block to measure I, and calculating the area Block measurement to solve the general flow balancing problem; and perform calculations in a computing system that uses a load-balancing deduction method in parallel with the entropy ratio function to generate on the virtual path—as much as possible Load distribution, where the block measurement used to model the provided flow spoon is an entropy ratio function, A (c) 'The entropy ratio function can be approximated by the probability method, and the random distribution follows Machine variable (greater than or equal to the pre-selected value (q), and the extraction ratio function is also a _ convex function to obtain the smallest zero value on its distribution average. 2: A method of measuring the size of a virtual path performed by a computer, The virtual path fe is defined on a telecommunication network carrying general traffic. The network has a plurality of interconnected links with limited transmission capacity. The measurement method includes the following steps: Select an applicable entropy ratio function to model Reduce the load on each virtual path of the telecommunication network; choose a deductive deduction method using an entropy ratio function, which should use -------- L k ----- ^-I 定 --- ---} line (please read the note $ on the back before filling out this page) Printed by the Central Standards Bureau, Ministry of Economic Affairs, Consumer Cooperatives 第851〇85〇4號專利申請案 申清專利範圍修正本(88年4月) 六、申請專利範圍 1.二以電腦執行之量測虛擬路徑大小的方法,該虛擬路 徑定義於—承載一般流量之電信網路上,該網路具有受 限傳輪容量之多個互連鏈路,該㈣大小方 步驟: J 挑選一適用的熵比率函數,以模型化該電信網路之 每個虛擬路;上的負冑’其中該摘比率函數是藉由理想 化一電信網路上所提供流量之特性來決定; 、定一化解演繹法,使用熵比率函數以作為一區塊 測I,運算該區塊測量以解出用於該一般流量 衡問題;及 於—計算系統中執行計算,此系統係利用併行該熵 比率函數之負載平衡演繹法,以在該虛擬路徑上產生— 儘可旎平均之負載分佈,其中該用於模型化所提供流量 勺區塊測量乃是熵比率函數,A(c) ’該熵比率函數可以 機率二4繹法的近似計算,而任意分佈之隨機變數( 大於等於預選定值(q,並且該摘比率函數另外也是_ 凸函數,以獲得在其分佈平均上之最小零值。 2·:種以電腦執行之量測虛擬路徑大小的方法,該虛擬路 fe定義於_承载一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小方法包括 步騾: 挑選一適用的熵比率函數,以模型化該電信網路之 每個虛擬路徑上的負載; 選疋使用熵比率函數之化解演繹法,其中該使用 --------L k-----^-I訂------}線 (請先閱讀背面之注$項再填寫本頁) 經濟部中央標準局負工消費合作社印製Amended Patent Scope of Patent Application No. 85108504 (April 88) 6. Application for Patent Scope 1.2 The method of measuring the size of a virtual path executed by a computer, the virtual path is defined in-bearing general On a telecommunication network with a large amount of traffic, the network has multiple interconnected links with limited wheel capacity. The steps are as follows: J Pick an applicable entropy ratio function to model each virtual path of the telecommunication network. ; On the negative 上 'where the extraction ratio function is determined by idealizing the characteristics of the traffic provided on a telecommunications network; and, a deductive deduction method, using the entropy ratio function as a block to measure I, and calculating the area Block measurement to solve the general flow balancing problem; and perform calculations in a computing system that uses a load-balancing deduction method in parallel with the entropy ratio function to generate on the virtual path—as much as possible Load distribution, where the block measurement used to model the provided flow spoon is an entropy ratio function, A (c) 'The entropy ratio function can be approximated by the probability method, and the random distribution follows Machine variable (greater than or equal to the pre-selected value (q), and the extraction ratio function is also a _ convex function to obtain the smallest zero value on its distribution average. 2: A method of measuring the size of a virtual path performed by a computer, The virtual path fe is defined on a telecommunication network carrying general traffic. The network has a plurality of interconnected links with limited transmission capacity. The measurement method includes the following steps: Select an applicable entropy ratio function to model Reduce the load on each virtual path of the telecommunication network; choose a deductive deduction method using an entropy ratio function, which should use -------- L k ----- ^-I 定 --- ---} line (please read the note $ on the back before filling out this page) Printed by the Central Standards Bureau, Ministry of Economic Affairs, Consumer Cooperatives A8 B8 C8 D8 申請專利範圍 摘比率函數作為— 量之負載平衡問題;:及畢法疋來解出該-般流 使用該併行熵比率函數之負載平衡演繹法於一計算 算’以在該虚擬路徑上產生儘可能平均的該負 载刀佈匕其中該下推式演繹法進—步包括下列步驟: 將要量測大小的虚擬路徑組合成—大小量測集合; 使㈣選定之㈣率函數,計算每—該所虛擬路握 横越4網路鏈路每一虚擬路徑上之區塊; 確赠每一網路鏈路上具有最大區塊之虛擬路徑;以 及 次.分配㈣容量給該經確認之虛擬路徑Μ違反網路 只源的限制,直到無最大區塊的存在。 —種以電職行之量_擬路徑大小的方法 徑定義於-承載-般流量之電信網路上,該網路二= :傳輸容量之多個互連鏈路,該量測大小方法包括下列 步驟: J 挑選一適用的摘比率函數,以模型化該電信網路之 每個虚擬路徑上的負载; 選足一使用熵比率函數之化解演繹法,其中該使用 摘比率函數作為—區塊測量,以運算來解出該-般流量 〈負載平衡問題之演繹法為_下推式演繹法,其是作為 -區塊測量’以運算來解出該一般流量之負載問 題;以及 於冲算系統中執行計算,此系統係利用併行該摘 ^紙張从適用巾關家縣(CNS )〜職 广k-----^-I訂.-----}線 (請先閎讀背面之注意事項再填寫本頁〕 經濟部中央標準局員工消費合作社印製 A8 B8 C8 D8 值 六、申請專利範圍 數之負載平衡演繹法,以在該虛擬路後上產生一 儘可忐平均之負載分佈;以及 其中該下推式演繹法進—步包括下列反覆步驟,這 二步驟會重覆直到該大小量㈣集合變成—空集合· 將所有虚擬路徑組合成一大小量測集合;σ. 確定每一實體鏈路之傳輸容量; 為每一虚擬路徑選定一相對較大之起始區塊; 選定-錯誤界限來評估—確認演繹法之關键鍵路的 收斂性; 使用一熵區塊量測來決定每一虚擬路徑上之區塊 只要實體鏈路仍具有可用之容量,就反覆確認具最 大區塊值之虛擬路徑; 只要是無實體鏈路達到完全的使用且該阻礙最大之 虛擬路徑之區塊縮減大於預選定之錯誤界限,就藉由在 該虚擬路徑間重新分配傳輸容量,減小具最大區塊值之 虛擬路徑的區塊。 確認欠缺可分配容量之實體鏈路為關鍵鏈路; 自大小量測集合中移除所有橫跨關鍵鏈路之虚擬路 徑;以及 反覆重新調整其餘實體鏈路之傳輸容量,以反應出 最近自大小量測集合中移除,分配給虛擬路徑的容量。 一種量測定義於一電信網路上之虚擬路徑大小的方法, 該電信網路具有與多個節點互連之多個實體鏈路,該方 -3 本紙張尺度適用中國國家標準(CNS ) Α4規格(210X297公釐) --------ak-----^-I訂.-----線 (請先閱讀背面之注意事項再填寫本頁) 經濟部中央標準局貝工消費合作社印製 經濟部中央標準局員工消費合作社印製 A8 B8 C8 D8 六、申請專利範圍 法包括下列步驟: 映射該多個實體鏈路至一或多個虚擬路徑,每一該虛 擬路徑在一電信網路上提供一可於一對節點之間個別切 換的連接; 確定每·-實體鍵路之傳輸容量; 组合所選之多個該虛擬路徑成為一個大小量測集合; 使用摘比率函數作為一區塊測量,分配傳輸容量的初 始值給該大小量測集合中之每一虛擬路徑,每一該初始 值均相等且經挑選以使區塊是大的; 藉由下列子步驟,將那些容量已全完分配於橫越它們 的虚擬路徑間之實體鏈路,重覆地確認為關鍵鏈結: 對熵比率函數估計值增加一固定值; 重新計算該大小量測集合中每一虚擬路徑之偏移參 數; 使用所增加的熵比率函數估計值和該經重新計算之 偏移參數,重新計算要分配給每一虚擬路徑的容量; 加總每一實體鏈路上分配給所有該虛擬路徑的容 量,以獲得每一實體鏈路上之所分配的總容量;以及 將該全部分配給每一實體鏈路之容量與其經確定的 容量相比較,以決定該實體鏈路之未分配容量是否為 零; 將橫越一實體鏈路(認定為一關键鏈路)上之每一虚擬 路徑上目前·所分配的容量輸出; 將橫越每一關鍵實體鏈路上之所有虚擬路徑自該大小 -4- 本紙張尺度適用中國國家標準(CNS ) A4規格(210X297公釐) --------l;k------—訂·-----}線 (請先閱讀背面之注意事項再填寫本頁) 、申請專利範圍 量測集合中移除;以及 重新疋義該貫體鏈路之確定實體容量,以 已移除虛擬路徑的容量。 刀配、.,。 5·二種以電腦執行之量測虚擬路&amp;大小的系統,該虛擬路 徑疋義於一承載一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小系統包括: I置,用於挑選一適合之熵比率函數來模型化在 忒電網路每一虚擬路徑上的負載; 、一裝置,用於使用一熵比率函數來選定—化解演繹 法,而其中該熵比率函數是決定於將一通信網路上所提 供心流量的特微理想化為一個為該一般流量解出一負載 平均問題的區塊測量;以及 一裝置,用於在一計算系統上,使用該併行熵比率 函數的負載平均演繹法,以在該虛擬路上產生一儘可能 平均的負載分佈來執行計算;以及 其中用來模型化所提供流量之熵比率函數,為熵比率函 數’ Ix(C),該熵比率函數可以機率負演繹法的近似計 算,而任意分佈之隨機變數(大於等於預選定值 (,並且孩熵比率函數另外也是一凸函數,以獲得在 其分佈平均上之最小值零。 6. /一種以電腦執行之量測虛擬路徑大小的系統,該虛擬路 徑定義於一承載一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小系統包括: 一裝置’用於挑選一適合之熵比率函數來模型化在 本紙張从適用申國國家標準(CNS ) Α4· (!i〇x297公董·] 經濟部中央標準局買工消費合作社印裝 A8 B8 C8 ______— ___D8 六、申請專利範圍 孩電信網路每一虛擬路徑上的負載; 裝置’用於使用-熵比率函數作為一區塊測量, 來選足—化解演繹法以解出該一般流量之負載平衡問 題;以及 裝i用於在一计算系統上,使用該併行摘比率 函數的負載平均演繹法,以在該虛擬路上產生—儘可能 平均:負載分佈來執行計算,並且其中該演繹法為一下 推弋秀澤法,其疋使用熵比率函數作為一區塊測量以解 出該一般流量之負載平衡問題;以及 裝置,用於將待大小測量之虛擬路徑組合一 小量測集合; 一裝置,用於使用該選定的熵比率函數來計算該虛 擬路徑所橫越之每一網路鏈路上每一虛擬路徑的區塊; 一裝置,用於確認在每一網路鏈路上具有最大區塊 的虚擬路徑。 一裝置,用於分配額外容量給經確認没有達反網路 資源限制的虚擬路徑,直到其不再具有最大的區塊為 止。 7. —種以電腦執行之量測虚擬路徑大小的系統,該虚擬路 徑定義於一承載一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小系統包括: 一裝置用於挑選一適合之搞比率函數來模型化在 該電信網路每一虛擬路徑上的負載; 一裝置’用於使用一熵比率函數作為一區塊測量, -6- 本紙張尺度適用中國國家標準(CNS ) A4規格(210X297公釐) ---------「.於-----„-I訂-----線 (請先閲讀背面之注意事項再填寫本頁) 經濟部中央標準局身工消費合作社印製 申請專利範圍 來選定-化解演繹法以解題; 股’此K負載平銜問裝置’用於在—斗豐玄β 函數的負载平均演°算系、无上’使用該併行熵比率 平均的負裁分佈來執行計算,並,H盡可能 推式演繹法,其异“ κ卫且其中m去為-下 出咳-般、、” Γ 率函數作為—區_量以解 出'^奴成I乏負載平衡問題;以及 % 裝置用於將待大小測量之虛擬路徑組合成 小量測集合; 仅、且σ成一大 裝置,用於確定每—實體鏈路的傳輸容量;區塊一裝置’用於挑選每一虛擬路徑之相對較大之初始 、一裝S,用㈣定一錯誤界限來估計一確認演 之關鍵鍵路的收叙性; ν 裝置,用於使用一熵區塊量測來決走每一虛擬路 徑上之區塊值; 一裝置,用於反覆確認具最大區塊值之虛擬路徑, 只要實體鏈路仍具有可用之容量; 一裝置,用於只要實體鏈路達到完全的使用且該阻 礙最大之虛擬路徑之區塊縮減大於預選定之錯誤界限, 就藉由在該虛擬路徑間重新分配傳輸容量,減小具最大 區塊值之虚擬路徑的區塊。 一裝置,用於確認欠缺可分配容量之實體鏈路為關 鍵鏈路; (請先閲讀背面之注意事項再填寫本頁} • I - - n .....I I .H · -訂 線. -7- 本紙張U適用中國國家標準(CNS ) Α4規格(210Χ297公釐) 六 、申請專利範圍 8. A8 68 C8 D8 經濟部中央楳隼局負工消費合作社印製 :裝置’用於自大小量測集合中移除所有橫跨關鍵 鍵路又虛擬路徑;以及 裝置,用於反覆重新調整其餘實體鏈路之傳輸容 量’一以反應出分配給最近自大小量測集合中移除 路担的容量。 ,種量測-受限電信網路大小的系統,該電信網路具有 與多個節點互連之多個實體鏈路,該系統包括: /襞置,用於映射該多個實體鏈路至—或多個虚擬路 徑,每一該虚擬路徑在一電信網路上提供一可於一對節 ,點之間個別切換的連接; /襞置,用於確定每一實體鏈路之傳輸容量; 一裝置,用於組合所選之多個該虛擬路徑成為一個大 小量測集合; 置,用於使用熵比率函數作為一區塊測量,分配 傳輸容量的初始值給該大小量測集合中之每一虚擬路 徑,每一該初始值均相等且經挑選以使區塊是大的; 一装置,用於為低於目前值之區塊選定一接續值; 一装置,用於計算橫越一單一實體鏈路之每一該虛擬 路猱上的區塊; /裝置,用於在橫越一單一實體鏈路間之分配可用的 傳輸容量,以回應不同虛擬路徑間區塊的變化,直到有 一個實體鏈路被確認為不再具有未分配的容量; 一裝置,用於自該大小量測集合移除所有橫越該經確 認實體鏈路的虚擬路徑; -8 本紙張从適财關) A4^m ( 210X297^ ) --------'哀-----^-I訂-----丨線 (請先閲讀背面之注意事項再填寫本頁) A8 Βδ C8 D8 9 經濟在中央標準局貝工消費合作社印製 六、申請專利範圍 一裝置,用於降低分配給每一實體鏈路之傳輪容量一 個先如分配給该移除虛擬路徑的量;以及 裝置,用於反覆重覆該計算、分配、移除和降低的 步驟直到在大小量測集合中無任何虛擬路徑留下。_ -種量測-受限電信網路大小的系統,該電信網路且有 與多個節點互連之多個實體鏈路,該系統包括:八 一裝置,用於映射該多個實體鏈路至一或多個虛擬路 徑,每一該虚擬路徑在—電信網路上提供—可於一對節 點之間個別切換的連接; P 一裝置,用於確定每一實體鏈路之傳輸容量; 一裝置,用於組合所選之多個該虛擬路徑成為一個大 小量測集合; 一裝置,用於使用熵比率函數作為一區塊測量,分配 傳輸容量的初始值給該大小量測集合中之每—虛擬路 徑,每一該初始值均相等且經挑選以使區塊是大的; 一裝置,用於將那些容量已全完分配於横越它們的虛 擬路徑間之實體鏈路重覆確認為關键鏈結,該確認裝置 包括: 一裝置,用於對熵比率函數估計值增加一固定 一裝置’用於重新計算該大小量測集合中每— 虚擬路徑之偏移參數; 裝置,用於使用所增加的摘比率函數估計值 和该經重新計算之偏移參數,重新計算要分配給每—虛 9- 本紙張尺度適用中國國家標準(CNS ) Α4規格(210X297公釐) ---------^&gt;'衣 ------:-I訂------線 (請先閱讀背面之注意事項再填寫本頁) 經濟部中央插準局員工消費合作社印製 A8 B8 C8 _________ D8 .夂、申請專利範圍 擬路徑的容量; 一裝置’用於加總每—實體鏈路上分配給所有 該虛擬路徑的容量,以獲得每一實體鏈路上之所分配的 總容量;以及 一裝置,用於將該全部分配給每一實體鏈路之 容量與其經確定的容量相比較,以決定該實體鏈路之未 分配容量是否為零; 一裝置,用於將橫越一實體鏈路(認定為一關键鏈路) 上之每一虛擬路徑上目前所分配的容量輸出; 一裝置,用於將橫越每一關鍵實體鏈路上之所有虛擬 路控自該大小量測集合中移除;以及 一裝置,用於重新定義該實體鏈路之確定實體容量, 以償還分配給已移除虛擬路徑的容量。 ίο. —種以電嘴執行之量測虛擬路徑大小的方法,該虚擬路 徑定義於一承載一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小方法包 步驟: 挑選一適用的熵比率函數,以模型化該電信網 路之每個虛擬路徑上的負載; 選定一使用熵比率函數之化解演繹法,其中該 使用墙比率函數作為_區塊測量之演繹法為推: Π;;:;:來解出該一般流量之負載平衡問題= 中及下推式次繹法進一步包括下列反覆步驟; 运些步驟會重覆直到該女丨 里復且W邊久小I測集合變成一空 —______________ -10- ( CNS ) )----: ----------------^-I訂------線 (請先閲讀背面之注意事項再填寫本頁) ABCD 經濟部中央標準局員工消費合作社印製 六、申請專利範圍 集合: 將所有虛擬路徑組合成一大小量測集合; 確定每一實體鍵路之傳輸容量; 為每一虛擬路徑選定一相對較大之起始區塊; 選定一錯誤界限來評估一確認演繹法之關键鏈 路的收斂性; 使用一熵區塊量測來決定每一虛擬路徑上之區 塊值; 只要實體鏈路仍具有可用之容量,就反覆確認 具最大區塊值之虚擬路徑; 只要是無實體鏈路達到完全的使用且該阻礙最 大之虛擬路徑之區塊縮減大於預選定之錯誤界限,就藉 由在該虚擬路徑間重新分配傳輸容量,減小具最大區塊 值之虚擬路徑的區塊。 確認欠缺可分配容量之實體鏈路為關键鏈路; 自大小量測集合中移除所有橫跨關键鏈路之虛 擬路徑;以及 反覆重新調整其餘實體鏈路之傳輸容量,以反 應出分配給最近自大小量測集合中移除之虚擬路徑的容 量; 作為一負載之區塊測量以解出該一般流量的平衡問 題;以及 於一計算系統上,使用併行該熵比率函數之該負載 平均演繹法執行計算,以在該虚擬路徑上產生一儘可能 _-11 -____ 本紙張尺度適用中國國家標準(CNS ) A4規格(210X297公釐) lv訂 線 (請先閱讀背面之注意事項再填寫本頁) 申請專利範圍 經濟部中央標隼局貝工消費合作社印製 平均之負載分佈;以及 其中減小於下推式演繹法 .塊的步驟,是 。有最大區塊值之虛擬區 11·-種以電腦執, 近似技術來執行。 徑定義於測ΐ擬路徑大小的系統,該虛擬路 限傳輸容量之夕Λ電信網路上’該網路具有受 輸多個互連鏈路,該量測大小系統包括: 化在該•俨J ’用於選定—適合之熵比率函數來模型 Μ私L .竭路每一虛擬路徑上的負載; 量,來:ί〜7於使用—摘比率函數作為-區塊測 繹法以解出該一般流量之負載平衡 …一裝置,用於在一計算系統上,使用該併行熵 比率函數的負載平均演繹法,以在該虛擬路上產生一儘 :能平均的負載分你來執行計算,並且其中該演绎法為 一下推式演、繹法,其是使關比率函數作為一區塊測量 以解出該一般流量之負载平衡問題; .一裝置,用於將待大小測量之虛擬路徑組合成一大小 用於確定每一實體鏈路之傳輸容量; 用於挑選每一虛擬路徑之相對較大之初始區A8, B8, C8, D8, apply for a range of patent functions to extract the ratio function as a load balancing problem; and Bi Fayu to solve the-general flow using the parallel entropy ratio function load balancing deduction method in a calculation ' The load knives are generated as evenly as possible on the path. The push-down deduction method further includes the following steps: Combining the virtual paths to be measured into a size measurement set; using the selected rate function to calculate Each—the virtual path grips the blocks on each virtual path of the 4 network links; the virtual path with the largest block on each network link is guaranteed; and the assigned virtual capacity is allocated to the confirmed virtual path The path M violates the limitation of the source of the network until no largest block exists. —A method based on the amount of electricity and the size of the pseudo-path is defined on a telecommunication network that carries the same amount of traffic. The network 2 =: multiple interconnecting links for transmission capacity. The method of measuring size includes the following: Steps: J pick an applicable extraction ratio function to model the load on each virtual path of the telecommunication network; choose a deductive deduction method using an entropy ratio function, where the extraction ratio function is used as a block measurement The deductive method of solving the-general flow <load balancing problem by calculation is the _ push-down deductive method, which is used as a -block measurement 'to solve the load problem of the general flow by operation; and in the imputation system To perform calculations in this system, this system uses the paper in parallel to extract the paper from Guanjia County (CNS) ~ Jiguang k ----- ^-I order .-----} line (please read the first Please fill in this page again.] A8 B8 C8 D8 value printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs 6. Load balancing deduction method for the number of patent applications to generate a load distribution on the virtual road. ; And where the push-down The method of step-by-step includes the following iterative steps. These two steps are repeated until the set of size and volume becomes an empty set. All virtual paths are combined into a size measurement set. Σ. Determine the transmission capacity of each physical link; Select a relatively large starting block for each virtual path; select-error bounds to evaluate-confirm the convergence of the key path of the deductive method; use an entropy block measurement to determine the block on each virtual path As long as the physical link still has available capacity, it repeatedly confirms the virtual path with the largest block value; as long as the non-physical link reaches full use and the block reduction of the largest blocked virtual path is greater than the preselected error limit, By reallocating the transmission capacity between the virtual paths, the blocks of the virtual path with the largest block value are reduced. Confirm that the physical link that lacks allocable capacity is a key link; remove all horizontal links from the size measurement set Virtual paths across critical links; and re-adjust the transmission capacity of the remaining physical links repeatedly to reflect the most recent self-size measurement set The capacity allocated to the virtual path. A method of measuring the size of a virtual path defined on a telecommunications network that has multiple physical links interconnected with multiple nodes. The scale is applicable to the Chinese National Standard (CNS) Α4 specification (210X297 mm) -------- ak ----- ^-I order .----- line (please read the precautions on the back before filling (This page) Printed by the Central Standards Bureau of the Ministry of Economics, printed by the Shellfish Consumer Cooperative, printed by the Central Standards Bureau of the Ministry of Economic Affairs, printed by the Consumers' Cooperative of the A8, B8, C8, and D8. 6. The scope of the patent application method includes the following steps: Map the multiple physical links to one or more Virtual paths, each of which provides a connection that can be individually switched between a pair of nodes on a telecommunications network; determines the transmission capacity of each physical key path; combines a plurality of selected virtual paths into a size Measurement set; use the cut ratio function as a block measurement, and assign an initial value of the transmission capacity to each virtual path in the size measurement set, each of which is equal and selected so that the block is large Borrow In the following sub-steps, those physical links between the virtual paths that traverse them have been fully allocated, and are repeatedly identified as key links: Add a fixed value to the estimated value of the entropy ratio function; recalculate the size measurement Offset parameters for each virtual path in the set; use the increased entropy ratio function estimate and the recalculated offset parameters to recalculate the capacity to be allocated to each virtual path; add up the allocation on each physical link Give the capacity of all the virtual paths to obtain the total capacity allocated on each physical link; and compare the capacity allocated to each physical link with its determined capacity to determine the physical link's capacity Whether the unallocated capacity is zero; Will traverse the currently allocated capacity on each virtual path on a physical link (identified as a critical link); Will traverse all virtual paths on each key physical link from The size -4- This paper size is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) -------- l; k ------- order · -----} line ( Please read first Note again fill the page surface), the measurement range set patent removed; and determining the entity to re-Cloth sense the through capacity of the link member to the capacity of the virtual path has been removed. Knife with ... 5. Two computer-based systems for measuring the size of a virtual path &amp; the virtual path is defined on a telecommunications network carrying general traffic, and the network has multiple interconnected links with limited transmission capacity. The measurement size system includes: I, for selecting a suitable entropy ratio function to model the load on each virtual path of the electric network; and, a device for using an entropy ratio function to select-deductive deduction method The entropy ratio function is determined by idealizing the characteristics of the cardiac flow provided on a communication network into a block measurement that solves a load-average problem for the general flow; and a device for calculating On the system, the load entropy deduction method using the parallel entropy ratio function is performed to generate a load distribution as even as possible on the virtual road to perform the calculation; and the entropy ratio function used to model the provided traffic is the entropy ratio function 'Ix (C), the entropy ratio function can be approximated by the negative deductive method of probability, and random variables of arbitrary distribution (greater than or equal to the preselected value), and the entropy ratio The function is also a convex function to obtain a minimum value of zero on its distribution average. 6. / A system for measuring the size of a virtual path performed by a computer, the virtual path is defined on a telecommunications network carrying general traffic. The network has a plurality of interconnected links with limited transmission capacity. The measurement size system includes: a device for selecting a suitable entropy ratio function to model the paper from the applicable National Standard (CNS) Α4 · (! I〇x297 public director ·] Printed by the Central Standards Bureau of the Ministry of Economic Affairs, Consumer Cooperatives, A8, B8, C8 ______— ___D8 VI. Scope of Patent Application Load on each virtual path of the telecommunications network; device 'for use- The entropy ratio function is taken as a block measurement to select the foot-resolve deduction method to solve the load balance problem of the general flow; and the load average deduction method installed on a computing system using the parallel abstraction ratio function, The calculation is performed by generating on the virtual road-as even as possible: load distribution, and the deductive method is the following push-show method, which uses the entropy ratio function As a block measurement to solve the load balancing problem of the general traffic; and a device for combining the virtual paths to be measured into a small measurement set; a device for calculating the entropy ratio function using the selected A block of each virtual path on each network link traversed by the virtual path; a device for identifying the virtual path with the largest block on each network link; a device for allocating additional capacity to It has been confirmed that the virtual path that does not reach the limit of anti-network resources has not reached the largest block. 7. —A system for measuring the size of a virtual path executed by a computer. The virtual path is defined as a path carrying general traffic. On a telecommunication network, the network has a plurality of interconnected links with limited transmission capacity. The measurement system includes: a device for selecting a suitable ratio function to model each virtual path in the telecommunication network A device 'is used to use an entropy ratio function as a block measurement, -6- This paper size is applicable to China National Standard (CNS) A4 specifications (2 10X297 mm) --------- 「. 于 -----„-I order ----- line (Please read the precautions on the back before filling out this page) Central Standards Bureau of the Ministry of Economic Affairs The Industrial and Consumer Cooperatives printed the scope of the patent application to select-resolve the deductive method to solve the problem; "This K load level questioning device" is used to calculate the load average calculation system, supreme 'of the Kung Fung Xuan β function using the parallel The entropy ratio averages the negative cut distribution to perform the calculation, and H is as far as possible a deductive method. The difference is "κ Wei and where m goes to-under cough-like," Γ The rate function is used as a zone-quantity solution. The problem of load balancing is solved; and the device is used to combine the virtual paths to be measured into a small measurement set; and σ is a large device for determining the transmission capacity of each physical link; A block-device is used to select a relatively large initial and one S for each virtual path, and a set of error bounds is used to estimate the narrative properties of a key path for confirmation. Ν A device is used to use a Entropy block measurement to determine the block value on each virtual path; a device for repeatedly confirming the maximum value A block-valued virtual path, as long as the physical link still has available capacity; a device for reducing the block size of the virtual path that is most obstructed as long as the physical link is fully used and the pre-selected error limit is reduced by Reallocation of transmission capacity between the virtual paths reduces the blocks of the virtual path with the largest block value. A device used to confirm that the physical link lacking allocable capacity is a critical link; (Please read the precautions on the back before filling out this page} • I--n ..... II .H · -Order. -7- This paper U applies the Chinese National Standard (CNS) A4 specification (210 × 297 mm) 6. Scope of patent application 8. A8 68 C8 D8 Printed by the Central Government Bureau of the Ministry of Economic Affairs and Consumer Cooperatives: Device 'for self-size Remove all virtual paths that cross key links from the measurement set; and a device for repeatedly readjusting the transmission capacity of the remaining physical links'-to reflect the allocation of the most recently removed road load from the size measurement set Capacity. A system for measuring the size of a restricted telecommunications network, the telecommunications network having multiple physical links interconnected with multiple nodes, the system includes: / settings for mapping the multiple physical chains Road to—or multiple virtual paths, each of which provides a connection on a telecommunications network that can be individually switched between a pair of nodes, points; / set, used to determine the transmission capacity of each physical link ; A device for combining selected A plurality of the virtual paths become a size measurement set; and is configured to use an entropy ratio function as a block measurement, and allocate an initial value of the transmission capacity to each virtual path in the size measurement set, each of the initial values. Are equal and selected so that the block is large; a device for selecting a continuation value for a block below the current value; a device for calculating each of the virtual paths across a single physical link Block on ;; / device for allocating available transmission capacity across a single physical link in response to changes in blocks between different virtual paths until one physical link is confirmed to no longer have Unallocated capacity; a device to remove all virtual paths across the confirmed physical link from the size measurement set; -8 papers from Shicaiguan) A4 ^ m (210X297 ^) --- ----- 'wailing ----- ^-I order ----- 丨 line (please read the notes on the back before filling this page) A8 Βδ C8 D8 9 Printed 6. Applicable scope of a device for reducing the allocation to each The capacity of the physical link is assigned to the removed virtual path as before; and a device for repeatedly repeating the steps of calculation, allocation, removal and reduction until there are no virtual paths in the size measurement set Stay. _-A system for measuring the size of a restricted telecommunications network, the telecommunications network having multiple physical links interconnected with multiple nodes, the system includes: a Bayi device for mapping the multiple physical chains To one or more virtual paths, each of which is provided on a telecommunications network and can be individually switched between a pair of nodes; P a device for determining the transmission capacity of each physical link; a A device for combining the selected plurality of virtual paths into a size measurement set; a device for using an entropy ratio function as a block measurement, and allocating an initial value of the transmission capacity to each of the size measurement set -Virtual paths, each of which has an initial value that is equal and selected so that the block is large; a device for repeatedly identifying the physical links between those capacity that have been fully allocated between the virtual paths that traverse them as key chains In conclusion, the confirmation device includes: a device for adding a fixed one to the estimated value of the entropy ratio function, for recalculating the offset parameter of each virtual path in the size measurement set; the device It is used to recalculate the value to be assigned to each-imaginary 9-using the increased extraction ratio function estimate and the recalculated offset parameter. This paper size applies the Chinese National Standard (CNS) Α4 specification (210X297 mm)- ------- ^ &gt; 'clothing ------: -I order ------ line (please read the precautions on the back before filling out this page) Staff Spending Bureau Cooperative prints A8 B8 C8 _________ D8. 夂, the capacity of the proposed path in the scope of patent application; a device 'used to aggregate the capacity allocated to all the virtual paths on each-physical link to obtain the allocated on each physical link The total capacity of the physical link; and a device for comparing the total allocated capacity of each physical link with its determined capacity to determine whether the unallocated capacity of the physical link is zero; a device for Crossing a physical link (identified as a key link) to the currently allocated capacity output on each virtual path; a device for controlling all virtual road links across each key physical link from the amount Remove from test set; A means for redefining the physical capacity of the physical link to determine the order assigned to repay removed virtual path capacity. ίο. —A method of measuring the size of a virtual path performed by an electric nozzle, the virtual path is defined on a telecommunication network carrying general traffic, the network has a plurality of interconnected links with limited transmission capacity, the measurement Large and small method steps: Select an applicable entropy ratio function to model the load on each virtual path of the telecommunication network; select a deductive deduction method using the entropy ratio function, where the wall ratio function is used as the _ block The deductive method of measurement is push: Π ;;:;: to solve the load balancing problem of the general flow = medium and push-down type deductive method further includes the following iterative steps; these steps will be repeated until the woman repeatedly And the W side long small I measurement set becomes empty — ______________ -10- (CNS)): ---------------- ^-I subscription ------ (Please read the notes on the back before filling this page) Printed by the ABCD Employee Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 6. Collection of Patent Application Ranges: Combine all virtual paths into a size measurement set; determine the size of each physical key Transmission capacity; select for each virtual path A relatively large starting block; selecting an error bound to evaluate the convergence of a key link of a deductive deduction method; using an entropy block measurement to determine the block value on each virtual path; as long as the physical link If there is still available capacity, the virtual path with the largest block value is repeatedly confirmed; as long as the non-physical link reaches full use and the block size of the virtual path that hinders the greatest reduction is greater than the pre-selected error limit, Redistribute transmission capacity between virtual paths to reduce the blocks of the virtual path with the largest block value. Confirm that the physical link lacking allocable capacity is a critical link; remove all virtual paths across the critical link from the size measurement set; and repeatedly readjust the transmission capacity of the remaining physical links to reflect the allocation to the most recent self-size Measure the capacity of the virtual path removed from the set; measure as a load block to solve the general flow balance problem; and perform calculations on a computing system using the load average deduction method in parallel with the entropy ratio function In order to generate as much as possible on this virtual path _-11 -____ This paper size applies Chinese National Standard (CNS) A4 specification (210X297 mm) lv order line (Please read the precautions on the back before filling this page) Apply The scope of patents is the average load distribution printed by the Central Bureau of Standards, Ministry of Economic Affairs, Shellfish Consumer Cooperatives; and the steps in which the block is reduced to the push-down deduction method, yes. The virtual zone with the largest block value 11 ·-is implemented by computer and approximate technology. The path is defined by a system that measures the size of the simulated path. On the evening of the transmission capacity of the virtual limit, the network has multiple interconnected links. The measurement size system includes: 'Used to select-the appropriate entropy ratio function to model the M private L. The load on each virtual path of the exhaust path; the amount, come: ί ~ 7 in the use-abstract ratio function as-block deduction method to solve the Load balancing for general traffic ... A device for a computing system that uses the load-average deduction method of the parallel entropy ratio function to produce everything on the virtual road: you can perform the calculation on average load points, and where The deductive method is a push-type deductive method, which uses the ratio function as a block measurement to solve the load balancing problem of the general traffic; a device for combining the virtual paths to be measured into a size Used to determine the transmission capacity of each physical link; used to select a relatively large initial area for each virtual path 關鍵鏈路的收斂性 —裝置,用於使用 用於選定一錯誤界限來估計一確認演繹法之 摘區塊量測來決定每一虛擬路徑 12 本紙張尺度適用中國國家標準(CNS &gt; A4規格(21〇χ297公釐) 請 先 閎 之 注 意 事 項 再 訂 線 經濟部中央標準局貝工消費合作社印製 A8 B8 C8 D8 申請專利範圍 上之區塊值; 一裝置’用於反覆確認具最大區塊值之虚擬路徑,只 要實體鏈路仍具有可用之容量; 一裝置,用於只要實體鏈路達到完全的使用且該陴礙 最大之虚擬路徑之區塊縮減大於預選定之錯誤界限,就 藉由在該虛擬路徑間重新分配傳輸容量,減小具最大區 塊值之虚擬路徑的區塊。 一裝置,用於確認欠缺可分配容量之實體鏈路為關鍵 鏈路; 一裝置’用於自大小量測集合中移除所有橫跨關鍵鏈 路之虛擬路徑;以及 一裝置,用於反覆調整其餘實體鏈路之傳輸容量,以 反應出最近自大小量測集合移出,分配給虚擬路徑的容 量;以及 其中減小於下推式演繹法中具有最大區塊值之虚擬區 塊的步騾’是以使用一連績近似技術來執行。 12. —種以電腦執行之量測虚擬路徑大小的方法,該虛擬路 徑定義於一承載一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小方法包括下列 步騾: 挑選一適用的熵比率函數,以模型化該電信網 路之每個虚擬路徑上的負載,其中該熵比率函數是藉由 理想化一電信網路上所提供流量之特性來決定; 選定一化解演繹法,使用熵比率函數以作為— 良紙張从適用中國國家標準(CNS) A4祕(21〇χ297公釐了 ---------A-----—訂------— 線 c请先閲讀背面之泣意事項存填寫本買)Convergence of critical links—A device for determining each virtual path using block measurements for selecting an error boundary to estimate a confirmation deduction method. This paper size applies Chinese national standards (CNS &gt; A4 specifications). (21〇χ297mm) Please note before you set the line printed by the Central Standards Bureau of the Ministry of Economic Affairs, Shellfish Consumer Cooperative, printed A8 B8 C8 D8 block value on the scope of patent application; a device 'used to repeatedly confirm the largest block Valued virtual path, as long as the physical link still has available capacity; a device for as long as the physical link reaches full use and the block of the virtual path that hinders the most is reduced by more than a preselected error limit, Reallocate the transmission capacity between the virtual paths to reduce the block of the virtual path with the largest block value. A device for confirming that the physical link lacking the allocable capacity is a key link; a device 'for self-size Test set to remove all virtual paths across critical links; and a device to repeatedly adjust the transmission of the remaining physical links To reflect the capacity that was recently removed from the size measurement set and allocated to the virtual path; and the step in which the virtual block with the largest block value in the push-down deduction method is reduced. 12. — A method of measuring the size of a virtual path performed by a computer. The virtual path is defined on a telecommunications network that carries general traffic. The network has multiple interconnected links with limited transmission capacity. The measurement size method includes the following steps: Select an applicable entropy ratio function to model the load on each virtual path of the telecommunication network, where the entropy ratio function is provided by idealizing a telecommunication network The characteristics of the flow rate are determined; a deductive deduction method is selected, and the entropy ratio function is used as a good paper—from the application of the Chinese National Standard (CNS) A4 secret (21〇χ297 mm) --------- A-- ----- Order -------- Line c, please read the Weeping Items on the back first and fill in this purchase) 申清專利範圍 經濟部中央榡準局員工消費合作社印製 :鬼測里’運算m區塊測量以解出用於該一般流量之負 载平衡問題;以及 r ^於—計算系統中執行計算,此系統係利用併行 I'同比率函數之負載平衡演繹法,以在該虚擬路徑上產 生一儘可能平均之負載分佈。 13’:申請專利範圍第12項^^之方法, 其中孩熵比率函數係決定於流量 ΐ4.:申請專利範圍第12項^^^之方法, -中及摘比率函數係決定於理想U信網路上供應流 量的特徵。 :/、1 f-'M i. 15.如申請專利範圍第14項之方法, 其中該璃比率函數係由—供應流章丄同質p〇iss〇nian分 佈加以理想化。 . 16·如申凊專利範園第14項之#測虛之方法, 其中該摘比率函數由一供應流量之多lP“SS0nian分伟 加以理想化。 Π.如申請專利範園第i 4項冬鮝之方法, 其中該熵比率函數由一供應流量之;常態身佈加以理想 化。 18. 如申請專利範圍第丨4項之之方法, 其中該熵比率函數由一供應流量之二分佈加以理想 化。 19. 「種以電腦執行之量測虚擬路徑大小的方法,該虛擬路 徑定義於一承載一般流量之電信網路上,該網路具有受 -14- 本紙張从家標準(21Gx297公釐) ---------,V衣-----^—訂-----I (請先閲讀背面之注意事項再填寫本頁) 經濟部中央標準局員Η消費合作社印製 A8 B8 C8 D8 申請專利範圍 限傳輸容量之多個互連鏈路,該量測大小方法包括下列 步騾: 挑選一適用的熵比率函數,以模型化該電信網 路之每個虚擬路徑上的負載,其中該熵比率函數是以在 —電信網路上所提供流量的理想化特徵來決定,其中该 =於模型化所提供流量的理想化熵值區塊測量乃是熵比 率函數,4(C),該熵比率函數可以機率負演繹法的近似 計算,而任意分佈之隨機變數大於等於預選定值 (CV ,並且?亥熵比率函數另外也是一凸函數,以獲得在 其分侔平均上之最小零值; 、選足一化解演繹法,使用熵比率函數以作為一區塊 測量,運算該區塊測量以解出用於該一般流量之負載平 衡問題;以及 ' 於计算系統中執行計算,此系統係利用併行該熵 比率函數之負載平衡演繹法,以在該虚擬路徑上產生一 儘可能平均之負載分佈。 .如申叫專利範園第〗2項測虚之方法, 其中孩演繹法為一下推式演繹法,其〆使用墙比率函數作 為一區塊測量,運算該區塊測量以踩:裘用於該一般流量 之負載平衡問题。 21‘ :種以電腦執行之量測虚擬路徑大小的方法,該虛擬路 徑定義於一承載一般流量之電信網路上,該網路具有成 限傳輸容量之多個互連鏈路,該量測大小方法包括下二 -15 210X297公釐 YDeclared the scope of patents printed by the Consumer Cooperatives of the Central Bureau of Standards and Quarantine Bureau of the Ministry of Economic Affairs: Ghost Surveying 'calculates m-block measurements to solve the load-balancing problem for this general traffic; The system uses a parallel load balancing deduction method with the same ratio function to produce a load distribution on the virtual path that is as even as possible. 13 ': Method for applying item 12 ^^ in the scope of patent application, where the entropy ratio function is determined by the flow rate ΐ 4 .: Method for applying item 12 ^^^ in the scope of patent application,-the neutral and abstract ratio function is determined by the ideal U letter Characteristics of the supply traffic on the Internet. : /, 1 f-'M i. 15. The method according to item 14 of the scope of patent application, wherein the glass ratio function is idealized by the supply flow chapter p homogeneous p0issoonian distribution. 16. The method of measuring falsehoods in item 14 of Rushen Patent Park, where the abstraction ratio function is idealized by a large amount of supply flow LPSSianian. Π. As for item 4 of the patent park Winter's method, where the entropy ratio function is idealized by a supply flow; the normal body is idealized. 18. For the method of item 4 of the patent application scope, wherein the entropy ratio function is added by a distribution of two supply flows Idealized. 19. "A method of measuring the size of a virtual path performed by a computer. The virtual path is defined on a telecommunications network carrying general traffic. This network has a standard of -14- ) ---------, V-shirt ----- ^-Order ----- I (Please read the notes on the back before filling this page) Printed by the Central Standards Bureau of the Ministry of Economic Affairs and Consumer Cooperatives A8 B8 C8 D8 The scope of the patent application for multiple interconnected links with limited transmission capacity includes the following steps: Select an applicable entropy ratio function to model each virtual path on the telecommunication network. Load, where the entropy ratio function is at — The idealized characteristics of the traffic provided on the Internet are determined, where the block measurement of the idealized entropy of the traffic provided by the model is an entropy ratio function, 4 (C), which can be a negative deductive method of probability The approximate calculation of the random variable is greater than or equal to the pre-selected value (CV, and the? H entropy ratio function is also a convex function to obtain the minimum zero value on its tiller average; , Using the entropy ratio function as a block measurement, computing the block measurement to solve the load balancing problem for the general traffic; and 'performing calculations in a computing system, which uses the load of the entropy ratio function in parallel Balanced deduction method to produce a load distribution as even as possible on the virtual path. For example, the second method of measuring virtuality in the patented patent park, where the child's deduction method is a push-based deduction method, which uses a wall The ratio function is taken as a block measurement, and the block measurement is calculated to step on the load balance problem of the general traffic. 21 ': a virtual path for computer-based measurement Small method, the virtual path defined in the general flow of a carrier of telecommunications network, the network having a plurality of interconnected links of limited transmission capacity, the method comprising measuring the size of the two Y -15 210X297 mm 申請專利範圍 經 濟 A 標 準 局 員 工 消 費 合 作 社 印 製 挑選—適用的搞比率函數, 個虛擬路徑上的負載; 生化该網路之每 選定一使用搞比率函數之化解演绎法 比率函數作為—F 其中邊使用摘 ' s塊測量之演繹法是來解出 U平相題,其中該使㈣: 量之演繹法為—下推 +③數作為—區塊測 騾: 下推式虎繹法,其進-步包含下列步 ^用料㈣比率函數之負载平衡演繹法於一 在該虛擬路徑上產生儘可能平均的該負 載刀佈,田其中該下推式演繹法進一步包括下列步驟: 將要量測大小的虛擬路徑組合成一大小量測集合; 吏用肩選足之熵比率函數,計算每一該虛擬路徑所 橫越之網路鏈路每一虛擬路徑上之區塊; 確認每一網路鏈路上具有最大區塊之虛擬路徑;以 及 刀配額外容量給該經確認之虚擬路徑而不達反網路 資源的限制,直到無最大區塊的存在。 於一·計算系統中執行計算,此系統係利用併行該熵比 率函數心負載平衡演繹法,以在該虚擬路徑上產生一儘 可能平均之負載分佈。 2.種以%知執行之量測虚擬路徑大小的方法’該虛擬路 t疋義於一承載一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小方法包括下列 步驟: 請 先 閱 之 注 項 再 4 頁 訂 線 個虛擬路::::::率函數’以模型化_網路之每 二—使用熵比率函數作為—區塊測量之化 比率山:出該一般流量之負载平衡問題,其中該睪 比丰函數之演繹法作為一區 使用麵 演繹法,其進一步包含下列反覆步下推式 量測集合變成一空集為止:驟’直重覆到該大小 將所有虛擬路徑组合成-大小量測集合; 確定每一實體鏈路之傳輸容量; 2-虛擬路徑選定—相對較大之起始區塊; 收錯誤界限來評估-確認料法之關鍵鍵路的 使用一搞區塊量測爽法今^ 值; 」术决疋母一虛擬路徑上之區槐 二要實體鏈路仍具有可用之容量,就反覆確認具最 大區塊值之虛擬路徑; 八要疋播實鏈路達到完全的使用且該阻礙最大之 虚擬路徑之區I縮減大於預収之錯誤界限,就藉由在 該虚擬路徑間重新分配傳輸容量,減小具最大區塊值之 虛擬路徑的區塊。 確認欠T可分配容量之實體鏈路為關键鏈路; 自大小f測集合中移除所有橫跨關鍵鏈路之虚擬路 徑;以及 反覆重新調整其餘實體鏈路之傳輸容量,以反應出 -17 A8 B8 C8 D8 申請專利範圍 分配給最近自大小量測集合中移除之虛擬路徑的容量。 於-計算系統中執行計算,此系統係利用併行該煩 率函數I負載平衡演繹法,以在該虛擬 儘可能平均之負裁分佈。 上屋生 23·-種量測-受限電信網路大小的方法,該電信網路具有 與多個節點互連之多個實體鏈路,該方法包括下列步 驟: ^ 映射該多個實體鏈路至一或多個虛擬路徑,每一該 虚擬路徑在-電信網路上提供—可對節點之間個別 切換的連接; 確定對每一虛擬路徑所提供的流量; 確定每一電信網路上實體鏈路之傳輸容量限制; 使用一熵區塊測量模型化電信網路上所提供的流 量;以及 分配容量給該多個受到鏈路傳輸容量限制之虛擬路 徑,使得不同虛擬路上之區塊機率'儘可能地平均。 24.—種量測一受限電信網路大小的方法,該電信網路具有 與多個節點互連之多個實體鏈路,該方法包括下 驟: V 映射該多個實體鏈路至一或多個虛擬路徑,每—該虛 擬路徑在一電信網路上提供一可於—對節點之間個^切 換的連接; 組合所選之多個該虛擬路徑成為一個大小量測集合; 分配傳輸容量的初始值給該大小量測集合中之每—虚 -18- 本紙張尺度適用中國國家橾準(CNS ) M現格(210x297公釐) -^------丨訂.-----線 (請先閱讀背面之注意事項再填寫本頁) 經濟部中央榡準局員工消費合作社印袋 經濟部中央標準局員工消費合作社印策 A8 B8 C8 D8 夂'申請專利範圍 擬路徑,使得區塊一開始是大的; 為低於目前值之區塊選定一接續值; 計算橫越一單一實體鏈路之每一該虚擬路徑上的 塊; ™ 在橫越一單一實體鏈路間之分配可用的傳輸容量,以 回應不同虛擬路徑間區塊的變化,直到有一個實體鏈路 被確認為不再具有未分配的容量; 自該大小量測集合移除所有橫越該經確認實體鏈路的 虛擬路徑; 降低分配給每一實體鏈路之傳輸容量一個先前分配給 該移除虚擬路徑的量;以及 反覆重覆該計算、分配、移除和降低的步驟直到在大 小量測集合中無任何虚擬路徑留下。 25. —種以電腦執行之量測虛擬路徑大小的系統,該虛擬路 徑定義於一承載一般流量之電信網路上,該網路具有受 限傳輪容量之多個互連鏈路,該量測大小系統包括: 一裝置,用於挑選一適合之熵比率函數來模型化在 該電信網路每一虛擬路徑上的負載; 裝置’用於使用一摘比率函數來選定—化解演繹 法而其中該摘比率函數是決定於將一通信網路上所提 供(流量的特微理想化為一個為該—般流量解出一負載 平均問題的區塊測量;以及 一裝置,用於在一計算系統上,使用該併行熵比率 函數的負載平均演繹法,以在該虛擬路上產生—儘可能 --------1 ^-----ί — 訂------''線 f碕先閲讀背面之注意事項再填寫本頁}Application for patent scope Economy A Standard Bureau employee consumer cooperative print selection—applicable ratio function, load on a virtual path; for each biochemical network selected, the ratio function is used to calculate the ratio function as —F where The deductive method using the measurement of the 's block is to solve the U-phase problem, where the deduction method: the deduction method of the quantity is-push-down + ③ number as-the block test: the push-down tiger deduction method, its progress -The steps include the following steps: ^ Use the load balancing deduction method of the material-ratio ratio function to produce the load knife cloth as evenly as possible on the virtual path. The push-down deduction method further includes the following steps: The size to be measured The virtual paths are combined into a size measurement set; the shoulder entropy ratio function is used to calculate the blocks on each virtual path of the network link traversed by each virtual path; The virtual path with the largest block; and additional capacity allocated to the confirmed virtual path without reaching the limit of anti-network resources until no maximum block exists. The calculation is performed in a computing system, which uses a parallel load-balanced deduction method of the entropy ratio function to generate a load distribution on the virtual path that is as even as possible. 2. A method for measuring the size of a virtual path performed in%. The virtual path t is defined on a telecommunications network carrying general traffic. The network has multiple interconnected links with limited transmission capacity. The sizing method includes the following steps: Please read the note first and then 4 pages to order a virtual path ::::::: Rate function 'modeling _ network of every two — use the entropy ratio function as — block measurement Conversion ratio mountain: The problem of load balancing of the general flow is introduced, in which the deduction method of the Biefeng function is used as a one-region deduction method, which further includes the following iterative push-down measurement set until it becomes an empty set: Repeat to this size to combine all virtual paths into a size measurement set; determine the transmission capacity of each physical link; 2- virtual path selection-a relatively large starting block; receive error bounds to evaluate-confirm the material The key method of the method is to use the block measurement method to determine the value of the current method; "The technical decision is that the physical link on the virtual path is still available, and it is repeatedly confirmed that it has the largest block value. Virtual path; Eighth, when the real link has reached full use and the area I of the virtual path that hinders the largest is reduced by more than the pre-received error limit, the transmission capacity is reduced by reallocating between the virtual paths. Block value of the block of the virtual path. Confirm that the physical link that has less than T allocable capacity is the key link; remove all virtual paths across the key link from the size test set; and repeatedly readjust the transmission capacity of the remaining physical links to reflect -17 A8 B8 C8 D8 The scope of the patent application is allocated to the capacity of the virtual path recently removed from the size measurement set. The calculation is performed in a -computing system, which uses a load-balancing deduction method in parallel with the annoyance function I to minimize the negative distribution as evenly as possible in the virtual.上 屋 生 23 · -A method for measuring the size of a restricted telecommunications network, the telecommunications network having multiple physical links interconnected with multiple nodes, the method includes the following steps: ^ mapping the multiple physical chains To one or more virtual paths, each of which is provided on the telecommunication network-a connection that can be individually switched between nodes; determine the traffic provided for each virtual path; determine the physical chain on each telecommunication network Transmission capacity limitation of the road; using an entropy block measurement to model the traffic provided on the telecommunication network; and allocating capacity to the multiple virtual paths that are limited by the transmission capacity of the link, so that the probability of blocks on different virtual roads is as much as possible Ground average. 24. A method for measuring the size of a restricted telecommunication network, the telecommunication network having a plurality of physical links interconnected with a plurality of nodes, the method comprising the following steps: V mapping the plurality of physical links to a Or multiple virtual paths, each of which provides a switchable connection between a pair of nodes on a telecommunication network; the selected multiple virtual paths are combined into a size measurement set; the transmission capacity is allocated The initial value is given to each of the size measurement sets—virtual-18- This paper size is applicable to the Chinese National Standard (CNS) M spot (210x297 mm)-^ ------ 丨 order .--- --Line (please read the notes on the back before filling out this page) Printed bags for employees' cooperatives of the Central Economic and Technical Bureau of the Ministry of Economic Affairs A8 B8 C8 D8 The block is initially large; a continuation value is selected for a block that is lower than the current value; a block on each of those virtual paths that traverses a single physical link is calculated; ™ between traversing a single physical link Allocate available transmission capacity to Block changes between different virtual paths until one physical link is confirmed to no longer have unallocated capacity; remove all virtual paths that traverse the confirmed physical link from the size measurement set; reduce allocation The transmission capacity of each physical link is an amount previously allocated to the removed virtual path; and the calculation, allocation, removal, and reduction steps are repeated until no virtual path remains in the size measurement set. 25. A system for measuring the size of a virtual path performed by a computer. The virtual path is defined on a telecommunications network carrying general traffic. The network has multiple interconnected links with limited wheel capacity. The measurement The size system includes: a device for selecting a suitable entropy ratio function to model the load on each virtual path of the telecommunication network; and a device for selecting and deducting the deductive method using a ratio function, wherein the The abstraction ratio function is determined by idealizing the characteristics of traffic provided on a communication network as a block measurement that solves a load-average problem for the general traffic; and a device for a computing system, A load-averaging deduction method using the parallel entropy ratio function to produce on the virtual road—as far as possible -------- 1 ^ ----- ί — order ------ '' line f 碕Read the notes on the back before filling out this page} A8 B8 C8 D8 平均的負載分佈來執行計算。 ,申:專利範圍第25項之纽系統, 中孩辆比率函數是決定於流量的測_ 其 » -U ::.ν 系統,其 4則系統,其 ss⑽朋 申請專利範圍 27. =專利範圍第25項之……其 商比率函數是決定於一電信網雜所提供流量的逕 想化特徵。 掛巧 〕逕 28. 如申請專利範園第27項之^系統,其 :=率函數由-供應η之同質。, 29. 如申請專利範圍第27項之虛系統,其 中孩墒比率函數由—供應流量;之多級分体加 以理想化。 30.如申請專利範圍第27項之處&amp; .......... 中孩熵比率函數由一供應流量之秦备佈加以理想化。 31_如申請專利範圍第27項之^_路_^_知肘系統,其 中茲熵比率函數由一供應流量之二項分佈加以理想化。 32. —種以電腦執行之量測虛擬路徑大小的系統,該虛擬蹲 徑走義於一承载一般流量之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路’該量測大小系統包括: 一I置’用於挑選一適用之熵比率函數來模型化在爲 電信網路每一虛擬路徑上的負載; 一裝置,用於來選定使用一熵比率函數之化解演釋 法’而其中該熵比率函數是決定於將一通信網路上所招 供之流量的特微理想化,其中該用於模型化所提供流j 20- 本紙張认適用中國國家標準(CNS )八顿洛(21〇&gt;&lt;297公董) ---------:衮-----^ —訂------線 (請先閱讀背面之注意事項再填寫本頁〕 經濟部中央標隼局員工消費合作社印製 A8 B8 C8 D8 經濟部中央標準局員工消費合作社印裝 、申請專利範圍 的理想化摘值區塊測量 4里乃疋熇比率函數,4(c),該熵比 =函數可以機率負料法的近似計算,而任意分怖之隨 機k數(幻大於等於預選定值並且該熵比率函數 另外也是一凸函數,以獲得在其分佈平均上之最小零 值;以及 -裝置,用於在-計算系統上,使用該併行墙比率函 數的負載平均演繹法,以在該虛擬路生-儘可能平 均的負載分佈來執行計算。 玖如申請專利範園第25項之系統, 其=該使用熵比率函數作為一區塊測秀、來解出該一般 流量之負載平衡問題的演繹法為下推式演繹法。 34.「種以電腦執行之量測虛擬路徑大小的系統,該虚擬路 走我於承載一般流里之電信網路上,該網路具有受 限傳輸容量之多個互連鏈路,該量測大小系統包括: 一裝置,用於挑選一適合之熵比率函數來模型化在 該電信網路每一虚擬路徑上的負載; 一裝置,用於選定一使用熵比率函數作為一區塊測 置,來解出該一般流量之負載平衡問題的化解演繹法, 其中該使用熵比率函數作為一區塊測量的該演繹法為下 推式演繹法,其進一步包含有: 一裝置’用於將待大小測量之虛擬路徑組合成 一大小量測集合; —裝置,用於計算該虛擬路徑橫越之每一網路 鍵路中個別虚擬路徑上的區塊,其是使用該所選的網比 21 - 本紙張尺度適用中國國家榡準(CNS ) A4規格(210X297公釐) ----------^ X------丨訂.-----線 (請先閱讀背面之注意事項再填寫本頁) 399S8GA8 B8 C8 D8 average load distribution to perform calculations. Application: The new system of the 25th item in the patent scope. The ratio function of the children's vehicle is determined by the measurement of the flow_its »-U ::. Ν system, its 4 systems, and its patent application scope 27. = patent scope Item 25 ... The quotient ratio function is determined by the radial characteristics of the traffic provided by a telecommunications network. Coincidentally] Path 28. For example, the 27th patent system of the patent application park, whose: = rate function is supplied by-homogeneity of η. 29. For example, the virtual system in the scope of patent application No. 27, in which the child ratio function is provided by-supply flow; the multi-level split is idealized. 30. For example, in the 27th place of the scope of application for patent & ......... The entropy ratio function is idealized by Qin Beibu, a supply flow. 31_ As in the 27th patent application of the ^ _ 路 _ ^ _ known elbow system, the z-entropy ratio function is idealized by a binomial distribution of a supply flow. 32. — A system for measuring the size of a virtual path performed by a computer. The virtual path is defined on a telecommunications network carrying general traffic. The network has multiple interconnecting links with limited transmission capacity. The sizing system includes: an I-set for selecting an applicable entropy ratio function to model the load on each virtual path for the telecommunication network; a device for selecting a solution interpretation using an entropy ratio function The entropy ratio function is determined by the idealization of the traffic advertised on a communication network, and the model is used to model the provided flow j 20- This paper is considered to be applicable to the Chinese National Standard (CNS) Octon Luo (21〇 &gt; &lt; 297 public directors) ---------: 衮 ----- ^ —Order ------ line (please read the precautions on the back before filling this page) 〕 Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs A8 B8 C8 D8 The idealized value block measurement of the printed and patented scope of the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economics measures 4 Li Nai ratio function, 4 (c) The entropy ratio = function can be approximated by the probability negative material method, and any Distributed random k-number (magic is greater than or equal to a preselected value and the entropy ratio function is also a convex function to obtain the smallest zero value on its distribution average; and-means for using on a computing system, using the The load-average deduction method of the parallel wall ratio function is used to perform calculations on the virtual road as far as possible to distribute the load as evenly as possible. For example, if the system of the 25th item of the patent application park, the entropy ratio function is used as a block. The deduction method to test the load balance problem of the general traffic is the push-down deduction method. 34. "A system implemented by a computer to measure the size of a virtual path. This virtual path takes me through the general flow. On a telecommunication network, the network has a plurality of interconnected links with limited transmission capacity. The measurement system includes: a device for selecting a suitable entropy ratio function to model each virtual network in the telecommunication network. Load on the path; a device for selecting a deductive deduction method that uses the entropy ratio function as a block measurement to solve the load balance problem of the general flow, wherein the entropy is used The deduction method with the ratio function as a block measurement is a push-down deduction method, which further includes: a device 'for combining the virtual paths to be measured into a size measurement set;-a device for calculating the virtual The blocks on the individual virtual paths in each network key that the path traverses are based on the selected network ratio 21-This paper size is applicable to China National Standard (CNS) A4 specification (210X297 mm)- -------- ^ X ------ 丨 Order .----- line (Please read the precautions on the back before filling this page) 399S8G 率函數 塊之虚擬路;以及 經濟·部中央標準局員工消費合作社印裝 .、 裝置’用於分配額外容量給經確認沒有達反 五路資源限制的虚擬路徑,直到其不再具有最大的區塊 4止;以及 裝置,用於在一計算系統上,使用該併行熵比率 2的負載平均演繹法,以在該虛擬路上產生一儘可能 肀均的負載分伟來執行計算。 =Γ電腦執行之量測虛擬路徑大小的系統,該虛擬路 疋我於一承載一般流量之電信網路上,該網路具有受 限傳輸谷量(多個互連鏈路,該量測大小的'系統包括: 、裝置$於挑選—適用的,摘比率函數,以模型化 孩電信網路之每個虛擬路徑上的負載; 裝置,用於選定一使用熵比率函數作為區塊測量 (化解演繹法,其是運算來解出該—般流量之 問題’其中該使㈣比率函數作為—區塊測量ς演绎法 為一下推式演繹法’其中該下推式演繹法進一步包括下 列反覆步驟,這些步驟會重覆直到該大小 一空集合為止: 禾》又成 一裝置,用於將所有虚擬路徑組合成—太小量 測集合; 裝置,用於確定每一實體鏈路之傳輸容量,· 裝置,用於為每一虛擬路徑挑選一相對較^ 22- 本紙張纽it财賴家縣(CNS )Α4^Μ·(210Χ297^Υ (請先閱讀背面之注意事項再填寫本頁) 訂 線 .- I I I— I --1 I A8 B8 C8 D8 申請專利範圍 之起始區塊; (請先閱讀背面之注意事項再填寫本頁) 一裝置,用於選定一錯誤界限來評估一確認演 繹法之關鍵鏈路的收斂性; 一裝置,用於使用一熵區塊量測來決定每一虛 擬路徑上之區塊值; 一裝置,用於只要實體鏈路仍具有可用之容 量,就反覆確認具最大區塊值之虚擬路徑; 一裝置,用於只要是無實體鏈路達到完全的使 用且該阻礙最大之虚擬路徑之區塊縮減大於預選定之錯 誤界限’就藉由在該虚擬路徑間重新分配傳輸容量,減 小具最大區塊值之虚擬路徑的區塊。 裝置,用於確§忍欠缺可分配容量之實體鍵路 為關鍵鏈路; 一裝置,用於自大小量測集合中移除所有橫跨 關鍵鏈路之虛擬路徑;以及 衣置,用於反覆重新綱整其餘實體鏈路之傳 輸谷里,以反應出最近自大小量測集合中移除,分配給 虛擬路徑的容量;以及 -裝置’用於在-計算系統中執行計算,此系統係 用併行該垌比率函數之負載%衡演繹法,以在該虚 路徑上產生一儘可能平均之負分.佈。 此 =申請專利範園第22項之^^方法,其 推式演繹法中具有最大區_之虛擬區塊的步辨, 疋使用一連續近似技術來執行Γ Λ S993SU 鉍 C8 D8 六、申請專利範圍 丨 37.如申請專利範圍第3 5項之太小量-測的系統,其中減小 於下推式演繹法中具有最大區虚擬區塊的步騾, 是使用一連續近似技術來執行。一‘ ---------^------—1Τ..:-----線 (請先閱讀背面之注意事項再填寫本頁) 經濟部中央標準局員工消費合作社印製 -24- 本紙張尺度適用中國國家標準(CNS ) Α4規格(210X297公釐)Rate function block of virtual roads; and printed by consumer consumption cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs; device 'for allocating additional capacity to virtual paths that have been confirmed to have not reached the anti-five-way resource limit until they no longer have the largest area Block 4; and means for performing calculations on a computing system using the parallel entropy ratio 2 load-averaging deduction method to generate a load distribution that is as uniform as possible on the virtual road. = Γ A computer-implemented system for measuring the size of a virtual path. The virtual path is on a telecommunications network that carries general traffic. The network has a limited transmission valley (multiple interconnected links, 'The system includes:, device $ for selection — applicable, extract ratio function to model the load on each virtual path of the child telecommunications network; device for selecting an entropy ratio function as a block measurement (resolving deduction) Method, which is an operation to solve the problem of general flow, 'where the ratio function is used as a block measurement, the deduction method is the following push-down method', where the push-down method further includes the following iterative steps, these The steps will be repeated until an empty set of this size: "He" becomes a device for combining all virtual paths into a too small measurement set; a device for determining the transmission capacity of each physical link. For each virtual path, choose a relative comparison ^ 22- This paper New Itai Laijia County (CNS) Α4 ^ Μ · (210 × 297 ^ Υ (Please read the precautions on the back before filling this page). -III— I --1 I A8 B8 C8 D8 The starting block of patent application scope; (Please read the notes on the back before filling this page) A device for selecting an error boundary to evaluate a confirmation deduction method Convergence of key links; a device for determining the value of a block on each virtual path using an entropy block measurement; a device for repeatedly confirming that the physical link has available capacity A virtual path with the largest block value; a device for reallocating between virtual paths as long as the physical link is fully used and the block reduction of the largest virtual path is reduced by more than a preselected error threshold Transmission capacity, reducing the block of the virtual path with the largest block value. Device for confirming that the physical link lacking allocable capacity is a key link; a device for removing from the size measurement set All virtual paths across critical links; and clothing that is used to iteratively reorganize the transmission valleys of the remaining physical links to reflect the recent removal from the size measurement set and allocate to The capacity of the virtual path; and-the device 'is used to perform calculations in a -computing system, which uses a load-% balance deduction method in parallel with the unitary ratio function to produce a negative score as average as possible on the virtual path. This is the ^^ method of item 22 of the applied patent park, and its step-by-step identification of the virtual block with the largest area _ in the push method, 疋 uses a continuous approximation technique to perform Γ Λ S993SU bismuth C8 D8 The scope of patent application 丨 37. The system of too small measurement-measurement according to item 35 of the patent application scope, in which the steps reduced in the push-down deduction method with the largest area of virtual blocks is a continuous approximation technique to Execution. I '--------- ^ ------— 1T .. :: ----- line (please read the precautions on the back before filling this page) Staff of Central Bureau of Standards, Ministry of Economic Affairs Printed by Consumer Cooperatives-24- This paper size applies to China National Standard (CNS) Α4 (210X297 mm)
TW85108504A 1995-08-11 1996-07-13 System and method for optimal virtual path capacity dimensioning with broadband traffic TW399380B (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US2779095A 1995-08-11 1995-08-11

Publications (1)

Publication Number Publication Date
TW399380B true TW399380B (en) 2000-07-21

Family

ID=21839801

Family Applications (3)

Application Number Title Priority Date Filing Date
TW85108505A TW317054B (en) 1995-08-11 1996-07-13
TW85108506A TW319944B (en) 1995-08-11 1996-07-13
TW85108504A TW399380B (en) 1995-08-11 1996-07-13 System and method for optimal virtual path capacity dimensioning with broadband traffic

Family Applications Before (2)

Application Number Title Priority Date Filing Date
TW85108505A TW317054B (en) 1995-08-11 1996-07-13
TW85108506A TW319944B (en) 1995-08-11 1996-07-13

Country Status (1)

Country Link
TW (3) TW317054B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2879339A1 (en) * 2013-11-27 2015-06-03 Thomson Licensing Method for distributing available bandwidth of a network amongst ongoing traffic sessions run by devices of the network, corresponding device.

Also Published As

Publication number Publication date
TW319944B (en) 1997-11-11
TW317054B (en) 1997-10-01

Similar Documents

Publication Publication Date Title
US6304639B1 (en) System and methods for controlling virtual paths within a network based on entropy rate function
US5727051A (en) System and method for adaptive routing on a virtual path broadband network
US5764740A (en) System and method for optimal logical network capacity dimensioning with broadband traffic
JP3720147B2 (en) A method for logical network design in multi-service networks.
JPH10506243A (en) Network operation and performance enhancement
Le Vu et al. Scalable performance evaluation of a hybrid optical switch
JP3184822B2 (en) Method, network apparatus and PNNI network for generating optimal PNNI compound node representation for limited cost
Hwang LLR routing in homogeneous VP-based ATM networks
TW399380B (en) System and method for optimal virtual path capacity dimensioning with broadband traffic
Tang et al. Topology design and bandwidth allocation of embedded ATM networks using genetic algorithm
Wong et al. Fairness in window flow controlled computer networks
Medhi Multi-hour, multi-traffic class network design for VP-based wide-area dynamically reconfigurable ATM networks
KR100459891B1 (en) Method for allocating bandwidth to the multi-users in a network and computer readable medium therefor
KR100343774B1 (en) Method of multicast route allocating for delay-sensitive service using VP branching node in ATM networks
Lin et al. A minimax utilization routing algorithm in networks with single-path routing
Larsson et al. Performance evaluation of a local approach for VPC capacity management
JP3786371B6 (en) System and method for optimal dimensioning of logical network capacity due to band traffic
Chang et al. Hierarchical QoS routing in ATM networks based on MDP cost function
Jiyapanichkul et al. Optimal resource management with dynamic virtual paths in ATM networks
Tanterdtid et al. Optimizing ATM network configuration based on virtual path concept using genetic algorithm
Park Modeling of multimedia communications in ATM networks
JP3804960B2 (en) Optical path setting method and optical path setting device
Ross et al. Multiservice Loss Systems
Choi et al. Reallocation of paths of ABR services in a multistage interconnection network (MIN)-type ATM switch network
KURINCKX et al. Rapportsde RCcheréhe

Legal Events

Date Code Title Description
GD4A Issue of patent certificate for granted invention patent
MM4A Annulment or lapse of patent due to non-payment of fees