200400720 玖、發明說明： 【發明所屬之技術領域】 本發明與排定在包括複數個連結之無線主/從組態微網之 網路中之訊息之方法有關，並進一步與適於施行該方法之 微網主裝置有關。 【先前技術】 短程無線網路具一主無線裝置，其隨後暫存或連結從無 線裝置’以形成更受關注之主/從微網。此類主/從裝置（或 網路節點）之交互操作性，係視具預定與標準化無線協定之 各裝置而定，諸如IEEE通過之無線標準之802系列。此協定 之一熟知範例為藍芽（Bluetooth)TM協定。另一與本專利申請 案同時發展之協定刻正由ZigBee Alliance公司群發展中 (www.zigbee.org)。ZigBee Alliance主要目標在訂出適於低 資料率、低功率應用之協定及無線堆疊，使得採用zigBee 標準之播線裝置得以低價並可交互操作。期望此類低價自 行配置無線裝置微網可開發許多家庭消費者及工業控制市 場，例如在加熱及照明應用上。ZigBee聯盟公司群目的在 製造爲入時，目標成本低於美金2元之無線裝置，而此類裝 置具備相當簡單之微控制器可充作微處理器之用，且可獲 得之電路板上記憶體量有限。 但ZigBee無線微網在施行此應用時，包括一主節點及相 連之從節點，推估對在2.4千兆赫ISM頻帶中介定之16頻道 之一中通連之ZigBee微網而言，直接與主節點無線廣播範 圍有關之播線涵蓋限制範圍在數十米。另一缺點在於自一 85589 200400720200400720 (1) Description of the invention: [Technical field to which the invention belongs] The present invention relates to a method for scheduling messages in a network including a plurality of connected wireless master / slave configuration micronets, and further relates to a method suitable for carrying out the method Related to the microgrid master device. [Prior art] The short-range wireless network has a master wireless device, which then temporarily stores or connects the slave wireless devices' to form a more-received master / slave micronet. The interoperability of such master / slave devices (or network nodes) depends on the devices with predetermined and standardized wireless protocols, such as the 802 series of wireless standards adopted by IEEE. One well-known example of this protocol is the Bluetooth ™ protocol. Another agreement developed concurrently with this patent application is being developed by the ZigBee Alliance group of companies (www.zigbee.org). The main goal of the ZigBee Alliance is to formulate protocols and wireless stacks that are suitable for low data rate and low power applications, so that cable feeders using the zigBee standard can be cheap and interoperable. It is expected that such low-cost self-configuring wireless device microgrids can develop many home consumer and industrial control markets, such as in heating and lighting applications. The purpose of the ZigBee Alliance group is to manufacture wireless devices with a target cost of less than US $ 2 when they are manufactured, and these devices have fairly simple microcontrollers that can be used as microprocessors, and available on-board memory Weight is limited. However, in the implementation of ZigBee wireless micronet, including a master node and connected slave nodes, it is estimated that for a ZigBee micronet connected to one of the 16 channels in the 2.4 GHz ISM band, it will directly communicate with the master node. The range of the broadcast line related to the radio broadcasting range is limited to tens of meters. Another disadvantage is that since 85589 200400720
ZigBee微網至另一已定出々 已知現行行動網路排定 般在各節點處需要大量資源能 万法（ Η Φ # /Μ ^ ^ ^儲存例如須***訊息 机中kU/排疋貝汛，此類資訊可 Γ夂扦〇 加的w曰 f 4中間及目的節點位址 (各位址一般均獨具六或八位 制。 、、、數）），並無排定訊息之機 【發明内容】 因此’本發明之—目的在消拜上述缺失。 依本發明之第一態樣，提供一 ^ 種用以排定在具複赵佃、 從組態微網之無線網路中之祝 、 數個王/ _ ^ 〈万〉去，各微網均且〜、 即點，其連結至另一微網之主節 、王 1 ”、、* 並可操作以與乏 及其中各連結主節點儲存有一 連， 、…、、 亍$排疋表，其中該排定表4 用以辨識任何與其微網相連之# i括 】邳埂爻坆即點之局部微網入口 及用以辨識其直接連結之微網 馬， θ <逑結微網入口碼，兮、 包括一微網之一主節點： ~万法 自-連結主節點接收—訊息，該訊息包含—排定 依-接收規則計算-來自該指標之局部參考碼，π ’ 依該局部參考碼及排定表資訊辨識一目標節 擇一 久兩者 、是否該目標節點經辨識為_與傳遞訊息至從節點 邯微網相連之從節點；或 是否藏目標節點經辨識為-可經由_連結微網進入之> 點’依-傳遞規則計算一新排定指標，以該新排定指：1 換在訊息中之排定指標，並將訊息傳遞至連結微網:^ 點。 王！ 85589 200400720 本U已⑧#出_種適於排定跨越數個經由主節點連 結之微網之訊息之古、、土 - , „ t 万去’其中具排足指標之訊息係依簡單 儲存之規則而跨越今遠沾# — ^ — μ I、、、口排疋，且僅藉由主節點與所儲存 之排定表資訊併用。妯令本：欠 排疋表貝詋包括與主節點直接連結之 微網有關之限制资祧，η益山、a上 、 且精由她加該方法之步騾，主節點 在需遞送訊息時，即可轔,ψ谈、兰 、 |_硪出遞迗訊息之連結主節點。 優點在於，在排定表φ >欠、击丨 衣甲 < 各連結微網入口碼均具一與連 結械網之主郎點之位址相途々失去 和逑心參考碼，及一經由主節點進 入之總結點數之計數。士卜步 > 二本缸士 此處 < 计數係表可經由連結之微網 主節點進入之總結點數，向本所鱼 、 已口所運、、、口之彳政網本身連纟士之任 何其它微網之節點。令這些人卩碼併存並確保在料產生 之網路拓樸中不具迴路，致使主節點得以儲存極緊密之排 定表’藉此併同規則（本身視表人口碼之順序而定）即可有效 率地排定訊息。 在依上述主節點之網路中具有處理、收發信機及記憶體， 依簡單規則及緊密儲存之排定表，併同施行所述之排定訊 息之方法之步驟。 依ZigBee AUiance所訂定之預定短程無線協定配置在網 路之微網内之主節點與從節點間之通連較佳。彳以直接連 結主節點之簡單料或電力線達成主節關之通連，或可 藉由無線或紅外線硬體達成連結。若以無線鏈路連結微網， 則優點在於使用已存在於節點中之收發信機及以邙“協 定，並於局部微網内操作通連，及在相異無線頻道上，跨 越微網鏈路自一微網至另一微網通連’俾將無線干擾降至 85589 200400720 最低。 【實施方式】 圖1顯示具連結微網丨〇、20、30之網路範例。各微網均具 典線裝置或模組，配置成微網之主節點，並以主節點作為 /、Η、、’同之中央v周郎為。在微網1 〇中，主節點包括具獨特辨 識碼Ml之主無線裝置。主節點使得其它無線裝置得以暫 存，並藉以與其微網相連成為從裝置，且在一操作模式下， 藉由提出周期性參考碼信號（無線信標）而於微網内協調通 詋，其中孩信號包含具有供其局部從裝置使用之辨識碼之 訊息。 在圖式中，主節·Μ1具三相連從裝置14、16、18，分具 辨識碼S11、S12&S13。故主裝置12協調其相連從裝置 16、18間之通訊。熟悉此技藝者常將此類微網稱之為主^足 為網或星型網路，或有時稱之為具星型或主/從拓樸之網 路。微網10主節點之無線廣播範圍有限。常維持供電予主 節點（有時稱之為“全功能裝置，，），而從裝置（有時稱之為‘‘縮 減功此衣置）則可由電池供電，並以併含睡眠及清醒模式 之各種功率節省無線機制程式化。ZigBee聯盟嘗試將此類 諸如照明及加解0料率及低功率應用之無線模組標準 化，並控制具備在例如2.4兆赫ISM頻帶中之丨6預定頻道之 中操作《微網之房舍周邊’及依直接序列分佈頻譜機制 及載波感測多重存取（CSMA)協定為之4主/從微網中之微 網通訊一般係以網路主無線信標（具網路辨識碼，例如主裝 置之辨識碼)同步化。為利闡釋之故’在從無線裝置Η、Μ 85589 200400720 及1 8之應用中，可併入照明開關或燈具，其中從裝置間彼 此途輯成對”’俾使開關之操作導致具資料封包之無線信 號傳遞土主裝置’查詢成對裝置（或有時稱之為端點）之辨識 碼並傳遞適當訊息至成對燈具’以開啟或關閉燈具。此類 主/從微網之組態詳述如2001年4月19曰發表之本申請人之 審理中之專利申請案冒00128156及^¥〇〇128157，供做讀者 指引，並將其中所揭以引用方式併入本文。 圖1亦顯示兩其它微網20及30，各具自有之個別主與相連 從裝置。微網20包括主節點22(具獨特辨識碼M2)，協調從 節點24(S21)及26(S22)。微網30包括主節點32(M3)，協調從 節點34(S3 1)及36(S32)。在一大型建物中可能常發生下列情 況，已安裝第一微網，而後安裝數個其它微網。例如可於 保全產品（例如窗/門感測器）中使用微網1〇之無線裝置，其 具後期安裝之第二20及第三微網3〇，以增加保全。在圖 所示各微網均相連，使得訊息得以利用本發明之方法自一 微網之協調節點傳送至另一處。 微網1〇已由鏈路4〇連結至M2之微網2〇，且當安裝μ) 之微；時，依序連結至M2之微網。遮謝直接鏈路可經由有 泉或播線廣播或光學裝置為《，甚至可以另依中間網路（例 如局部LAN或網際網路-未示於此圖中）中以虛擬方式為之。 仏在下列較佳具體實施例中，假設主裝置間具備利用與其 :罔中〈各王裝置所採用者相同之無線協定及硬體所為之 鏈路。但熟悉此技藝者將知，“鏈路，，之實際實體具體實施 例了具其B熟知通訊裝置，諸如光學紅外線通訊或電力線 85589 -10- 200400720 信號傳遞或直接電報。會耍# 夏要處在於鏈路於微網間攜載特定 型式訊息，現將簡述之。 班在考量網路拓樸時存在—附加考量，尤其是具有迴路或 %時。一般瞭解應避免迴路或環拓樸(以避免訊息“陷，，於無 盡環型路徑中）’且所述系統亦無例外。確保在網路拓樸中 《連結主裝置本身不構成一迴路之簡單方法，在於納入一 王裝置’接收來自供應—辨識之另_主裝置之請求，並請 求建上直接鏈路，先檢查其它連結裝置以辨識請求裝置， 並且若未找到請求裝置之辨識，則僅建立一鏈路。參閱圖卜 並假叹U、..罔1G、20、3G<主節點均落在其它無線涵蓋區域 範圍内（未示於圖中），如所示建立MW與M2_M3間之鏈路 40考里M3&出印求至具M1之鏈路。在准予連結前n 會傳遞-請求至其直接連結之節點（M2)，以詢問⑷是否已 知。M2復以-肯定確認，因此M1拒絕來自⑷之連結請求。 右M2本身連結至另一主裝置，則應將詢問傳遞至主裝置， 其須將結果經M2回覆至M1等。 現將描述致使如圖丨所示無線微網依本發明之一態樣遞送 訊息之主節點裝置部件、程式碼、排定表及訊息格式之一 較佳具體實施例，其中微網係依ZigBee無線標準操作。 圖2係適於施行並協調微網1〇並跨越連結之微網操作訊息 排足之主無線裝置12之方塊圖。該裝置具電源裝置6〇，可 將以微處理器50(或微控制器）型式存在之處理裝置充電，其 中該處理器耦合至供傳遞與接收無線訊息用之無線收發信 機54。戎處理器亦耦合至儲存操作碼以及與相連之從裝置 85589 -11- 200400720 有關之微以貝a (例如辨識不I、成對端，點資料及從服務能力） 之記憶體裝置56。圖中顯示與處理器分開之記憶體包括 ROM(或EEPROM)及RAM，但亦可依成本及晶 片設計考量， 將之與處理器整合在一起（以虛線表之）。 批在此具體實施例巾，以收發信機54充作主裝置間之連結 衣置較佳❼纟衣置_主裝置傳遞/接收係於用以服務其微網 内〈相連k裝置用〈主裝置採用之不同頻道上發生。注意 :裝置邓T包括光學連結I置58，β連結主節點至另一主 節點，如後具體實施例所將述。 處理裝置依預定ZigBee無線堆疊及協定運作微網，如圖3 所不此堆璺70位於記憶體56之R〇m部中，並由較低實體 (PHY)72及中間進入控制（MAC)74層組成，經由這些層，所 接收4無線訊息即可依序通至網路層（nwk)76。此層與本 身即具應用碼（例如若該碼係併於燈具中，則屬照明應用）之 較高層78相介面。程式指令被處理器5〇執行時，會導致本 發明内（万法之步驟施行，並以部分Nwk層為之較佳。 、圖4係報頭/商務負載資料封包或訊息8〇之結構範例。該 訊息具有各種欄位82，該欄位82所具報頭資訊包括傳送者 獨特辨識碼（一般為代表裝置製造廠及傳送無線資料封包之 «之8位（组獨特數），例如—欄位84，其中在此較佳具 體實施例中。將簡述兩位元組排定指標（RI)之插人或改良。 亦定出訊息資料87及總和檢查欄位89。如熟悉封包無線系 統領域者所周知’部分訊息（服務及協定資料單元SDU、PDU) 係於無線標準中定義之各層±操作或與之㈣卜本發明之 85589 -12- 200400720 NWK層在自MAC層接收一訊息時，將由主裝置⑽作具訊 息傳送者位址及RI資料欄位84之報頭資料，以及儲存於排 定表中之資料。 圖5闡釋由記憶體56中各主節點12、& 42配置及儲存之 排定表之範例，其中NWK層在其上操作以排定訊息8〇。主 裝置建互排定表62，其包括具相關從裝置之參考碼（ref)及 位址（ADDR)之局部微網入口碼63，以及其連結之主節點之 連結微網入口碼65。表62對應於圖i所示網路之主mi儲存 之排定表，而表64則對應於M2所儲存，及表66對應於“3所 儲存。連結微網入口碼65係於局部微網入口碼63下參考， 並包括直接連結之主節點之位址及可經由該鏈路進入之裝 置/節點總數（CNT)。例如：當微網1〇先被安裝表62時，將 僅包括微網内相關裝置之局部參考碼。後期M2連結至Mi， Ml將更新其排定表而具REF 4，該時間之辨識碼位址，M2, 及計數將僅包括M2之微網之裝置（亦即3)。 後期某時間微網M3連結至M2，M2加入對M3之入口碼而 具自先前參考碼（REF)入口碼及相關計數計算之參考碼（亦 即REF 3加上CNT 4給定對M2表64之REF行7中之入口碼）。 M2接著通知新微網（屬於M3)已連結而具3節點之所有先前 連結之微網（亦即Ml)。Ml因而更新M2之計數入口碼為6(此 係表經由M2進入之裝置數）。在一類似方式中，由連結至具 4節點之另一微網（Ml)之M2通知M3，因而M3更新與M2之 計數（CNT)有關之表66入口碼為3+4 = 7。 在這些表中之資料為主裝置所用’併同接收與傳遞規則 85589 -13- 200400720 (演算法)以計算後續用以將訊息自源節點排定至目的節點之 排定指標。#收主t置施行接收規Μ ;辨識此舉之結果為 目標節點；若該目標節點係另一微網之主裝置則***依傳 遞規則計算之新排定指標，並將訊息傳遞至該目標節點。 圖6闡釋在接收規則11〇中所含操作之範例虛擬碼，圖^ 闡釋傳遞規則之對應虛擬碼12〇。在圖6所示碼中，幻係一 排定指標，_SRC)則係在供辨識連結之主裝置已傳遞訊自 至操作接收規則之接收主裝置之排Μ中之參考碼，τ〇τα: 係在網路中之裝置和（由主裝置加總計數表中人口碼並將之 加入其微網内之相關裝置數後計算所得），LR係計算結果之 局部參考碼。圖7闡釋在傳遞規則中所含操作之範例虛擬 碼，其中ref(TN)係在施行傳遞規則之主裝置之排定表中辨 識I目標節點之參考碼，CNT(TN)係與目標節點有關之計 數’ LR則係自接收規則計算並為傳遞演算法用以計算一新 規則指標之參考碼。 為心整之故，現描述一發現過程，其中主裝置針對特定 裝置查詢規則指標，俾供未來傳遞之用。在圖丨中，假設裝 置S 11而要與裝置S2丨之配對（例如s丨丨可為照明控制單元， "使用者希土在M2之控制下將以與已安裝之燈具S2 1配 對）。S1向其主M1提出請求，需要與S21產生一端點。Ml檢 查其排疋表62且發現並無對S21之入口碼。Ml，故提出一 搜寻請求至其連結之主％2。M2接收該請求，檢查其表並尋 找舁其相關且具局邵參考碼丨之裝置S2丨。M2現施加傳遞規 則至此參考碼而產生排定指標RI為5，將之送回主Ml。Ml 85589 -14 - 200400720 接收對S21之RI=5之參考碼，並施加接收規則’在此例中造 成所得參考碼5。M1儲存此結果並將之連 之初始參考碼，以徂去才Λ你4、、 心 、 ，、未耒傳込之足址S21，並儲存將裝置S11 與儲存碼配對之端點資料。 接收主M1現依圖8之流程圖中所示一般方法排定訊息， 其中接收訊息（步驟90)具一排定指標，施加接收規則11〇(步 驟92)，自接收規則計算之結果辨識一目標碼（步驟94)，且 右孩目標碼經辨識為可經由直接連結之微網進入，則施加 傳遞規則120(步驟96)以產生一新排定指標（步驟ι〇〇)，將之 ***U〇《搁位84並傳遞訊息（步驟！ 〇2)否則若在步驟％ 中’目標節誠辨識為接收主微網之相關構件（亦即接收規 則之結果給定落在排定表之局部入口碼63範園内之局部參 考碼LR)，則王裝置之處理器經由路徑97指向步驟⑻，將 訊息（步騾102)傳遞至局部節點。 當然需以排定指標計算多跳程（multi_h〇p)路徑（例如川_ ⑷-M2-S21)中之第—主裝置，接著以各接收主節點路由操 作圖8之方法。 例如在自S11接收無線訊息中，M1將其先前自尋找處理 儲存之參考碼恢復，並自參考碼值及排定表資訊辨識一目 標碼。參考碼5並未儲存於表62中，但5位於M2之參考碼4 與對應於M2加上其可進入之裝置數之參考碼（4 + 6=ι㈠間。 將瞭解對此範例而言雖屬瑣事，在具有許多連結之主節點 及具相關計數人π碼之排^表之網路中，參考碼（先前料 找處理中計算所得)可用以辨識遞送訊息予哪一個連結主： 85589 200400720 置。M2儲存之排定表之檢杳 ^ ^ 員現對應於Ml可抵達之節點， 靶圍在3至6之參考碼；而範圍在 、 国在7土 10〈參考碼則係對應於 王M2可抵達之節點。因此計 上參考碼資料一般即可辨 滅目標郎點供遞送之用。 回到多跳程路徑範例，主M1 、 辨滅M2為目標節點，接著 於參考碼5上施行傳遞規則，給 、。疋徘私指標以為1。將此插 入訊息80中’並傳遞至所辨識之目標節點μ]。M2現施加圖 ㈣示接㈣訊息（步驟9G)即施行接收 規則（步驟92)’以給定LR41。藉由比較此參考碼與其儲存 (排足表M2而辨識（步驟94)對應於此參考碼之目標節點係 -局部相關裝置（步驟95)，因而傳遞訊息至係⑵之裝置。 對具有許多連結主裝置之更廣域網路而言，接收與傳遞 規則於路徑之各跳程給定所接收參考碼之轉換，並且這些 規則併同儲存於各排定表中之結構化參考碼資料，致使接 收主裝置得以辨識一訊息是否欲供其微網之局部構建使 用，或是否該訊息係欲供一連結之主裝置可進入之節點之 用，而主裝置計算併同訊息傳送之新排定指標。 熟悉此技藝者將瞭解，當一新微網連結至一既有網路時， 需更新整個網路之排定表，且若微網或節點脫離網路時亦 然。 ^述自動配置程序中，在微網與經配置以反映這些鏈路 之焦金排定表62、64間形成直接鏈路40。此外，微網拒絕 請求連結至其它已知微網，以避免迴路或環拓樸。在以圖i 所述範例中，主節點M3提出請求與Ml連結，Ml與其它直 85589 -16- 200400720 接連結之主節點（M2)檢查決定在請求前是否已知有M3。由 於M2《響應為確認，故連結請求M3 — M1遭拒絕以避免迴路 產生。 在另具體貫施例中，鏈結請求雖遭拒絕但被記錄為次 級鏈路。如前，來自M3之訊息經由具M2之直接鏈路排定至 Ml。但監視跨鏈路之服務品質（平均潛伏期、下降或未受承 涊封包/訊息數等）。在M3與M2間鏈路4〇閉鎖或失敗，或因 過度負載而變差之情況下，則先前記錄之次級鏈路可提昇 至一直接鏈路，再計算排定表及排定指標，並利用如前所 述（昇級鏈路於後排定訊息。此具致使網路在諸如直接連 結失敗或因某些因素而閉鎖之特定情況下自行療治。此外， 其可為直接連結之情況，但高負載仍可作業，且接著可利 用次級鏈路幫助分攤跨越具許多微網之密集網路之傳輸負 載。將於下列具體實施例中描述一種可達成上述之範例。 在此具體實施例中，在初始配置期間於各節點定出靜態 參考碼’而參考碼隨加入一微網之各節點而增。例如：參 閱圖1，可針對M1、SU、Sn及Sn分別指定參考碼〇、}、 2及3予微網1 〇。接著將具有如μ i之表62中所示計數之參考 碼4 ’經由鏈路4〇(M1-M2)連結至微網1〇之第二微網2〇。但 微網20在表64中之啟始參考碼，對M2為4，對S21為5，對S22 為6等。在表64中對Ml之參考由Ml供應予M2，在此情況下， M2紀錄之入口碼為〇。類似地，微網3〇之參考碼將包括分 別對M3、S31及S32之7至9，及對M2之4。 在此具體實施例中，傳遞與接收規則簡化而無需轉換， 85589 -17- 200400720 =即排#2僅為網路巾之節點參考數H對傳遞及 2收規則而τ，rI=lr。在表中之計數仍可用以決定目標 :點’俾如前述般遞送訊息。例如參考碼5並未儲存於㈣ 中，但5介於M2之參考碼4及對應細之參考碼加上可進入 裝置數（4制0)間。故計數加上參考碼資料一般仍可如前 述般辨識目標節點供遞送之用，同時維持緊密排定表。 此具體實施例之優點在於決定何時施行次級鏈路為宜之 有效機制。例如：各主/協同節點與次級連結之主節點交換 特定’、訊此貞A包括傳②者自有靜態RI/LR(自傳送者排 定表之ref(SRC ADDR))及傳送者計數（包含利用直接鍵路可 經由傳送者之微網抵達之其它節點）。例如利用圖卜⑷轉 移其LR 7及計數3 m兄下係表微網3〇上之局部節點 數。Ml現可利用先前紀錄之次級鏈路⑷-⑷，決定直接傳 送所欲參考碼範圍為7至9之目標至河3之訊息。在此範例 中’此舉可降低m程數。可以記憶體56中之各主節點儲存 八級鏈路及計數資料為次級排定表，其與原始直接鏈路排 定表—併被檢視，以決定目標微網，因而連結使用。將瞭 ^具有許多微網之密集複雜網路中，使用次級鍵路及靜 態參考碼可更有效率操作，同時仍維持相當緊密之排定表。 在另一進一步具體實施例中，致使網路得以於發生改變 時（例如一節點加入或離開一微網時）自動重新配置。在此背 景自動配置程序（ACP)中，各微網主/協同節點（M丨、M2、M3) 均具一替代配置表，可於發生變化時更新。繫住訊息8〇之 網路PDU(NPDU)部中之一位元，以指示接收協同應使用何 85589 -18 - 200400720 配置表，決定目標節點供遞送訊息之用。 例如：新節點加入主節點M1協同之微網1〇，導致mi開始 ，直接連結之相鄰主節點之通訊過程。在該過程中，⑷計 算與變化有關之新局部參考碼’並將之儲存於替代配置表 中。⑷接著與其已計算供與該鄰近者相鄰者（M2)用之新局 部參考碼通訊。當⑷已完成替代表之建立時，其接著採用 此表為排定表。如此藉由在訊息中之卿⑽元指示使用該 新表產生之訊息。亦即’在網路有變化之前，例如隱含值 NPDU位元為〇時。在改變後，⑷傳送具有被設定為1之卿口 位元之訊息。 相鄰節點（例如M2)類似與其連結之相鄭節點（例如㈣下 游通訊，並建立替代排定表。只要—建構完成，M2即藉由 NPDU位元之設定為丨而表示新表之使用。 在此方<中，？文變於網路中激起連冑，並隨時 有主節點。注意由未更新之節點產生之訊息以npdu位元示 之’而已更新但仍接收此訊息之節點僅暫時回復至原始組 態表，以決定適當排定。網路層76(圖3)具有描述暫停期之 碼’其係表自動配置程序後，可利用原始表之時間長和。 在：例模式中，假設訊息1〇秒内無法達其目的地，但原:始 組態表維持有效-分鐘，以確保許多遠離目標之跳程產生° 之訊息仍得以正確遞送。因Λ，此背景Acp致使網路操作 並重新配置，而無大量中斷服務之情況。 在另-具體實施例中，選用連結裝置58包括一光學紅外 線收發信機’其具相關硬碟以依本發明轉換無線訊息為紅 85589 -19- 200400720 外線訊息，並以紅外線型式接收與傳遞這些訊息。 在又-具體實施例中，連結裝置58包括系列資料境線及 相關插頭與插座。 在又一進一步具體實施例中，連結裝置58包括利用電源 線為訊息載體，以於微網間傳遞訊息。此施行恰適於家用 ZigBee無線微網之具體實施例，其中設定蓋盒具無線裝置 並係王要在冢中之—層樓上之裝置主體，且個人電腦戋電 :見可能為主要在另一層樓上之裝置主體。可採用熟悉此技 蟄者已知用以轉播穿越諸如χι〇或TC2〇5(Cenele幻等電力線 之訊息之機制。 上述方法及程序一般併同互連微網運作，其中定律避免 迴路並通知所有主節點，且依不論何時—節點離開一=結 t微網或一新節點加入一連結之微網或一微網本身離開該 網路，即更新其排定表。概言之，排定表應為各主裝置更 新，移除或加入對該微網之相關入口碼，並根據計數值再 計算微網之新參考碼。 在具本發明之一態樣之上述方法中，每當主節點一接收 到訊息中之排定指標，即以接收規則轉移之，且每當其轉 移至另一連結之主節點時，即亦由傳遞規則轉移之。欲以 施行上述之方法即主裝置特別適於，但不以之為限，^於 例如ZigBee無線協定之低功率、低資料率無線網路。預期 此類ZigBee無線裝置在寫入時，記憶體及處理資源有限： 此外，裝置一般獨由8位元組辨識碼所辨識，且透過空氣傳 遞之訊息受限於1 0 〇位元組之報頭及資料資訊。因此，戈、丨 85589 -20- 200400720 述建構之依序排定表’併同簡單接收與傳遞規則之使用， 致使訊息為2位元組排定址表所排定。藉以釋放受限之訊息 空間，並可用以改善無線頻帶寬使用。 此外，配置程序之組合，包含次級鏈路之表格化，致使 網路得以表現堅定’並至一定範圍施行背景重新配置，無 須大量停止操作。 閱讀本揭示，熟悉此技藝者即可明瞭其它變化及改良。 :類變化及改良可包含此技藝中已知之等效品及其它特 徵’可用以取代或加入此處已述之特徵中。 【圖式簡單說明】 以上藉由範例，並參考隨附圖式描述本發明，其中·· 圖1顯示具連結微網之網路圖； 圖2係依本發明所製主節點無線裝置方塊圖； 圖3係表主節點採用之無線堆疊； 圖4闡釋供本發明之方法使用之具排定辨識碼欄位之無線 訊息（封包）； 圖5闡釋供本發明之方法使用之範例排定表； 圖6列舉依本發明供接收規則用之範例虛擬碼； 圖7列舉依本發明供傳遞規則用之範例虛擬碼； 圖8係闡釋依本發明之方法之示例性步驟之流程圖。 應注意以上圖式僅係圖解，並未按比例繪製。所顯示之 這些圖式之邵件之相對尺寸及比例已經大小擴張或縮減， 俾清楚及便於闡釋圖式。一般採用相同代號代表在改良與 相異具體實施例中之類似部件。 85589 -21 - 200400720 【圖式代表符號說明】 10,20,30 微網 12 主裝置 14，16，18 從裝置 Sll，S12, S13, 辨識碼 S21，S22, S31， S32, Ml，M2， M3 22, 32 主節點 24, 26, 34, 36 從節點 40 鏈路 50 微處理器 54 收發信機 56 記憶體 58 連結裝置 60 電源裝置 62 排定表 63 局部微網入口碼 65 連結微網入口碼 64, 66 表 70 堆疊 72 較低實體 74 中間進入控制 76 網路層 85589 -22- 200400720 78 較高層 80， 100 訊息 82, 84 爛位 87 訊息資料 89 總和檢查欄位 90 接收訊息 92 施加接收規則 94 辨識目標節點 96 施加傳遞規則 97 路徑 100 產生新排定指標 102 傳遞訊息 110 接收規則 120 傳遞規則 85589 - 23 -ZigBee micronet to another already established. Known current mobile network schedules generally require a lot of resources at each node. Η Φ # / Μ ^ ^ ^ Storage, for example, must be inserted into the message machine kU / row In the flood season, such information can be added to the middle and destination node addresses of w 4 and f 4 (each address generally has a unique six- or eight-bit system.). There is no opportunity to schedule information. SUMMARY OF THE INVENTION Therefore, 'the purpose of the present invention is to eliminate the above-mentioned defects. According to the first aspect of the present invention, there are provided ^ kinds of wishes for scheduling in a wireless network with a complex network, several kings / _ ^ <10,000>, each micro network Both, ~, point, which is connected to the main section of another micronet, Wang 1 ",, *, and is operable to store a link with Liao and each of its connected master nodes, ... ,,, 亍 $ 排 疋 表, Among them, the schedule table 4 is used to identify any micronets that are connected to its microgrid. 邳 埂 爻 坆 Click the local microgrid entry and the micronet horse to identify its direct connection. Θ < Code, including one of the master nodes of a micro-mesh: ~ Wanfa self-connected master node receives—the message contains—scheduled according to—receive rule calculation—local reference code from the indicator, π 'according to the local The reference code and schedule information identify a target that is selected for a long time, whether the target node is identified as _ and the slave node that transmits the message to the slave node connected to the Hanwei network; or whether the hidden target node is identified as-accessible via _ Connected to the micronet > Click on the 'by-pass rule' to calculate a new scheduled indicator, Definite index: 1 Change the scheduling index in the message, and pass the message to the link micronet: ^ point. Wang! 85589 200400720 This U has been #out_types suitable for scheduling across several links connected by the master node The ancient, the earth, and the earth of the message of the net, "t Wan Qu", among which the message with the row index is based on the rules of simple storage, which spans this distance. # — ^ — Μ I, ,, and row row, and only by The master node uses the stored schedule information. Script: The list of deficiencies includes restricted resources related to the micronet that is directly connected to the master node. Η Yishan, a on, and she adds the steps of this method. When the master node needs to deliver the message, That is, 辚, 节点 ,、, 兰, | _ 硪 link out the main node of the message. The advantage is that in the scheduling table φ > owe, hit 丨 armor < each link micronet entry code has a path that is the same as the address of the master point of the link mechanical network, and the reference code is lost, and a Count of summary points entered through the master node. Shibubu > Here are two counting tanks. ≪ The counting system is a summary of points that can be entered through the connected micronet master node. A node of any other Wechat. Make these people coexist and ensure that there is no loop in the network topology that is expected to be generated, so that the master node can store a very tight schedule. This is the same as the rules (it depends on the order of the population code). Schedule messages efficiently. In the network according to the above-mentioned master node, there are processing, transceivers and memory, a simple schedule and tightly stored schedule, and the same steps as the method of implementing the described schedule information. The communication between the master node and the slave nodes in the micro-network of the network according to the predetermined short-range wireless protocol determined by ZigBee AUiance is better.简单 The connection of the main node can be achieved by simple materials or power lines directly connected to the master node, or the connection can be achieved by wireless or infrared hardware. If the micronet is connected by a wireless link, the advantage lies in the use of the transceiver and the "protocol" already existing in the node, and the operation of the local micronet connection, and on different wireless channels, across the micronet chain The connection from one micronet to another micronet will reduce the wireless interference to the lowest level of 85589 200400720. [Embodiment] Figure 1 shows an example of a network with connected micronets 〇, 20, and 30. Each micronet has a code Line device or module, configured as the master node of the microgrid, and using the master node as /, Η ,, 'same central v Zhoulang as the master node. In the microgrid 10, the master node includes a master with a unique identification code M1 Wireless device. The master node allows other wireless devices to be temporarily stored and connected to their micronets to become slave devices. In an operating mode, it coordinates the communication within the micronet by proposing periodic reference code signals (wireless beacons). Alas, the child signal contains a message with an identification code for its local slave device. In the figure, the master section M1 has three connected slave devices 14, 16, 18, and the identification code S11, S12 & S13. Therefore, The master device 12 coordinates its connected slave devices 16 18 communication. Those familiar with this technology often refer to this type of micronet as a master or a star network, or sometimes a star or master / slave topology. Micronet 10 The radio broadcast range of the master node is limited. Power is often maintained to the master node (sometimes called a "full-function device,"), while the slave device (sometimes called a "reduced power device") can be powered by a battery. , And programmed with various power saving wireless mechanisms that include sleep and awake modes. The ZigBee Alliance attempts to standardize such wireless modules such as lighting and extended data rate and low-power applications, and controls the operation of the "peripheries of the microgrid's house" and According to the direct sequence distribution spectrum mechanism and the Carrier Sense Multiple Access (CSMA) protocol, the micronet communication in the master / slave micronet is generally based on the network master wireless beacon (with a network identification code, such as the master device). Identification code). For the sake of explanation, 'In the application of slave wireless devices, M 85589 200400720, and 18, it can be incorporated into a light switch or a lamp, where the slave devices are paired with each other.' 'The switch operation results in a data packet. The wireless signal transmission of the master device 'inquires the identification code of the paired device (or sometimes called the endpoint) and transmits the appropriate message to the paired lamps' to turn the lamps on or off. This kind of master / slave micro-network configuration For details, such as the applicant's pending patent application published on April 19, 2001, 00128156 and ^ ¥ 00128157 are provided for readers' guidance, and the disclosure thereof is incorporated herein by reference. Figure 1 also Two other microgrids 20 and 30 are shown, each with its own individual master and connected slave devices. The microgrid 20 includes a master node 22 (with a unique identification code M2), which coordinates the slave nodes 24 (S21) and 26 (S22). The network 30 includes a master node 32 (M3) and coordinated slave nodes 34 (S3 1) and 36 (S32). The following situations may often occur in a large building, where the first microgrid has been installed, and then several other microgrids have been installed. For example, we can use microgrid 10 in security products (such as window / door sensors) Line device, which has a second 20 and a third microgrid 30 installed later to increase security. Each microgrid shown in the figure is connected, so that the information can be transmitted from the coordination node of a microgrid to the microgrid using the method of the present invention. In another place, the microgrid 10 has been connected to the microgrid 20 of M2 by link 40, and when the microgrid of μ) is installed, it is connected to the microgrid of M2 in sequence. The direct link can be passed through The spring or cable broadcast or optical device is ", and it can even be virtualized in an intermediate network (such as a local LAN or Internet-not shown in this figure). 仏 In the following preferred embodiments It is assumed that there is a link between the main device and the wireless protocol and hardware used by the same device used by each of the king devices. However, those skilled in the art will know that "link", the actual entity of the specific embodiment. With its B well-known communication devices, such as optical infrared communication or power line 85589 -10- 200400720 signal transmission or direct telegram.会 戏 # The main point of Xia is that the link carries a specific type of information between the micronets, which will now be briefly described. Classes exist when considering network topology—additional considerations, especially when there are loops or% s. It is generally understood that loops or ring topologies should be avoided (to avoid the message "trapped, in an endless loop path") and the system described is no exception. Make sure that "connecting the master device itself does not constitute a loop in the network topology The simple method is to include a king device 'receive the request from the supply-identification of another_master device, and request to establish a direct link, first check other connected devices to identify the requesting device, and if the identification of the requesting device is not found, then Only one link is established. Refer to the figure and sigh U, .. 罔 1G, 20, 3G < the master nodes are all within the range of other wireless coverage areas (not shown in the figure), establish the MW and M2_M3 as shown The link 40 of the test is M3 & print to the link with M1. Before granting the connection, n will pass-request to its directly connected node (M2) to ask if ⑷ is known. M2 reply with-positive confirmation Therefore, M1 refuses the connection request from ⑷. Right M2 itself is connected to another master device, then it should pass the inquiry to the master device, which must reply the result to M1, etc. via M2. The description will now be made to make the wireless as shown in Figure 丨Microgrid according to one aspect of the present invention One of the preferred embodiments of the master node device component, code, schedule and message format for sending messages, where the micronet operates according to the ZigBee wireless standard. Figure 2 is suitable for implementing and coordinating the micronet 10 and spanning links The block diagram of the main wireless device 12 of the micronet operation message is sufficient. The device has a power supply device 60, which can charge a processing device existing in the form of a microprocessor 50 (or a microcontroller), wherein the processor is coupled to A wireless transceiver 54 for transmitting and receiving wireless messages. The processor is also coupled to store opcodes and to communicate with connected slave devices 85589-11-200400720 (e.g., Identifiers, paired ends, Point data and service capabilities) memory device 56. The figure shows that the memory separate from the processor includes ROM (or EEPROM) and RAM, but it can also be integrated with the processor according to cost and chip design considerations (It is shown by the dotted line.) This embodiment is approved. The transceiver 54 is used as the connection between the main devices. The better device is set. The main device transmits / receives it to serve its micronet. 〈Connected It takes place on different channels used by the main device. Note: The device D includes an optical link I set 58 and β links the master node to another master node, as will be described in a specific embodiment later. The processing device is based on a predetermined ZigBee wireless stack and The protocol operates the micronet. As shown in Figure 3, the stack 70 is located in the ROM section of the memory 56 and is composed of the lower entity (PHY) 72 and the intermediate access control (MAC) 74 layers. After receiving 4 wireless messages, it can pass to the network layer (nwk) 76 in sequence. This layer interfaces with the higher layer 78 which has an application code itself (for example, if the code is in a lamp, it is a lighting application). Program When the instructions are executed by the processor 50, the steps in the present invention will be executed, and a part of the Nwk layer is preferred. Figure 4 is an example of the structure of a header / commercial payload data packet or message 80. The message has various fields 82. The header information of the field 82 includes the unique identifier of the sender (usually the «8 digits (group unique number) representing the device manufacturer and the wireless data packet transmitted, for example-field 84 Among them, in this preferred embodiment, the insertion or improvement of the two-tuple scheduling index (RI) will be briefly described. The message data 87 and the sum check field 89 are also determined. For those familiar with the field of packet wireless systems, Known 'part of the message (service and protocol data unit SDU, PDU) is in the layers defined in the wireless standard ± operation or related to 85589 -12-200400720 of the present invention When receiving a message from the MAC layer, it will be composed of The master device works with the header address of the sender and the RI data field 84, and the data stored in the schedule table. Figure 5 illustrates the configuration and storage of each master node 12, & 42 in the memory 56. An example of a scheduling table in which the NWK layer operates to schedule a message 80. The master device builds an mutual scheduling table 62, which includes a local micro-code with the reference code (ref) and address (ADDR) of the relevant slave device Net access code 63, and its linked master Node link micronet entry code 65. Table 62 corresponds to the schedule of the main mi storage of the network shown in Figure i, while table 64 corresponds to the storage of M2, and table 66 corresponds to "3 storage. Link micro The network entry code 65 is referenced under the local micronet entry code 63, and includes the address of the directly connected master node and the total number of devices / nodes (CNT) that can be accessed through the link. For example, when the microgrid 10 was first When the table 62 is installed, it will only include the local reference code of the relevant devices in the microgrid. Later M2 will be linked to Mi, M1 will update its schedule table with REF 4, the identification code address at that time, M2, and the count will only be Including the device of M2's microgrid (ie 3). At some later time, the microgrid M3 is connected to M2, and M2 adds the entry code to M3 and has the reference code calculated from the previous reference code (REF) entry code and related counts (also That is, REF 3 plus CNT 4 gives the entry code in REF row 7 of M2 table 64). M2 then informs the new microgrid (belonging to M3) of all previously linked microgrids with 3 nodes (ie Ml ). M1 therefore updates the count entry code of M2 to 6 (this is the number of devices entered via M2). In a similar manner, M2 connected to another micronet (M1) with 4 nodes notified M3, so M3 updates the table 66 entry code related to the count (CNT) of M2 as 3 + 4 = 7. The data in these tables is the master device Used 'and receive and transfer rules 85589 -13- 200400720 (algorithm) to calculate the subsequent scheduling indicators used to schedule messages from the source node to the destination node. # 收 主 t 置 Implement the receiving rules; identify this The result is the target node; if the target node is the master device of another microgrid, a new scheduled indicator calculated according to the transfer rule is inserted, and the message is passed to the target node. FIG. 6 illustrates an example virtual code for the operations contained in the receiving rule 110, and FIG. ^ Illustrates the corresponding virtual code 12 for the delivery rule. In the code shown in Figure 6, the magic system is a scheduling indicator, _SRC) is the reference code in the row M of the receiving host device that has transmitted the message from the master device for identification to the receiving rule, τ〇τα: It is the device in the network and (calculated after the main device adds the population code in the total count table and adds it to the relevant device number in its micronet), and LR is the local reference code of the calculation result. Figure 7 illustrates an example virtual code for the operations contained in the delivery rule, where ref (TN) is a reference code that identifies the I target node in the scheduling table of the master device implementing the delivery rule, and CNT (TN) is related to the target node The count LR is a reference code calculated from the receiving rule and used by the delivery algorithm to calculate a new rule index. For the sake of reorganization, a discovery process will now be described, in which the master device queries the rule index for a specific device for future transmission. In the figure 丨, it is assumed that the device S 11 is to be paired with the device S2 丨 (for example, s 丨 丨 can be a lighting control unit, " the user hopes that under the control of M2, it will be paired with the installed lamp S2 1) . S1 makes a request to its master M1 and needs to generate an endpoint with S21. M1 checks his list 62 and finds that there is no entry code for S21. M1, so a search request is made to its linked master% 2. M2 receives the request, checks its list and finds its related device S2, which has the local reference code. M2 now applies the transfer rule to this reference code to generate a scheduling index RI of 5 and sends it back to the master M1. Ml 85589 -14-200400720 Receive the reference code of RI = 5 for S21, and apply the receiving rule 'to generate the resulting reference code 5 in this example. M1 stores this result and connects the initial reference code to it, and then removes your 4, 4, heart,, and unseen foot address S21, and stores the endpoint data that pairs device S11 with the storage code. The receiving master M1 now schedules the message according to the general method shown in the flowchart in FIG. 8, where the receiving message (step 90) has a scheduling index, and a receiving rule 11 (step 92) is applied. The target code (step 94), and the right child target code is identified as being accessible through the directly connected micro-mesh, then the delivery rule 120 (step 96) is applied to generate a new scheduled index (step ι〇〇), and inserted U〇 "Seat 84 and pass the message (step! 〇2) Otherwise, if in step% 'target sincerity is identified as receiving the relevant components of the main micronet (that is, the result of the receiving rule is given to be part of the schedule table The entry reference code 63 is the local reference code LR), the processor of the king device points to step ⑻ via path 97, and transmits the message (step 骡 102) to the local node. Of course, it is necessary to calculate the first master device in a multi-hop path (for example, Kawa __- M2-S21) according to the scheduling index, and then operate the method of FIG. 8 with each receiving master node route. For example, in receiving a wireless message from S11, M1 recovers the reference code it previously stored from the search process, and identifies a target code from the reference code value and schedule information. The reference code 5 is not stored in Table 62, but 5 is between the reference code 4 of M2 and the reference code (4 + 6 = ι㈠) corresponding to M2 plus the number of devices that can be accessed. It will be understood that for this example, although This is a trivial matter. In a network with many connected master nodes and a list of π codes with related counting people, the reference code (calculated in the previous material finding process) can be used to identify which link master to deliver the message to: 85589 200400720 The checklist of the schedule stored in M2 corresponds to the reachable node of M1, and the target range is from 3 to 6; the range is, and the country is at 7 and 10. The reference code is corresponding to Wang. M2 can reach the node. Therefore, the reference code data can generally be used to identify the target Lang point for delivery. Back to the multi-hop path example, the main M1 and M2 are identified as the target node, and then the transmission is performed on reference code 5. Rule, give,. ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, ', and the identified target node μ]. M2 is now applying a message to show the receiving message (step 9G), and then the receiving rule (step 92 ) 'To give LR41. By comparing this reference code with its stored (Identify the table M2 and identify (step 94) the target node system corresponding to this reference code-the local related device (step 95), thus transmitting the message to the system's device. For a wider area network with many connected master devices , The receiving and transmitting rules in each path of the path give the conversion of the received reference code, and these rules are also stored with the structured reference code data in each schedule, so that the receiving host device can identify whether a message is intended for The micro-network is used for local construction, or whether the message is intended for a node that can be accessed by a connected master device, and the master device calculates and sends a new scheduled indicator with the message. Those familiar with this art will understand that when a When a new micronet is connected to an existing network, the schedule of the entire network needs to be updated, and if the micronet or node is off the network. ^ In the automatic configuration procedure, the micronet and The focal gold scheduling tables 62 and 64 of these links form a direct link 40. In addition, the microgrid refuses to request to connect to other known microgrids to avoid loops or ring topologies. In the example described in Figure i, Master node M3 proposes Request to connect with M1, M1 and other directly connected master nodes (M2) 85589 -16- 200400720 Check whether the M3 is known before the request. Because M2 "The response is a confirmation, the connection request M3-M1 is rejected to avoid Loop generation. In another specific embodiment, the link request is rejected but recorded as a secondary link. As before, messages from M3 are scheduled to M1 via a direct link with M2. But monitoring cross-links Service quality (average latency, decreased or unaccepted packets / messages, etc.). In the case where the link between M3 and M2 is blocked or fails, or deteriorated due to excessive load, the previously recorded secondary The link can be upgraded to a direct link, and then the scheduling table and scheduling indicators are calculated, and as described above (upgrade the link to schedule the message later). This can cause the network to heal itself in certain situations, such as when a direct connection fails or is blocked for some reason. In addition, it can be a direct connection case, but high loads are still operational, and then secondary links can be used to help spread the transmission load across dense networks with many microgrids. An example that can achieve the above will be described in the following specific embodiments. In this specific embodiment, a static reference code 'is set at each node during the initial configuration, and the reference code increases with each node joining a microgrid. For example, referring to Fig. 1, reference codes 0,}, 2 and 3 can be assigned to M1, 0 for M1, SU, Sn and Sn, respectively. A reference code 4 'having a count as shown in Table 62 of μi is then linked to the second microgrid 20 of the microgrid 10 via a link 40 (M1-M2). However, the starting reference code of the microgrid 20 in Table 64 is 4 for M2, 5 for S21, 6 for S22, and so on. The reference to M1 in Table 64 is supplied by M1 to M2. In this case, the entry code of the M2 record is zero. Similarly, the reference code of WeChat 30 will include 7 to 9 for M3, S31 and S32, and 4 for M2, respectively. In this specific embodiment, the transmission and reception rules are simplified without conversion, 85589 -17- 200400720 = that row # 2 is only the node reference number of the network towel H pair transmission and two reception rules and τ, rI = lr. The count in the table can still be used to determine the goal: points': deliver the message as before. For example, the reference code 5 is not stored in ㈣, but 5 is between the reference code 4 of M2 and the corresponding detailed reference code plus the number of accessible devices (4 to 0). Therefore, the count plus the reference code data can still identify the target node for delivery as described above, while maintaining a tight schedule. The advantage of this embodiment is that it is an effective mechanism to decide when to implement the secondary link. For example: each master / cooperative node exchanges with the master node of the secondary link. The message A includes the static RI / LR (ref (SRC ADDR) from the sender schedule) and the sender count. (Including other nodes that can be reached via the sender's micronet using direct links). For example, use Figure ⑷ to transfer its LR 7 and count the number of local nodes on the micro-grid 30 of the 3 m sibling table. Ml can now use the previously recorded secondary link ⑷-⑷ to decide to directly transmit the information from the target with the reference code range 7 to 9 to River 3. In this example, 'this will reduce the number of m passes. Each master node in the memory 56 can store the eight-level link and count data as a secondary schedule, which is compared with the original direct link schedule—and viewed to determine the target microgrid, and therefore used in conjunction. In dense and complex networks with many micronets, the use of secondary bonds and static reference codes can operate more efficiently, while still maintaining a fairly tight schedule. In another further specific embodiment, the network can be automatically reconfigured when a change occurs (for example, when a node joins or leaves a piconet). In this background automatic configuration program (ACP), each microgrid master / cooperative node (M 丨, M2, M3) has an alternative configuration table that can be updated when changes occur. A bit in the network PDU (NPDU) part of the message 80, to indicate which configuration should be used by the receiving team 85589 -18-200400720 to determine the target node for message delivery. For example, a new node joins the micro-network 10 coordinated by the master node M1, causing mi to start a communication process with directly connected neighboring master nodes. In the process, the new local reference code related to the change is calculated and stored in the replacement configuration table. ⑷ Then communicate with the new local reference code that it has calculated for use with that neighbor (M2). When ⑷ has completed the establishment of the replacement table, it then uses this table as the scheduling table. Thus, the message generated by the new watch is instructed by the secretary in the message. That is, 'before the network changes, such as when the hidden value NPDU bit is zero. After the change, ⑷ sends a message with the key port set to 1. Adjacent nodes (such as M2) are similar to the associated nodes (such as downstream communication and establish alternative schedules). As long as the construction is completed, M2 indicates the use of the new table by setting the NPDU bit to 丨. In this way, the text change arouses flail in the network, and there is a master node at any time. Note that the message generated by the unupdated node is indicated by the npdu bit, and the node has been updated but still receives this message. Only temporarily return to the original configuration table to determine the appropriate schedule. The network layer 76 (Figure 3) has a code describing the suspension period. 'It is the automatic configuration procedure of the table, and the time of the original table can be used. In the mode, it is assumed that the message cannot reach its destination within 10 seconds, but the original: initial configuration table is kept valid for one minute to ensure that many messages that are far away from the target produce ° are still delivered correctly. Because of Λ, this background Acp As a result, the network is operated and reconfigured without a large number of service interruptions. In another embodiment, the connection device 58 is selected to include an optical infrared transceiver, which has an associated hard disk to convert the wireless message to red according to the present invention. 85589 -19- 200400720 external line information, and receive and transmit these messages in infrared type. In another embodiment, the connection device 58 includes a series of data lines and related plugs and sockets. In yet another specific embodiment, the connection device 58 Including the use of a power cord as a message carrier to transfer messages between microgrids. This implementation is suitable for a specific embodiment of a home ZigBee wireless microgrid, in which the cover box is equipped with a wireless device and the king wants to be in the middle of the mound-upstairs The main body of the device, and the power of the personal computer: See the main body of the device that may be mainly on another floor. Known by those skilled in this technology can be used to relay messages across power lines such as χι〇 or TC205 (Cenele magic) The above methods and procedures generally operate in conjunction with interconnected microgrids, where the law avoids loops and notifies all master nodes, and is based on whenever-nodes leave a = micro-grid or a new node joins a linked microgrid or A micronet itself updates its schedule when it leaves the network. In short, the schedule should be updated for each master device, remove or add the relevant entry code for the micronet, The new reference code of the microgrid is recalculated according to the count value. In the above method with one aspect of the present invention, whenever the master node receives the scheduled index in the message, it is transferred by the receiving rule, and whenever When it is transferred to another connected master node, it is also transferred by the transfer rule. To implement the above method, the master device is particularly suitable, but not limited to it. For example, the low power and low power of the ZigBee wireless protocol Data rate wireless network. It is expected that such ZigBee wireless devices have limited memory and processing resources when writing: In addition, devices are generally uniquely identified by 8-byte identifiers, and the information transmitted through the air is limited to 10 〇 bytes header and data information. Therefore, Ge, 丨 85589 -20- 200400720 described the construction of the sequential scheduling table 'and the use of simple reception and transmission rules, making the message a 2-byte addressing table Scheduled. This frees up limited message space and can be used to improve wireless frequency bandwidth usage. In addition, the combination of configuration procedures, including the tabularization of secondary links, has enabled the network to perform steadily 'and perform background reconfiguration to a certain extent without having to stop operations in large numbers. After reading this disclosure, other changes and improvements will become apparent to those skilled in the art. Variations and improvements may include equivalents and other features known in the art, which may be used in place of or in addition to the features already described herein. [Brief description of the drawings] The above describes the present invention by way of example and with reference to the accompanying drawings, in which FIG. 1 shows a network diagram with a connected micronet; Figure 3 is a wireless stack used by the master node of the table; Figure 4 illustrates a wireless message (packet) with a scheduled identifier field used by the method of the present invention; Figure 5 illustrates an example scheduling table used by the method of the present invention Figure 6 lists example virtual codes for receiving rules according to the present invention; Figure 7 lists example virtual codes for delivery rules according to the present invention; Figure 8 is a flowchart illustrating exemplary steps of the method according to the present invention. It should be noted that the above figures are only illustrations and are not drawn to scale. The relative sizes and proportions of the figures shown have been expanded or contracted, making them clear and easy to interpret. Generally, the same code is used to represent similar parts in modified and different embodiments. 85589 -21-200400720 [Illustration of Symbols in the Drawings] 10,20,30 Microgrid 12 Master Device 14,16,18 Slave Device Sll, S12, S13, Identification Codes S21, S22, S31, S32, Ml, M2, M3 22, 32 Master node 24, 26, 34, 36 Slave node 40 Link 50 Microprocessor 54 Transceiver 56 Memory 58 Link device 60 Power supply device 62 Schedule 63 Local microgrid entry code 65 Link microgrid entry code 64, 66 Table 70 Stack 72 Lower entity 74 Intermediate access control 76 Network layer 85589 -22- 200400720 78 Higher layer 80, 100 Message 82, 84 Bad bits 87 Message data 89 Sum check field 90 Receive message 92 Apply receive rule 94 Identify target node 96 Apply delivery rule 97 Path 100 Generate new scheduled indicator 102 Pass message 110 Receive rule 120 Pass rule 85589-23-