基于網(wǎng)絡(luò)演算的隨機(jī)型網(wǎng)絡(luò)延時(shí)模型的研究與仿真

《基于網(wǎng)絡(luò)演算的隨機(jī)型網(wǎng)絡(luò)延時(shí)模型的研究與仿真》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫。

1、摘要網(wǎng)絡(luò)控制系統(tǒng)是指通過實(shí)時(shí)網(wǎng)絡(luò)形成閉環(huán)的反饋控制系統(tǒng)，網(wǎng)絡(luò)控制系統(tǒng)將不同地域的監(jiān)控計(jì)算機(jī)、控制器、智能I／O設(shè)備等分布對象通過網(wǎng)絡(luò)連接起來，網(wǎng)絡(luò)控制系統(tǒng)具有布線少、診斷和維護(hù)方便、系統(tǒng)靈活的特點(diǎn)。但是網(wǎng)絡(luò)的引入帶來了信息傳輸延遲和信息丟失等問題，對網(wǎng)絡(luò)控制系統(tǒng)性能產(chǎn)生重要影響的因素是點(diǎn)到點(diǎn)的延時(shí)，延時(shí)會(huì)使系統(tǒng)性能降低甚至不穩(wěn)定。網(wǎng)絡(luò)控制系統(tǒng)中的信源和服務(wù)分為確定型和隨機(jī)型兩種，因此網(wǎng)絡(luò)延時(shí)也分為確定延時(shí)和隨機(jī)延時(shí)。利用隨機(jī)型網(wǎng)絡(luò)演算理論分析延時(shí)能更好的利用統(tǒng)計(jì)復(fù)用，提高資源利用率。本文著重分析優(yōu)先

2、級調(diào)度下，隨機(jī)型業(yè)務(wù)流進(jìn)入就緒隊(duì)列等待服務(wù)器空閑的延時(shí)，對如何構(gòu)建隨機(jī)型業(yè)務(wù)流模型進(jìn)行了探討，采用gSBB(廣義型隨機(jī)界限的突發(fā)性)模型描述隨機(jī)型業(yè)務(wù)流，并以常見的泊松業(yè)務(wù)流為例，推導(dǎo)出了嚴(yán)格優(yōu)先級調(diào)度下兩個(gè)泊松業(yè)務(wù)流在帶寬瓶頸限制下的延時(shí)估算公式。同時(shí)，對NS2進(jìn)行擴(kuò)展編程，添加了泊松業(yè)務(wù)流應(yīng)用程序和嚴(yán)格優(yōu)先級調(diào)度隊(duì)列，然后利用NS2仿真平臺(tái)，在有帶寬瓶頸的節(jié)點(diǎn)下，對兩個(gè)不同優(yōu)先級泊松業(yè)務(wù)流的傳輸進(jìn)行仿真，并分析它們的延時(shí)，得出統(tǒng)計(jì)結(jié)果，與理論分析作比較。關(guān)鍵詞：隨機(jī)型業(yè)務(wù)流，泊松業(yè)務(wù)流，延時(shí)，g

3、SBB模型，隨機(jī)型網(wǎng)絡(luò)演算AbstractABSTRACTNetworkControlSystemfNCS)isakindoffeedbackcontrolsystemwhereinthecontrolloopsareclosedviathecommunicationnetwork．Informationandcontrolsignalsareexchangedbetweenspatiallydistributedsystemcomponents，likesupervisorycomputers，c

4、ontrollersandintelligentI／Odevices(e．g．，smartsensorsandactuators)whichmadetheNCShavetheprimaryadvantagesofreducedsystemwiring,easeofsystemdiagnosisandmaintenance，andincreasedsystemagility．InNCS，certainissuesneedtobeaddressed，includingtimedelaysandinfor

5、mationlossandetc．ThefactorwhichplaysanimportantroleontheperformanceofNCSistimedelays，whichCanleadtosystemperformancedegradationandevensysteminstability．BothofthetrafficsandtheserviceguaranteesinNCSCanbedividedintodeterministicandstochastic，SOdoesthedel

6、aytimesofflowsinNCS．StochasticnetworkcalculusCanreapthebenefitsofstatisticalmultiplexing,whichCanresultinahighutilizationofnetworkresources．Thisarticlemainlyanalysesthetimedelaysofwaitingforwhentheserverisidleafterastochastictrafficflowentersintoaready

7、queue．Itisdiscussedthathowtostructurethemodelofstochastictraffic，andgSBB(generalizedstochasticallyboundedburstiness)modelisadopted．TakethecommonPoissontrafficasanexample，thedelayestimatingformula,ofwhichtwoPoissontrafficflowspassingtroughabandwidth-lim

8、itednodeundertheschedulingbypriorities，isderived．ThemodulesofNS2areextended，byaddingPoissontrafficgeneratingprogramandSPqueues．ThensimulationiScardedoutonNS2thattwoPoissontra伍Cflowspassingtroughabandwidth-limitednodeundertheschedulingby