【摘要】：近年來,隨著移動(dòng)通信網(wǎng)絡(luò)的發(fā)展,移動(dòng)網(wǎng)絡(luò)結(jié)構(gòu)越來越復(fù)雜,用戶對(duì)網(wǎng)絡(luò)質(zhì)量要求也越來越高,因此移動(dòng)網(wǎng)絡(luò)維護(hù)變得越來越重要。傳統(tǒng)的響應(yīng)式故障處理已經(jīng)無法滿足運(yùn)營(yíng)商對(duì)網(wǎng)絡(luò)質(zhì)量的要求,如何掌握網(wǎng)絡(luò)實(shí)時(shí)運(yùn)行狀況,將網(wǎng)絡(luò)性能隱患進(jìn)行主動(dòng)預(yù)警,將用戶投訴前移是當(dāng)前運(yùn)營(yíng)商急需解決的重要問題。本文針對(duì)中國(guó)移動(dòng)某市MSC Pool的移動(dòng)網(wǎng)絡(luò)結(jié)構(gòu),設(shè)計(jì)一套移動(dòng)核心網(wǎng)性能分析預(yù)警系統(tǒng),用于監(jiān)測(cè)核心網(wǎng)性能指標(biāo)變化,對(duì)異常的指標(biāo)進(jìn)行預(yù)警,以便網(wǎng)絡(luò)維護(hù)人員盡早分析、處理,減少故障發(fā)生。移動(dòng)核心網(wǎng)性能分析預(yù)警系統(tǒng)主要由數(shù)據(jù)獲取模塊、KPI計(jì)算模塊、預(yù)警預(yù)判模塊、郵件分發(fā)模塊和數(shù)據(jù)呈現(xiàn)模塊組成。數(shù)據(jù)獲取模塊定時(shí)通過FTP協(xié)議下載移動(dòng)核心網(wǎng)設(shè)備MSC Server的數(shù)據(jù)文件,保存到本地目錄中。KPI計(jì)算模塊讀取下載完成的文件,并根據(jù)配置的KPI公式進(jìn)行KPI計(jì)算,將指標(biāo)值保存到數(shù)據(jù)庫(kù)中,KPI公式可由用戶進(jìn)行自由配置。預(yù)警預(yù)判模塊用于將實(shí)時(shí)計(jì)算的KPI指標(biāo)與設(shè)定的預(yù)警門限比較,超過門限則產(chǎn)生預(yù)警,并按照網(wǎng)絡(luò)結(jié)構(gòu)將預(yù)警匯總為預(yù)判,用于反映網(wǎng)絡(luò)整體狀況。本系統(tǒng)采用兩級(jí)預(yù)警,即高級(jí)別預(yù)警和低級(jí)別預(yù)警,采用靜態(tài)門限和動(dòng)態(tài)門限結(jié)合作為預(yù)警門限。靜態(tài)門限由用戶進(jìn)行配置,動(dòng)態(tài)門限采用時(shí)間序列分解及K中心點(diǎn)聚類算法計(jì)算得到。郵件分發(fā)模塊用于將網(wǎng)絡(luò)狀況通過郵件分發(fā)給相應(yīng)工作人員,以便其及時(shí)分析處理。數(shù)據(jù)呈現(xiàn)模塊用于人機(jī)交互,主要供用戶進(jìn)行KPI指標(biāo)、預(yù)警預(yù)判查詢和KPI配置、系統(tǒng)配置。本系統(tǒng)采用Visual Studio 2008作為開發(fā)平臺(tái),VC++作為開發(fā)語言,數(shù)據(jù)庫(kù)采用SQL Server 2008,采用Matlab R2013a作為算法仿真平臺(tái)。本文實(shí)現(xiàn)的移動(dòng)核心網(wǎng)性能分析預(yù)警系統(tǒng),在中國(guó)移動(dòng)某市進(jìn)行現(xiàn)場(chǎng)測(cè)試調(diào)整后,已經(jīng)正式上線運(yùn)行2年左右,系統(tǒng)每15分鐘進(jìn)行一次性能獲取,每次計(jì)算用時(shí)小于2分鐘。運(yùn)行結(jié)果表明,各個(gè)模塊均運(yùn)行正常,對(duì)于系統(tǒng)異常情況能夠自動(dòng)處理；計(jì)算指標(biāo)正確,為用戶分析網(wǎng)絡(luò)狀況提供了可靠的依據(jù)；本文提出的動(dòng)態(tài)門限算法能夠準(zhǔn)確預(yù)警,為用戶提前發(fā)現(xiàn)網(wǎng)絡(luò)故障提供了依據(jù),適合用于移動(dòng)網(wǎng)絡(luò)性能預(yù)警。
[Abstract]:In recent years, with the development of mobile communication network, the mobile network structure is becoming more and more complex, and users are demanding more and more high quality of the network, so the maintenance of mobile network becomes more and more important. The traditional response fault processing has been unable to meet the network quality requirements of operators, how to grasp the real-time operation of the network, the network performance hidden dangers for active early warning, It is an important problem for operators to move forward customer complaints. Aiming at the mobile network structure of MSC Pool in a certain city of China Mobile, this paper designs a set of mobile core network performance analysis and warning system, which is used to monitor the change of core network performance index, to warn the abnormal index, so that network maintainers can analyze it as soon as possible. Deal with, reduce the occurrence of failure. The early warning system for mobile core network performance analysis is mainly composed of data acquisition module, KPI computing module, early warning prejudgment module, mail distribution module and data presentation module. The data acquisition module periodically downloads the data files of the mobile core network device MSC Server through the FTP protocol, saves them to the local directory. The KPI computing module reads the downloaded files, and calculates the KPI according to the configured KPI formula. Save the index value to the database, the KPI formula can be freely configured by the user. The early warning prejudgment module is used to compare the KPI index calculated in real time with the set early warning threshold, and when the threshold is exceeded, the early warning is generated, and the early warning is summarized into pre-judgment according to the network structure, which is used to reflect the overall situation of the network. The system adopts two levels of early warning, namely, high level early warning and low level early warning, and uses static threshold and dynamic threshold as early warning threshold. The static threshold is configured by the user, and the dynamic threshold is calculated by time series decomposition and K-center clustering algorithm. The mail distribution module is used to distribute the network status to the corresponding staff through mail so that it can be analyzed and processed in a timely manner. Data presentation module is used for man-machine interaction, mainly for users to carry out KPI index, early warning query and KPI configuration, system configuration. The system uses Visual Studio 2008 as the development platform, VC as the development language, SQL Server 2008 as the database and Matlab R2013a as the algorithm simulation platform. The performance analysis and early warning system of mobile core network realized in this paper has been formally put into operation for about 2 years after field test and adjustment in a certain city of China Mobile. The system can be obtained once every 15 minutes, and the time of calculation is less than 2 minutes each time. The running results show that each module is running normally and can automatically handle the abnormal situation of the system, and the calculation index is correct, which provides a reliable basis for the user to analyze the network condition. The dynamic threshold algorithm proposed in this paper can accurately warn users, which provides a basis for users to detect network faults in advance, and is suitable for early warning of mobile network performance.
【學(xué)位授予單位】：大連理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TP311.52;TN929.5

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 ;ZTE Achieves TM Forum's Frameworx 12.0 Conformance Certification[J];ZTE Communications;2013年04期

相關(guān)碩士學(xué)位論文 前1條

1 宋蔚騰;GSM移動(dòng)網(wǎng)絡(luò)優(yōu)化技術(shù)的研究及應(yīng)用[D];北京郵電大學(xué);2008年

，

本文編號(hào)：2383173