發(fā)布時(shí)間：2018-01-25 19:13

  本文關(guān)鍵詞： 云安全 云平臺(tái)安全監(jiān)控 可視化技術(shù) 離群節(jié)點(diǎn)自識(shí)別　出處：《山東大學(xué)學(xué)報(bào)(理學(xué)版)》2017年06期 　論文類型：期刊論文

【摘要】：通過總結(jié)目前云平臺(tái)安全監(jiān)控的數(shù)據(jù)可視化技術(shù),結(jié)合具體的多維監(jiān)控?cái)?shù)據(jù)探討可視化技術(shù)的應(yīng)用方法,從時(shí)間、節(jié)點(diǎn)號(hào)、性能指標(biāo)類型三個(gè)維度出發(fā),提出了基于維度壓縮與維度切面的性能數(shù)據(jù)集可視化方法,并在此基礎(chǔ)上,應(yīng)用動(dòng)態(tài)時(shí)間規(guī)劃和卷積神經(jīng)網(wǎng)絡(luò)實(shí)現(xiàn)離群節(jié)點(diǎn)自識(shí)別,豐富擴(kuò)展了警報(bào)系統(tǒng)的語義。經(jīng)實(shí)驗(yàn)驗(yàn)證方法可行,能夠更直觀地展現(xiàn)有效信息,提高云管理員的決策效率。
[Abstract]:Through summarizing the current cloud platform security monitoring data visualization technology, combined with specific multi-dimensional monitoring data to explore the application of visualization technology, from the time, node number, performance index type of three dimensions. Based on dimension compression and dimension tangent, a new method of performance data set visualization is proposed. Based on this, outlier node self-recognition is realized by using dynamic time planning and convolution neural network. It enriches and expands the semantics of the alarm system and proves that the method is feasible and can show the effective information more intuitively and improve the decision efficiency of the cloud administrator.
【作者單位】： 武漢大學(xué)空天信息安全與可信計(jì)算教育部重點(diǎn)實(shí)驗(yàn)室;武漢大學(xué)計(jì)算機(jī)學(xué)院;
【基金】：國家自然科學(xué)基金資助項(xiàng)目(U1536204,61373169) 國家科技支撐計(jì)劃項(xiàng)目(2014BAH41B00) 國家高技術(shù)研究發(fā)展(863)計(jì)劃項(xiàng)目(2015AA016004) 信息保障技術(shù)重點(diǎn)實(shí)驗(yàn)室開放基金資助項(xiàng)目(KJ-14-110,KJ-14-101)
【分類號(hào)】：TP277;TP393.08
【正文快照】： 0引言隨著云計(jì)算技術(shù)的發(fā)展,云平臺(tái)逐漸演進(jìn)為具有超大規(guī)模節(jié)點(diǎn),節(jié)點(diǎn)內(nèi)又包含海量的、層次服務(wù)復(fù)雜的網(wǎng)絡(luò)。從各類節(jié)點(diǎn)、服務(wù)中收集到的基礎(chǔ)數(shù)據(jù)通過云監(jiān)控系統(tǒng)匯總的監(jiān)控?cái)?shù)據(jù)集具有層次多、數(shù)據(jù)量大、數(shù)據(jù)類型復(fù)雜的特點(diǎn)[1]。隨著用戶數(shù)量的大幅提升和任務(wù)種類、規(guī)模及難度的，

本文編號(hào)：1463497