基于人工免疫的入侵檢測方法研究.pdf

1、Y1211479分類目：——uDC：——工學(xué)碩士學(xué)位論文密級：——編號：基于人工免疫的入侵檢測方法研究碩士研究生指導(dǎo)教師學(xué)科、專業(yè)學(xué)位論文主審人郭春陽張汝波教授計算機(jī)應(yīng)用技術(shù)李雪耀教授哈爾濱工程大學(xué)2007年5月哈爾濱下稃大學(xué)碩十學(xué)位論文AbstractWjthwideUSeofcomputernetworkpeoplearepayingmorcandmoreattentiontonetworksecurityproblemIntrus

2、iondetectionisallessentialcomponentofinformationsecurityprotectionsystem，whichguaranteesthesecurityofthenetworkresourceneproblemofintmsiondetectionmayboildowntothedifferentiationbetweenthevalidauthorizeduseofcomputersySt

3、em(sel0andtheInvalidauthorizedu∞ofcomputersystem(nonself)whichissimilartoimmunesystemInrecentyearstheresearcherswhoapplythemselvestointrusiondetectionstudyhavefocusedonintrusiondetectiontechnologybasedonimmunology，whosep

4、rominentcharacteristicistodetectintrusionbyimmunesystemtheories，architectureandalgorithmsThemainshortcomingsofcurrentintrusiondetectionstudyIncludepoorexpansibility，pooradaptabilityandIncapabilityofdetectingunknownattack

5、sAimingatthesedeficiencies，thispaperappliesnegativeselectionalgorithminimmunesystemtointrusiondetectionfieldInthispaper，anewmethodbasedonnegativeselectionisstudiedtodetectintrusionThismethodincludesthreepaas：(1)Definea“s

6、elf’set；(2)Initializethesetofvaliddetectorswhoseelementcannotmatchanyelementofthe‘‘self’set；0)Usethevaliddetectorsinitializedabovetodetectintrusion，Becausethemethodproposedinthispaperapplystheideaofthecompositedetection，

7、ithastheadvantagesofMisuseDetectionandAnomalyDetectionTheexperimentresultsshowthatthemethodproposedinthispaperhashighTreePositiveRateandlowFalsePositiveRate，isabletodetecttheunknownintrusionwhosetypeisknownwithpertinency