Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Computer network > General issues > Network equipment

Research on Site Selection of DNS Server Based on Partition-weighted Voronoi Diagrams

Author ZhangHaiLong
Tutor ChiLeJun
School Harbin Institute of Technology
Course Computer technology
Keywords DNS requests quantity DNS anomaly testing DNS server performance partition weighted V diagram weight
CLC TP393.05
Type Master's thesis
Year 2012
Downloads 10
Quotes 0
Download Dissertation

With the rapid development of Internet,the network has become an indispensablepart of people’s lives. However,network service performance and quality of servicebecome an issue of concern.Having such a huge user base,DNS service as animportant part of internet,performance and distribution of it play a key role in thequality of service of internet.To solve the problem above,this paper research the location of DNS service.Thepurpose is to lay DNS service at a more reasonable position to achieve the purposeof providing the best network service for all users.For the nearest division of theregion,we can through the use of the mathematical theory of weighted partition Vdiagram combined with ALA algorithm. To generate the partition weighted Vdiagram,we can statistic the amount of DNS requests and simulate the performancetesting of DNS server in order to get the parameter values which is needed incalculating the weight.For the statics of DNS requests,the inva lid DNS requests should be excludedfrom the normal DNS requests.we can detect DNS attacks by method of accordingto the flow of information through DNS service. We detect DNS attacks on all DNSrequest packet to filter out invalid DNS request packet from the normal DNS requestpackets.After that,the source IP address of normal DNS request packets should beclassified.For the performance testing of DNS service,the main purpose is to test themaximum number of DNS requests that DNS service can deal with in the case ofunexpected network traffic. We test the DNS server in both cases. After finishing thetest of both cases,we select the average value of the test results in both cases asultimate DNS server’s maximum processing capacity value.For the location algorithm of DNS server,we use the partition-weighted Vdigram theory in mathematics,and combine with the ALA algorithm and theminimum covering circle algorithm. We can get weight to generate the partitionweighted V diagram,and so on until the termination conditions meets the calibration.Finally,the DNS server location result obtained through the above algorithmcan be optimizated by the load moment criterion.

Related Dissertations
More Dissertations