Title :
Network Topology Inference Based on Delay Variation
Author :
Hong-hua Zhao ; Ming Chen
Author_Institution :
Inst. of Command Autom., PLA Univ. of Sci. & Tech., Nanjing
Abstract :
Network topology could be inferred based on one way delay in unicast networks, but time synchronization was needed between source node and destination nodes when measuring one way delay. In order to reduce the limit of time synchronization, a measurement method named 4-packets train was proposed which did not need time synchronization and could measure delay variation. Then a topology inference algorithm based on delay variation was put forward. Second the feasibility and correctness of topology inference algorithm based on delay variation were analyzed theoretically. At last the topology inference algorithm and 4-packets train measurement method were validated through simulation tool of NS2, the results of simulation indicated that the topology inference algorithm based on delay variation and 4-packets train measurement could infer network topology efficiently and correctly.
Keywords :
telecommunication network topology; telecommunication traffic; 4-packets train measurement method; NS2 simulation tool; delay variation; network topology inference; time synchronization; unicast networks; Computer networks; Delay effects; Hamming distance; Inference algorithms; Multicast algorithms; Network topology; Probes; Time measurement; Tomography; Unicast; 4-packets train; Delay Variation; Topology Inference;
Conference_Titel :
Advanced Computer Control, 2009. ICACC '09. International Conference on
Conference_Location :
Singapore
Print_ISBN :
978-1-4244-3330-8
DOI :
10.1109/ICACC.2009.29