TY - CHAP
T1 - Packet analysis in congested networks
AU - Fukushima, Masaki
AU - Goto, Shigeki
PY - 2002
Y1 - 2002
N2 - This paper proposes new methods of measuring the Internet traffic. These are useful to analysing the network status, especially when the traffic is heavy, i.e. the network is congested. Our first method realizes a light weight measurement which counts only TCP flags, which occupies 6 bits in a TCP packet. Based on the simple flag counts, we can tell whether the network is congested or not. Moreover, we can estimate the average throughput of a network connection based on the flag count. Our second method analyses a sequence of TCP packets based on an automaton, or a protocol machine. The original automaton has been used in the formal specification of TCP protocol. However, it is not applicable to the real Internet traffic. We have improved the automaton in various ways, and established a modified machine. Using the new machine, we can analyse the Internet traffic even if there are packet losses.
AB - This paper proposes new methods of measuring the Internet traffic. These are useful to analysing the network status, especially when the traffic is heavy, i.e. the network is congested. Our first method realizes a light weight measurement which counts only TCP flags, which occupies 6 bits in a TCP packet. Based on the simple flag counts, we can tell whether the network is congested or not. Moreover, we can estimate the average throughput of a network connection based on the flag count. Our second method analyses a sequence of TCP packets based on an automaton, or a protocol machine. The original automaton has been used in the formal specification of TCP protocol. However, it is not applicable to the real Internet traffic. We have improved the automaton in various ways, and established a modified machine. Using the new machine, we can analyse the Internet traffic even if there are packet losses.
UR - http://www.scopus.com/inward/record.url?scp=23044533726&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=23044533726&partnerID=8YFLogxK
M3 - Chapter
AN - SCOPUS:23044533726
SN - 3540433384
SN - 9783540433385
VL - 2281
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 600
EP - 615
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ER -