This paper proposes a new technique to estimate available bandwidth in the channel named as Stream of Octal packet pair dispersion (SOOPPD). Available bandwidth between two points in the Internet is generally unknown and the time varying. It is observed that if the sender transmits faster than the available bandwidth, then it may cause congestion, packet loss leading to a severe drop in video quality. If the sender transmits slower than the available bandwidth then the receiver produces sub-optimal video quality. This chapter aims to overcome the problem by estimating the available bandwidth and to match the sending video bit rate to the available bandwidth; accurate estimation of the available bandwidth is a challenging task taken up in this chapter. Various techniques used to estimate available bandwidth can be classified as, ‘passive measurement' and ‘active probing'. Passive measurement tools use the trace history of existing data transfers and it is observed that these are very efficient, accurate, and their scope is limited to network path that have recently carried user traffic. Active probing on the other hand could explore the entire network.
[...] Through the experimental results it is observed that, the available bandwidth estimated by SOOPPD is accurate than other bandwidth estimation techniques. This proposed method calculates dispersion at the client and this information is piggybacked to the server INTRODUCTION It is very important to monitor available bandwidth for quality of service(QoS) guarantees[1] of different types of Internet flows and various network applications like rate based streaming media transport, end-to-end admission control, server selection, optimal rate selection in overlay networks, congestion control in streaming media and service level agreement verification. [...]
[...] PROPOSED AVAILABLE BANDWIDTH ESTIMATION TECHNIQUE: STREAM OF OCTAL PACKET PAIR DISPERSION (SOOPPD) Let us suppose a network path p = L2, is a sequence of H inter connected FCFS, store-and forward hops. Two packets of the same size ‘L' are sent from sender ‘S' to the receiver ‘R'. These packets are called probing packets. The sequence of steps in packet pair sampling is a follows. Send back-to-back two probe packets faster than the capacity from sender ‘S' where ‘C' is the bottleneck capacity, we can measure capacity through UDP-ICMP messages or TCP-Fin or Cap probe tool 18] or any other method. [...]
[...] Once the cross traffic rate (CTr) is known it becomes easy to measure the available bandwidth L C AB = C CTr = LP Lp , where∆c = C Depa B L P 1 X t P2 Where ‘C' = capacity Lp = the size of the packet and CTr = the cross traffic But it is not that easy to calculate the cross traffic rates, because the cross traffic might be in the form of one hop persistent cross traffic, or path persistent cross traffic, consider the figure 8 and 9. [...]
[...] In this technique we take the two packets of equal size; because dispersion of unequal packet of a packet pair may not measure the narrow link capacity or available bandwidth accurately, consider the proof shown in figure 5 for this statement. ti P2 Size=L P1 Size=2L L 2L > C 3C Capacity = c P 2 Capacity = 3C P 1 = L 2L C C t1 + L L + L t C 3C1 + t1 + 2 L t C 3C 1 Figure 5. [...]
[...] Bottle neck of a channel The bottleneck link of the figure 2 can be identified through the following Algorithm: If (AB (S1 > AB Then the bottleneck link is S2 R Else If (AB (S1 [...]
APA Style reference
For your bibliographyOnline reading
with our online readerContent validated
by our reading committee
