東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Queuing analysis and control of long...

Yao, Lei.

FindBook

Google Book

Amazon

博客來

Queuing analysis and control of long range dependent traffic: Applications to Internet traffic engineering.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Queuing analysis and control of long range dependent traffic: Applications to Internet traffic engineering./
作者:	Yao, Lei.
面頁冊數:	100 p.
附註:	Source: Dissertation Abstracts International, Volume: 64-01, Section: B, page: 0354.
Contained By:	Dissertation Abstracts International64-01B.
標題:	Engineering, Electronics and Electrical. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3077262
ISBN:	0493976701

Queuing analysis and control of long range dependent traffic: Applications to Internet traffic engineering.
Yao, Lei.

Queuing analysis and control of long range dependent traffic: Applications to Internet traffic engineering. - 100 p.

Source: Dissertation Abstracts International, Volume: 64-01, Section: B, page: 0354.

Thesis (D.Sc.)--The George Washington University, 2003.

The thesis presents a statistical analysis of Internet backbone traffic, based on traces with levels of aggregation 10 times larger and timestamp accuracy 1000 times better than in previous studies. The first three moments, marginal distributions and correlation structures of packet size, packet inter-arrival time, byte count and packet count, are studied. It is found that highly aggregated Internet backbone traffic is still long-range dependent (LRD) and self-similar (SS). In fact, all time series examined (packet size, inter-arrival time, byte count, packet count) exhibit long-range dependency and self-similarity. In addition to the now classical analysis at large time-scales (>100ms), the research reports the first statistically relevant results on the short-term correlation ([50μs, 10ms]) of byte and packet count processes. The fit of various analytical models to the traffic traces is also studied. The empirical queuing analysis (i.e., feeding a simulated queue with real traces) confirms the long-range dependence detected through direct analysis by showing that the queue behavior at high level of aggregation still diverges greatly from that predicted by Poisson model, but converges to that predicted by the fractional Brownian motion model (fBm). However, the traffic variability decreases as the level of aggregation increases, as shown by the decreased ratio of standard deviation over mean of byte and packet count processes, and by the lightening of the tails of their marginal distributions. With the less variable traffic, a low loss rate can be achieved using a buffer size in the order of milliseconds of link bandwidth.

ISBN: 0493976701Subjects--Topical Terms:

626636
Engineering, Electronics and Electrical.

Queuing analysis and control of long range dependent traffic: Applications to Internet traffic engineering.
LDR:03713nmm 2200313 4500 001 1857969
005 20040823111712.5
008 130614s2003 eng d
020 $a 0493976701
035 $a (UnM)AAI3077262
035 $a AAI3077262
040 $a UnM $c UnM
100 1 $a Yao, Lei. $3 1945678
245 1 0 $a Queuing analysis and control of long range dependent traffic: Applications to Internet traffic engineering.
300 $a 100 p.
500 $a Source: Dissertation Abstracts International, Volume: 64-01, Section: B, page: 0354.
500 $a Director: Milos Doroslovacki.
502 $a Thesis (D.Sc.)--The George Washington University, 2003.
520 $a The thesis presents a statistical analysis of Internet backbone traffic, based on traces with levels of aggregation 10 times larger and timestamp accuracy 1000 times better than in previous studies. The first three moments, marginal distributions and correlation structures of packet size, packet inter-arrival time, byte count and packet count, are studied. It is found that highly aggregated Internet backbone traffic is still long-range dependent (LRD) and self-similar (SS). In fact, all time series examined (packet size, inter-arrival time, byte count, packet count) exhibit long-range dependency and self-similarity. In addition to the now classical analysis at large time-scales (>100ms), the research reports the first statistically relevant results on the short-term correlation ([50μs, 10ms]) of byte and packet count processes. The fit of various analytical models to the traffic traces is also studied. The empirical queuing analysis (i.e., feeding a simulated queue with real traces) confirms the long-range dependence detected through direct analysis by showing that the queue behavior at high level of aggregation still diverges greatly from that predicted by Poisson model, but converges to that predicted by the fractional Brownian motion model (fBm). However, the traffic variability decreases as the level of aggregation increases, as shown by the decreased ratio of standard deviation over mean of byte and packet count processes, and by the lightening of the tails of their marginal distributions. With the less variable traffic, a low loss rate can be achieved using a buffer size in the order of milliseconds of link bandwidth.
520 $a Expedited Forwarding (EF) Per-Hop Behavior (PHB) in Differentiated Services (Diffserv) is designed to support voice and other real-time applications over the Internet. This thesis studies the steady-state queue overflow probabilities of priority and first-come-first-served (FCFS) queuing systems with hybrid inputs: the Poisson input process modeling voice and the fBm input process modeling general Internet traffic. It is shown that the two queuing systems can achieve almost the same maximum allowed link utilization on OC48 or higher-speed links. The result suggests that Diffserv may not be necessary on properly traffic engineered high-speed backbone links to support VoIP over Internet.
520 $a The thesis proposes a new approach to linear minimum-mean-square-error (MMSE) prediction of LRD traffic arrival process. The approach fits a discrete-time fBm (dt-fBm) model to a traffic arrival process.
520 $a Finally, the research proposes a prediction-based active queue management mechanism, prediction-based random early detection (<italic>PRED</italic>). (Abstract shortened by UMI.)
590 $a School code: 0075.
650 4 $a Engineering, Electronics and Electrical. $3 626636
650 4 $a Computer Science. $3 626642
690 $a 0544
690 $a 0984
710 2 0 $a The George Washington University. $3 1017405
773 0 $t Dissertation Abstracts International $g 64-01B.
790 1 0 $a Doroslovacki, Milos, $e advisor
790 $a 0075
791 $a D.Sc.
792 $a 2003
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3077262