A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic
碩士 === 國立中山大學 === 資訊工程研究所 === 87 === Internet Protocol (IP) conformed the ATM network, the complicated way is IP packet and the transform of ATM cell, while the IP packet segment to cells, and the cells reassemble to IP packet make the Internet Protocol link up with messages. Under contin...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
1999
|
Online Access: | http://ndltd.ncl.edu.tw/handle/04332639241440554711 |
id |
ndltd-TW-087NSYSU392012 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-087NSYSU3920122016-07-11T04:13:19Z http://ndltd.ncl.edu.tw/handle/04332639241440554711 A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic 新的ATM交換機緩衝區管理法以改善IP封包傳送良率 Mu-Rung Shiu 許木榮 碩士 國立中山大學 資訊工程研究所 87 Internet Protocol (IP) conformed the ATM network, the complicated way is IP packet and the transform of ATM cell, while the IP packet segment to cells, and the cells reassemble to IP packet make the Internet Protocol link up with messages. Under continual reassemble and segmentation, the completely received of packets have the closely relationship with cell in buffer allocation scheme of ATM backbone switches and how to control it and proceed to the next step to find out the relation of goodput of packet is the point of this research. In this paper, we will focus on shared-buffering schemes. As the multicast traffic increases and multimedia becomes popular, shared-buffering schemes are better choices. Researchers have studied the effects of Per-VC queueing. However, only considering queues of each VC is not enough. As we will show, the buffers needed should be also dependent on output rates. In addition, the number of VCs could be a lot and the management of them could be a problem. Instead, we suggest the idea of per-port queueing here. By per-port queueing, we can lower the complexity by managing ports as well as take into account the channel rate of each output port. By combing per-port queuing with packet discarding schemes, we show the goodput of ATM switches for different proposed schemes. The paper is organized as follows. In section II, we introduce advantages and disadvantages of dedicated-buffering and shared-buffering queueing, and different packet discarding schemes. Several shared-buffering schemes are proposed in section III. Simulation results and comparisons are discussed in section IV. Three different topologies are simulated. Section V is the conclusions. Wei -Kuang Lai 賴威光 1999 學位論文 ; thesis 54 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立中山大學 === 資訊工程研究所 === 87 === Internet Protocol (IP) conformed the ATM network, the complicated way is IP packet and the transform of ATM cell, while the IP packet segment to cells, and the cells reassemble to IP packet make the Internet Protocol link up with messages. Under continual reassemble and segmentation, the completely received of packets have the closely relationship with cell in buffer allocation scheme of ATM backbone switches and how to control it and proceed to the next step to find out the relation of goodput of packet is the point of this research.
In this paper, we will focus on shared-buffering schemes. As the multicast traffic increases and multimedia becomes popular, shared-buffering schemes are better choices. Researchers have studied the effects of Per-VC queueing. However, only considering queues of each VC is not enough. As we will show, the buffers needed should be also dependent on output rates. In addition, the number of VCs could be a lot and the management of them could be a problem. Instead, we suggest the idea of per-port queueing here. By per-port queueing, we can lower the complexity by managing ports as well as take into account the channel rate of each output port. By combing per-port queuing with packet discarding schemes, we show the goodput of ATM switches for different proposed schemes.
The paper is organized as follows. In section II, we introduce advantages and disadvantages of dedicated-buffering and shared-buffering queueing, and different packet discarding schemes. Several shared-buffering schemes are proposed in section III. Simulation results and comparisons are discussed in section IV. Three different topologies are simulated. Section V is the conclusions.
|
author2 |
Wei -Kuang Lai |
author_facet |
Wei -Kuang Lai Mu-Rung Shiu 許木榮 |
author |
Mu-Rung Shiu 許木榮 |
spellingShingle |
Mu-Rung Shiu 許木榮 A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic |
author_sort |
Mu-Rung Shiu |
title |
A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic |
title_short |
A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic |
title_full |
A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic |
title_fullStr |
A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic |
title_full_unstemmed |
A new shared-buffer scheme for goodputs of IP packets under ATM UBR traffic |
title_sort |
new shared-buffer scheme for goodputs of ip packets under atm ubr traffic |
publishDate |
1999 |
url |
http://ndltd.ncl.edu.tw/handle/04332639241440554711 |
work_keys_str_mv |
AT murungshiu anewsharedbufferschemeforgoodputsofippacketsunderatmubrtraffic AT xǔmùróng anewsharedbufferschemeforgoodputsofippacketsunderatmubrtraffic AT murungshiu xīndeatmjiāohuànjīhuǎnchōngqūguǎnlǐfǎyǐgǎishànipfēngbāochuánsòngliánglǜ AT xǔmùróng xīndeatmjiāohuànjīhuǎnchōngqūguǎnlǐfǎyǐgǎishànipfēngbāochuánsòngliánglǜ AT murungshiu newsharedbufferschemeforgoodputsofippacketsunderatmubrtraffic AT xǔmùróng newsharedbufferschemeforgoodputsofippacketsunderatmubrtraffic |
_version_ |
1718342423221895168 |