樹狀結構為基礎之行動隨意網路IP位址分配機制; A tree-based IP assignment scheme in mobile ad hoc network

NCU Institutional Repository > 管理學院 > 資訊管理研究所 > 博碩士論文 > Item 987654321/13294

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/13294

题名:	樹狀結構為基礎之行動隨意網路IP位址分配機制;A tree-based IP assignment scheme in mobile ad hoc network
作者:	吳文志;Wen-zhi Wu
贡献者:	資訊管理研究所
关键词:	位址分配;行動隨意網路;ip assignment;mobile ad hoc network
日期:	2008-03-27
上传时间:	2009-09-22 15:28:19 (UTC+8)
出版者:	國立中央大學圖書館
摘要:	行動隨意網路為一個多點跳躍的網路，因此裝置間需要時常通訊。在有線網路中裝置可透過DHCP自動取得獨一無二的IP位址，然而行動隨意網路是臨時且隨意形成，因此IP位址分配工作無法由特定單一裝置負責。有鑑於此，本研究設計出一套適用於行動隨意網路中的IP位址分配機制，使所有節點進入網路後均可獲得一個獨一無二的IP位址。目前行動隨意網路上IP位址分配機制可分為兩類：氾濫式與位址池，本論文以位址池方式為主。在本論文中，只有部份節點可分配IP位址(稱之為分配者)，而無法分配IP位址之節點，彼此間可透過週期性廣播得知目前最近分配者資訊，因此當新節點進入網路後便可使用較少的頻寬資源且不必浪費過多的時間來取得IP位址。行動隨意網路之拓撲是持續變動，使得網路分離與合併的情況時常出現，因而發生IP位址遺失與衝突，本研究也針對網路分離與合併設計出一套偵測方式，並且解決節點IP位址遺失與衝突情況的發生。在進行網路與合併分離時需要確認目前網路IP位址使用狀況，以往的研究是節點逐一回報IP位址資訊給合併/分離者，本機制利用氾濫訊息建立位址回報樹，因此可減少節點回報IP位址資訊所造成之流量。本研究透過ns2模擬証明所提出機制之可行性，並且分析不同參數設定值對機制效能所造成之影響，進而從中歸納機制之優缺點與可能改進之處，作為未來改進方向。 Mobile Ad-hoc Network (MANET) is a multi-hop network for which the communication between mobile nodes is persistent. The IP addresses can be assigned automatically by DHCP server in the wired network. However, it cannot be applied to MANET because of its distributed and dynamic nature. In this paper, we propose a tree-based IP assignment scheme which assigns a unique IP address for each node in MANET. There are two ways of IP assignment scheme for MANET, flooding based and address-pool based. The presented scheme is based on an address-pool way in which only some of the nodes are assigned as allocators which are responsible for assigning an IP address to a new node in the network. Mobile nodes exchange hello messages to find the closest allocators, hence new incoming node can occupy less bandwidth resource to obtain an IP address quickly. Because of the rapid change and unpredictability of MANET topology, networks merging and partitioning are occurred frequently. The IP addresses of the nodes may be lost or collide when networks merge or partition. We design methods for detecting and solving such problems. In previous IP assignment schemes, nodes report their IP addresses information to the node which detects networks merging and partitioning through unicast. In our scheme, we design an address-reported-tree in order to reduce the communication overhead occurred by all nodes reporting their IP address at the same time. In this paper, we use ns-2 simulators to evaluate our scheme by modifying various parameters. According to the simulation results, the advantages and disadvantages of our scheme are discussed, and the future works of our scheme is figured out.
显示于类别:	[資訊管理研究所] 博碩士論文

文件中的档案:

档案	大小	格式	浏览次数

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....