An Efficient Multicast Routing Scheme with Minimal Interference Path in Mobile Ad hoc Networks

• Main Authors: Kuan-yu Lin, 林冠佑
• Other Authors: San-Yuan Wang
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/56057878921047999271

碩士 === 義守大學 === 資訊工程學系碩士班 === 93 === Multicasting is a highly efficient and one-to-many communication method. It is usually implemented by setting up a multicasting tree or mesh. Because of the battery limited and transmission bandwidth constrained in the Mobile Ad Hoc Networks (MANETs), Multicasting can significantly improve the performance of this network. Since the topology is time-vary, the maintenance of multicasting tree which to ensure its availability will be a difficult work. This thesis improves the multicasting performance in MANETs by setting minimal Interference paths in multicasting tree or mesh. Unlike the tree or mesh schemes in wired network, the nodes of MANET edge compete the bandwidth for transmission even they do not belong to a mesh edge or tree edges of the multicasting tree or mesh. The interference characteristic leads time division effect of transmission and the effect will propagate down the multicasting tree or mesh. Hence, the multicasting performance can be proved by reducing the interference between subtree of multicasting tree or nodes near the source of the multicasting mesh. In this thesis, we use QualNet simulator to evaluate our idea and compare with ODMRP under vary node densities, mobility, number of groups, and number of group sizes. For simulation convenient and efficient, we also develop a user interface in QualNet for generating a large number of simulated result systematically to manifest the effects of simulation and efficient analysis. This AP will save much time for the need of producing a lot of samples and gathering statistic of result efficiently. Simulation results show that the proposed MIPMRP outperforms ODMRP while the working size (the number of group times group size) is small. The smaller the working size, the better the performance of MIPMRP is. When the working size is larger than the network size, our scheme has a tiny performance down since our path length of mesh is larger than that of ODMRP.