dc.description.abstract |
Mobile Ad-hoc networking (MANET) is an evolving field of wireless networks where nodes form a temporary network without any central administration. To establish communication with a distant node in a multi-hop fashion, intermediate nodes must work as router to forward data packets for other nodes. A routing protocol is therefore needed to find routes for end-to-end communication. The topology of the network changes frequently and unpredictably due to high mobility of nodes, hence making routing a very challenging task. Typically MANETs were supposed to be a closed group of nodes communicating with each other. The group members were well-defined and pre-registered; new nodes may not be allowed to join the network due to strict group management. However the situation has changed and MANETs are getting popular in open groups e.g. VANETs and local area social networks(Stieglitz & Fuchß, 2011).Now any node can join or leave the network without permission. Much of the research work has been done for close group MANETs while a little has been done for the open group MANETs.
While unicasting (one-to-one communication) and broadcasting (one-to-all communication) are two extreme modes of communications, but often a subgroup of nodes is the only target of a communication, e.g. teleconferencing or live media streaming, such communication is termed as Multicasting. An efficient multicast protocol tries to deliver data to all the target recipients with a minimum load on the network. This is achieved by establishing a distribution structure that avoids unnecessary data duplication.
Several routing protocols have been proposed to achieve multicasting in MANETs. However these protocols are typically designed for many-to-many- teleconferencing type multicast applications and neglect the need of one-to-many TV/Radio streaming media type multicast applications. Further for open-groups, it is difficult for these protocols to maintain the group membership for such a highly volatile environment and hence result in degraded performance.
In literature, only SLIM protocol was proposed to cater this need. However, we identified two shortcomings of SLIM -viz. (1) its dependence upon the underlying unicast routing protocol and (2) lack of advertisement mechanism that makes the new comers aware about the availability of a live media stream - a much needed feature in the open groups. In this thesis, we propose SLIM+ as an improvement that promises to fulfill the said gaps, implement it in NS2 simulator and evaluate its performance in comparison with the contemporary multicast routing protocols like PUMA and MAODV.
The simulation results show that SLIM+ is indeed scalable, lightweight and simple protocol that outperforms its counterparts with respect to all the evaluated QoS parameters particularly PDR was observed above 75% and NRL was observed below 5%. |
en_US |