Efficient algorithms for real-time multicasting

Abstract

This thesis addresses the critical requirement of the total network cost and network delay of the various existing multicast tree algorithms. Existing multicast tree algorithms are static in nature where reconstructing or regenerating is required when a node is removed or added so as to maintain a consistent optimal network cost and network delay. In addition, the network traffic is disrupted whenever the multicast tree requires reconstruction. In this thesis, our proposed efficient dynamic algorithm for real-time multicast tree helps to alleviate the above two problems by its ability to handle dynamic multicast group without reconstructing the existing multicast tree and also maintain a consistent optimal network performance. This thesis also shows that our algorithm performs slightly better than some existing multicast tree algorithms in terms of total network cost and network delay in our simulation.