Energy-Efficient for Multicast Networks: A New Approach to Efficiency Measure

This paper considers a new approach to energy efficiency in wired and wireless multicast networks. Two major multicasting algorithms are analyzed using energy-efficiency as metric. The performance of these algorithms is evaluated and optimized using combined methods: simulation and Data Envelopment Analysis (DEA) method. These approaches are combinations of methods from different disciplines. The first method has been considered for measuring network effectiveness through simulations, while the second method is based on data envelopment analysis for measuring network relative efficiency. This paper focuses on the DEA method, which is a non-parametric technique that is based on linear programming. The technique has been used in the Operations Research (OR) and economics communities to study the relative efficiency of systems with multiple inputs and outputs. Two types of DMUs are assumed for multicast networks. First, the network size is varied. and second the multicast group is varied. It is found that RLNC algorithm show better energy efficiency performance than MIP algorithm for all DMUs that are considered.