摘要
Application layer multicast routing is a multiobjective optimization problem.Three routing constraints,tree’s cost,tree’s balance and network layer load distribution are analyzed in this paper.The three fitness functions are used to evaluate a multicast tree on the three indexes respectively and one general fitness function is generated.A novel approach based on genetic algorithms is proposed.Numerical simulations show that,compared with geometrical routing rules,the proposed algorithm improve all three indexes,especially on cost and network layer load distribution indexes.
基金
supported by the National Natural Science Foundation of China (Grant No.60432030).
