Article Preview
Top2. Background
The multi-objective multicast routing problem is a topical issue that has been the subject of very important research, this kind of optimization problems has attracted the attention of researchers in computer communication and operational research. Several meta-heuristics have been developed around this topic including Genetic Algorithms, Tabu Search, Variable Neighborhood Search, Simulated Annealing, Immune Systems, Particle Swarm Optimization, Ant Colonies…
Cui et al. (Cui, Lin, & Wei, 2003) proposed a routing approach based on a genetic algorithm and Pareto dominance, their approach optimizes three qualities of service parameters. The simulations show that the used modeling can found a set of non-dominated multicast trees with a good quality to approximate the Pareto front.
Crichingo and Barán (Crichigno & Barán, 2004) proposed a new multicast algorithm based on the SPEA algorithm optimizing simultaneously the cost of the tree, the maximum end-to-end delay, and the average delay. They used an evolutionary population (P) and a Pareto set (Pnd), their proposed algorithm starts with a random population (P), individuals evolve to optimal solutions, and these are included in (Pnd).
They proposed also other multi-objective multicast routing algorithm (MMA) based on SPEA that simultaneously optimizes the maximum link utilization, the cost of the multicast tree, the end-to-end delay, and the average time (Crichigno & Barán, 2004). In the MMA1 algorithm, they used the binary tournament selection; in the MMA2 algorithm, they used the roulette selection.
Koyama et al. (Koyama, Barolli, Matsumoto, & Apduhan, 2004) have proposed a multi-objective genetic algorithm optimizing the cost and delay of the routing tree, their proposed approach is a source-based routing method and has a flexible and adaptive behavior. The experimentations demonstrate the performance of their proposed solution
Donoso et al. (Donoso, Fabregat, & Marzo, 2004) proposed a new a multi-objective traffic engineering scheme using different distribution trees to multicast several flows by combining into a single aggregated metric the different objective functions of multicast routing.
Fabregat et al. (Fabregat, Donoso, Baran, Solano, & Marzo, 2005) proposed a new traffic engineering load balancing taxonomy, they developed a new algorithm inspired by the SPEA Algorithm.