A fast and scalable technique for constructing multicast routing trees with optimized quality of service using a firefly based genetic algorithm

We are seeing an explosive proliferation of multimedia content made available on the Internet. Multimedia applications have a multipartite nature where content has to be disseminated to multiple parties using group communication. Overheads involved in distributing the content through unicast can be overcome by using multicast mode of transmission. Many multimedia applications such as audio-video news streams, stock quotes, live conferences and online gaming require strict Quality-of-service (QoS) guarantees. Optimizing a multicast tree for multiple QoS constraints is a multi-objective NP-hard optimization problem. In this work we propose an optimization that uses modified Genetic Algorithm (GA), a branch of evolutionary computation, to determine near-optimal multicast trees to satisfy multiple QoS constraints such as bandwidth, delay and packet loss. Our modification uses the Firefly effect to reduce the convergence time as well the premature convergence of the GA. Our simulation results show that our proposed algorithm is capable of finding a set of near-optimal multicast trees in computationally feasible time within a few iterations and is much faster than other optimization techniques proposed in research literature. Moreover we show that the protocol is scalable and exhibits a linear increase in processing overhead with the increase in group size.