Genetic-based randomized network coding for dynamic satellite multicast

Abstract

The multicast rate of satellite communication networks (SCN) can be maximized by using randomized network coding (RNC). To optimize the RNC-based multicast for SCN, the number of coding links should be minimized to reduce the computational complexity on satellite, while the dynamic topology needs to be considered. To this end, in this paper, we propose an improved genetic algorithm (IGA) to minimize the number of coding links and provide successful multicast for dynamic network topology. In order to support dynamic multicast, the erasure protection (EP) is carried out with the IGA in fitness test process to select child generations with a better-fitness. It shows that the IGA method provides good performance for RNC-based multicast in dynamic SCN scenarios, where the number of coding links is minimized to reduce the complexity. Through simulation results, we can confirm that our proposed method outperforms the existing ones, where dynamic multicast is not considered.