Research On Congestion Avoidance Multicast Routing Based On Swarm Intelligence For LEO Satellite Network

Low Earth Orbit(LEO)satellite network can achieve real global seamless coverage by constructing LEO satellite communication system with multiple LEO satellites.It has the outstanding advantages of low transmission loss,short communication delay,strong antidestruction ability and terminal miniaturization.Its high bandwidth characteristic makes it an ideal carrier for multicast service.One of the key technologies for satellite multicast communication is to construct a multicast tree with good performance.In order to transmit multiple multicasts,the consumption of network bandwidth should be minimized.However,there are network congestion in LEO satellite network,which will lead to the performance degradation of the whole network.The goal of existing multicast routing algorithms is usually to minimize the overhead or transmission delay of the multicast tree,and there is almost no optimal method that can achieve congestion avoidance and save network bandwidth at the same time.Based on the rectilinear Steiner minimum tree theory in the inter-satellite grid topology,this thesis propose a congestion avoidance multicast routing algorithm based on the ant colony and genetic joint optimization algorithm for LEO satellite network to achieve the goal of congestion avoidance and bandwidth saving.In this thesis,the LEO satellite network is modeled as a multi-hop network with a rectangular grid topology,and the multicast routing problem of congestion avoidance and bandwidth saving is transformed into the congestion-avoiding rectilinear Steiner minimum tree(CARSMT)problem.Considering that the CARSMT problem is NP hard,an ant colony optimization algorithm is designed to obtain its approximate optimal solution,and congestion-avoiding rectilinear Steiner minimum tree algorithm based on ant colony optimization algorithm(ACO-CARSMT)is proposed.The experimental results show that the ACO-CARSMT algorithm proposed in this thesis has obvious advantages over the traditional multicast routing algorithm in terms of computing time and bandwidth saving.In view of the lack of initial pheromone in the early stage of ant colony algorithm,which leads to the slow solution of the problem,considering the advantages of genetic algorithm with fast global search ability,a congestion-avoiding rectilinear Steiner minimum tree algorithm based on the combination of ant colony algorithm and genetic algorithm(GAACO-CARSMT)is proposed.The experimental results show that the GA-ACO-CARSMT algorithm proposed in this thesis not only has more advantages in computing time than ACO-CARSMT algorithm,but also improves the bandwidth saving performance by 0.08%to 0.62%,which proves that GA-ACO-CARSMT algorithm can not only achieve congestion avoidance,but also save network bandwidth and computing time.Compared with other algorithms,GA-ACO-CARSMT algorithm has obvious advantages.