Research On Multicast Routing Algorithm Of LEO Satellite Networks Based On Subtree Connection

In recent years,video traffics show an exponential growth.It is critical to employ multicast to achieve bandwidth multiplexing and save video traffics.Compared with high orbit satellite TV and traditional content distribution network(CDN),Network layer multicast has lower cost and larger scale than satellite TV,and can greatly save the unicast traffics of CDN with higher efficiency and no additional operation and maintenance costs,which is obviously more advanced.Network layer multicast cannot be implemented across multiple Internet Service Provider(ISP)networks,and it is difficult to give full play to the advantages of network layer multicast.As a promising civil and commercial network,low earth orbit(LEO)satellite networks are expected to achieve excellent video distribution.Due to the limitation of technology and cost in this field,the mature massive and low-cost LEO satellite networks have not been established in the past.Thus multicast routing in inter satellite links(ISL)have not been paid attention in the existing studies.However,with the continuous development of space technologies,various LEO satellite constellation projects are recently proposed by groups and enterprises at home and abroad.It is highly urgent to investigate multicasting in LEO satellite networks.Although Network layer multicast is hard to be widely deployed in the conventional Internet,it fits well with the LEO satellite networks,and there are a lot of gaps in the research of multicast routing on satellite networks.Therefore,applying network layer multicast to achieve high-performance multicasting in LEO satellite networks is proposed in this thesis.we focus on congestion in network,and design more advanced and efficient multicast routing algorithms.Based on the minimal rectilinear Steiner trees and with the core idea of subtree connection,we investigate multicast tree construction to satisfy the routing requirements of inter satellite network,and creatively propose two efficient congestion-aware multicast routing algorithms.We first propose a novel congestion-aware minimal rectilinear steiner tree(CA-MRST)which approachs the global optimal solution gradually with the local optimal solution and effectively saves bandwidth.Secondly,we further optimizes CA-MRST algorithm,and propose another congestion-aware minimal rectilinear steinertree algorithm based on ant colony algorithm(CA-ACO-MRST)which solves many potential challenges in the process of finding the best treelength of each subtree.Experimental results show that compared with the control groups,bandwidth savings performance of CA-MRST canincrease by at least 8%and up to 2 to 4 times.CA-MRST sacrifices a certain amount of algorithm running time in exchange for excellent performance in bandwidth saving.However,CA-ACO-MRST is more adaptive to small-scale multicast groups.When the multicast group sizesare less than 20,CA-ACO-MRST can save 6% more bandwidth than CA-MRST on average.