Research On Layered Multicast Transmission With User Cluster In Heterogeneous Network

How to achieve high speed video transmission is one of the research hotspots pursued by both industry and academia. On the one hand, improving video transmission rate is a significant impetus for the development of information service, on the other hand, service diversity and network heterogeneity put forward new requirements for the video transmission research. Layered multicast can primely meet the differential demands of users in heterogeneous network. Therefore, in order to improve the system throughput and bandwidth utilization respectively, the thesis mainly analyzes the layer rate optimization for layered multicast and the intra-cluster D2 D multicast retransmission algorithm based on the user cluster. The specific research contents are summarized as follows:1. Aiming at the fact that the problem of layer rate optimization for layered multicast in heterogeneous network is that of NP-Hard, the optimal layer rate allocation algorithm based on user cluster is proposed, in which layering is performed according to users' number and theoretical bandwidth, and each layer rate as well as its corresponding transmission sub-graph is determined. Afterwards, the link bandwidth is allocated for receivers of each layer. On the basis of reusing links allocated to others, this algorithm selects the path with the maximum node degree to perform reasonable bandwidth allocation, through which to reserve more resource for higher layer. Finally, combined with user clustering, multi-layers are merged into a scheduled layer to optimize the layer rate and allocate link bandwidth from the global perspective. Simulation results show that the proposed algorithm can effectively improve the system throughput, compared with the existing layer rate optimization algorithms.2. Aiming at the problem of lost packets recovery of intra-cluster user, an intra-cluster two-stage D2 D multicast retransmission algorithm is proposed. With full consideration of difference in D2 D link quality, this algorithm models the system link cost and employs the two-stage multicast retransmission to recover all of the lost packets of intra-cluster users, which reduces the probability of single multicast and extra signaling overheads caused by the increase of multicast times. Meanwhile, the average link cost is defined to assist in selecting the multicast sub-cluster and the repeated multicast is eliminated. Moreover, the user who has recovered its lost packets via one time multicast is converted into the alternative multicast transmitter to further enhance the system bandwidth efficiency. Simulation results show that the proposed intra-cluster two-stage multicast retransmission algorithm has superiority performance in the bandwidth efficiency over the conventional D2 D multicast retransmission algorithm.