The Coverage Model Of Nodes’Relative Leaving Probability Of Application Layer Multicast Based On NICE And Load Balancing Mechanism

In recent years, along with the Internet users and Internet penetration are improving quickly, the new media industry is also developing rapidly in the world, especially the video business. Because the data transmission of video business is large, so it needs amounts of network bandwidth, low delay and good stability, and it gives the bigger challenges to the quality of network communication and terminal handling capacity. However, the data transmission of the traditional unicast is one-to-one, it can’t satisfy with the multimedia data transmission obviously, so the researchers promote the technology of IP multicast. The early IP multicast needs the special routers, so its extensibility and practical scope are restricted, the application layer multicast is produced in this environment.Application layer multicast uses the hosts to do these functions such as forward data packet, replication, and management instead of traditional routers. And the performance differences of these hosts is great, when we use the poor hosts as Intermediate nodes to forward data may affects the reliability, latency, scalability and transmission efficiency of the system. How to construct the optimal model of the application layer multicast to meet the real-time multimedia transmission has been a hot spot of the research.This paper mainly researches model building of the current application layer multicast and load balancing mechanism, and we aims at the deficiency of the present study to do some improving. Then, basing on these factors, we propose an application layer multicast algorithm (NRLP-NICE) concerning the probability of the nodes’ relative leaving which bases on NICE and a mechanism (ALBS-SOCS) named adaptive load balancing strategy based on score of customer satisfactionFirst of all, this paper analyzes the distribution of the rule of the nodes’ session time in the application layer multicast. It combines the idea of hierarchical clustering of NICE with rule of the nodes’ session time to propose NRLP-NICE. After simulating algorithm by the framework of Omnet++4.0and comparing with NICE, the results show that NRLP-NICE has low control cost, at the same time, reduces the percentage of delay error.At second, this paper researches the imbalance of the overall system utilization characteristics caused by the differences of hardware and software applications in the actual life due to factors and proposes the concept of average MUS of based on the system of quality of service, bandwidth utilization and channel capacity. ALBS-SOCS is proposed base on these concepts. We simulate ALBS-SOCS with the same environment to NRLP-NICE. Comparing with NICE, the results show that ALBS-SOCS has low average transmission delay, low average link pressure, and high average MUS. ALBS-SOCS improves the overall system performance and gives the better experience to users.