Researches On Application Layer Multicast Strategy Based On Overlay

As the Internet grows up, the development of application requirements impels the production of new communication means, of which multicasting is one. At first, the multicast functionality which was a basic communication means was proposed to be added to the network layer. Due to a variety of technical and non-technical reasons, the deployment of IP Multicast in Internet has been limited and sparse. Therefore some researchers in the recent past have revisited the issue whether the network layer is the best layer for implementing multicast functionality and have proposed application layer multicast (ALM) as a new approach. The basic advantage of ALM is that it requires no infrastructure support and can be easily deployed in the Internet.With the detailed and in-depth research for application layer multicast, we found that there are a variety of problems about ALM in technology and applications, such as the intermediate nodes overloading, network bottlenecks and asynchronous and mobile users etc. Therefore, this paper analyzes the current several relatively popular ALM protocols and their classification. We also do a detailed evaluation for their performance, especially for their respective limitations . Meanwhile for several existing problems of the ALM on the overlay networks, we suppose some corresponding network optimization mechanisms :For the overloading problem of intermediate nodes in ALM networks, this paper analyzes the two important factors that impact the ALM node load in Chapter three: the number of multicast groups and the number of recipients of the group. And that combining the idea of Explicit Multicast (Xcast) respectively with the dynamic strategy and the greedy algorithm ,we propose two kinds of low load overlay-based ALM tree optimization strategy,namely LLDP\ALMT LLGP\ALMT. The analysis reveals that our mechanisms for easing nodes would enable network resources to be used more efficiently.In view of the bottleneck problem of application layer multicast , we research it from two aspects respectively ofr a general network and a heterogeneous one in chapter four and propose two kinds of mechanisms to eliminate bottlenecks in the ALM networks , namely BNFP\ALM and BNFP\HALM. Through analysis and evaluation we found that our mechanisms eliminate bottlenecks in the ALM networks effectively,and that they reduce the waste of network resources and improve the network's reliability and efficiency of data transmission.For the problem of asynchronous and mobile users on the overlay, we research it from two aspects respectively of the time-related of requestion from the users and the cache ability of overlay nodes in chapter five and propose an adaptive ALM algorithm based on asynchronous and mobile users on overlay to solve the problem, namely AALMP\AMU. This algorithm takes into account not only the ALM mobile receiver, which also takes into account the movement of data sources, and that members of the multicast tree nodes could create a logical multicast tree to transmit data to meet the mobile nodes, which try to avoid duplicate data and save network resources. The strategy has strong feasibility and effectiveness.Furthermore, in order to evalue the performance of our mechanisms,we use algorithm complexity analysis and network simulator NS-2 to conduct some detailed experiment and analysis from different perspectives of the proposed strategies.The simulation results indicate that our mechanisms are effective to improve the performance of application layer multicast, especially the low load dynamic programming for ALM Tree (LLDP\ALMT) and the adaptive ALM programming for asynchronism and mobile users(AALMP/AMU) which have shown satisfactional results in the data transmission efficiency and expansibility of networks etc. We hope that this paper will be able to play a certain role in promoting the research and development for application layer multicast based on overlay.