Tree based structured management of delivery path for the multi-path streaming

Recently, the real-time streaming application has gained great interest of engineers, because of the increasing requirements for transferring massive volumes data, such as high definition video and audio. Therefore, various architectures have been proposed to enhance the efficiency and quality of the transmission. The proposals based on the Application Layer Multicast (ALM) usually applying on the peer-to-peer protocol is one of those. The previous work of this paper designed a Multi-Path based ALM, which utilized the transmission ability of free peers existing in the overlay networks. The design focused on an unstructured overlay network, where peers are connected randomly. Therefore, the flooding query method, which checked transmission ability of each peer, is used to discover and construct the desired multiple delivery path. This unstructured management made the communication cost much higher than expected. This paper proposed a tree-based structured management way to provide efficient discovery and construction. Based on the K-tree model, this method is also available to promote the current ALM structures. The improvement of the management cost has been illustrated in our experiment with an instance of Multi-Path transmission, which is based on the well-developed project DVTS (Digital Video Transport System). The improved transmission speed could be seen from our result.