This paper describes new challenges in the use (design) of peer-to-peer (P2P) mesh streaming. It contextualizes and presents all the advantages and disadvantages of using P2P technology to download three-dimensional (3D) meshes. It explains the problems with using P2P in certain types of applications, and shows the reasons for using it in 3D mesh streaming.

In order to show the advantages and disadvantages of using the P2P approach, the paper compares and contrasts P2P mesh streaming, P2P video streaming, and P2P file download. All P2P techniques used are presented in detail.

The use of P2P technology is different for each type of application. In the case of the 3D mesh, there is a particular interest in chunks. Since each peer wants to see (receive) a certain part of the image, one peer may request information from another peer that never got that chunk; thus, the authors propose a chunk hierarchical structure based on a binary tree that embeds the vertex hierarchy--splitting the image in vertices allows for a better exploration of the use of chunks. This structure provides a way to find other peers that can provide the chunk.

Cheng, Liu, and Ooi present two programs in C++ based on OpenGL and OpenMesh (client/server), and use them in an experiment on images with five million vertices. They use the OMNeT++ simulator to perform a statistical analysis of the transmission of the 147,897 chunks generated.

The analysis of the results shows a 90 percent reduction in bandwidth of the server, compared to a client/server architecture. The authors present several graphs that illustrate the stability of this technology with a large number of online peers. This suggests that the P2P technology can and should be used in accordance with the advantages presented for this application; to support this, the paper includes a discussion on server overhead, response time, and control overhead.

The paper contains some graphs and a detailed discussion of the experimental results--the centralized lookup and hierarchical P2P lookup results. There is also a detailed discussion of P2P technology and progressive mesh streaming. The paper is suited for both beginners and senior researchers in the networking area.