A user interest-aware server selection in hybrid CDN-P2P mesh-based architecture for VDO-on-demand streaming | |
| Author | Arnon Saengarunwong |
| Call Number | AIT Thesis no.ICT-18-01 |
| Subject(s) | Computer network architectures Streaming technology (Telecommunications) Peer-to-peer architecture (Computer networks) |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Information and Communication Technology, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | Thesis ; no. ICT-18-01 |
| Abstract | Video content takes most part of the traffic nowadays because of the evolution of the internet. Alongside with the video content, the number of video watchers is also increasing significantly due to the same reason. Hybrid Peer-to-Peer Content Delivery Network (CDNP2P) is one of the most promising system to satisfy the huge amount of video content demand. The server and neighbor selection play an important role in improving the performance of content delivery. However, most of the existing algorithms consider only load balancing and network proximity and do not take user interest characteristics into account. Therefore, this thesis proposes a user interest-aware server selection in Hybrid CDN-P2P mesh-based architecture for Video-on-Demand (VoD) streaming. The algorithm proposed in this thesis applies Vector Space Model (VSM) and K-means clustering algorithm to extract groups of interest in the video and developed a server and neighbor selection algorithm to assign a client to an optimal overlay which hosts for a group of neighbors with the same interest. The main objective is to enhance the delivery performance in P2P overlay and encourage peers to contribute more to the system. The performance of the algorithms is evaluated in terms of startup delay, discontinuity, distortion, server direct connections, server data bandwidth consumption and peer data bandwidth upload contribution. The results show that the algorithm proposed in this thesis provides better performance than load balancing and closest-based CDN-P2P regarding almost all of evaluation metrics. |
| Year | 2018 |
| Corresponding Series Added Entry | Asian Institute of Technology. Thesis ; no. ICT-18-01 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Information and Communication Technology (ICT) |
| Chairperson(s) | Teerapat Sanguankotchakorn; |
| Examination Committee(s) | Attaphongse Taparugssanagorn;Poompat Saengudomlert; |
| Scholarship Donor(s) | Royal Thai Government Fellowship; |
| Degree | Thesis (M. Eng.) - Asian Institute of Technology, 2018 |