Key research themes
1. How can peer selection policies be optimized in BitTorrent-like protocols to improve performance in interactive on-demand streaming scenarios?
This research area investigates the adaptation of BitTorrent peer selection algorithms to better suit interactive multimedia streaming contexts, where users have temporal and spatial access patterns that affect content access. Considering client interactivity is crucial to enhance streaming quality, reduce latency, and maintain scalability amidst diverse viewing behaviors.
2. What are the design trade-offs and protocol modifications needed to optimize BitTorrent performance and fairness in heterogeneous peer environments?
This theme addresses the inherent tension in BitTorrent protocol design between maximizing system-wide performance (i.e., minimizing average download times) and preserving fairness, where peers contributing more bandwidth receive better service. The heterogeneity of peer upload capacities necessitates strategies balancing incentive-compatible behaviors with efficiency, influencing piece selection, peer connection establishment, and flow management.
3. How can routing and content discovery mechanisms in peer-to-peer networks be optimized to reduce overhead and improve efficiency in BitTorrent-like systems?
This research direction focuses on the development of indexing and routing strategies that enable efficient content lookup in decentralized P2P systems without resorting to costly flooding or centralized indices. Effective routing indices and query forwarding heuristics can substantially reduce network bandwidth consumption while preserving high success rates in locating content, critical for the scalability of BitTorrent overlays with dynamically joining/leaving peers.