Liquidstream II - scalable P2P overlay optimization with adaptive minimal server assistance for stable and efficient video on demand

Abstract

In recent years, a number of research efforts have focused on effective and stable P2P architectures aiming at large scale and low bandwidth cost, real time video streaming systems. In our work, we consider a BitTorrent-like VoD system and try to answer the following questions: (1) how we can dynamically manage a P2P overlay using distributed-scalable algorithms in order to exploit the available upload bandwidth of highly heterogeneous participating peers and (2) how media servers (cloud) can minimize the amount of the bandwidth they offer while ensuring uninterrupted video playback. To illustrate the success of our approach, we have developed a network level P2P VoD packet simulator for demonstrating the efficiency, scalability and stability of our work under variant and dynamic conditions.

Distributed balancer of mobile nodes' bandwidth and battery resources in P2P VoD systems

“…The peer discovery issue has been widely studied in our previous work [7], [4]. We developed a distributed gossip protocol for peer discovery, which enables the peers to be informed about other peers with low control overhead.…”

“…As in each iteration of the Algorithm 1 the worst incoming neighbor i is exchanged with the best candidate j, the ratio u(i)/|Out(i)| becomes higher and the u(j)/|Out(j)| lower. This cost function was developed in our previous work [7], [4] and it was proven that the ratio u(j)/|Out(j)| converges to a mean value. In Eq.…”

“…Peers with scarce bandwidth resources straggle to serve the requests and often are not able to do so. In our previous work [4] we successfully tackled the heterogeneity issue by targeting requests to the neighbor with shortest serving queue proportional to its upload bandwidth (qs(i)/u(i)).…”

“…Therefore, the senders need to evolve their serving queue's algorithm and prioritize the requests. In all recent works [4], [13] the sender serves the requests in its outgoing queue based on the deadline of every block (Earliest Deadline First), while those requests that are bound to miss their deadlines are dropped. Media server(s) are only being contacted to provide some urgent blocks.…”

“…Small blocks are more easily diffused, however they increase significantly the control overhead of the system. A more detailed analysis for the block size, as well as the factors that affect the DBTS is presented in our previous work [7], [4]. Peers enter the system in a uniform arrival rate and after a short buffering period, they start viewing the video from its beginning.…”