Mesh or Multiple-Tree: A Comparative Study of Live P2P Streaming Approaches

Abstract

Abstract-Existing approaches to P2P streaming can be divided into two general classes: (i) tree-based approaches use pushbased content delivery over multiple tree-shaped overlays, and (ii) mesh-based approaches use swarming content delivery over a randomly connected mesh. Previous studies have often focused on a particular P2P streaming mechanism and no comparison between these two classes has been conducted. In this paper, we compare and contrast the performance of representative protocols from each class using simulations. We identify the similarities and differences between these two approaches. Furthermore, we separately examine the behavior of content delivery and overlay construction mechanisms for both approaches in static and dynamic scenarios. Our results indicate that the meshbased approach consistently exhibits a superior performance over the tree-based approach. We also show that the main factors attributing in the inferior performance of the tree-based approach are (i) the static mapping of content to a particular tree, and (ii) the placement of each peer as an internal node in one tree and as a leaf in all other trees. I. INTRODUCTIONUsing Peer-to-Peer overlay has become an increasingly popular approach for streaming live media over the Internet due to its potential scalability and ease of deployment. This approach is generally referred to as P2P streaming. In P2P streaming, participating end-systems (or peers) actively contribute their resources (mainly outgoing bandwidth) by forwarding their available content to their connected peers. Since the aggregate available resources in this approach organically grow with the user population, this approach can potentially scale with the number of participating peers in a session.Existing approaches for live P2P streaming can be generally divided into two classes: tree-based approaches and meshbased approaches. The tree-based P2P streaming approach expands on the idea of end-system multicast [1] by organizing participating peers into multiple diverse trees. Then, each description of a Multiple Description Coded (MDC) content is pushed through a separate tree [2], [3]. The mesh-based P2P streaming approach is inspired by file swarming mechanisms (such as BitTorrent) where participating peers form a randomly connected mesh and employ a swarming content delivery mechanism over a recent window of content [4]. However, the limited availability of new content in live streaming coupled with the notion of "quality" for the delivered stream (i.e., number of descriptions) introduce new dimensions in the design and evaluations of mesh-based P2P streaming. Most of the previous studies on P2P streaming have focused on a particular mechanism and evaluated certain aspects of its

Enabling broadcast of user-generated live video without servers

“…This largely reduces the chunk delays and improves distribution efficiency at the base layer. The collection of edges used for the delivery of a single chunk from the source to all participating peers form a source-rooted delivery tree [21] in either mesh-based or tree-based approaches. Suppose the source and all peers have homogeneous bandwidth of u = 2r, each node can have two children in the chunk delivery tree.…”

Video Streaming with Interactive Pan/Tilt/Zoom

“…P2P streaming systems can be broadly classified into mesh-pull vs. tree-push systems [43]. The design of mesh-pull systems evolved from P2P file-sharing systems.…”

On the impact of playout scheduling on the performance of peer-to-peer live streaming

“…A pair of peers (v i , v j ) is deemed compatible when each of them may serve as a parent peer for the other without the client peer experiencing any discontinuity or loss (see Fig.5) given the following assumptions: (1) connection to a parent peer is instant, (2) the parent peer is supplied with new frames at least at the nominal rate, and (3) the available bandwidth between client and parent peers is at least equal to the nominal video rate . It is clear that due to these assumptions A is an upper bound of the real availability that can exist in practice.…”

“…is the "right-one" has been going-on for some time, and although it seems that recent mesh-based systems using coding have several advantages, 1 it all depends in the end on the assumed operating environment and the desired cost/complexity of building and maintaining the system: well-behaving environments (e.g., dedicated cable networks) can benefit from the simplicity/economy offered by tree-based distribution; uncontrolled/variable environments (e.g., under-provisioned parts of the current Internet) can benefit from the redundancy offered by mesh topologies and coding.…”