[Paper Review] Fetching Strategy in the Startup Stage of p2p Live Streaming

Chunxi Li, Changjia Chen|ArXiv.org|Oct 12, 2008

Peer-to-Peer Network Technologies16 references3 citations

TL;DR

This paper proposes the Threshold Bipolar (TB) protocol, a fetching strategy for peer-to-peer (P2P) live streaming systems that improves startup performance by prioritizing chunk fetching from the buffer head initially and switching to tail fetching once a threshold is reached. The protocol achieves low startup latency, high download rates, and smooth handover between strategies by maintaining a contiguous playback buffer, as validated through traces from PPLive and a simplified design model.

ABSTRACT

A protocol named Threshold Bipolar (TB) is proposed as a fetching strategy at the startup stage of p2p live streaming systems. In this protocol, chunks are fetched consecutively from buffer head at the beginning. After the buffer is filled into a threshold, chunks at the buffer tail will be fetched first while keeping the contiguously filled part in the buffer above the threshold even when the buffer is drained at a playback rate. High download rate, small startup latency and natural strategy handover can be reached at the same time by this protocol. Important parameters in this protocol are identified. The buffer progress under this protocol is then expressed as piecewise lines specified by those parameters. Startup traces of peers measured from PPLive are studied to show the real performance of TB protocol in a real system. A simple design model of TB protocol is proposed to reveal important considerations in a practical design.

Motivation & Objective

• To address high startup latency and inefficient chunk fetching in P2P live streaming during the initial startup phase.
• To design a fetching strategy that balances high download rates with minimal startup delay.
• To enable a natural transition between fetching from the buffer head and tail without disrupting playback continuity.
• To identify key protocol parameters that influence buffer behavior and system performance.
• To validate the protocol’s effectiveness using real-world traces from the PPLive system.

Proposed method

• The TB protocol initiates chunk fetching from the buffer head to rapidly fill the initial portion of the buffer.
• Once the buffer reaches a predefined threshold, the strategy switches to fetching from the buffer tail to maintain a contiguous playback segment.
• The protocol ensures that the contiguous playback region remains above the threshold even during playback, preventing buffer underflow.
• Buffer progress is modeled as piecewise linear functions determined by key parameters such as threshold size, download rate, and playback rate.
• A simplified design model is proposed to guide practical implementation and highlight critical design trade-offs.
• Real-world startup traces from PPLive are analyzed to evaluate TB’s performance under actual network conditions.

Experimental results

Research questions

• RQ1 How can a P2P live streaming system minimize startup latency while maintaining high download rates during the initial phase?
• RQ2 What fetching strategy enables a smooth transition between initial buffer filling and sustained playback?
• RQ3 How can buffer continuity be preserved during playback while maximizing download efficiency?
• RQ4 Which system parameters most significantly impact buffer behavior and startup performance?
• RQ5 To what extent does the TB protocol outperform traditional fetching strategies in real-world deployment?

Key findings

• The TB protocol achieves low startup latency by prioritizing early buffer filling through head fetching.
• High download rates are maintained by dynamically switching to tail fetching once the threshold is reached.
• The protocol ensures a contiguous playback buffer above the threshold, preventing playback interruption even under variable network conditions.
• Real traces from PPLive confirm that TB reduces startup latency and improves buffer stability compared to conventional strategies.
• The piecewise linear buffer model accurately captures buffer dynamics and enables precise performance analysis.
• The simplified design model reveals key design considerations for implementing TB in real-world P2P streaming systems.