ReputeStream: Mitigating Free-Riding through Reputation-Based Multi-Layer P2P Live Streaming

TL;DR

This work targets free-riding and churn in single-layer P2P live streaming by introducing a reputation-based multi-layer architecture. The framework combines joining/authentication, topology formation, and per-stream tables (RT/NT/BRT/OMT/BPT) with a reputation layer that updates via $R(t)=\frac{R(t-\\tau) e^{-\\alpha \\tau}+R_r L}{1+R_r L}$, where $R_r$ is the sender's reputation and $L$ the number of layers. It uses a beacon-driven overlay and a request-to-join mechanism to handle flash crowds and to position high-reputation peers near sources to improve efficiency. The analysis uses Nash equilibrium to show that altruistic peers are favored, free-riders are deterred, and the system stabilizes, enabling scalable, QoS-aware P2P live streaming under heterogeneous networks.

Abstract

This paper presents a novel algorithm for peer-to-peer (P2P) live streaming that addresses the limitations of single-layer systems through a multi-layered approach. The proposed solution adapts to diverse user capabilities and bandwidth conditions while tackling common P2P challenges such as free-riding, malicious behavior, churn, and flash crowds. By implementing a reputation-based system, the algorithm promotes fair resource sharing and active participation. The algorithm also incorporates a request-to-join mechanism to effectively manage flash crowds. In addition, a dynamic reputation system improves network efficiency by strategically positioning high-reputation peers closer to video sources or other significant contributors.

Figures (7)

• Figure 1: Layered encoding
• Figure 2: Flowchart of live streaming working
• Figure 3: Streaming Setup and Management for Source
• Figure 4: Streaming Setup and Management for Subscriber/forwarder
• Figure 5: Topology