Viewing experience optimization for peer-to-peer streaming networks with credit-based incentive mechanisms

摘要

Effective in fighting against "free-riding" and stimulating the cooperation between peers, credit-based incentive mechanisms are widely adopted in today's peer-to-peer (P2P) streaming networks. This work considers a P2P multimedia streaming system that relies on credits for incentivizing peers to upload. The main problem of focus is to derive the optimal strategy for a peer, in terms of allocating its credits across different time slots, to maximize its long-term viewing experience. Especially, the dynamic changing feature of credits is taken into consideration when we formulate the problem, and the optimal credits allocation is shown to be a staircase-like function over time. Then, based on the characteristics of the optimal credits allocation strategy, an effective double-loop iterative algorithm is proposed. For the consideration of practical implementation, three low-complexity credits allocation strategies are proposed. It is shown that each of the strategies has its own feature and is suitable for a specific scenario. Then, as an extension, the proposed credits allocation schemes are reinvestigated for P2P streaming networks that adopt dynamic-pricing credits-based incentive mechanisms. It is shown that the previously obtained credits allocation strategies and algorithms can be easily applied to these systems with minor modifications. (C) 2017 Elsevier B.V. All rights reserved.