A Dynamic Bandwidth Pricing Mechanism For P2P Streaming Systems

P2P streaming systems are popular on the Internet.In P2 P streaming systems,peers share their bandwidth with each other and improve users' experience.To incentivize users share more bandwidth,credit-based incentive mechanisms are widely used: peers share bandwidth to gain credits and pay credit to buy bandwidth.Thus,to get bandwidth from the system,peers have to share more.A critical issue in credit-based mechanisms is to set reasonable prices for purchasing bandwidth.In this paper,we propose a robust dynamic bandwidth pricing mechanism based on Stackelberg game and trust management.A Stackelberg game is formulated to describe interactions and obtain peers' optimal bandwidth pricing and purchasing strategies.This mechanism can dynamically adjust the bandwidth price based on peers' trust values: in each transaction,the lower trust a peer has,the higher price the peer gets.Through theoretical analysis,we obtain peers' optimal pricing and purchasing strategies and an algorithm for obtaining optimal pricing strategies is proposed to fit in real systems.Simulation results show that our proposed mechanism can ensure system fairness and defend typical data pollution attack.By introducing risk evaluation,unsuccessful transaction rate can be obviously decreased.However,peers may send false information to bandwidth provider to disturb normal pricing mechanism and gain benefits from it.Such behaviors are called misreporting attack.To induce rational misreporting downloaders send true information,we introduce an auditing mechanism based on reinforcement learning: the auditing probability of each peer is dynamically adjusted according to the system environment,and transactions risks are reduced.Vice versa,misreporting peers adjust their misreporting rate according to bandwidth providers' behaviors.Extensive simulations show that our proposed mechanism can effectively separate malicious peers,reduce misreporting attack rate and reduce auditing cost.To conclude,this paper proposes a robust P2 P streaming bandwidth pricing mechanism which can assure system fairness,defend misreporting attack,separate malicious peers and improve system efficiency.