Continuous data availability in instantaneous streaming networks

Being one of the important factors of evaluation in peer-to-peer streaming networks, availability of data is the point of our focus. Efficient data sharing in global peer-to-peer system is complicated by node failure, unreliable network connectivity and limited bandwidth. Using a caching system can improve availability of data. Yet, determining when and where to put the data in the cache to met performance goals in system with many users and files, dynamic network characteristics and changing users behavior, is difficult. We propose an approach to face these problems using a probabilistic model queue M/M/1/K and an optimization method called "Pareto method of dominance and optimality" to maximize the availability and access of data in the network. We evaluate the accuracy and performance of the proposed approach using simulations which showing the effectiveness of our approach.