In order to improve the availability and reliability of data in P2P networks by using a dynamic replica management mechanism, a dynamic replica management mechanism based on trend prediction is proposed. The time series smoothing algorithm in economics is introduced to predict hot files, and three different strategies are used to dynamically manage operations such as placement, deletion, and replacement of replicas. Simulation experiments and analysis of relevant data have proved that this research can effectively improve the hit rate of resource search in P2P networks, reduce the overhead of the entire network, and achieve load balance among nodes in the network. Keywords: unstructured P2P; replica management; trend prediction; resource search; network performance
You Might Like
Recommended ContentMore
Open source project More
Popular Components
Searched by Users
Just Take a LookMore
Trending Downloads
Trending ArticlesMore