Abstract
In a publish/subscribe network, message delivery is guaranteed for all active subscribers at publish time. However, in a dynamic scenario where users join and leave the network, a user may be interested in content published before the subscription time. In this paper, we introduce mechanisms that enable caching in such networks, while maintaining the main principle of loose-coupled and asynchronous communication. Furthermore we investigate two caching policies; caching in all candidate brokers (basic caching) which yields high survivability and low delay and caching in leaf brokers (leaf caching) which maintains low overhead and querying complexity. The comparison is performed via simulations and testbed measurements and insights are given for future work.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have