Reinforcement Learning Based Real-Time Pricing in Open Cloud Markets

Abstract

Recently, with the rise in demand for reliable and economical cloud services, there is a rise in the number of cloud providers competing among each other. In such a competitive open market of multiple cloud providers, providers aim to model the selling prices of their requested resources in real-time to maximise their revenue. In this regard, there is a pressing need for an efficient real-time pricing mechanism, that effectively considers a change in the supply and demand of the resources in a certain open cloud market. In this research, we propose a reinforcement learning-based real-time pricing mechanism for dynamically modelling the prices of the requested resources. In specific, the proposed real-time pricing mechanism in a reverse-auction based resource allocation paradigm, which utilises the supply/demand of the resources and undisclosed preferences of the cloud users. Further, we compare the proposed approach with two state-of-the-art resource allocation approaches and the proposed approach outperforms the other two resource allocation approaches.