Enhanced Smart Home Architecture using Deep Reinforcement Learning and Blockchain

Abstract

The Advent of the technology has raised the living standard of the society and we have incorporated smart home appliances in to our lives. These smart devices communicate with each other to help us ease the household demands of the residents. These appliances can be in public, or a private network and they share crucial data and information which further helps the residents of a Smart Home. This data management among devices becomes challenging because they are all implemented on a centralized mechanism which is more prone to attack and single point of breakdown. So, with such centralized architectures, service availability can never be assured. Technologies such as combination of Machine learning, Deep Reinforcement learning and Blockchain can help reduce this risk and further helps managing these appliances. This paper will explore the scope of Deep Reinforcement and Blockchain technique in smart home application. We have designed, implemented, and assess the Deep Reinforcement learning technique and Blockchain for managing appliances of smart homes. Furthermore, we evaluate the effectiveness of our approach with that of the conventional system of smart home applications.