Deep Reinforcement Learning for Blockchain-Enabled Mobile Edge Computing Systems

Abstract

With the rapid development of mobile edge computing (MEC), users' data can be shared anywhere at any time. However, data privacy in MEC systems cannot be ensured, which impedes the growth of MEC. Blockchain is proposed as a promising technology, which is suitable for guaranteeing the security and traceability of data sharing. Nonetheless, existing blockchain schemes do not apply to MEC systems due to the dynamic characteristics of wireless channels and traffic loads. In this paper, we propose a secure data sharing framework for the blockchain-enabled MEC system. To meet the privacy requirements of different users, we present an adaptive privacy-preserving mechanism according to the available resources of the system and the level of users' privacy. A secure data sharing scheme is introduced to maximize the performance of the system in terms of the decreased energy consumption of the MEC system and the increased throughput of the blockchain system by employing deep reinforcement learning. Numerical results demonstrate the advantages of the proposed secure data sharing scheme in MEC systems.