Secure Computation Offloading in Blockchain based IoT Networks with Deep\n Reinforcement Learning

Abstract

For current and future Internet of Things (IoT) networks, mobile edge-cloud\ncomputation offloading (MECCO) has been regarded as a promising means to\nsupport delay-sensitive IoT applications. However, offloading mobile tasks to\nthe cloud is vulnerable to security issues due to malicious mobile devices\n(MDs). How to implement offloading to alleviate computation burdens at MDs\nwhile guaranteeing high security in mobile edge cloud is a challenging problem.\nIn this paper, we investigate simultaneously the security and computation\noffloading problems in a multi-user MECCO system with blockchain. First, to\nimprove the offloading security, we propose a trustworthy access control using\nblockchain, which can protect cloud resources against illegal offloading\nbehaviours. Then, to tackle the computation management of authorized MDs, we\nformulate a computation offloading problem by jointly optimizing the offloading\ndecisions, the allocation of computing resource and radio bandwidth, and smart\ncontract usage. This optimization problem aims to minimize the long-term system\ncosts of latency, energy consumption and smart contract fee among all MDs. To\nsolve the proposed offloading problem, we develop an advanced deep\nreinforcement learning algorithm using a double-dueling Q-network. Evaluation\nresults from real experiments and numerical simulations demonstrate the\nsignificant advantages of our scheme over existing approaches.\n