Fuzhou University

Wang Zeng

Junjie Zhang

Zheyi Chen

Sijin Huang

Zhiqin Huang

A computation offloading approach in blockchain-enabled MCS systems is provided to reach a lower total cost in computation offloading. Firstly, building a consortium blockchain-based framework to guarantee secure transactions in MCS systems. Secondly, designing a novel credit-based proof-of-work (C-PoW) algorithm instead of PoW to confirm transactions and add new blocks to the chain, thereby relieving the complexity of POW while keeping the reliability of blockchain. Thirdly, using a scalable deep reinforcement learning based computation offloading (DRCO) method to handle the computation-intensive tasks of C-PoW; by integrating PPO and DNC, the DRCO executes differentiable read-write operations on structured external memories by following an objective-oriented way; the DRCO uses a clipped surrogate objective to control the update of offloading policy, in order to improve the decision-making efficiency; the DRCO uses the DNNs to address the problem of high-dimensional state space.

Growing community of inventors

Fuzhou, China

Wang Zeng