Terminal Cooperative Interdependent Computing Task Offloading for 6G

Abstract

Towards the sixth generation (6G) mobile network, the increasingly powerful communication and computation capacity of mobile terminals enables distributed cooperative computation among nearby mobile terminals. Terminal cooperative computing is proposed to leverage the potential of numerous mobile terminals to promisingly support emerging applications and enhance the service performance of the existing computing network. One great obstacle to achieving high efficiency in the terminal cooperative network is that the neglect of task interdependency for simplicity leads to performance degradation of existing offloading optimization algorithms in practice applications. To address this issue, we study interdependent task offloading in a device-to-device (D2D)-enabled multi-helper terminal cooperative computing network in this paper. Facing the difficulty of exploring and deriving the offloading properties with the complicated coupling among tasks, we formulate a combinatorial optimization problem to minimize the weighted sum of latency and energy consumption under latency and energy constraints by jointly optimizing offloading selection, computing frequency control and transmission power control. We design a convex-based algorithm and derive closed-form solutions to this combinatorial optimization problem. The tradeoff between energy consumption and latency is studied. Simulation results demonstrate the performance of the proposed algorithm.