Energy-Aware Dynamic Computation Offloading in High-Speed Railway Networks with D-TDD

摘要

This paper investigates the energy-aware dynamic resource allocation and computation offloading in high-speed railway networks with dynamic time division duplex (D-TDD), where all tasks of passengers in the high-mobility train are aggregated to be a "big task", and the vehicle server (VS) deployed on the train completes the "big task" by two processing modes, i.e., local computing and vehicular edge offloading. For such a system, an optimization problem is formulated to minimize the energy consumption by jointly optimizing the size of task block, the offloading ratio, the transmit power and the central processing unit (CPU) speed of VS with multiple system constraints. As the problem is non-convex with unknown solution, a bi-level optimization method is proposed to solve it. Via one-dimensional searching and Lagrangian multiplier method, the global optimal solution is obtained with high computational complexity. Then, a suboptimal approach is proposed with low computational complexity. Numerical results show that the performance of our proposed scheme significantly outperforms some other benchmark schemes, and a great performance gain is obtained via employing D-TDD. Besides, the effect of Doppler shift on the system performance is also discussed.