Energy-Efficient Power Allocation with Imperfect CSI in MEC Networks

By introducing mobile edge computing (MEC) technology into the Internet of Things (Io'T), computing-intensive tasks can be offloaded to edge servers deployed at base stations (BS) for processing, thereby improving user quality of service. The conflict between low delay requirement and high power consumption have become key issue. This paper focuses on the energy-efficient resource allocation with considering delay constraints and imperfect CSI. The problem is formulated as a stochastic optimization model to optimize EE while satisfying delay, system queue stability, and peak power constraints. To solve the formulation, we propose a dynamic traffic admission control and resource allocation algorithm (DACRA) by employing fractional programming, Lyapunov optimization techniques, and the Lagrange dual decomposition method. Finally, its performance, which is further verified by simulation results.