Adaptive Scheduling of Stochastic Task Sequence for Energy-Efficient Mobile Cloud Computing

Mobile cloud computing (MCC) is an attractive technique to improve energy efficiency and computing performance by offloading complex applications from resource-constrained mobile devices. This paper presents an adaptive scheme for scheduling randomly arriving heterogeneous tasks in MCC to minimize energy consumption while satisfying quality-of-service (QoS) requirements. A queuing with trigger mechanism is introduced to characterize the sequential dynamics of interdependent tasks. The energy-efficient QoS-guaranteed task scheduling is formulated as a constrained stochastic optimization problem based on continuous-time Markov decision processes. An adaptive policy iteration algorithm that combines potentials estimation and stochastic approximation is proposed to find the optimal scheduling policy online in unknown dynamic environments. Simulation results demonstrate the effectiveness of the proposed method.