Optimizing Energy Efficiency in Heterogeneous Task-Oriented IRS-Aided Wireless-Powered Mobile Edge Computing Systems

The integration of mobile edge computing (MEC) and wireless power transfer (WPT) holds significant promise, providing a robust solution to address the limitations imposed by the computing and energy resources in wireless devices (WDs) operating within various low-power networks. In this context, intelligent reflecting surface (IRS) technology emerges as a noteworthy communication innovation that not only conserves energy but also optimizes spectrum resources. With the aid of IRS, wireless-powered MEC systems can considerably enhance the efficiency of radio frequency (RF) energy transmission while simultaneously improving wireless information transmission performance. This paper delves into the long-term energy efficiency of IRS-aided multiuser wireless-powered MEC systems. Recognizing the stochastic nature of task arrivals, we introduce a stochastic optimization problem aimed at addressing the energy efficiency challenge. This problem’s objective is to optimize the system’s long-term energy efficiency while adhering to constraints related to network stability, energy stability, task offloading policies, IRS phase shift vectors, device central processing unit (CPU) frequencies, transmission power, and temporal causality. Subsequently, we transform this long-term optimization problem into a short-term deterministic problem using Lyapunov optimization theory. However, the transformed problem remains highly coupled and involves discrete variables. Consequently, we have developed an algorithm based on the penalty dual decomposition (PDD) method to effectively address this challenge. Simulation results prove conclusively that with the assistance of IRS, system energy efficiency is effectively improved.