Joint Optimization for Cooperative Computing Framework in Double-IRS-Aided MEC Systems

This letter investigates a cooperative task computing framework, where the source node partially offloads its computational task to multiple user equipments (UEs) aided by double intelligent reflecting surfaces (IRSs). With the aim of maximizing the total amount of computing task subject to latency and power constraints, we highlight an interesting tradeoff between the transmit power and the computing power at the source node and optimize the computing frequency resources as well as phase shift matrices for double IRSs. Numerical results verify the power allocation tradeoff and demonstrate the superiority of our double-IRS-aided solution in terms of maximizing the total amount of computing task over other benchmark strategies.

Automated summary

This letter investigates a cooperative task computing framework utilizing double intelligent reflecting surfaces (IRSs) to aid multiple user equipments (UEs) in offloading computational tasks from a source node. By optimizing the power and computing frequency resources, interesting tradeoffs are highlighted between transmit power and computing power. Numerical results demonstrate the superiority of our double-IRS-aided solution in maximizing the total amount of computing tasks, compared to other benchmark strategies.