Wireless network

Nonlinear programming

Electrical engineering

Physics

Channel (broadcasting)

Nonlinear system

Maximization

Computation offloading

Computer network

Mathematics

Wireless

Transmitter power output

Server

Algorithm

Mathematical optimization

Geodesy

Engineering

Enhanced Data Rates for GSM Evolution

Fractional programming

Benchmark (surveying)

Geography

Efficient energy use

Mobile edge computing

Edge computing

Wireless power transfer

Computer science

Computation

Transmitter

Quantum mechanics

Telecommunications

IEEE Internet of Things Journal

Emerging smart Internet-of-Things (IoT) applications are increasingly relying on mobile-edge computing (MEC) networks, where the energy efficiency (EE) of computation is one of the most pertaining issues. In this article, considering the limited computation capacity at the MEC server and a practical nonlinear energy harvesting (EH) model for IoT devices, we propose a scheme to maximize the system computation EE (CEE) of a wireless power transfer (WPT) enabled nonorthogonal multiple access (NOMA)-based MEC network by jointly optimizing the computing frequencies and execution time of the MEC server and the IoT devices, the offloading time, the EH time and the transmit power of each IoT device, as well as the transmit power of the power beacon (PB). We formulate the joint optimization into a nonlinear fractional programming problem and devise a Dinkelbach-based iterative algorithm to solve it. By means of convex theory, we derive closed-form expressions for parts of the optimal solutions, which reveal several instrumental insights into the maximization of the system CEE. In particular, the system CEE increases as the optimal computing frequencies of both the IoT devices and the MEC server decrease, and the system CEE is maximized when the MEC server and the IoT devices use the maximum allowed time to complete their computing tasks. Simulation results demonstrate the superiority of the proposed scheme over benchmark schemes in terms of system CEE.

Computation Energy Efficiency Maximization for a NOMA-Based WPT-MEC Network