Enhancing MEC Offloading Performance With NOMA-CDRT in the Presence of Impaired Hardware

Recently, mobile edge computing (MEC) has been proposed to improve wireless devices’ computational capabilities by offloading computation-intensive tasks to a network-edge server. Analytical research indicates that the application of non-orthogonal multiple access (NOMA) can significantly reduce the latency and energy consumption of MEC offloading. In order to capture the potential gains of NOMA in the context of MEC, this paper proposes an edge computing-aware coordinated direct and relay transmission (CDRT) based NOMA technique which can enjoy the benefits of uplink NOMA in reducing MEC users’ uplink energy consumption and computational delay. By lowering the likelihood of uplink outages for MEC users, the NOMA-assisted MEC system can benefit from uplink NOMA’s advantages. In addition, CDRT-based NOMA has the sum capacity scaling of
$\log \mathop \rho \nolimits _{b}$
as signal-to-noise-ratio
$\mathop \rho \nolimits _{b} $
increases, but
$\frac {1}{2}\log \mathop \rho \nolimits _{b} $
for non-CDRT-based NOMA. Hence, applying CDRT-based NOMA in MEC increases the computation capability and decreases communication outage probability by extending the cell coverage of the MEC users. Based on this idea and the superior performance requirements in future wireless networks, we derive new closed-form expressions for the exact ergodic sum capacity and offloading outage probability under both residual hardware impairment and perfect/imperfect successive interference cancellation. Furthermore, to set a benchmark, accurate analytical expressions and numerical results are provided to demonstrate the effectiveness of the NOMA-assisted MEC system over the orthogonal multiple access-based scheme. At the end, we provide the basic guidelines for choosing transceiver hardware that will meet the practical requirements of a CDRT-based uplink-NOMA transmission system for efficient offloading, where the performance of the MEC server is evaluated in terms of latency and energy consumption. The Monte Carlo simulations validate the accuracy of the derived analytical expressions.