Abstract
Non-orthogonal multiple access (NOMA) assisted mobile edge computing (MEC) has recently attracted significant attention due to its superior capability to reduce the energy consumption and the latency of MEC offloading. In this paper, a general hybrid NOMA-MEC offloading strategy is proposed, which includes conventional orthogonal multiple access (OMA) and pure NOMA based offloading as special cases. A multi-objective optimization problem is formulated to minimize the energy consumption for MEC offloading, and a low-complexity resource allocation solution is derived and shown to be Pareto-optimal. Furthermore, by analyzing the properties of the obtained resource allocation solution, important insights regarding NOMA-MEC offloading are obtained. For example, it is proved that pure NOMA-MEC offloading cannot outperform hybrid NOMA-MEC. In addition, a precise condition under which NOMA-MEC outperforms OMA-MEC is established, and shown to match the one previously developed for the two-user special case. Furthermore, the developed analytical results also establish an interesting analogy between the proposed hybrid NOMA-MEC power allocation scheme and the well-known water-filling strategy.
Original language | English (US) |
---|---|
Pages (from-to) | 5377-5391 |
Number of pages | 15 |
Journal | IEEE Transactions on Wireless Communications |
Volume | 21 |
Issue number | 7 |
DOIs | |
State | Published - Jul 1 2022 |
Externally published | Yes |
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Electrical and Electronic Engineering
- Applied Mathematics
Keywords
- Non-orthogonal multiple access (NOMA)
- Pareto optimality
- mobile edge computing (MEC)
- multi-objective optimization