Efficient Resource Allocation for Mobile-Edge Computing Networks With NOMA: Completion Time and Energy Minimization

Zhaohui Yang, Zhaohui Yang 0001, Cunhua Pan, Jiancao Hou, Mohammad Shikh-BahaeiIEEE Trans. Communications 2019 · DOI

TL;DR

This paper tackles the challenge of optimizing resource allocation in mobile-edge computing (MEC) networks that utilize Non-Orthogonal Multiple Access (NOMA). The authors propose an iterative algorithm to simultaneously minimize task completion time and energy consumption, offering a more efficient solution for future wireless applications.

What Problem Does This Paper Address?

Mobile-edge computing (MEC) offers the promise of enhanced computation capabilities for mobile devices by offloading tasks to nearby edge servers. However, limited radio and computation resources in these networks pose a significant challenge for efficient operation. Existing research has focused on minimizing either completion time or energy consumption separately. This paper addresses the critical need for a unified approach that balances both objectives, particularly in the context of NOMA, a promising technology for next-generation wireless networks that allows multiple users to share the same frequency band. The authors highlight that previous NOMA-based MEC studies often overlooked the complexities of user grouping and time allocation across different groups, which this work aims to rectify.

Methods

The paper proposes an iterative algorithm to solve the general problem of minimizing a weighted sum of completion time and total energy consumption. This problem is first transformed into an equivalent form. The iterative algorithm works by alternately optimizing different sets of variables, leading to a closed-form solution at each step.

For a special case where only completion time minimization is considered (weight $\omega=1$), a bisection-based algorithm is developed. This algorithm efficiently searches for the optimal completion time by repeatedly narrowing down a feasible interval.

Furthermore, for the scenario with infinite edge cloud capacity, the problem is shown to be equivalent to a convex optimization problem, which can be solved efficiently. The authors also leverage techniques like successive interference cancellation (SIC) inherent in NOMA for decoding user signals.

Datasets

Not discussed in the paper. The paper focuses on theoretical modeling and algorithm development, with numerical results generated based on simulated network parameters.

Key Insights and Ideas

  • Unified Optimization: The core insight is the necessity and feasibility of jointly optimizing for both completion time and energy consumption in NOMA-enabled MEC networks, moving beyond single-objective approaches.
  • NOMA's Advantage: NOMA's ability to allow simultaneous transmission within groups significantly contributes to reduced completion times and energy savings compared to traditional TDMA and FDMA schemes, especially under high edge cloud capacity.
  • User Pairing Matters: The choice of user pairing strategy (e.g., strong-strong, strong-weak) significantly impacts performance, with strong-strong pairing showing superior results in the conducted simulations.
  • Time vs. Energy Tradeoff: Transmitting with longer completion times is shown to be energy-efficient, a key observation for designing energy-aware MEC systems.
  • Iterative Algorithm Efficiency: The proposed iterative algorithm offers a low-complexity approach to find near-optimal solutions for the complex general problem.

Key Results

  • The proposed NOMA-based approach outperforms conventional TDMA and FDMA schemes in terms of both completion time and total energy consumption, particularly when edge cloud capacity is high.
  • The iterative algorithm achieves near-globally optimal solutions, with a small gap compared to exhaustive search methods.
  • For the completion time minimization case ($\omega=1$), the bisection-based algorithm effectively finds the optimal solution.
  • In the infinite cloud capacity scenario, the problem is transformed into a convex one, allowing for efficient global optimization.
  • Numerical results demonstrate that increasing edge cloud capacity generally leads to lower completion time and energy consumption.

Limitations and Future Work

  • User Grouping: The paper assumes that user grouping is pre-determined and focuses on resource allocation within these groups. The optimization of user grouping itself is identified as a direction for future research.
  • Centralized Approach: The proposed algorithms are centralized, requiring the base station (BS) to collect information from all users. This can lead to significant overhead, especially with a large number of users. Future work could explore distributed or semi-distributed approaches.
  • Channel State Information: The algorithms assume perfect knowledge of uplink channel conditions. In real-world scenarios, imperfect channel state information could affect performance.

Summary updated

BibTeX
@article{yang2019efficient,
  author = {Zhaohui Yang and Zhaohui Yang and Cunhua Pan and Jiancao Hou and Mohammad Shikh-Bahaei},
  title = {{Efficient Resource Allocation for Mobile-Edge Computing Networks With NOMA: Completion Time and Energy Minimization}},
  journal = {IEEE Trans. Communications},
  year = {2019},
  doi = {10.1109/tcomm.2019.2935717},
}
Figure