Spectral-Energy Efficiency Trade-Off-Based Beamforming Design for MISO Non-Orthogonal Multiple Access Systems

Haitham Moffaqq Al-Obiedollah, Kanapathippillai Cumanan, Jeyarajan Thiyagalingam, Jie Tang, Alister G. Burr, Zhiguo Ding, Octavia A. DobreIEEE Trans. Wireless Communications 2020 · DOI

TL;DR

This paper tackles the challenge of simultaneously optimizing spectral efficiency (SE) and energy efficiency (EE) in MISO-NOMA wireless systems. They propose a novel beamforming design that formulates this as a multi-objective optimization problem and solves it using a sequential convex approximation technique, offering a flexible trade-off between these crucial performance metrics.

What Problem Does This Paper Address?

Future wireless networks, particularly 5G and beyond, face a dual challenge: delivering higher data rates (spectral efficiency, SE) while minimizing power consumption (energy efficiency, EE). Traditional approaches often optimize one metric at the expense of the other, leading to suboptimal performance. Non-Orthogonal Multiple Access (NOMA) is a promising technology for improving SE and supporting massive connectivity, but its integration with multiple-input single-output (MISO) systems requires careful resource allocation. This paper addresses the need for a beamforming design that can achieve a good balance between SE and EE, allowing base stations to adapt to varying energy sources (like renewables) and operational requirements.

Methods

The core of this paper is a novel beamforming design that tackles the SE-EE trade-off.

  • Multi-Objective Optimization (MOO) Formulation: The problem is initially framed as a multi-objective optimization problem, aiming to maximize both SE and EE simultaneously. This is a departure from traditional single-objective approaches.
  • A Priori Articulation and Weighted Sum: Since MOO problems have multiple Pareto-optimal solutions, the authors employ an "a priori articulation" scheme. This involves assigning weights to each objective (SE and EE) to reflect their relative importance. These weighted objectives are then combined into a single objective function using a weighted-sum approach, transforming the MOO problem into a Single Objective Optimization (SOO) problem.
  • Sequential Convex Approximation (SCA): The resulting SOO problem is non-convex, making it difficult to solve directly. The authors leverage the Sequential Convex Approximation (SCA) technique. This iterative method approximates the non-convex functions with lower-bounded convex functions in each iteration, allowing for the application of convex optimization solvers.
  • Slack Variables and Constraint Reformulation: To handle the non-convexity within the constraints, various slack variables and mathematical transformations (like first-order Taylor series approximations and second-order cone (SOC) formulations) are introduced. This process converts complex, non-convex constraints into a series of convex ones that can be handled by the SCA algorithm.

Datasets

Not discussed in the paper. The evaluation is based on simulated scenarios.

Key Insights and Ideas

  • The Inherent Trade-off: The paper clearly demonstrates that beyond a certain "green power" threshold, increasing spectral efficiency comes at the cost of decreasing energy efficiency, and vice-versa. This highlights the necessity of a joint optimization approach.
  • Flexibility through Weighting: The proposed weighted-sum approach provides significant flexibility. By simply tuning the weight factor ($\alpha$), the base station can prioritize either SE or EE, or strike a balance between them, adapting to different operational scenarios and energy availability.
  • Pareto-Optimality: The method guarantees finding Pareto-optimal solutions, meaning that no improvement can be made to one objective (SE or EE) without degrading the other. This ensures that the achieved solutions represent the best possible trade-offs.
  • SCA as a Powerful Tool: The paper showcases the effectiveness of the SCA technique in solving complex, non-convex optimization problems that arise in wireless resource management.

Key Results

  • Achieved SE-EE Trade-off: Simulation results (Figures 1, 2, 3, and Table I) demonstrate that the proposed design effectively achieves a trade-off between EE and sum rate (SE) by adjusting the weight factor $\alpha$.
  • Performance Comparison: Table II shows that the proposed SCA algorithm achieves performance comparable to a benchmark semi-definite programming (SDP) solution for the power minimization problem, validating its effectiveness.
  • Feasibility Analysis: The paper emphasizes the importance of a feasibility check, as the problem can become infeasible if the available power is insufficient to meet minimum SINR requirements.
  • Impact of SINR Threshold: Figure 4 illustrates how the minimum SINR threshold can influence the feasibility of the SE-EE trade-off design, potentially requiring an alternative SE-Max design if infeasible.

Limitations and Future Work

  • Perfect CSI Assumption: The current design assumes perfect channel state information (CSI) is available at the base station. In real-world scenarios, CSI is often imperfect, which could impact performance. Future work could explore robust beamforming designs under channel uncertainties.
  • Computational Complexity: While the SCA algorithm is iterative, its computational complexity, though analyzed, could still be a factor in highly dynamic or dense network environments. Further optimization of the algorithm or exploration of alternative, potentially faster, methods could be beneficial.
  • Hybrid Power Sources: The paper mentions hybrid power sources as a motivation but focuses on the SE-EE trade-off itself. Future work could delve deeper into how the proposed design can be dynamically adapted based on the real-time availability of different power sources.

Summary updated

BibTeX
@article{alobiedollah2020spectral,
  author = {Haitham Moffaqq Al-Obiedollah and Kanapathippillai Cumanan and Jeyarajan Thiyagalingam and Jie Tang and Alister G. Burr and Zhiguo Ding and Octavia A. Dobre},
  title = {{Spectral-Energy Efficiency Trade-Off-Based Beamforming Design for MISO Non-Orthogonal Multiple Access Systems}},
  journal = {IEEE Trans. Wireless Communications},
  year = {2020},
  doi = {10.1109/twc.2020.3004292},
}
Figure