Optimal Fairness-Aware Time and Power Allocation in Wireless Powered Communication Networks

Zhaohui Yang, Zhaohui Yang 0001, Wei Xu, Yi-Jin Pan, Cunhua Pan, Ming ChenIEEE Trans. Communications 2018 · DOI

TL;DR

This paper tackles the challenge of balancing network throughput and user fairness in wireless powered communication networks (WPCNs). They introduce a novel framework for joint time and power allocation that optimizes a "sum a-fair utility," allowing for a tunable tradeoff between overall network performance and equitable resource distribution among users.

What Problem Does This Paper Address?

Wireless powered communication networks (WPCNs) are an exciting frontier, enabling devices to harvest energy wirelessly from dedicated transmitters. This allows for extended device lifetimes, especially in the Internet of Things (IoT). However, a critical challenge arises in these networks: how to efficiently allocate limited resources (time and power) to maximize both the overall data throughput (sum rate) and ensure that all users receive a fair share of these resources. Prior work often focused on optimizing one metric at the expense of the other, or made simplifying assumptions about energy harvesting. This paper addresses this by proposing a method to jointly optimize time and power allocation to achieve a flexible tradeoff between sum rate and fairness, catering to various fairness definitions.

Methods

The core of this paper lies in formulating and solving a "sum a-fair utility maximization problem." This is achieved by:

  • Convexification: The original problem, which is generally nonconvex, is transformed into an equivalent convex problem. This is crucial for finding globally optimal solutions.
  • Iterative Optimization: For general fairness parameters, an alternating optimization approach is employed. This involves iteratively optimizing time allocation and power allocation strategies, with each step yielding a closed-form solution.
  • Special Case Solutions: For specific fairness scenarios (zero fairness, $\alpha=0$, and max-min fairness, $\alpha=+\infty$), closed-form solutions for optimal time and power allocation are derived.
  • Sum $\alpha$-Fair Utility: The paper introduces a generalized utility function that allows users to tune the fairness level by adjusting the parameter $\alpha$. $\alpha=0$ represents zero fairness (maximizing sum rate), $\alpha=1$ represents proportional fairness, and $\alpha=+\infty$ represents max-min fairness.

Datasets

Not discussed in the paper. The paper focuses on theoretical analysis and simulation-based evaluation of the proposed algorithms.

Key Insights and Ideas

  • Tunable Fairness: The introduction of the sum $\alpha$-fair utility function is a key contribution, providing a flexible mechanism to balance sum rate and fairness. This allows network operators to choose the desired operating point on the rate-fairness tradeoff curve.
  • Convexity is Key: Transforming the problem into a convex one is essential for guaranteeing globally optimal solutions, a significant step beyond heuristic approaches.
  • Zero Fairness Simplicity: For the zero fairness case ($\alpha=0$), the optimal solution is remarkably simple: time allocation is proportional to available power, and at most one user transmits in the uplink per epoch. This highlights the efficiency gains possible when fairness is not a primary concern.
  • Iterative Efficiency: The proposed iterative algorithms for common fairness levels offer a practical approach to solving the problem with lower computational complexity compared to standard convex optimization solvers like interior-point methods.

Key Results

  • The proposed algorithms (ZFPI for zero fairness, CFPI for common fairness) achieve sum rates comparable to the optimal solution obtained via a standard interior-point method (IPM), but with significantly lower computational complexity.
  • The sum rate increases with the maximal transmission power of the base station (BS).
  • Jain's index, a measure of fairness, increases monotonically with the fairness parameter $\alpha$, confirming that higher $\alpha$ values lead to more equitable resource distribution.
  • The tradeoff curve between sum rate and Jain's index shows that maximizing sum rate (at $\alpha=0$) results in the least fairness, while achieving maximum fairness (at $\alpha=+\infty$) leads to the lowest sum rate.
  • The proposed algorithms outperform simpler schemes like equal time allocation, equal transmission power, and equal power-splitting factor (ETEPES).

Limitations and Future Work

  • Idealized Assumptions: The paper assumes perfect channel state information (CSI) at the BS and ideal energy harvesting efficiency. In real-world scenarios, these assumptions may not hold.
  • Synchronization: The synchronization of DL and UL slots is assumed to be perfect.
  • Non-ideal Energy Harvesting: The paper acknowledges that non-ideal energy harvesting characteristics are not considered.
  • Future Work: The authors suggest extending the work to scenarios with multi-antenna BS, where obtaining globally optimal time and power allocation becomes more challenging. They also propose exploring alternating methods for beamforming and considering non-ideal energy harvesting.

Summary updated

BibTeX
@article{yang2018optimal,
  author = {Zhaohui Yang and Zhaohui Yang and Wei Xu and Yi-Jin Pan and Cunhua Pan and Ming Chen},
  title = {{Optimal Fairness-Aware Time and Power Allocation in Wireless Powered Communication Networks}},
  journal = {IEEE Trans. Communications},
  year = {2018},
  doi = {10.1109/tcomm.2018.2807443},
}
Figure