Joint Spectral Efficiency Optimization of Uplink and Downlink for Massive MIMO Enabled Wireless Energy Harvesting Systems

Wenfeng Sun, Chen Liu, Mujun Qian, Yan Chen, Shu Xu

Abstract

This paper proposes a unified system model and uses a general wireless energy harvesting (EH) protocol in massive multiple-input multiple-output (MIMO) systems, in which the terminals have no fixed power supply and thus need to replenish energy via the received signals from the base station (BS). On the downlink, the BS delivers radio frequency (RF) signals to all terminals and each terminal coordinates the processes of EH and information decoding (ID) with wireless EH protocol. On the uplink, a fraction of the harvested energy is used for uplink pilot transmission and the remaining fraction is used for uplink data transmission. Closed-form lower bound expressions for each terminal are first obtained on the uplink and downlink, respectively. Based on these expressions, we then formulate the joint spectral efficiency (SE) of uplink and downlink maximization problem with some practical constraints. As the formulated optimization problem is non-linear and non-convex, it is hard to solve directly. To provide a solution, an iteration algorithm is proposed by utilizing one-dimensional search technique and successive approximation method based on geometric program (GP). Additionally, the convergence and complexity of the proposed algorithm are discussed as well. Finally, the feasibility of the proposed algorithm and analytical results are demonstrated by extensive computer simulations.

Introduction

In specialized environments like medical implants, mines, or military zones, charging or replacing batteries in wireless terminals is often impossible. Wireless Energy Harvesting (EH) allows these devices to "sip" energy from the radio signals sent by a Base Station (BS). However, there is a fundamental trade-off: every millisecond or milliwatt spent on harvesting energy is a resource taken away from transmitting data.

While previous research looked at optimizing either the uplink (terminal to BS) or the downlink (BS to terminal) in isolation, this paper tackles the more complex "joint" problem. The challenge is to find the "Goldilocks zone" for several variables—time allocated for charging, the ratio of power split between data and energy, and how much harvested energy to spend on channel estimation versus actual data transmission—all while using the massive antenna arrays (Massive MIMO) typical of 5G and beyond.

System model of the massive MIMO enabled wireless energy harvesting system featuring a hybrid protocol.

Proposed Methods and Hybrid Protocol

The authors propose a hybrid wireless EH protocol that combines Time-Switching (TS) and Power-Splitting (PS) mechanisms. This protocol operates within a four-phase transmission frame:

  1. Uplink Pilots: Terminals send signals so the BS can learn the channel.
  2. Downlink WPT: The BS sends pure energy (Wireless Power Transfer).
  3. Downlink SWIPT: The BS sends data and energy simultaneously (Simultaneous Wireless Information and Power Transfer).
  4. Uplink Data: Terminals use their stored energy to send data back.
A general transmission frame structure for the hybrid wireless EH protocol.

To solve the resulting non-convex optimization problem, a two-tier algorithm was developed:

  • Outer Loop: Employs a one-dimensional search to find the optimal time durations for channel estimation and energy transfer.
  • Inner Loop: Utilizes a Successive Approximation method based on Geometric Programming (GP) to transform non-linear variables into a solvable convex format.

Simulation and Performance Analysis

The model was validated through extensive computer simulations in MATLAB, modeling a system with a Base Station equipped with $M = 300$ to $360$ antennas and $K = 3$ single-antenna terminals at varying distances (10m to 20m) under Rayleigh fading conditions.

Key Insights

  • Hybrid Advantage: The hybrid protocol acts as a "super-set" of TS and PS. While PS generally outperforms TS, the hybrid approach provides the necessary flexibility to reach peak efficiency.
  • The "Sweet Spot": Spectral efficiency exhibits a quasi-concave relationship with the power-splitting ratio and energy allocation. There is a clear optimal point; deviations lead to sharp performance degradation.
  • Energy Beamforming: Massive MIMO focuses RF energy into narrow beams, allowing terminals to harvest significantly more power even at a distance.
  • Channel Hardening: In massive MIMO, the channel stabilizes. Spending excessive time on channel estimation (CSI) can be counterproductive, as the time loss outweighs the CSI gain.

Results

  • Efficiency: The iterative algorithm typically converges within 31 iterations.
  • Performance: The joint optimization model achieved a net sum spectral efficiency of approximately 18.8 bit/s/Hz in baseline scenarios.
  • Scaling: Increasing antennas from 300 to 360 boosted spectral efficiency from ~19.2 to ~20.1 bit/s/Hz at 4.5W transmit power.

Conclusion and Future Work

This research demonstrates that energy and data in massive MIMO systems must be managed as a single ecosystem. While providing a robust mathematical foundation, future work could address the non-linearities of real-world EH circuits, more complex channel models like Rician fading, and hardware impairments such as RF-to-DC conversion losses.

Summary updated

Abstract

<title>Abstract</title> This paper proposes a unified system model and uses a general wireless energy harvesting (EH) protocol in massive multiple-input multiple-output (MIMO) systems, in which the terminals have no fixed power supply and thus need to replenish energy via the received signals from the base station (BS). On the downlink, the BS delivers radio frequency (RF) signals to all terminals and each terminal coordinates the processes of EH and information decoding (ID) with wireless EH protocol. On the uplink, a fraction of the harvested energy is used for uplink pilot transmission and the remaining fraction is used for uplink data transmission. Closed-form lower bound expressions for each terminal are first obtained on the uplink and downlink, respectively. Based on these expressions, we then formulate the joint spectral efficiency (SE) of uplink and downlink maximization problem with some practical constraints. As the formulated optimization problem is non-linear and non-convex, it is hard to solve directly. To provide a solution, an iteration algorithm is proposed by utilizing one-dimensional search technique and successive approximation method based on geometric program (GP). Additionally, the convergence and complexity of the proposed algorithm are discussed as well. Finally, the feasibility of the proposed algorithm and analytical results are demonstrated by extensive computer simulations.
BibTeX
@inproceedings{sun2021joint,
  author = {Wenfeng Sun and Chen Liu and Mujun Qian and Yan Chen and Shu Xu},
  title = {{Joint Spectral Efficiency Optimization of Uplink and Downlink for Massive MIMO Enabled Wireless Energy Harvesting Systems}},
  booktitle = {EURASIP},
  year = {2021},
  doi = {10.1186/s13638-021-02057-4},
}
Figure