Training-Based SWIPT: Optimal Power Splitting at the Receiver

TL;DR

This paper proposes an optimal power-splitting strategy for SWIPT systems with training, balancing channel estimation and energy harvesting to maximize ergodic capacity.

Contribution

It introduces a joint optimization of power-splitting ratios during training and data phases for improved SWIPT performance.

Findings

Optimal power-splitting ratios enhance ergodic capacity.

The method balances energy harvesting and information transfer effectively.

Performance gains are demonstrated through theoretical analysis.

Abstract

We consider a point-to-point system with simultaneous wireless information and power transfer (SWIPT) over a block fading channel. Each transmission block consists of a training phase and a data transmission phase. Pilot symbols are transmitted during the training phase for channel estimation at the receiver. To enable SWIPT, the receiver adopts a power-splitting design, such that a portion of the received signal is used for channel estimation or data detection, while the remaining is used for energy harvesting. We optimally design the power-splitting ratios for both training and data phases to achieve the best ergodic capacity performance while maintaining a required energy harvesting rate. Our result shows how a power-splitting receiver can make the best use of the received pilot and data signals to obtain the optimal SWIPT performance.