Resource Allocation in SWIPT Networks under a Non-Linear Energy Harvesting Model: Power Efficiency, User Fairness, and Channel Non-Reciprocity

TL;DR

This paper introduces a novel resource allocation framework for multi-user SWIPT systems with non-linear energy harvesting, addressing power efficiency, user fairness, and channel non-reciprocity through multi-level optimization and closed-form solutions.

Contribution

It proposes a structured multi-level optimization approach for resource management in SWIPT systems considering non-linear energy harvesting and channel non-reciprocity, including both optimal and suboptimal solutions.

Findings

Achieves significant performance gains with the proposed schemes.

Provides closed-form solutions for practical implementation.

Validates effectiveness through numerical simulations.

Abstract

This paper considers a multi-user simultaneous wireless information and power transfer (SWIPT) system with a non-linear energy harvesting model, in which a multi-antenna base station (BS) estimates the downlink channel state information (CSI) via uplink pilots. Each single-antenna user is equipped with a power splitter. Three crucial issues on resource management for this system include: (i) power-efficient improvement, (ii) user-fairness guarantee, and (iii) non-ideal channel reciprocity effect mitigation. Potentially, a resource allocation scheme to address jointly such issues can be devised by using the framework of multi-objective optimization. However, the resulting problem might be complex to solve since the three issues hold different characteristics. Therefore, we propose a novel method to design the resource allocation scheme. In particular, the principle of our method relies on structuralizing mathematically the issues into a cross-layer multi-level optimization problem. On this basis, we then devise solving algorithms and closed-form solutions. Moreover, to instantly adapt the CSI changes in practice while reducing computational burdens, we propose a closed-form suboptimal solution to tackle the problem. Finally, we provide numerical results to show the achievable performance gains using the optimal and suboptimal solutions, and then validate the proposed resource allocation scheme.