[Paper Review] Backscatter-Assisted Wireless Powered Communication Networks Empowered by Intelligent Reflecting Surface

Parisa Ramezani, Abbas Jamalipour|arXiv (Cornell University)|Apr 26, 2021

Energy Harvesting in Wireless Networks33 references43 citations

TL;DR

This paper proposes an intelligent reflecting surface (IRS)-assisted backscatter-enabled wireless powered communication network (BS-WPCN) to maximize network throughput by jointly optimizing IRS reflection coefficients, transmit/receive beamforming, power and time allocation. Using a practical energy harvesting model and a two-stage algorithm with alternating optimization and successive convex approximation, the scheme achieves significant spectral efficiency gains over conventional WPCN and IRS-aided baselines, especially under realistic discrete phase shift constraints.

ABSTRACT

Intelligent reflecting surface (IRS) has recently been emerged as an effective way for improving the performance of wireless networks by reconfiguring the propagation environment through a large number of passive reflecting elements. This game-changing technology is especially important for stepping into the Internet of Everything (IoE) era, where high performance is demanded with very limited available resources. In this paper, we study a backscatter-assisted wireless powered communication network (BS-WPCN), in which a number of energy-constrained users, powered by a power station (PS), transmit information to an access point (AP) via backscatter and active wireless information transfer, with their communication being aided by an IRS. Using a practical energy harvesting (EH) model which is able to capture the characteristics of realistic energy harvesters, we investigate the maximization of total network throughput. Specifically, IRS reflection coefficients, PS transmit and AP receive beamforming vectors, power and time allocation are designed through a two-stage algorithm, assuming minimum mean square error (MMSE) receiver at the AP. The effectiveness of the proposed algorithm is confirmed via extensive numerical simulations. We also show that our proposed scheme is readily applicable to practical IRS-aided networks with discrete phase shift values.

Motivation & Objective

• To address the throughput limitations in energy-constrained wireless networks by integrating backscatter communication with IRS-aided WPCN.
• To overcome the long energy harvesting delay in traditional WPCNs by enabling backscatter transmission with minimal energy consumption.
• To maximize total network throughput through joint optimization of IRS reflection coefficients, beamforming, power, and time allocation.
• To design a practical algorithm compatible with discrete-phase shift IRS hardware, ensuring real-world applicability.

Proposed method

• Proposes a two-stage optimization framework: first optimizing IRS reflection coefficients during backscatter transmission using alternating optimization (AO) and successive approximation (SA), then optimizing beamforming, power, and time allocation using semidefinite relaxation (SDR) and SA.
• Employs a minimum mean square error (MMSE) receiver at the access point to minimize signal distortion and improve spectral efficiency.
• Models realistic energy harvesting with a non-linear EH model to reflect practical harvester behavior.
• Uses a continuous-phase IRS reflection model initially, then validates performance with 2-bit resolution discrete phase shifts to ensure hardware feasibility.
• Applies the block coordinate descent (BCD) technique to jointly optimize IRS reflection coefficients and AP receive beamforming for active transmission.
• Derives analytical expressions for SINR and MSE to enable efficient optimization under the MMSE receiver framework.

Experimental results

Research questions

• RQ1 How can IRS be leveraged to enhance the throughput of backscatter-assisted wireless powered communication networks?
• RQ2 What is the optimal joint design of IRS reflection coefficients, beamforming vectors, power allocation, and time scheduling to maximize network throughput?
• RQ3 How does the performance of the proposed scheme compare to conventional IRS-aided WPCN and random/constant IRS reflection schemes?
• RQ4 To what extent does discrete-phase shift IRS hardware impact the performance, and can the proposed algorithm be effectively applied to such systems?
• RQ5 What is the impact of IRS position on system throughput, and where should it be deployed for optimal performance?

Key findings

• The proposed IRS-empowered BS-WPCN achieves significantly higher throughput than conventional IRS-aided WPCN, especially at high transmit power, due to the energy-efficient backscatter mode.
• The scheme with 2-bit phase shift resolution closely matches the performance of the continuous-phase case, proving the algorithm's practicality for real-world IRS hardware.
• Random or pre-configured IRS reflection matrices lead to unstable and inferior performance due to destructive signal combining, highlighting the need for dynamic optimization.
• Throughput initially decreases with increasing IRS x-coordinate due to longer AP distance, but improves again as IRS gets closer to users, indicating optimal deployment near users or AP.
• Optimizing time allocation yields a substantial performance gain over equal time allocation, confirming its importance in system design.
• The proposed algorithm achieves near-optimal performance across all evaluated metrics, including throughput, energy efficiency, and robustness to hardware constraints.