Source-linked AI summary
Wireless Information and Power Transfer for Multi-Relay Assisted Cooperative Communication
Yuan Liu
TL;DR
The paper addresses multi-relay SWIPT, where prior work often assumed one assisting relay or fixed splitting coefficients. It optimizes relay power splitting for DF and AF, and simulations report gains over traditional best relay selection, with the preferred protocol varying by SNR region.
Problem
Most related SWIPT relay studies assume one assisting relay, while the cited multi-relay work uses fixed power-splitting coefficients rather than optimizing them.
Source: §I · p. 1Method
The paper optimizes power-splitting ratios for multiple relays using distributed space-time coding in both DF and AF two-hop transmission.
Source: §I · p. 1Results
The proposed DF and AF schemes significantly outperform traditional best relay selection schemes, and power splitting outperforms time switching at high SNR while time switching is better at low SNR.
Source: §IV · pp. 3–4, §V · p. 4Takeaways & Limitations
The simulations favor power splitting over time switching in the high-SNR region and time switching over power splitting in the low-SNR region.
Source: §IV · p. 4
Abstract
from arXiv · showhide
In this paper, we consider simultaneous wireless information and power transfer (SWIPT) in multi-relay assisted two-hop relay system, where multiple relay nodes simultaneously assist the transmission from source to destination using the concept of distributed space-time coding. Each relay applies power splitting protocol to coordinate the received signal energy for information decoding and energy harvesting. The optimization problems of power splitting ratios at the relays are formulated for both decode-and-forward (DF) and amplify-and-forward (AF) relaying protocols. Efficient algorithms are proposed to find the optimal solutions. Simulations verify the effectiveness of the proposed schemes.
I. INTRODUCTION
The paper addresses multi-relay SWIPT as a response to limited relay energy and prior work focused mainly on single-relay assistance or fixed power-splitting coefficients. It optimizes relay power splitting for DF and AF transmission using distributed space-time coding.
I. INTRODUCTION
- RF signals can carry information and energy simultaneously, motivating SWIPT as an approach to energy harvesting in wireless networks.Unlike natural energy sources, RF energy is not subject to the stated availability and controllability limitations of solar, wind, or vibration sources.Source: §I · p. 1
- The introduction identifies time switching and power splitting as two practical SWIPT receiver schemes.Time switching alternates between decoding and harvesting; power splitting divides received signal power between them.Source: §I · p. 1
- Prior multi-relay SWIPT work included fixed power-splitting analysis, while most related relay studies assumed a single assisting relay.The paper motivates coordinating the signals received by multiple relays to use their spatial diversity without extra source power.Source: §I · p. 1
- The paper optimizes relay power-splitting ratios for both DF and AF in a two-hop, multi-relay system using distributed space-time codes.Relays use harvested source-signal energy to forward information, and the objective is to maximize end-to-end rate.Source: §I · p. 1
II. SYSTEM MODEL
The system is a two-hop network in which multiple half-duplex relays harvest energy from the source signal and jointly forward it to the destination. The model assumes a blocked direct link and specified channel and CSI conditions.
II. SYSTEM MODEL
- The direct source–destination link is blocked, so the relay set must forward transmission between the source and destination.Nodes have single omnidirectional antennas; the model assumes independent unit-variance AWGN, path loss and Rayleigh fading, frame-constant fading, and perfect CSI at all nodes.Source: §II · p. 2
- Each relay splits received signal energy, using α_i for information decoding and the remainder for energy harvesting and forwarding.Transmission is half-duplex: the source broadcasts in phase one, then the relays simultaneously forward using distributed space-time codes.Source: §II · p. 2
III. OPTIMIZATION FRAMEWORK
The optimization framework develops problem formulations and optimal solutions for both DF and AF relaying.
III. OPTIMIZATION FRAMEWORK
- The framework treats problem formulation and optimal-solution derivation separately for DF and AF relaying.Source: §III · p. 2
A. Decode-and-Forward Relaying
For DF relaying, the paper maximizes end-to-end rate under first- and second-hop decoding constraints by optimizing relay power-splitting ratios. The optimum equalizes the relevant SNRs, and the authors characterize the problem as convex with a closed-form optimal solution.
A. Decode-and-Forward Relaying
- Power splitting makes the DF source rate depend on the minimum received SNR across relays, rather than only the worst source-to-relay channel gain in traditional DF.The paper identifies this change as complicating the optimization.Source: A. Decode-and-Forward Relaying · p. 2
- The DF end-to-end rate is bounded by reliable decoding at every relay and at the destination, with a factor of one-half for the two transmission slots.Source: A. Decode-and-Forward Relaying · p. 2
- The DF objective is to maximize end-to-end rate by optimizing the relays’ power-splitting ratios.Source: A. Decode-and-Forward Relaying · p. 2
- At the DF optimum, the first- and second-hop SNRs match, and the received SNR is equal across relays.The paper states that the resulting problem is convex and has a closed-form optimal solution.Source: A. Decode-and-Forward Relaying · pp. 2–3
B. Amplify-and-Forward Relaying
The AF formulation is nonconvex, but the paper decomposes it into relay-specific subproblems and converts them into a convex form that can be solved efficiently. The resulting optimal power-splitting ratios can be determined in a distributed way.
B. Amplify-and-Forward Relaying
- AF’s nonconcave rate makes P2 nonconvex, but its summation separates into K independent relay subproblems with identical structure.Each subproblem corresponds to one relay.Source: B. Amplify-and-Forward Relaying · p. 3
- Each relay subproblem is a concave fractional program, with a nonnegative concave numerator and a positive convex denominator.The paper introduces transformed functions and variables to obtain the convex formulation.Source: B. Amplify-and-Forward Relaying · p. 3
- The transformed problem is convex and can be solved efficiently using the interior-point method.After solving for x_i* and y_i*, the relay obtains its optimal power-splitting ratio.Source: B. Amplify-and-Forward Relaying · p. 3
- The AF subproblems use each relay’s own channel gains, allowing optimal power-splitting ratios to be found without a centralized controller.The paper contrasts this with DF, where relays broadcast their channel gains.Source: B. Amplify-and-Forward Relaying · p. 3
IV. SIMULATION RESULTS
The proposed multi-relay schemes significantly outperform best relay selection, with DF slightly ahead of AF. Against time switching, power splitting performs better at high SNR while time switching performs better at low SNR.
IV. SIMULATION RESULTS
- The simulations place five relays uniformly in a 2 m square centered midway between source and destination, with path-loss exponent 3 and energy-conversion efficiency 60%.The source and destination are fixed at (0, 0) and (10, 0) m.Source: §IV · p. 3
- The proposed DF and AF schemes significantly outperform best relay selection, with DF slightly superior to AF.The reported gain over best relay selection requires no extra source power because the relays use the source’s same signals.Source: §IV · p. 3, Figure 2 · p. 4
- Power splitting outperforms time switching at high SNR, while time switching is better at low SNR.The time-switching allocation is exhaustively searched over t in [0, 1].Source: §IV · p. 4, Figure 3 · p. 4
V. CONCLUSION
The paper concludes that optimized power splitting in multi-relay two-hop SWIPT significantly outperforms best relay selection. Its comparison with time switching favors power splitting at high SNR and time switching at low SNR.
V. CONCLUSION
- The proposed multi-relay power-splitting schemes significantly outperform traditional best relay selection in simulation.The schemes optimize relay power-splitting ratios for both DF and AF.Source: §V · p. 4
- Power splitting performs better than time switching in the high-SNR region, while time switching is favored in the low-SNR region.Source: §V · p. 4