Wireless energy harvesting based relaying : a diversity multiplexing tradeoff perspective

This thesis presents a radio frequency (RF)-based energy harvesting (EH) coupled with relaying-based communication as a solution to address the end-to-end quality-of-service (QoS) requirements of future wireless communication systems. Typically, an increase in data rate leads to a rise in error prob...

Full description

Bibliographic Details
Main Author: Rao, Yepuri Sudhakara
Other Authors: A. S. Madhukumar
Format: Thesis-Doctor of Philosophy
Language:English
Published: Nanyang Technological University 2019
Subjects:
Online Access:https://hdl.handle.net/10356/136482
Description
Summary:This thesis presents a radio frequency (RF)-based energy harvesting (EH) coupled with relaying-based communication as a solution to address the end-to-end quality-of-service (QoS) requirements of future wireless communication systems. Typically, an increase in data rate leads to a rise in error probability. Conversely, to improve reliability, the system must sacrifice the achievable data rate. Thus, there is an inherent tradeoff between these two metrics. For quasi-static fading channels, Zheng and Tse proposed a fundamental performance metric, diversity-multiplexing tradeoff (DMT), to formalize the relationship between rate and reliability at asymptotically high signal-to-noise ratios. A direct characterization of the DMT for RF-EH-based relay channels is in general difficult because the transmission power of RF-EH-based relay is a function of source-to-relay channel gain. In this thesis, the DMT framework has been adapted to suit this class of channels. The essence of this work is in deriving the closed-form expressions for the DMT in the high-SNR regime. Specifically, the focus is on the time-switching- based scheme in which the intermediate relay switches between energy harvesting (EH), information decoding (ID), and the information retransmission (IR). In the single-relay case, the DMT has been characterized for dynamic decode-and- forward (DDF) and non-orthogonal amplify-and-forward (NAF) protocols. An investigation of these RF-EH-based protocols yields a detailed characterization of the limitations of energy harvesting concerning optimal performance. In particular, an interesting conceptual relation between the diversity gain and the energy- harvesting interval has been found. It is observed that the achievable diversity gain improves with the reduction in energy-harvesting interval. Moreover, a comparison with various benchmarking schemes has established the supremacy of RF-EH-based DDF among all the proposed relaying protocols. Next, a multiple-access relay system (MARS), where the relay node powered by RF-EH, is analyzed. For this system, the DMT is characterized for two variants of EH-based DDF protocols: without feedback (EH-DDF) and with limited feedback (EH-DDF-LF). Analysis of these protocols reveals that the diversity gain of EH- DDF scheme is inferior to that of traditional non-EH-based DDF protocol. On the other hand, EH-DDF-LF achieves optimal diversity gain in the high-multiplexing- gain regime. Subsequently, the DMT study has been extended to an interference- limited relay-aided communication system in the moderate-interference regime. Unfortunately, the asymptotic (high-SNR) DMT does not reveal the tradeoff between diversity and multiplexing gains at low to moderate SNRs. Thus, there is a need to characterize the DMT for finite SNR regimes to assess the complete picture on the effects of various design parameters. In this work, an analytical framework to derive closed-form expressions for finite SNR DMT (f-DMT) for RF- EH-based amplify-and-forward (AF) and decode-and-forward (DF) protocols is also proposed. The results of this investigation suggest that the DF offers superior performance in the low-SNR regime, whereas AF performs better in high-SNR scenarios. Furthermore, an analytical study is presented to evaluate the effect of the fraction of time devoted to RF-EH (time-sharing parameter, E) on the f-DMT. For both AF and DF, it is shown that the optimum E depends on the operating SNR. However, it is observed that the influence of E is nominal in the high-SNR regime. As a result, it is concluded that the asymptotic DMT has its relevance, even though it hides a few finer details.