Date of Award

August 2017

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Electrical Engineering and Computer Science


Mustafa C. Gursoy


energy harvesting communication networks, finite alphabets input, Non-orthogonal multiple access, simultaneous information and power transfer, statistical queuing constraints, wireless powered cellular network

Subject Categories



Energy harvesting is considered as a prominent solution to supply the energy demand for low-power consuming devices and sensor nodes. This approach relinquishes the requirements of wired connections and regular battery replacements. This thesis analyzes the performance of energy harvesting communication networks under various operation protocols and multiple access schemes. Furthermore, since the radio frequency signal has energy, in addition to conveying information, it is also possible to power energy harvesting component while establishing data connectivity with information-decoding component. This leads to the concept of simultaneous wireless information and power transfer. The central goal of this thesis is to conduct a performance analysis in terms of throughput and energy efficiency, and determine optimal resource allocation strategies for wireless information and power transfer.

In the first part of the thesis, simultaneous transfer of information and power through wireless links to energy harvesting and information decoding components is studied considering finite alphabet inputs. The concept of non-uniform probability distribution is introduced for an arbitrary input, and mathematical formulations that relate probability distribution to the required harvested energy level are provided. In addition, impact of statistical quality of service (QoS) constraints on the overall performance is studied, and power control algorithms are provided.

Next, power allocation strategies that maximize the system energy efficiency subject to peak power constraints are determined for fading multiple access channels. The impact of channel characteristics, circuit power consumption and peak power level on the node selection, i.e., activation of user equipment, and the corresponding optimal transmit power level are addressed. Initially, wireless information transfer only is considered and subsequently wireless power transfer is taken into account. Assuming energy harvesting components, two scenarios are addressed based on the receiver architecture, i.e, having separated antenna or common antenna for the information decoding and energy harvesting components. In both cases, optimal SWIPT power control policies are identified, and impact of the required harvested energy is analyzed.

The second line of research in this thesis focuses on wireless-powered communication devices that operate based on harvest-then-transmit protocol. Optimal time allocation for the downlink and uplink operation interval are identified formulating throughput maximization and energy-efficiency maximization problems. In addition, the performance gain among various types of downlink-uplink operation protocols is analyzed taking into account statistical QoS constraints.

Furthermore, the performance analysis of energy harvesting user equipment is extended to full-duplex wireless information and power transfer as well as cellular networks. In full-duplex operation, optimal power control policies are identified, and the significance of introducing non-zero mean component on the information-bearing signal is analyzed. Meanwhile, SINR coverage probabilities, average throughput and energy efficiency are explicitly characterized for wireless-powered cellular networks, and the impact of downlink SWIPT and uplink mmWave schemes are addressed.

In the final part of the thesis, energy efficiency is considered as the performance metric, and time allocation strategies that maximize energy efficiency for wireless powered communication networks with non-orthogonal multiple access scheme are determined. Low complex algorithms are proposed based on Dinkelbach’s method. In addition, the impact of statistical QoS constraints imposed as limitations on the buffer violation probabilities is addressed.


Open Access

Included in

Engineering Commons