STUDENT PROJECTS
This page provides some examples of the projects I have supervised for undergraduate/master thesis.
Resource Allocation for Communications and Power Transfer under a Practical RF Energy Harvesting Model
Master Thesis, Chalmers University of Technology, 2016
By Xiaowei Xu
Following paper is based on this thesis: X. Xu, A. Ozcelikkale, T. McKelvey, and M. Viberg, “Simultaneous Information and Power Transfer under a Non-Linear RF Energy Harvesting Model", 2017 IEEE International Conference on Communications (ICC) -Workshop on Emerging Energy Harvesting Solutions for 5G Networks, 2017.
Energy harvesting (EH) has been an emerging topic in the recent years. This technology promises to increase the energy efficiency and extend the working time of energy-constrained wireless networks as well as free the devices from wires and batteries. Wireless power transfer (WPT) refers to transmission of electrical power from a source to another device wirelessly. Among different approaches, WPT via radio frequency (RF) is a viable method to implement. On the other hand, RF signal has been widely used for transmitting information, namely wireless information transfer (WIT). Recently, the idea of using RF signals to carry information as well as energy at the same time, i.e., simultaneous wireless information and power transfer (SWIPT) has emerged.
In the area of SWIPT research, it has been typically assumed that energy conversion efficiency is independent from the level of the input power at the energy receiver. On the other end, in practice the energy conversion efficiency exhibits a non-linear behaviour and depends on the input power. This leads to a disperancy between the SWIPT designs made and practical EH circuits. This thesis addresses this issue.
We propose a practical quadratic model for the power conversion efficiency using curve fitting tools. Comparisons with the conventional linear models as well as another non-linear model proposed in the literature are made using mean-square based metrics. Using the proposed model, the problem of resource allocation for a multi-user Orthogonal Frequency-Division Multiple Access (OFDMA) system is investigated. In particular, the problem of allocating the power optimally among different sub-channels assigned to information and energy users is considered. Analytic solutions for some illustrative scenarios are provided. Numerical results are presented for the general case. Our investigations show that compared to using the traditional simple linear efficiency model, the proposed model provides more realistic harvested energy estimates at the energy receivers with minimal energy waste at the transmitter.
Measurement and Characterization of Solar Energy Availability for Low-Power Energy Harvesting Devices
Master Thesis, Chalmers University of Technology, 2016
By Rani El-Maalouf
Many of today's smaller devices can be liberated from their dependence of a constant power source by solar power, allowing for a more efficient use of current energy resources. The aim of this thesis is to measure the available light energy in an area and use this data for characterisation and prediction. This allows for better insight into what devices can be powered by solar energy, and the requirements for solar panel and battery.
For measurements, a device was constructed by connecting light sensors to a micro-controller. The micro-controller would then process the signals received from the light sensors and store this data on a micro SD card. Measurements were taken at various offices in the Department of Signals and Systems in Chalmers University of Technology.
Once the measurements were taken, they were used to obtain a profile for the character of sunlight. Each day contained a total of 48 time-slots, where each time-slot represents the mean value of the past half hour. By calculating the mean value for all days in each time-slot, a profile was constructed that gave a reasonable estimate for the character of sunlight. It was concluded that, while the mean gave an illustration as to how the character of available light energy typically changed during a day, values representing how the light varied for each time-slot were also necessary.
For determining the requirements of low-power devices, variations of solar panel sizes, intervals for active periods and initial battery charges were considered. It was concluded that the estimated available energy provided sufficient energy for indefinite use of a low-power wireless sensor during the month of May. This had a strong dependence on the size of the solar panel and the frequency of active periods chosen for the Machine-to-Machine (M2M) model. It was also established that an initial charge of the battery was required in order to avoid initial down-time.
Four algorithms, Exponentially Weighted Moving Average (EWMA), Weather Conditioned Moving Average (WCMA), Pro-Energy and a Weighted Min-Max Profile algorithm, were chosen for prediction. Min-Max was a proposed algorithm which was inspired by Pro-Energy, and was designed to require less memory and computational resources. All four algorithms provided the ability to perform a prediction for the next time-slot, but only Pro-Energy and Min-Max allowed for predictions of larger prediction horizons. It was concluded that the Min-Max algorithm provided most benefits when computational and memory constraints were present. However, it was also determined that this algorithm was not preferred when larger prediction horizons, with highly varying weather conditions, were present.
Simultaneous Information and Power Transfer in Multiple Antenna Relay Channels
Master Thesis, Chalmers University of Technology, 2015
By Kooros Moabber
Simultaneous wireless information and power transfer (SWIPT) is an emerging technology that has received significant attention recently. In a communication system with SWIPT capabilities, both of these tasks, information and power transfer is done over wireless medium simultaneously instead of doing these separately. In addition to providing a reliable alternative to powering communication networks solely with batteries or with cables, SWIPT capabilities bring increased mobility and prolong network up time.
In this work, SWIPT in a multiple-antenna relay system with multiple users is studied. The relay is used for increasing communication range for the information receivers (IR) as well as powering the energy harvesting receivers (ER) that demand power. We focus on the design of amplify and forward strategy at the relay (relay precoder) and the precoder at the base station. The minimum mean-square error (MMSE) is used as the performance criterion for information transfer. We consider two main scenarios: i) one-way relay system and ii) two-way relay system.
For the one-way relay system, the relay amplify and forward strategy (relay precoder) to minimize the MMSE from the source to the information user while satisfying energy harvesting constraints at the energy harvesting users is investigated. For the two-way relay system, minimizing the MMSE from the information users to the main source while satisfying the signal-to-interference-plus-noise ratio (SINR) constraints at the information users and the energy harvesting constraints at the ERs is considered. Here the precoder at the main source and the relay precoder is optimized jointly.
These problem formulations lead to non-convex optimization problems. Due to limitations of available approaches to solve these problems, a novel numerical approach is proposed. Using the proposed technique, the trade-offs between the MMSE and the energy harvesting constraints are found. The results show that it is important to design novel transmission techniques in order to deliver information and energy simultaneously most efficiently. The performance of the proposed designs is also compared with the performance of the some existing techniques in the literature that is developed for some special scenarios and it is observed that the proposed designs provide the same performance with these existing techniques in these special cases. The rate of the convergence of the proposed method is also investigated and it is found suitable for solving the mentioned problems.