Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


LeeJihyun2020.pdf (1.38 MB)

ETD Abstract Container

Abstract Header

Sequential decision in wireless networks under uncertainty

Lee, Jihyun
http://orcid.org/0000-0002-0224-9358
http://rave.ohiolink.edu/etdc/view?acc_num=osu1587481709887626

Abstract Details

2020, Doctor of Philosophy, Ohio State University, Electrical and Computer Engineering.
This thesis is concerned with the design and analysis of new algorithms for optimization problems under uncertainty with limited feedback on the channel in wireless systems. Broadly, our work is divided into two parts. In the first part, we study a user scheduling problem in a millimeter-wave (mmWave) network where the beam search can be performed only for a limited number of users at a time. Unlike traditional cellular networks in sub-6GHz, mmWave networks employ narrow directional beams to compensate for high propagation loss, by which the beam searching can cause an unaffordable overhead to the system especially in a mobile environment. Therefore, under the assumption that the beam search is performed for the scheduled users at each time, we posed the scheduling problem as an infinite horizon average cost constrained Markov decision process (CMDP) with the goal of minimizing the average beam searching overhead subject to the average rate constraint on each user. By using a structural result derived from the Lagrangian formulation of the CMDP, we show that the complexity of the problem can be reduced from exponential to polynomial in the number of users. Based on this result, a heuristic deterministic scheduling algorithms is proposed. In the second part, we study a sensor scheduling problem in a cognitive radio network where sensing on the licensed user channel can be performed at a limited number of sensors at a time. In cognitive radio (CR) networks, to protect licensed users (primary users, PU) and achieve high spectrum utilization by the unlicensed users (secondary users, SU), it is important to improve the sensing performance of the SU network by using sensors with high sensing accuracy. However, due to the presence of shadow fading, each sensor experiences different channel conditions from the PU transmitter and the channels between the PU transmitter and the sensors are not known to the SU network controller a priori and can only be learned through sensing feedback. To cope with such uncertainties resulting from lack of prior knowledge and limited feedback, we show that the problem can be considered as an instance of the multi armed bandit (MAB) model in which each sensor is considered as an arm. Under a MAB framework, we develop an algorithm for sequential sensor selection and channel access decision with the objective of maximizing total number of correct decisions on channel availability over finite time periods. Our index based heuristic algorithms is shown to achieve sub-linear accumulated regret in time period T and the performance of the proposed algorithms is also evaluated with numerical examples.
Eylem Ekici (Advisor)
153 p.
Electrical Engineering

Recommended Citations

Citations

  • Lee, J. (2020). Sequential decision in wireless networks under uncertainty [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1587481709887626

    APA Style (7th edition)

  • Lee, Jihyun. Sequential decision in wireless networks under uncertainty. 2020. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1587481709887626.

    MLA Style (8th edition)

  • Lee, Jihyun. "Sequential decision in wireless networks under uncertainty." Doctoral dissertation, Ohio State University, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=osu1587481709887626

    Chicago Manual of Style (17th edition)

osu1587481709887626
183
© 2020, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.