Optimal Wireless Rate and Power Control Using Reinforcement Learning

dc.contributor.advisorMiao, Lei
dc.contributor.authorRaji, Fadlullah Adeola
dc.contributor.committeememberSbenaty, Saleh
dc.contributor.committeememberFaezipour, Misagh
dc.date.accessioned2021-12-09T17:05:21Z
dc.date.available2021-12-09T17:05:21Z
dc.date.issued2021
dc.date.updated2021-12-09T17:05:27Z
dc.description.abstractIncreased throughput and energy efficiency of wireless devices are achievable when factors such as interference reduction, power management, and data transmission rate selection are considered. This thesis proposes an algorithm for optimizing the performance of wireless networks using reinforcement learning. The algorithm, a.k.a. the agent, observes the state of a wireless device’s battery, packet queue, transmission medium, etc. and establishes the optimal policy for joint control of transmission power and speed. Using the NS3 network simulation software, we implement agents focusing on three different reward functions: throughput-critical, energy-critical, and throughput and energy balanced. We compare their performance to the conventional Minstrel rate adaptation algorithm: our approach can achieve (i) higher throughput when using the throughput-critical reward function; (ii) lower energy consumption when using the energy-critical reward function; and (iii) higher throughput and roughly the same energy when using the throughput and energy balanced reward function.
dc.description.degreeM.S.
dc.identifier.urihttps://jewlscholar.mtsu.edu/handle/mtsu/6586
dc.language.rfc3066en
dc.publisherMiddle Tennessee State University
dc.source.urihttp://dissertations.umi.com/mtsu:11529
dc.subjectDeep Learning
dc.subjectMachine Learning
dc.subjectRate Control
dc.subjectReinforcement Learning
dc.subjectWireless Communication
dc.subjectWireless Network
dc.subjectComputer science
dc.subjectElectrical engineering
dc.subjectComputer engineering
dc.thesis.degreelevelmasters
dc.titleOptimal Wireless Rate and Power Control Using Reinforcement Learning

Files

License bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
2.27 KB
Format:
Item-specific license agreed upon to submission
Description:

Collections