ATTENTION: The works hosted here are being migrated to a new repository that will consolidate resources, improve discoverability, and better show UTA's research impact on the global community. We will update authors as the migration progresses. Please see MavMatrix for more information.
Show simple item record
dc.contributor.advisor | Liang, Qilian | |
dc.contributor.advisor | Pan, Chenyun | |
dc.creator | Bhole, Dheeral Naresh | |
dc.date.accessioned | 2023-01-26T16:14:12Z | |
dc.date.available | 2023-01-26T16:14:12Z | |
dc.date.created | 2022-12 | |
dc.date.issued | 2022-12-05 | |
dc.date.submitted | December 2022 | |
dc.identifier.uri | http://hdl.handle.net/10106/31019 | |
dc.description.abstract | Machine learning (ML) has recently been used to solve critical problems. This dissertation focuses on developing systems using Ultra-Wideband (UWB) wireless sensor networks and machine learning to solve critical tasks such as target detection in various challenging scenarios. These tasks have been researched for several years and efforts have been made to achieve universal solutions.
In the first part of this dissertation, we have proposed a system to detect metallic targets in foliage environment. Mission critical systems need to be ready for the harsh working environment such as dense foliage, water bodies, rain, heavy winds and other natural challenges. Extreme engineering excellence is needed to achieve a faultless system during such critical tasks and routines. To solve this problem, we come up with a Machine Learning system trained on wireless sensor network dataset. Our work consists of four main parts: First, we clean and standardize the dataset. Second, we transform the dataset using IFFT and PCA. Third, we train a XGBoost model on this dataset and make predictions on the test dataset. Finally, we calculate errors by comparing the predicted and actual values and obtain high accuracy with our method.
In the second part of this dissertation, we have proposed a system to detect humans through walls. Human detection through walls, doors and corridors is critical in applications such as hostage rescue situation, surveillance, activity recognition, etc. Our work consists of three main parts: First, we clean and standardize the dataset. Second, we train a Neural Network model on this dataset and make predictions on the test dataset. Finally, we calculate errors by comparing the predicted and actual values and obtain high accuracy with our method.
In the last part of this dissertation, we have proposed an ensemble ML system to optimize the first task of target detection in foliage environment. We apply generalized stacked machine learning system to harness the power of different ML models and we achieve the best accuracy with this method. | |
dc.format.mimetype | application/pdf | |
dc.language.iso | en_US | |
dc.subject | Machine learning | |
dc.subject | Wireless sensor networks | |
dc.subject | Digital signal processing | |
dc.subject | Neural networks | |
dc.subject | XGBoost | |
dc.title | MACHINE LEARNING FOR TARGET DETECTION USING UWB RADAR SENSOR NETWORKS | |
dc.type | Thesis | |
dc.date.updated | 2023-01-26T16:14:12Z | |
thesis.degree.department | Electrical Engineering | |
thesis.degree.grantor | The University of Texas at Arlington | |
thesis.degree.level | Doctoral | |
thesis.degree.name | Doctor of Philosophy in Electrical Engineering | |
dc.type.material | text | |
dc.creator.orcid | 0000-0003-1834-6109 | |
Files in this item
- Name:
- BHOLE-DISSERTATION-2022.pdf
- Size:
- 21.22Mb
- Format:
- PDF
This item appears in the following Collection(s)
Show simple item record