{"id":{"repo_id":"ttu","oai_identifier":"oai:ttu-ir.tdl.org:2346/102322"},"canonical_url":"https://search.dev.ndltd.org/etd/ttu/oai:ttu-ir.tdl.org:2346/102322","repository":{"repo_id":"ttu","name":"Texas Technology University","base_url":"https://ttu-ir.tdl.org/server/oai/request"},"display":{"title":"High Frequency Transmission Spectroscopy","abstract":"High Frequency Transmission Spectroscopy is a novel technique for the characterization of a wide range of materials in a diverse set of problem domains. HFTS has potential implications for problem domains such as semiconductor manufacturing and counterfeit detection, disease/infection screening, and environmental monitoring. HFTS is rooted in traditional spectroscopy techniques such as optical and impedance spectroscopy but offers several key advantages. A relatively simple and affordable experimental apparatus may be developed by utilizing existing instrumentation for work in the VHF – UHF frequency bands. Additionally, keeping the material of interest electrically isolated from the measurement instrumentation can avoid the electrochemical effects commonly occurring at the electrode/material interface during impedance spectroscopy. This enables insight into the properties of the material without the rigor of accounting for these effects. When coupled with machine learning techniques such as supervised classification and unsupervised anomaly detection, HFTS truly shines as a low-cost, problem-agnostic solution for material classification and identification. This work explores the viability of HFTS along two axes: classification of semiconductor devices and classification of chemicals in a liquid sample. For semiconductor devices, custom device interface boards were developed and scattering parameter measurements were taken on identically packaged devices across a range of distinction levels. These scattering parameter measurements were then used to train machine learning classification models to perform multiclass classification. Ultimately, application of the S21 scattering parameter with a k-nearest neighbors clustering approach achieved an accuracy of 88% when classifying devices based on the individual manufacturing lot. Similarly, a test setup for performing HFTS on liquid samples was developed. By applying both k-nearest neighbors clustering and deep learning models, HFTS is able to accurately classify contaminants in water at concentrations as low as 0.15µM, track the change in concentration of carbon dot nanoparticles, and identify the size of plastic microparticles.","abstract_html":"High Frequency Transmission Spectroscopy is a novel technique for the characterization of a wide range of materials in a diverse set of problem domains. HFTS has potential implications for problem domains such as semiconductor manufacturing and counterfeit detection, disease/infection screening, and environmental monitoring. HFTS is rooted in traditional spectroscopy techniques such as optical and impedance spectroscopy but offers several key advantages. A relatively simple and affordable experimental apparatus may be developed by utilizing existing instrumentation for work in the VHF – UHF frequency bands. Additionally, keeping the material of interest electrically isolated from the measurement instrumentation can avoid the electrochemical effects commonly occurring at the electrode/material interface during impedance spectroscopy. This enables insight into the properties of the material without the rigor of accounting for these effects. When coupled with machine learning techniques such as supervised classification and unsupervised anomaly detection, HFTS truly shines as a low-cost, problem-agnostic solution for material classification and identification. This work explores the viability of HFTS along two axes: classification of semiconductor devices and classification of chemicals in a liquid sample. For semiconductor devices, custom device interface boards were developed and scattering parameter measurements were taken on identically packaged devices across a range of distinction levels. These scattering parameter measurements were then used to train machine learning classification models to perform multiclass classification. Ultimately, application of the S21 scattering parameter with a k-nearest neighbors clustering approach achieved an accuracy of 88% when classifying devices based on the individual manufacturing lot. Similarly, a test setup for performing HFTS on liquid samples was developed. By applying both k-nearest neighbors clustering and deep learning models, HFTS is able to accurately classify contaminants in water at concentrations as low as 0.15µM, track the change in concentration of carbon dot nanoparticles, and identify the size of plastic microparticles.","abstract_has_math":false,"creators":["Johnston, Derek"],"institution":"Texas Tech University","degree_name":"Doctor of Philosophy","degree_level":"Doctoral","degree_discipline":"Electrical Engineering","degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":["Karp, Tanja"],"committee_members":["Pappas, Dimitri","Christopher, Gordon","Pal, Ranadip"],"year":2025,"date_issued":"2025-05","date_published":"2025-05","updated_at":"2026-07-24T05:04:47Z","subjects":["Network Analysis","Machine Learning"],"languages":["English"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2346/102322","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.committeechair","label":"Committee Chair","values":["Karp, Tanja"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Pappas, Dimitri","Christopher, Gordon","Pal, Ranadip"]},{"key":"dc:creator","label":"Author","values":["Johnston, Derek"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2025-05-19T01:58:02Z"]},{"key":"dc:date.issued","label":"Date","values":["2025-05"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Doctoral"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["Texas Tech University"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Network Analysis","Machine Learning"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["English"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/2346/102322"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["High Frequency Transmission Spectroscopy is a novel technique for the characterization of a wide range of materials in a diverse set of problem domains. HFTS has potential implications for problem domains such as semiconductor manufacturing and counterfeit detection, disease/infection screening, and environmental monitoring. HFTS is rooted in traditional spectroscopy techniques such as optical and impedance spectroscopy but offers several key advantages. A relatively simple and affordable experimental apparatus may be developed by utilizing existing instrumentation for work in the VHF – UHF frequency bands. Additionally, keeping the material of interest electrically isolated from the measurement instrumentation can avoid the electrochemical effects commonly occurring at the electrode/material interface during impedance spectroscopy. This enables insight into the properties of the material without the rigor of accounting for these effects. When coupled with machine learning techniques such as supervised classification and unsupervised anomaly detection, HFTS truly shines as a low-cost, problem-agnostic solution for material classification and identification. This work explores the viability of HFTS along two axes: classification of semiconductor devices and classification of chemicals in a liquid sample. For semiconductor devices, custom device interface boards were developed and scattering parameter measurements were taken on identically packaged devices across a range of distinction levels. These scattering parameter measurements were then used to train machine learning classification models to perform multiclass classification. Ultimately, application of the S21 scattering parameter with a k-nearest neighbors clustering approach achieved an accuracy of 88% when classifying devices based on the individual manufacturing lot. Similarly, a test setup for performing HFTS on liquid samples was developed. By applying both k-nearest neighbors clustering and deep learning models, HFTS is able to accurately classify contaminants in water at concentrations as low as 0.15µM, track the change in concentration of carbon dot nanoparticles, and identify the size of plastic microparticles."]},{"key":"dc:format.mimetype","label":"Dc Format Mimetype","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["High Frequency Transmission Spectroscopy"]}]}],"canonical_facts":{"dc:contributor.committeechair":["Karp, Tanja"],"dc:contributor.committeemember":["Pappas, Dimitri","Christopher, Gordon","Pal, Ranadip"],"dc:creator":["Johnston, Derek"],"dc:date.accessioned":["2025-05-19T01:58:02Z"],"dc:date.issued":["2025-05"],"dc:description.abstract":["High Frequency Transmission Spectroscopy is a novel technique for the characterization of a wide range of materials in a diverse set of problem domains. HFTS has potential implications for problem domains such as semiconductor manufacturing and counterfeit detection, disease/infection screening, and environmental monitoring. HFTS is rooted in traditional spectroscopy techniques such as optical and impedance spectroscopy but offers several key advantages. A relatively simple and affordable experimental apparatus may be developed by utilizing existing instrumentation for work in the VHF – UHF frequency bands. Additionally, keeping the material of interest electrically isolated from the measurement instrumentation can avoid the electrochemical effects commonly occurring at the electrode/material interface during impedance spectroscopy. This enables insight into the properties of the material without the rigor of accounting for these effects. When coupled with machine learning techniques such as supervised classification and unsupervised anomaly detection, HFTS truly shines as a low-cost, problem-agnostic solution for material classification and identification. This work explores the viability of HFTS along two axes: classification of semiconductor devices and classification of chemicals in a liquid sample. For semiconductor devices, custom device interface boards were developed and scattering parameter measurements were taken on identically packaged devices across a range of distinction levels. These scattering parameter measurements were then used to train machine learning classification models to perform multiclass classification. Ultimately, application of the S21 scattering parameter with a k-nearest neighbors clustering approach achieved an accuracy of 88% when classifying devices based on the individual manufacturing lot. Similarly, a test setup for performing HFTS on liquid samples was developed. By applying both k-nearest neighbors clustering and deep learning models, HFTS is able to accurately classify contaminants in water at concentrations as low as 0.15µM, track the change in concentration of carbon dot nanoparticles, and identify the size of plastic microparticles."],"dc:format.mimetype":["application/pdf"],"dc:identifier.uri":["https://hdl.handle.net/2346/102322"],"dc:language.iso":["English"],"dc:subject":["Network Analysis","Machine Learning"],"dc:title":["High Frequency Transmission Spectroscopy"],"dc:type":["Thesis"],"thesis:degree_discipline":["Electrical Engineering"],"thesis:degree_level":["Doctoral"],"thesis:degree_name":["Doctor of Philosophy"],"thesis:institution_name":["Texas Tech University"]},"updated_at":"2026-07-24T05:04:47Z"}