BEGIN:VCALENDAR VERSION:2.0 X-WR-CALNAME:EventsCalendar PRODID:-//hacksw/handcal//NONSGML v1.0//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York LAST-MODIFIED:20240422T053451Z TZURL:https://www.tzurl.org/zoneinfo-outlook/America/New_York X-LIC-LOCATION:America/New_York BEGIN:DAYLIGHT TZNAME:EDT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZNAME:EST TZOFFSETFROM:-0400 TZOFFSETTO:-0500 DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT CATEGORIES:College of Engineering,Thesis/Dissertations DESCRIPTION:Topic: Deep Learning-Based Multi-Source Angle-of-Arrival Estima tion and Spectrum Awareness Abstract: Angle-of-arrival (AoA) estimation en ables spatial localization, beamforming, and interference mitigation in wi reless communication and sensing systems. Recent machine learning approach es have improved robustness compared to classical subspace-based methods, but most existing work is evaluated under simplified conditions such as si ngle-source scenarios, fixed observation windows, and covariance-domain in puts. In practical environments, multiple simultaneous transmitters, limit ed data samples, and heterogeneous signal types introduce additional chall enges for reliable AoA estimation. A learning-based framework is developed for multi-source AoA estimation and spectrum awareness using raw IQ input s. The approach extends AoA estimation to multiple concurrent transmitters and two-dimensional spatial estimation, while incorporating protocol clas sification and evaluation across different antenna geometries. A decision- making module integrates spatial and signal-level features to detect jammi ng and estimate the number of active transmitters. Preliminary results dem onstrate that learning-based models can scale to multi-source scenarios an d maintain competitive performance using IQ-domain representations. Adviso r: Dr. Ruolin Zhou, Associate Professor, Department of Electrical & Comp uter Engineering, UMass Dartmouth Committee Members: Dr. Dayalan P. Kasil ingam, Professor, Department of Electrical & Computer Engineering, UMass D artmouth; Dr. Long Jiao, Assistant Professor, Department of Computer & Inf ormation Science, UMass Dartmouth; Dr. Nathaniel D. Bastian, Assistant Pro fessor, United States Military Academy, Department of Systems Engineering, Department of Mathematical Sciences, Department of Electrical Engineering and Computer Science, West Point, NY Note: All ECE Graduate Students a re encouraged to attend. All interested parties are invited to attend. Ope n to the public. *For further information, please contact Dr. Ruolin Zho u via email at rzhou1@umassd.edu.\nEvent page: https://www.umassd.edu/even ts/cms/elee-oral-comprehensive-exam-for-doctoral-candidacy-by-jin-feng-lin .php\nEvent link: https://umassd.zoom.us/j/92675106134 X-ALT-DESC;FMTTYPE=text/html:

Topic: Deep Learning-Based Mult i-Source Angle-of-Arrival Estimation and Spectrum Awareness

Abstra ct: Angle-of-arrival (AoA) estimation enables spatial localization\, beamf orming\, and interference mitigation in wireless communication and sensing systems. Recent machine learning approaches have improved robustness comp ared to classical subspace-based methods\, but most existing work is evalu ated under simplified conditions such as single-source scenarios\, fixed o bservation windows\, and covariance-domain inputs. In practical environmen ts\, multiple simultaneous transmitters\, limited data samples\, and heter ogeneous signal types introduce additional challenges for reliable AoA est imation.

A learning-based framework is developed for multi-source AoA estimation and spectrum awareness using raw IQ inputs. The approach ex tends AoA estimation to multiple concurrent transmitters and two-dimension al spatial estimation\, while incorporating protocol classification and ev aluation across different antenna geometries. A decision-making module int egrates spatial and signal-level features to detect jamming and estimate t he number of active transmitters. Preliminary results demonstrate that lea rning-based models can scale to multi-source scenarios and maintain compet itive performance using IQ-domain representations.

Advisor: Dr. Ruolin Zhou\, Associate Professor\, Department of Electrical & Computer En gineering\, UMass Dartmouth

Committee Members:

Dr. Dayalan P. Kasilingam\, Professor\, Department of Electrical & Computer En gineering\, UMass Dartmouth\;
Dr. Long Jiao\, Assistant Professo r\, Department of Computer & Information Science\, UMass Dartmouth\;
Dr. Nathaniel D. Bastian\, Assistant Professor\, United States Milita ry Academy\, Department of Systems Engineering\, Department of Mathematica l Sciences\, Department of Electrical Engineering and Computer Science\, W est Point\, NY

Note: All ECE Graduate Students are encou raged to attend. All interested parties are invited to attend. Open to the public.

*For further information\, please contact Dr. Ruolin Zh ou via email at rzhou1@umassd.edu.

Event page: https://www.umassd.edu/events/cms/elee-oral-comprehen sive-exam-for-doctoral-candidacy-by-jin-feng-lin.php
Event link: https://umassd.zoom.us/j/9267 5106134

DTSTAMP:20260406T173753 DTSTART;TZID=America/New_York:20260417T140000 DTEND;TZID=America/New_York:20260417T160000 LOCATION:Lester W. Cory Conference Room, Science & Engineering Building (SENG), Room 213A SUMMARY;LANGUAGE=en-us:ELEE Oral Comprehensive Exam for Doctoral Candidacy by Jin Feng Lin UID:a1e6ad7be4ad2378e53e841ddd38417d@www.umassd.edu END:VEVENT END:VCALENDAR