Presented by: Nicholas Bruce

Date: Monday, June 1, 2026
Time: 11:00 am
Place: Zoom - see below.

Zoom link: https://uvic.zoom.us/j/81199921912

Abstract:

This two-part seminar series discusses the philosophy of next-generation radio frequency interference (RFI) monitoring within radio astronomy observatories, presenting several components of an end-to-end solution. Following the receiver and direction-of-arrival techniques introduced in Part 1, Part 2 dives deeper into the digital signal processing pipeline. First, a signal down-conversion analysis/synthesis filterbank pair is instroduced, which is used to reconstruct baseband timeseries from spectra for detected signals. Second, the seminar introduces SigCLR — a contrastive learning architecture that acts as a modulation and novelty detector. This unsupervised method enables the efficient processing and classification of large, unlabelled datasets of detected RFI signals.

Presented by: Nagwa Shaghluf

Date: Monday, June 15, 2026
Time: 10:00 am - 10:45 am
Place: Zoom - see below.

Abstract:

Future wireless networks are expected to support massive connectivity, high data demand, and increasingly dynamic spectrum environments. These requirements create major challenges in spectrum sensing, interference management, and resource allocation. This seminar presents an intelligent and adaptive framework for improving spectrum and resource management in cognitive radio networks. First, predictive cooperative spectrum sensing is introduced to improve spectrum awareness by combining spectrum prediction with cooperative sensing, reducing unnecessary sensing while improving spectrum and energy efficiency. Next, multi-agent deep reinforcement learning is used for cognitive device-to-device resource allocation, enabling distributed agents to learn resource block and power-control decisions while protecting cellular users. Finally, the framework is extended to NOMA-enabled cognitive D2D networks, where observation information, constraint-aware reward design, SIC feasibility, and fairness shaping are investigated. The seminar highlights how prediction, distributed learning, and information-aware resource allocation can support more efficient and autonomous spectrum management for 5G and beyond networks.

Location: Remote via Zoom

Join Zoom Meeting

https://uvic.zoom.us/j/89486738812?pwd=TzkXO0dB9ClDMYeRSKW8dWYxaQ6oXn.1

Meeting ID: 894 8673 8812

Password:  915308

Note: Please sign in to Zoom with your UVic account