On joint communication and sensing in multiple-input multiple-output (MIMO) wireless systems
| dc.contributor.author | Perera, Thakshila | |
| dc.contributor.examiningcommittee | Mezghani, Amine (Electrical and Computer Engineering) | |
| dc.contributor.examiningcommittee | Sherif, Sherif (Electrical and Computer Engineering) | |
| dc.contributor.supervisor | Hossain, Ekram | |
| dc.date.accessioned | 2025-04-30T16:31:59Z | |
| dc.date.available | 2025-04-30T16:31:59Z | |
| dc.date.issued | 2025-04-04 | |
| dc.date.submitted | 2025-04-25T18:10:08Z | en_US |
| dc.degree.discipline | Electrical and Computer Engineering | |
| dc.degree.level | Master of Science (M.Sc.) | |
| dc.description.abstract | This thesis advances the understanding of integrated beamforming design for joint communication and sensing (JCAS) systems, where a multi-input multi-output (MIMO) base station (BS) supports downlink communication users while simultaneously estimating parameters of a sensing target. The focus is on characterizing the Pareto boundary between communication performance (measured by mutual information (MI), sum-rate capacity) and sensing performance (measured by Fisher information (FI)), subject to practical system constraints. Two key scenarios are investigated: (1) a single-user, single-sensing-object system with and without equivalent isotropic radiated power (EIRP) constraints, and (2) a multi-user, single-sensing-object system, both subject to a total transmit power budget. In the single-user scenario without EIRP constraints, a closed-form solution is derived for the Pareto-optimal beamforming design, which is parameterized by a single scalar variable governing the trade-off between MI/sum-rate and FI. This solution is validated against iterative algorithms (e.g., projected gradient descent) and convex optimization (CVX) frameworks, demonstrating its optimality. For the EIRP-constrained case, the problem is reformulated using Lagrangian and Karush-Kuhn-Tucker conditions and solved via a customized gradient descent algorithm. The results confirm that joint beamforming outperforms independent strategies, with the Pareto boundary influenced by the user’s and target’s angles of departure (AoD), antenna configurations, and EIRP limits. In the multi-user scenario, uplink-downlink duality is used to decouple the multi-objective optimization problem. Through Lagrangian methods and block-coordinate ascent, an efficient algorithm is developed to maximize the weighted sum of MI/sum-rate across users and FI for sensing. This approach shows that joint beamforming remains Pareto-optimal, efficiently allocating resources to mitigate interference while enhancing sensing accuracy. Numerical evaluations across various configurations validate the theoretical findings, revealing that EIRP constraints tighten the Pareto boundary, highlighting the need for adaptive power allocation strategies in practical systems. This work provides a foundation for designing next-generation JCAS systems, emphasizing integrated beamforming as a means to harmonize communication and sensing objectives. The derived frameworks and numerical tools offer valuable insights for optimizing MIMO-based dual-functional networks in applications such as vehicular radar-communication coexistence and 6G perceptive networks. Keywords: Joint Communication and Sensing (JCAS), Pareto Boundary, Mutual Information (MI), Sum-rate, Fisher Information (FI), Equivalent Isotropic Radiated Power (EIRP), joint beamforming, uplink-downlink duality. | |
| dc.description.note | October 2025 | |
| dc.identifier.uri | http://hdl.handle.net/1993/39054 | |
| dc.language.iso | eng | |
| dc.subject | JCAS | |
| dc.title | On joint communication and sensing in multiple-input multiple-output (MIMO) wireless systems | |
| local.subject.manitoba | no | |
| oaire.awardTitle | University of Manitoba Graduate Fellowship | |
| project.funder.identifier | https://doi.org/10.13039/501100000038, https://doi.org/10.13039/100010318 | |
| project.funder.name | Natural Sciences and Engineering Research Council of Canada, University of Manitoba |