Candidate: Jin Gyu Lim
Date: April 2, 2026
Time: 11:00 AM
Location: Online
Supervisor: Prof. Slim Boumaiza
All are welcome!
Abstract:
The development of fifth-generation (5G) wireless communications standards is driven by the increasing demand for higher data rates, lower latency, and improved coverage as the number of connected devices continues to grow. To meet these requirements, massive multiple-input, multiple-output (MIMO) transmitter architectures have emerged as a key enabling technology. These systems employ multiple transmitter chains, each consisting of digital signal processing modules, power amplifiers (PAs), and antenna elements. By generating multiple directive beams simultaneously within the same frequency band, massive MIMO systems can serve multiple users at once and significantly improve spectral efficiency compared to conventional fourth-generation (4G) single-input single-output transmitters. However, the large number of transmitter chains also increases system complexity.
Within the transmitter, the PA plays a critical role because its efficiency strongly influences overall energy consumption, while its output power determines coverage. In conventional 4G systems, isolators are placed between the PA and the antenna to maintain a constant load impedance, ensuring stable performance. In 5G massive MIMO arrays, however, isolators are often impractical due to cost, size, performance degradations, and the high level of integration required. Their absence causes dynamic load impedance variations at each PA due to antenna mismatches and mutual coupling, which can degrade PA behavior and overall system performance. Existing mitigation strategies in digital signal processing, PA design, and antenna engineering have often been developed independently, without fully considering cross-domain interactions.
This seminar addresses these challenges through a holistic system-level design approach. First, a multidisciplinary co-simulation framework integrating digital signal processing, RF circuitry, and electromagnetic antenna effects is developed to analyze massive MIMO transmitter behavior under realistic modulated signals. The framework is experimentally validated using a four-channel fully digital MIMO transmitter and later extended to a sixteen-channel testbed. Second, it is investigated how precoding scheme in digital signal processing affects PA behavior, proposing an alternative processing order. Finally, an emulation platform is introduced to evaluate and design PAs for massive MIMO transmitters using a single PA test setup under realistic loading conditions.