Abstract:
Modern commercial sensing and communication systems that support high mobility primarily operate in the sub-6 GHz frequency band. While robust, this limits achievable data rates to tens of Mbps and leads to latencies exceeding several tens of milliseconds. To overcome these constraints, the mmWave IEEE 802.11ad protocol emerges as a promising candidate, offering significantly higher bandwidth for vehicle-to-vehicle and vehicle-to-infrastructure communication. However, mmWave signals are highly susceptible to atmospheric attenuation, confining their operation to line-of-sight scenarios and necessitating the use of narrow, directional beams. Fast and accurate beam alignment is therefore critical. Recent research has demonstrated that radar sensing can facilitate rapid beam identification while simultaneously supporting environmental perception. Designing a versatile hardware platform to evaluate the over-the-air performance of joint communication and radar sensing is essential. Such a platform enables rigorous assessment of how well existing waveforms perform in realistic environments, particularly in the presence of RF impairments. The first contribution of this thesis is the design and development of a wideband wireless physical layer using Orthogonal Frequency Division Multiplexing (OFDM) on the AMD RF-SoC platform, supporting an instantaneous bandwidth of up to 2.4576 GHz. The proposed architecture follows a hardware–software co-design methodology, wherein baseband signal processing and synchronization algorithms (covering time, frequency, and phase synchronization) are partitioned between the ARM Cortex-A53 processing system and the UltraScale+ FPGA fabric. These components are tightly integrated with the RFSoC’s high-speed on-chip data converters. To enable mmWave transmission, the sub-6 GHz RFSoC output is inter-faced with a multi-antenna mmWave analog front end (AFE) from Sivers Semiconductors, supporting over-the-air communication in the 24.25–29.5 GHz band. Additionally, a custom control interface is developed on the RFSoC to manage AFE parameters such as gain and beam direction dynamically. We demonstrate real-time, end-to-end over-the-air communication and provide a comprehensive bit error rate (BER) analysis under varying modulation schemes, coding rates, bandwidths, and realistic radio channel conditions, including RF impairments and beam misalignments. The second contribution focuses on radar-based sensing using monostatic radar configurations and explores both single-carrier and multi-carrier waveforms commonly employed in radar systems. A key challenge addressed in this work is achieving precise synchronization between the transmitter and receiver, which is critical for accurate target range estimation. Future work will focus on extending the current platform to support Integrated Sensing and Communication (ISAC), a key research priority for 6G networks. By enabling the convergence of communication and radar functionalities on a unified hardware platform, the system will serve as a versatile testbed for evaluating ISAC algorithms in real-world scenarios. Additionally, the platform facilitates over-the-air performance evaluation of artificial intelligence (AI)-enabled physical layer (PHY) techniques, which is an emerging area of interest in both academia and industry. Beyond communication and sensing, the platform also supports advanced channel sounding for in-depth characterization of mmWave channels, taking into account practical impairments such as fading, interference, and RF non-idealities.