Speaker
Description
Software Defined Radios (SDRs) are ubiquitous in modern wireless communications, offering flexibility and reconfigurability across various protocols, such as IEEE 802.11 and 4G LTE. Industry-standard SDR platforms such as the Ettus USRP and Xilinx RFSoC provide considerable wireless capabilities by combining FPGA- and software-based digital signal processing (DSP). However, the high cost of these SDR platforms can make them impractical for student or hobbyist use. In contrast, many entry-level SDRs lack FPGAs entirely. The software-only processing approach reduces cost, but limits both their maximum processing capability and their usefulness as educational tools for FPGA-based DSP.
This research leverages the Analog Devices Pluto SDR, which contains a Xilinx Zynq-7000 System-on-Chip (SoC), to create an accessible and versatile platform for FPGA-based implementation of SDR systems. By integrating Xilinx PYNQ, an embedded Linux framework for simplifying hardware/software co-design on Zynq SoCs, this research provides an isolated environment for developing hardware-accelerated signal processing functions. PYNQ provides a Python interface and Linux kernel drivers for memory-mapped peripherals and direct memory access, so developers can focus on FPGA implementation of DSP algorithms and avoid the overhead of Linux integration. The Pluto-PYNQ system is both an educational tool for learning FPGA signal processing and a platform for practical SDR application development.
The products of this research include the creation of the open-source Pluto-PYNQ Linux environment and an example DSP hardware accelerator. The open-source SDR community benefits from this research through the creation of a low-cost, educationally valuable SDR platform, and the Pluto-PYNQ can be used in advanced collegiate coursework to develop and demonstrate DSP functionality on real over-the-air radio frequency signals.
| Talk Length | 15 Minutes |
|---|
