Find more on our web: hipeac.

net/jobs/11189

CentraleSupélec
Rennes, France

Acceleration of GNU Radio Blocks

Using General Purpose GPUs
• Deadline: March 1, 2020
• Career levels: Assistant Researcher, Engineer, Master, Software Engineer
• Keywords: Computer architecture, Cyber-physical systems, Design Space Exploration, GPUs, Multicore /
Manycore, Parallel computing, Runtime performance

Software Defined Radio (SDR) is a wireless communication system in which components of transmitters and
receivers are mostly implemented in software. GNU Radio is an open source software package designed for SDR
application development. It provides a graphical user interface framework to create an interactive environment for the
user, and is well supported by SDR platforms such as USRP, HackRF, and BladeRF. GNU Radio uses a combination of
Python and C++, where Python handles the high level interface and C++ is used to implement drivers and low level
interfaces with hardware. In order to accelerate the baseband processing of the SDR, the main goal of this internship
project is to identify opportunities for compute acceleration using GPUs and move some processing blocks from the
CPU to the GPU accelerator. We will use NVIDIA GPUs, so you will explore options including CUDA C/C++, PyCUDA,
CuPy, cuSignal for accelerated signal processing of wireless communications algorithms.

The objectives of this internship project are:

• To become familiar with GNU Radio and GNU Radio-companion (GRC)

• To create custom out-of-tree (OOT) modules in C/C++/Python and to make them available in GRC
• To become familiar with development for NVIDIA GPGPUs for compute acceleration (CUDA C/C++, PyCUDA,
CuPy, cuSignal, other GPU-accelerated libraries)
• To accelerate the previous OOT modules in a GPU using the CUDA ecosystem
• To deploy the accelerated application in practical environments and evaluate the achieved speedup

Upon successful completion of tasks, a possibility to continue with a PhD around the topic of heterogeneous
computing architectures for acceleration of wireless communications algorithms in 5G and beyond (Tbps signal
processing, spectrum sensing, artificial intelligence for radio communications) could be settled.

Please send us:

• Your CV along with your academic records and marks

• A motivational text
• Any additional documents/links that you think can show your experience

Additional information:

• Hosting laboratory: IETR (Institute of Electronics and Telecommunications of Rennes)

• Hosting institution: CentraleSupélec
• Research group: SCEE (Signals, Communications and Embedded Electronics)
• Starting date: February/March 2020
• Duration: 5 months

© 2023 HiPEAC, High Performance, Edge And Cloud computing.

The HiPEAC project has received funding from the European Union's Horizon Europe research and innovation funding programme under grant agreement number 101069836.