Last week, I was out of the office visiting two important on-going research projects that involve Nutaq. In this blog post, I discuss one of them, the AVIO-505. Back in 2013 one of my colleagues wrote an introduction to this project. If you don’t want to read the complete blog post, here’s a quick summary:
The goal of the AVIO-505 project is to integrate the multiple navigation and telecommunication systems used within an airplane into a single system through the use of software-defined radio (SDR) technology. This will dramatically reduce the number of on-board devices and the avionic weight associated with them. This will also reduce greenhouse gas emission by increasing the efficiency of new space management systems.
AVIO-505 is a very interesting project that demonstrates the strong relationship between the avionic industry and Quebec’s universities. My visit, however, was focussed mainly on one university, the Laboratory of Specialized Embedded System, Navigation and Avionics (LASSENA) at the École de technologie supérieure (ETS). The LASSENA is in charge of implementing the different avionic technologies on Nutaq’s ZeptoSDR and PicoSDR platforms.
For the moment, the LASSENA researchers are following the logical development process. They develop their waveforms in GNU Radio and once validated in simulation, they plug in the ZeptoSDR/PicoSDR for a hardware-in-the-loop (HIL) test. By using the default bitstream, which treats the FPGA as pass-through only, they were able to perform the HIL test and validate that the algorithms behaved as expected:
“It only took us 3 days to validate that our algorithms were behaving well with real test vectors, through HIL simulations,” said Omar Yeste, researcher at LASSENA.
The researchers went a step further by using Nutaq’s GNU Radio feature to embed their completed waveforms in the ARM processor of the ZeptoSDR (which is based on the Xilinx Zynq processor). They simply generated the Python code for their model and transferred it onto the ZeptoSDR’s Secure Digital (SD) card.
Their next step is to move their waveforms onto their 2×2 PicoSDR-Embedded system in order to achieve improved performance. The embedded version of the PicoSDR includes an i7 processor running Linux and uses Gigabit Ethernet or PCIe to exchange data with the FPGA card, the Perseus. Their plan is to repeat the same process: perform the HIL testing at first and then embed the design directly onto the embedded processor.
However, the final goal is to implement, chunk by chunk, their algorithms into the FPGA using Nutaq’s Model-Based Design Kit (MBDK). This way, they will have a powerful embedded platform that is field-deployable.
This is all the information I have right now but I will be closely supporting their progress in the upcoming months. So stay tuned!