Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
9 years ago

Digital signal processing capabilities of DE2-115

Hello all,

I am new to this forum and am a third year engineering student. My final year project is based on digital signal processing and I am currently doing simple audio processing on the DE2-115 board with noise filtering as practice. It's laboratory 12 on the Altera website and I've pretty much completed it.

I decided to make my approach more research based but am struggling immensely! I genuinely have no idea where to go with anything. My project proposal is 'An investigation of the digital signal processing capabilities of field programmable gate-arrays'. I find audio processing boring but my teachers have already shut me down saying any other form of signal processing on an FPGA is extremely difficult and I am supposedly already behind so I shouldn't make any drastic changes to my project. I just feel so uninspired and overwhelmed all at the same time.

- What are future trends for audio processing? I think speech processing is quite cool but what sort of physical hardware/software can I demonstrate? I have some knowledge on audio compression, equalising and FIR filtering.

- I am also interested in OpenCL and OpenCV but can I implement this on an FPGA? (Apologies if I sound stupid) and am I restricting myself by sticking with the DE2-115??

I am struggling because I don't have a set final goal for my demo and dissertation. I understand that I am only an undergraduate and that I shouldn't overcomplicate or that my project doesn't have to be unprecedented but I can't focus when I am not inspired.

I am sorry for the rant but I am feeling hopeless, It's Christmas vacation and I should have done a substantial amount of research by now.

Can anyone out there give me inspiration, can anybody help?

Thank you in advance

5 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    9 years ago

    Hello,

    Long time ago i had a similar problem with my engineering project.

    I received a title and did not know what to do with it, it ended up well anyway.

    If you google your subject, you will end up with a list of books.

    First thing to do: study literature to see what your ecosystem looks like.

    http://www.springer.com/us/book/9783540726135 can be interesting.

    In your project book: first thing is to give overview of the technology, and the possible alternatives, in this case there a competition between dedicated DSP processors and FPGA's, what are the advantage of both technologies and what to use when. What is the current evolution towards cpu-fpga based solutions doing (soc things) with the ecosystem?

    Then you can decide to build a current application: for example, take a DE2-115, attach a microphone to the thing and decide to build something practical. Can you build something that can count the number of cars passing through a street, is it possible to determine the difference between cars, bikes and trucks?

    Best Regards,

    Johi
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    9 years ago

    Hello Johi

    Firstly thank you for your response I appreciate it.

    I've already submitted a formative report in which I've already discussed the competition between DSP processors and fpga's.

    I've looked at Altera's university ip cores obviously including the audio codec as well as the video/image ip core. I've found some neural networks/fuzzy logic/motion tracking projects on youtube that sound cool. I reckon I could hook up a camera and a vga monitor and do some video processing (maybe use some blob detection algorithm) on the DE2, I am just not sure what yet and I am worried I may struggle with it. The algorithms look quite scary.

    Thanks for the book recommendation.

    Shereen
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    9 years ago

    I recommend to start with something easy to get familiar with FPGAs and their electronic problems. Audio is easy on FPGAs as soon as you knwo what to do. Signalprocessing methods can be taken from various books.

    An exampe on what can be done with the DE regarding audio: www.96khz.org (Altera Audio DSP).
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    9 years ago

    --- Quote Start ---

    I genuinely have no idea where to go with anything.

    --- Quote End ---

    Welcome! Let's change that!

    --- Quote Start ---

    'An investigation of the digital signal processing capabilities of field programmable gate-arrays'. I find audio processing boring but my teachers have already shut me down saying any other form of signal processing on an FPGA is extremely difficult

    --- Quote End ---

    Ignore your teachers. Prove them wrong. DSP on FPGAs is not difficult. You just need people who know what they are doing to help you when you get stuck.

    Read the EELive! and ESC tutorial/slides/source code here ...

    https://www.ovro.caltech.edu/~dwh/correlator/cobra_docs.html

    Then keep asking questions ...

    --- Quote Start ---

    I can't focus when I am not inspired

    --- Quote End ---

    This is absolutely normal - and the fault of your teachers - they have not been able to inspire you, or inspire you with their own excitement.

    Cheers,

    Dave
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    9 years ago

    My advice to you, Propaste, is to follow the advice of Dave Hawkins (the post above) closely, as he is an FPGA/DSP expert and has written some excellent tutorials on the subject. As an undergraduate you shouldn't pick something overly complex, you're not trying to break new ground. You need to demonstrate to your instructor that you have a solid grasp of the fundamentals and can apply them to a practical problem. Good Luck.