discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Google Summer of Code 2014 applicant : Optimizati


From: Abhishek Bhowmick
Subject: Re: [Discuss-gnuradio] Google Summer of Code 2014 applicant : Optimization with VOLK
Date: Sat, 15 Mar 2014 15:07:17 +0530

Here is the link for my first proposal draft :
https://github.com/abhowmick22/GSoc14-Proposal

I will keep revising it. Seeking feedback in meantime. Thanks all.

Abhishek


On Sat, Mar 15, 2014 at 3:37 AM, Martin Braun <address@hidden> wrote:
On 14.03.2014 19:27, Abhishek Bhowmick wrote:
Hi,
So, according to some suggestions,  I looked into how I can potentially
use better signal processing for the OFDM receiver. I was thinking of a
LS estimator with higher order interpolation or an MMSE estimator for
the channel estimator part. Also, a MMSE-DFE or Viterbi equalizer. These
will need matrix operations and other computations, which can
potentially be developed into new volk kernels.
1. Are the computational complexities involved feasible in the current
framework ?
2. Though they can give better BER in adverse channel conditions, can
they do deliver more in terms of throughput/performance?
3. Is it a good idea to include such implementations alongside doing new
volk kernels in the same proposal ?

Abishek,

at this point, please just put together a proposal and upload it so we can make sure it gets into Melange in time.

M


Abhishek


On Wed, Mar 12, 2014 at 3:38 AM, Florian Kaltenberger
<address@hiddenfr
<mailto:Florian.Kaltenberger@eurecom.fr>> wrote:

    Hi Nathan and Abhishek,


    On 10/03/2014 23:22, West, Nathan wrote:
    Ah! So there was a slight miscommunication. Yes, porting the OpenAirInterfaces
    SIMD code to VOLK is a good option as well. The turbo channel coder/decoder
    is part of that. I've**briefly**  looked at the code to see what is

    currently there, and
    it's my understanding that the work involved will be to write generic
    C implementations
    of vectorized code where the generic version does not exist. Beyond
    that porting to
    newer/different ISAs (AVX or NEON depending on your preference and hardware
    availability). I think Florian is on the gr-discuss mailing list, but
    I've CCed him to
    hopefully provide more details as he's more familiar with the original
    code base.
    I only joined this mailing list recently, so I probably missed a
    part of the discussion. Let me summarize briefly what
    OpenAirInterface can provide. We have optimized SIMD (SSE4)
    implementations of the LTE turbo encoder and decoder as well as the
    LTE tail-biting Viterbi encoder and decoder. We also have the 802.11
    Viterbi encoder and decoder. The only functions for which we have
    generic non-vectorized functional equivalents is the LTE turbo decoder.
    I am not sure I understand why it is necessary to write generic
    versions for the already optimized SIMD code. My idea was to port
    the optimized SIMD code from OpenAirInterface to VOLK, such that is
    can be used by GR applications. I am not familiar with VOLK (yet)
    but this might just be as easy as writing a wrapper function.
    As Nathan suggested, the more interesting part is probably to
    upgrade the code to AVX2 or similar.

    Cheers,
    Florian.




--
Regards;
Abhishek Bhowmick,
Senior Undergraduate,
Department of Electrical Engineering,
IIT Bombay.


_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio





--
Regards;
Abhishek Bhowmick,
Senior Undergraduate,
Department of Electrical Engineering,
IIT Bombay.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]