Re: [10GBASE-T] matlab ad hoc
Hi Bill,
I agree with Bijit's point that MATLAB program needs
to be modular and accept different parameters such
as line codes.
Regarding transformer model the upper -3dB point should
be high enough to cover line codes such as PAM5. I would
suggest to use transformer bandwidth with f-3dBL=10 MHz
and f-3dBU=700 MHz.
Please add me to your MATLAB distribution list.
Regards,
Joseph N. Babanezhad
Plato Labs.
> Hi Bill,
>
> DFE is a good point to start. We suggest that since the core
> function to calculate optimal SNR will be used over and over again
> for different line code, transformer, transmit PSD it is better to
> develop this in a modular form as a MATLAB function. The function
> can take various models, such as, channel, noise, transformer, and
> transmit PSD as input. Of course we can have default values for
> those parameters. As far as the calculation is concerned we suggest
> that we start with sampled spaced DFE. Just a comment about the
> proposed transformer model, we feel that the 3dB cutoff point is too
> high and we should add a 3dB high pass cut off point as well, say around
> 350MHz.
>
> Please add me on your distribution list and I am planning to
> contribute in the Matlab code development.
>
> Thanks.
> -Bijit
>
> Telicos Corporation
>
> Participants
>
> To get this kicked off, I suggest developing some code to compute the
> optimal DFE SNR. I would like to see several participants independently
> develop some code. For the purposes of testing the code only, we
> need to agree upon a transmit PSD, transformer, channel and noise
> models. I do not want to use DSL models. I suggest something from
> TIA 568b or ISO 11801. I am open to specific proposals. For the
> transformer, I suggest a simple first order high pass filter with
> low frequency 3dB point in the range of 10 to 15 MHz. For the
> simulation results I have present, I used 12.5 MHz. For initial
> testing purposes, I suggest a PAM10 PSD with 10dBm power.
>
> I am sending this to the reflector. But, since several people are having
> problems getting to the reflector, I am thinking to carry on the ad
> hoc via direct distribution to interested participants. Let me know
> if this creates a problem for anyone. So, I will be sending a
> separate email directly to those participants that have contacted me
> either by email or business card by the end of business today
> (1/16/03). If you do not receive a second email and are interested
> in participating, kindly send me an email.
>
> Also, I have just received the 1000BASE-T code from Chris which I
> will be distributing.
>
> Regards
>
> Bill
>
> William W. Jones, Ph.D.
> Director of Systems Engineering
> SolarFlare Communications, Inc.
> 949-581-6830, ext. 2550
> mobile: 619-405-2445
> fax: 949-581-4695
> wjones@solarflare.com
--
Open WebMail Project (http://openwebmail.org)