Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.3_100GCU] Updated Minutes



At the Interim Meeting, during the Q&A session after our presentation (http://www.ieee802.org/3/bj/public/jan12/cole_02a_0112.pdf), Chris DiMinico suggested that we extend our analysis to include skin loss effects.

Since the result leads to interesting observations, we would like to share these with the Task Force.

 

The Salz SNR bound as a function of dielectric loss from the presentation is:

 

SNR = rho – ½ * IL(NRZ Nyquist)                 (for NRZ)

 

SNR = rho – ¼ * IL(NRZ Nyquist)                 (for PAM-4)

 

The newly derived Salz SNR bound as a function of skin loss is:

 

SNR = rho – 2/3 * IL(NRZ Nyquist)                      (for NRZ)

 

SNR = rho – sqrt(2)/3 * IL(NRZ Nyquist)          (for PAM-4)

 

On the enclosed graph, we plotted the Delta Salz SNR = PAM-4_Salz_SNR – NRZ_Salz_SNR - 7dB.

 

Positive Delta means PAM-4 BER is higher and negative Delta means NRZ BER is higher.

 

Since loss is caused by both contributors, the actual Delta lies somewhere in between the two plotted curves.

 

Notice how even for such a simple model, the cross-over point between the modulation schemes can shift by 8dB just by varying the ratio of channel loss contributors.

 

Chris & Ilya

 

 

From: John D'Ambrosia [mailto:jdambrosia@xxxxxxxxxxxxxxx]
Sent: Monday, February 06, 2012 12:44 PM
To: STDS-802-3-100GCU@xxxxxxxxxxxxxxxxx
Subject: [802.3_100GCU] Updated Minutes

 

All,

There were some modifications made to the minutes.  Please see the latest update at http://www.ieee802.org/3/bj/public/jan12/minutes_01a_0112_unapproved.pdf.

 

John

Attachment: Salz_SNR_bound.pdf
Description: Salz_SNR_bound.pdf