Re: Issues concerning 10GbE speed standards
- To: Drew Perkins <drew.perkins@xxxxxxxxxxxx>
- Subject: Re: Issues concerning 10GbE speed standards
- From: Dae Young KIM <dykim@xxxxxxxxxxxxxxxxxx>
- Date: Tue, 29 Jun 1999 09:44:47 +0900
- CC: "'nuss@xxxxxxxxxx'" <nuss@xxxxxxxxxx>, "'Paul Bottorff'" <pbottorf@xxxxxxxxxxxxxxxxxx>, "'Peter_Wang@xxxxxxxx'" <Peter_Wang@xxxxxxxx>, "'rabynum@xxxxxxxxxxx'" <rabynum@xxxxxxxxxxx>, "'stds-802-3-hssg@xxxxxxxx'" <stds-802-3-hssg@xxxxxxxx>
- Organization: Chungnam Nat'l Univ., InfoCom Eng. Dept.
- References: <2072E1221F1DD211848C00104B938E7B733574@xxxxxxxxxxxx>
- Sender: owner-stds-802-3-hssg@xxxxxxxxxxxxxxxxxx
I had been in the line coding field for years, but I don't readily see why
balancedness of a code should offer any cost savings. Normally, in order to
balance your code, you have to use more redundancy, thus more line bandwidth,
If there should be some DC-blocking component along the path, so that
DC-restoring is necessary at each regenerator when using a non-balanced codes,
then the balanced codes could somehow win over the non-balanced in cost saving.
Whether using a balanced codes or not should rather more depend on the
channel(path) charactersitics; is the path DC-blocking or not. In DC-blocing (or
AC-coupled) channels, balanced codes generally would give better BER performance
at the same implementation cost.
Drew Perkins wrote:
> Can some people who are more expert than myself
> please comment on the cost penalties for not having a balanced code? If you
> assume the same raw baud/bit rate (after coding), then how much less
> expensive is 8B/10B because it's balanced? Now if you factor in the coding
> inefficiency penalty (i.e. you raise the baud/bit rate for 8B/10B by 25%),
> how are the costs affected?