RE: Equalization
Henning,
I can't presume to educate you, but can point to a useful source
that will answer many of your questions: "Adaptive Equalization" by
S. Qureshi, IEEE Communications Magazine, March 1992, pages 9-16.
Let me take this opportunity to eliminate a potential area of
confusion. Every multimode fiber has DMD; only some fibers have
abnormal DMD resulting from severe anomalies in refractive index
profile. Normal DMD can be equalized using classical spectral
compensation techniques. Abnormal DMD is trickier to equalize. To
add one more twist, the above statements are true for launch types
(OFL, ROFL) that excite a large number of modes. But if you restrict
your launch to a subset of modes, perhaps we can use the simpler
spectral compensation... This is where we enter an unchartered
territory.
Regards,
Vipul
======================
> -----Original Message-----
> From: owner-stds-802-3-hssg@xxxxxxxx
> [mailto:owner-stds-802-3-hssg@xxxxxxxx]On Behalf Of
> Lysdal, Henning
> Sent: Monday, July 24, 2000 5:25 AM
> To: stds-802-3-hssg@xxxxxxxx
> Subject: RE: Equalization
>
>
>
> Vipul, all,
>
> When evaluating the feasibility of integrated equalizers,
> I think it's
> important to know exactly which gain-vs-frequency (and
> phase-vs-frequency)
> curves you are looking for.
>
> In my experience from low frequency (<3GHz) equalizers,
> it's one thing to
> get the gain you want at one frequency, but if you want
> specific gain (and
> phase) over a range of frequencies, it's a whole other
> story. Since we are
> dealing with broadband signals, I would imagine that this
> level of detail
> would be necessary. But please educate me on this.
>
> Therefore, could you please provide more detailed infomation
> (gain/phase-vs-frequency curves) about the equalization
> you're talking
> about. Or tell me why I don't need these details.
>
> Thanks,
> Henning
>