Dan,
 
I certainly won't defend the business decisions that were made by the
Clarity/Cisco leadership, and I'm really reluctant to challenge you
since you were very nice to take me to lunch today (thanks!), but I've
got a couple of nits to pick about your points!

> 1]  Cisco in those days was arrogant enough to bypass the IEEE, just as 
> the Bluetooth SIG had done several years earlier, in both cases to 
> their sorrow.  The customers I spoke to preferred to wait for IEEE 
> 802.16 so that they could have some confidence in upgradeability and 
> multi-vendor compatibility.

I don't believe that being 802.16 compliant was ever really a driving
factor in any of the big player's decision-making processes.
For example, Sprint was active in the 802.16 standards committees and
made noises about wanting standards-compliant gear, but they're still
not buying 802.16 gear even now, years later, when the standard has
finally been completed.
Instead the same Sprint folks are now even more active in the 802.20
standards committees.  
Sprint invested over a billion dollars in MMDS spectrum and has sat on
it for over five years.  There's something other than technology driving
their decisions.

Likewise for Worldcom/MCI, but we now know the reason they were dragging
their feet is that they weren't really making all the money they claimed
they were.

 
> 2]  The BWIF radios were rather expensive (we made them so I know), but 
> they were just the tip of the iceberg.  In order to make the system 
> work, a WISP customer had to hang a $70,000 Cisco router off of the 
> $500-$1000 radio, and then pay for expensive customer premises 
> equipment and siting.   If you get 100% uptake on your roughly 50 Mbps 
> capacity [my recollection -- I don't have any docs on this any more]   
> and you can sell it for $1K/year/Mbps, you get a revenue of $50K/year 
> from a basestation.  That puts a small WISP in the position of spending 
> perhaps $100K to $150K to gain a revenue stream of $50K per year 
> exposed to competition and requiring some support expenditures as well. 
>   The economics are much better for an 802.11-based solution.

First, the initial pt-mpt radios were explicitely targetted at business
customers who frequently pay around $1k/month for T1 service (I know
this is still true because I'm currently paying close to that amount for
two different T1 lines right now!  And I had to pay $10k to get the
lines installed!)

Second, don't forget to add an oversubscription factor (I've been told
by some operators that an oversubscription factor of eight or so is not
unreasonable for small-to-medium businesses, and an oversubscription
factor of 60 or so is not unreasonable for residential users).  This is
all highly subjective and quite variable from operator to operator.

Third the downlink capacity of the pt-mpt gear was about 20 Mbps
(slightly less at the network layer due to MAC overhead).  So assuming
100% uptake from business users subscribing to 1-Mbps service at
$1k/month/Mbps you get a revenue of ($1k/mo)x(20 simultaneous 1-Mbps
users)x(8x oversubscription factor)x(12 mo/yr)= $1.92M/yr.

Now, that's a pretty darn rosey scenario, so reduce the number of
subscribers, reduce the amount you can charge for service (because the
local telephone companies are pretty good about dropping their
exorbitant prices whenever they get a whiff of competition!),and factor
in $50k-$100k for cell-site planning/zoning/construction work.  I think
you still come out with a reasonable ROI, and we haven't even considered
what additional ROI could be derived from adding lower-rate best-effort
service for residential customers (after the lower-cost ASIC-based
products materialized of course!)


Cheers,

Greg


> 
> 
> Daniel M. Dobkin
> Enigmatics
> 1-408-314-2769
> [EMAIL PROTECTED]
> 
> BAWUG:  if this is still coming in as an HTML stream, I apologize (I 
> can't find a menu selection to turn it off in my email program);  you 
> can just throw it away; oh well. --DMD
> 
> --
> general wireless list, a bawug thing <http://www.bawug.org/>
> [un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless


--
general wireless list, a bawug thing <http://www.bawug.org/>
[un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless

Reply via email to