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

[EFM] RE: Relative OSP Costs of PON vs. P2P

Hello Gerry and all,
The slide that was presented at Network Interop by Ross Ireland of SBC that
Gerry refers to presents the infrastructure cost(materials and labor) only
(fiber interface at the CO to fiber interface at the Home-no electronics).
Compared are  point to point using a single fiber(not a pair) and PON with a
1:32 way split. SBC worked with Corning to model the infrastructures that
helped produce the results.  Each operator has different fiber costs, fiber
construction procedures, and different labor costs that will vary the
results somewhat. One thing is certain, PON grows in cost savings over point
to point as your deployment distance increases. 

Best Regards, Kent
Kent McCammon

-----Original Message-----
From: gerry.pesavento@xxxxxxxxxxxx [mailto:gerry.pesavento@xxxxxxxxxxxx]
Sent: Wednesday, June 06, 2001 7:48 PM
Cc: DuX@xxxxxxxxxxx; FengW@xxxxxxxxxxx; JayJA@xxxxxxxxxxx;
KunziAL@xxxxxxxxxxx; MusgroveKD@xxxxxxxxxxx;
JPropst@xxxxxxxxxxxxxxxxxxxx; ShanemanK@xxxxxxxxxxx;
Subject: RE: Relative OSP Costs of PON vs. P2P

Hi Robert,

Thanks for the PTP and PTMP numbers.  As you may know, SBC also completed a
PTP vs PTMP study that can be found here (see Ross Ireland's presentation):,2951,35,00.html

As you can see from the graph, the PTP vs PTMP costs are sensitive to
distance - SBC calculated PTMP is close to the same at short distance, and
50% the cost at >5 km.  I'd like to know more about what is behind SBC's
data (if it includes equipment costs, I think so). I noticed that neither
you nor Martin Adams mentioned distance; an important variable. 

Another general comment; equipment and OSP deployment costs are important,
but for the carrier, operation and maintenance costs are probably more

It may be tough to estimate hardware costs for PTP vs. PTMP.  A few speakers
said 2-3x for the PTMP transceiver. That might be right, but it is up to
volume and the component mfgs; that is what makes it difficult to estimate.
The rest of the hardware equipment should be on par with carrier-class
Ethernet switching gear.  

On the optical budget, I agree with Meir. The ONU will likely be 1310 nm, at
1 GBE off-the-shelf detector sensitivity is -24 dBm, but -26 dBm is
Gerry Pesavento

-----Original Message-----
From: Carlisle, Robert S [mailto:CarlisleRS@xxxxxxxxxxx]
Sent: Wednesday, June 06, 2001 12:03 PM
To: ''
Cc: Du, Xueping; Feng, Weiwei; Jay, John; Kunzi, Anne; Musgrove,
Kendall; Propst, Jeanne; Shaneman, Keith; Sweazey, Chad
Subject: FW: Relative OSP Costs of PON vs. P2P

The other reflector set up for discussing PON vs P2P costs doesn't seem to
be working, so I'm sending this to this forum for dissemination and


Rob Carlisle

> ----------
> From: 	Carlisle, Robert S
> Sent: 	Wednesday, June 06, 2001 12:58 PM
> To: 	'efmoptics@xxxxxxxxxxxxxxxxxxxxxxx'
> Subject: 	Relative OSP Costs of PON vs. P2P
> All,
> I wanted to share some relative Outside Plant (OSP) costs based on
calculations from analysts here at HQ and my colleagues at Corning Cable
Systems.  Based on fixed electronics cost at the Head End (HE) and in the
customer premises (CPE), the relative cost of the OSP is about 41% of the
total cost for 1 fiber to the home and 52% of total cost for 2 fibers for a
Point to Point network.  For a PON, the relative cost of the OSP is about
13-18% of the total cost.  In other words, OSP costs for a P2P network are
roughly 2X the cost for a PON.  OSP costs include the feeder and
distribution cables, splices and connectors, trenching and/or drops, and
splice enclosures.  Of course, as the cost of the electronics at the HE and
at the CPE drop, the relative cost of the OSP will go up.  What I don't know
and would like to see some discussion on, is the difference in the cost of
the electronics in the HE for a P2P vs. PON.  I expect that since the
management functions of a PON are more complex, the costs of the electr
> I also wanted to share an Excel graph showing the upstream reach of a 1500
nm FP in a 1X16 split EPON.  Assumptions made:  
> *	21 dBm power budget between the TX and RX (0 dBm TX and -21 dBm
receiver sensitivity @ 1 GBPS).
> *	Losses to splices, split, connectors, and 4 dB link margin - 20.4 dB
> *	Spectral width of FP laser - 5 nm
> *	Attenuation of fiber @ 1500 nm - .21 dB/km (vice .34 at 1310)
> Take aways from this:
> Dispersion limited using standard single mode fiber so consider using
dispersion shifted fiber if desirable to use cheaper FP laser in CPE.
> Does not reach the sweet spot of 10 km @ 1 GBPS so:
> 	Must live with lower link margin or
> 	Increase power of laser or
> 	Increase sensitivity of receiver
> Is this technically feasible?
> >  <<exportablelink.xls>> 
> Best Regards,
> Rob Carlisle
> Rob Carlisle
> Senior Market Development Engineer
> Optical Fiber
> v: 607 974-6806
> f: 607 974-7522
> c:607 368-5442