Verizon leaping ahead with EV-DO

[email protected] wrote in message news:<[email protected]>...
> On Thu, 8 Apr 2004 01:20:55 -0400, "Donkey Agony" <root@[127.0.0.1]>
> wrote:
>
> >http://ptech.wsj.com/ptech.html
>
> Rolling out EV-DO is an interesting and risky venture for multiple
> reasons.
>
> Part of the carrier's spectrum is dedicated to data only with EV-DO.
> This implies that, even under "emergency loads", which the FCC is
> encouraging carriers to handle, that EV-DO spectrum is "untouchable"
> for voice.
>
> The carriers using EV-DO hope that that data-only spectrum will be
> well utilized. That remains to be seen and is another risk.

Currently, VZW's trial deployments in D.C. & San Diego are both
CDMA1xEV-DO 1900. Undoubtedly, D.C. & San Diego were selected as test
markets due to two factors: VZW holds a PCS D or PCS E 10 MHz license
overlapping its Cellular B-side license in each market, and D.C.'s BAM
Lucent infrastructure versus San Diego's AirTouch Nortel
infrastructure allows VZW to test both vendor's EV-DO solutions. The
more important conclusion, however, is that VZW has currently
relegated EV-DO to PCS spectrum that was of little if any consequence
to its voice or 1xRTT data capacity. However, VZW does not possess
coincident PCS licenses across all of its major Cellular markets (no
supplementary PCS spectrum in Charlotte, Denver, Detroit,
Indianapolis, Phoenix, Portland, St. Louis, San Francisco, Seattle, et
al.) Thus, particularly in those markets, voice/1xRTT capacity could
potentially suffer if existing CDMA1x 800 carrier channels are
subsequently
devoted to EV-DO.

FYI, if you are interested to know in which markets VZW does hold PCS
spectrum, the maps that XFF & I have created are well-documented:

http://people.ku.edu/~cinema/wireless/vzw_pcs.html
http://people.ku.edu/~cinema/wireles...pcs_block.html

> One would think that Verizon Wireless would, eventually, like to
> migrate their especially heavy peak-load voice areas to EV-DV for the
> gains in voice capacity. Since EV-DV is also data capable,
> simultaneously, in the same spectrum, does EV-DO have a long term
> future for carriers that don't have entirely separate frequency ranges
> for it?

Assuredly, 1xEV-DV - in the true spirit of CDMA - is the more
efficient technology at balancing both voice & data capacity. Though
there is some irony involved in the following statement - since both
EV-DV & EV-DO use time-division scheduling techniques to dramatically
improve Ec/Io thus also increase instantaneous data throughput - the
hubristic flaw of EV-DO reminds me much of the inefficiency of GSM or
IS-136 TDMA. In CDMA, unused capacity - in the form of power - can be
fully distributed to other active users. But in the TDMA
air-interfaces, unused timeslots sit inefficiently vacant. Since
EV-DO dedicates carrier channels entirely to data, any unusued data
capacity also sits inefficaciously disengaged and unavailable to
voice, regardless of proportional demand.

My primary concern - for both EV-DO & EV-DV - is that neither will
ultimately be able to deliver upon promised broadband-like speeds.
Shannon's classic equation shows us that capacity cannot simply be
manufactured via technology alone. Increased capacity can only be
distributed via fundamental improvements in bandwidth,
signal-to-noise, or to a lesser degree also modulation agility - the
last of which is closely-tied to signal-to-noise.

The 1.2288 MHz channel at hypothetically 6 dB S/N can support a
maximum capacity of approximately 2.85 Mbps. If the S/N can somehow
be refined to 9 dB, theoretical capacity jumps to about 3.89 Mbps.
Regardless, however, that capacity will be divided amongst all users
of the channel. The greater the number of users, the less throughput
available to each individual user. The overall channel capacity may
be in the multi-megabit range - which is an impressive result over
little more than a 1 MHz wireless channel - but no single user will
approach that speed.

In a nutshell, there is no alchemy involved in either EV-DO or EV-DV.
Neither is a magic bullet. As the number of users of 3G data services
will increase, the multi-megabit capacity will be continually
subdivided into smaller & smaller individual pieces. As usage
multiplies, only increased bandwidth - additional 3G spectrum
allocation - or improved signal-to-noise - more frequent spatial
re-use (i.e. cell-splitting) - will maintain broadband-like 3G data
capacity per user.

Andrew
--
Andrew Shepherd
[email protected]
[email protected]
http://www.ku.edu/home/cinema/

---

› See More: Verizon leaping ahead with EV-DO

[email protected] wrote in message news:<[email protected]>...
>
> One would think that Verizon Wireless would, eventually, like to
> migrate their especially heavy peak-load voice areas to EV-DV for the
> gains in voice capacity. Since EV-DV is also data capable,
> simultaneously, in the same spectrum, does EV-DO have a long term
> future for carriers that don't have entirely separate frequency ranges
> for it?

Another reason why Sprint PCS - unlike VZW - will be bypassing 1xEV-DO
and instead evolving directly to the greater efficiencies &
equilibrium of 1xEV-DV is the character of SPCS' spectrum assets.

VZW, as the majority of its primary licenses are either Cellular 25
MHz or PCS 30 MHz, averages about 30-35 MHz of combined Cellular &/or
PCS spectrum per market. VZW has no major markets where it controls
less than 25 MHz, no major markets where it holds only a 10 MHz PCS
license.

In comparison, SPCS averages approximately 20 MHz of purely PCS
spectrum per market, including several notable markets (Cincinnati,
Dayton, Norfolk, Richmond, et al.) where SPCS still controls only a
single 10 MHz PCS D or PCS E license. The 10 MHz "Achilles' heel"
markets were previously even more prevalent, but SPCS has been
continually leveraging some of its PCS A or PCS B 30 MHz strongholds,
partitioning &/or disaggregating 10 MHz portions of those licenses in
spectrum transactions w/ primarily AT&TWS, receiving in exchange
supplemental 10 MHz blocks upping the spectrum ante to a more
consistent 20 MHz in notable markets (Atlanta, Charlotte, Cleveland,
Houston, Jacksonville, Memphis, Raleigh-Durham, et al.) for which
SPCS' had won at FCC auction only one 10 MHz license.

The primary point, however, is that w/ some 10 MHz markets yet
remaining, SPCS likely could not deploy EV-DO truly nationwide across
those markets. Only three 1.25 MHz CDMA carrier channels can be
deployed w/in a 10 MHz block. SPCS would have to devote or even
reassign to EV-DO at least one of the three potential CDMA carriers in
those markets at the consequential expense of 1xRTT voice/data
capacity. EV-DV is the solution. All protocols - IS-95A, 1xRTT,
EV-DV - and all services - voice, CSD, packet-data, VoIP - are
supported over each & every EV-DV carrier, providing the necessary
spectral efficiency even in those 10 MHz "Achilles' heel" markets.

Andrew
--
Andrew Shepherd
[email protected]
[email protected]
http://www.ku.edu/home/cinema/

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

O/Siris <NON-ASCII EMAIL> wrote:
>
> Interesting. 4 months old, but interesting. And I'll believe it
> when it's rolled out fully. It's easy to say you're doing it. it's
> another to actually do it.
>

I do believe that Sprint PCS also has plans to roll out this technology.
As we already know, Sprint likes to do all cities at once before
announcing a service rather than city by city as Verizon does.

- --

Thomas T. Veldhouse
Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (FreeBSD)

iD8DBQFAdp0u1p0e3NXsrtERAtayAJsGPQMwOP+EbaqrN0//oElpl20kpQCfdLSv
aSLKUYoCqIY+K/OKE530hxU=
=iMH6
-----END PGP SIGNATURE-----

NNTP-Posting-Host: thrush.cc.ku.edu
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Newsreader: Forte Free Agent 1.92/32.572
Xref: news.newshosting.com alt.cellular.sprintpcs:138219

On Thu, 08 Apr 2004 18:23:44 GMT, "Robert M." <[email protected]>
wrote:

>In article <[email protected]>,
> [email protected] wrote:
>
>> Why would a carrier "push" the much higher bit rate of VoIP on your
>> their wireless data network when they have an optimized solution
>> already on their wireless voice network?
>
>one word answer
>
>Capacity

So, they have to replace everyone's phone, re-engineer their entire
data network to give priority to VoIP (thereby potentially knocking
out data users) and up the capacity on their land-side data lines from
each tower.

I think it will be much cheaper to go EV-DV, and they'll get a larger
voice capacity than your alternative!

On Fri, 09 Apr 2004 04:09:44 GMT, O/Siris <0siris@sprīntpcs.com>
wrote:

>In article <[email protected]>,
>[email protected] says...
>> Why would a carrier "push" the much higher bit rate of VoIP on your
>> their wireless data network when they have an optimized solution
>> already on their wireless voice network?
>>
>
>You *do* remember, don't you, that PTT and Ready Link are both IP
>voice services?

And it "plays well" with voice, since it shares the same spectrum via
cdma2000 1x. Another "vote" for EV-DV.

In article <[email protected]>,
[email protected] wrote:

> On Thu, 08 Apr 2004 18:23:44 GMT, "Robert M." <[email protected]>
> wrote:
>
> >In article <[email protected]>,
> > [email protected] wrote:
> >
> >> Why would a carrier "push" the much higher bit rate of VoIP on your
> >> their wireless data network when they have an optimized solution
> >> already on their wireless voice network?
> >
> >one word answer
> >
> >Capacity
>
> So, they have to replace everyone's phone, re-engineer their entire
> data network to give priority to VoIP (thereby potentially knocking
> out data users) and up the capacity on their land-side data lines from
> each tower.
>
> I think it will be much cheaper to go EV-DV, and they'll get a larger
> voice capacity than your alternative!

Who said replace everything with VoIP? They can just add **some** VoIP
phones to even out capacity issues.

On 8 Apr 2004 21:57:14 -0700, [email protected] (Andrew Shepherd) wrote:

>[email protected] wrote in message news:<[email protected]>...
>The 1.2288 MHz channel at hypothetically 6 dB S/N can support a
>maximum capacity of approximately 2.85 Mbps. If the S/N can somehow
>be refined to 9 dB, theoretical capacity jumps to about 3.89 Mbps.
>Regardless, however, that capacity will be divided amongst all users
>of the channel. The greater the number of users, the less throughput
>available to each individual user. The overall channel capacity may
>be in the multi-megabit range - which is an impressive result over
>little more than a 1 MHz wireless channel - but no single user will
>approach that speed.

Hi, Andrew. The difference between the "old" TDMA (regular Mux) and
EV-DV (stat-Mux) is that if a mobile doesn't have anything "to say"
TDMA will still reserve the slot. EV-DV doesn't "reserve" the slot.
There are up to 64 ACTIVE users (not total users) per "frame". So, if
there aren't 64 active users other mobiles with "more to say" can take
up those slots within the frame.

Now, whether the internet path to the customer's data from the
wireless carrier's infrastructure can support a sustained
multi-megabit throughput will be a major factor in determining whether
the fastest throughput is available. This is not the responsibility of
the wireless carrier, however.

Migrating verizons? Get my shotgun!!


John R. Copeland wrote:
> "Steven J Sobol" <[email protected]> wrote in message news:[email protected]...
>
>>Both Sprint and Verizon will be migrating to EV-DV, and you can expect
>>a flurry of press releases when that happens.
>>
>>--
>>Steven J. Sobol
>
>
> I always try to say that Verizon is migrating to EV-DV,
> but Sprint is *leapfrogging* to EV-DV.
> ---JRC---
>

On Fri, 09 Apr 2004 13:13:02 GMT, "Robert M." <[email protected]>
wrote:

>> So, they have to replace everyone's phone, re-engineer their entire
>> data network to give priority to VoIP (thereby potentially knocking
>> out data users) and up the capacity on their land-side data lines from
>> each tower.
>>
>> I think it will be much cheaper to go EV-DV, and they'll get a larger
>> voice capacity than your alternative!
>
>Who said replace everything with VoIP? They can just add **some** VoIP
>phones to even out capacity issues.

If you want to do your best to try to get the most voice users through
during an emergency, you'll have to replace as many phones as you can,
if you want to use the data carrier (via VoIP) in an emergency.

With your suggestion the re-engineering of the entire data network
still would have to be done, along with some increase in capacity of
the land-side data lines.

In article <[email protected]>,
Jerome Zelinske <[email protected]> wrote:

> > "Steven J Sobol" <[email protected]> wrote in message
> > news:[email protected]...
> >
> >>Both Sprint and Verizon will be migrating to EV-DV, and you can expect
> >>a flurry of press releases when that happens.
> >>
> >>--
> >>Steven J. Sobol

With Sprint likely a flurry of press releases long before its released;
kinda like the press releases announcing the release of a Nokia phone
that was no where to be found, (even now, 3 weeks later)

In article <[email protected]>,
[email protected] wrote:

> On Fri, 09 Apr 2004 13:13:02 GMT, "Robert M." <[email protected]>
> wrote:
>
> >> So, they have to replace everyone's phone, re-engineer their entire
> >> data network to give priority to VoIP (thereby potentially knocking
> >> out data users) and up the capacity on their land-side data lines from
> >> each tower.
> >>
> >> I think it will be much cheaper to go EV-DV, and they'll get a larger
> >> voice capacity than your alternative!
> >
> >Who said replace everything with VoIP? They can just add **some** VoIP
> >phones to even out capacity issues.
>
> If you want to do your best to try to get the most voice users through
> during an emergency, you'll have to replace as many phones as you can,
> if you want to use the data carrier (via VoIP) in an emergency.
>
> With your suggestion the re-engineering of the entire data network
> still would have to be done, along with some increase in capacity of
> the land-side data lines.

If the data network takes IP, then its not any re-engineering. IP is IP.

On Fri, 09 Apr 2004 16:21:58 GMT, "Robert M." <[email protected]>
wrote:

>In article <[email protected]>,
> [email protected] wrote:
>

>If the data network takes IP, then its not any re-engineering. IP is IP.

Voice demands more stringent network service than best effort data.
$$$ (These things need to be thought about, beforehand!)

From the customer service side: think of the support nightmare:

Customer: "Hi, my call just droppedl".
Service Rep: "Where were you, and what time was it?"
Customer gives appropriate info.
Service Rep: "Was your phone doing VoIP (for load balancing) or was it
using our regular voice network?"

The customer probably won't understand the question, no less know how
to respond.

In article <[email protected]>,
[email protected] wrote:

> Service Rep: "Was your phone doing VoIP (for load balancing) or was it
> using our regular voice network?"
>
> The customer probably won't understand the question, no less know how
> to respond.

Thats doing it your way; which is not what I suggested.

What I suggested is selling maybe one model phone of 20 that would do
VoIP.

On Fri, 09 Apr 2004 16:52:32 GMT, "Robert M." <[email protected]>
wrote:

>In article <[email protected]>,
> [email protected] wrote:
>
>> Service Rep: "Was your phone doing VoIP (for load balancing) or was it
>> using our regular voice network?"
>>
>> The customer probably won't understand the question, no less know how
>> to respond.
>
>Thats doing it your way; which is not what I suggested.
>
>What I suggested is selling maybe one model phone of 20 that would do
>VoIP.

Yet the costly land-side network issues remain. And it still wouldn't
solve the "emergency capacity" problem.

It complicates matters more than it's worth!

In article <[email protected]>,
[email protected] wrote:

> On Fri, 09 Apr 2004 16:52:32 GMT, "Robert M." <[email protected]>
> wrote:
>
> >In article <[email protected]>,
> > [email protected] wrote:
> >
> >> Service Rep: "Was your phone doing VoIP (for load balancing) or was it
> >> using our regular voice network?"
> >>
> >> The customer probably won't understand the question, no less know how
> >> to respond.
> >
> >Thats doing it your way; which is not what I suggested.
> >
> >What I suggested is selling maybe one model phone of 20 that would do
> >VoIP.
>
> Yet the costly land-side network issues remain. And it still wouldn't
> solve the "emergency capacity" problem.
>
> It complicates matters more than it's worth!

Of course, you say that and then its a self fulfilling prophecy.