> I didn't place this in context until I read the news. I see, in their
> document, AT&T is complaining about lack of spectrum. Is there spectrum
> available, other than the band that LightSquared wants to develop? That
> band raises definite interference with many GPS devices, including those
> used for aircraft navigation.
Well, regarding the second issue, let me repeat: NEVER underestimate the
power of a number of well-placed geeks. Unfortunately, in this case, it
appears that the destruction of the GPS network is being pushed from the
White House. This is a cornerstone in the rural broadband initiative and he
isn't giving up on this.
As to spectrum, well, that is a major issue in regards to 4G (from here on,
I'll refer to LTE "Long Term Evolution", "LightSquared" and "Wimax" as
"4G"). Within the company I worked for, I was a huge advocate that we bid
on the AWS spectrum when it was being auctioned off. We had no ability to
do anything with it, but I knew it was going to be gold. We eventually sold
it at a decent profit. Verizon just spent a chunka change on AWS from Cox
which did the exact same thing. (Not that we didn't want to do anything
with it, but no equipment manufactures had anything to light it with, yet.
We had a time-table to do something with it. Catch-22.
Spectrum is indeed limited. And the data usage caps that most of the
carriers have won't be going away anytime soon. They MUST put those caps in
place because there just isn't enough bandwidth there to allow everybody to
be streaming from Netflix onto their smartphones, iPads and other 3G/4G
devices. This is a matter of physics. You can only squeeze so much brown
stuff through the colon.
Acquiring T-Mobile was all about two things: Customers and spectrum. There
is practically no increased footprint, although, it would have altered
the PRN tables a little bit. In high-density corridors, it would have given
AT&T more spectrum to work with. Essentially, what has happened is
that with 4G there are more customers than the network can support. AT&T
was trying to sell this to the authorities that it was going to allow them
to serve "Rural America" (see comments on LightSquared). For those of us in
the industry, this was a howler and thigh-slapper. Fortunately, the mass
ridicule forced them to back off on that assertion rather quickly.
As spectrum represents a real and fixed cap on capacity in what we call
"access", there are several ways of addressing this. Do note that I'm
specifying the "access" portion of wireless service. From the tower to
the world is what we call "backhaul" and that is scalable thanks to the
miracles of fiber-optics. Data traffic doubles about every 18-20 months.
This trend is historical back to the days of the telegraph, so I feel it is
safe to continue to use this figure in planning. With today's technology,
we can get as far as 2016 and we'll run out capability. Fortunately,
instead of 10GbE, we'll be deploying 40Gbe and 100GbE backhauls so no
worries.
The ways of addressing limited radio spectrum for 4G is as follows:
1. Using AWS and other spectrums. This includes the forklifting of much of
the 2G technology for all but the basic required-by-law services. Reuse and
upgrade of 2G/3G gear frees spectrum up for the 4G gear.
2. Narrower coverage antennas on the cell-towers. We've long since gone
from omni antennas to sectored antennas to hyper-sectored antennas which
allow the carriers to target specific streets or even blocks within a
street. The ability to reuse the same spectrum three times from the same
tower has been around a long time, but that will progress to dozens of
times. Not effective everywhere--especially where signal reflections are
present--i.e. in cities.
3. More towers. More towers with antennas of limited range.
4. Picocells. Already deployed in convention centers, schools, office
buildings, shopping malls, airports, train-stations and even airliners,
these are essentially cell towers that cover a small area with high-density
users. It is very easy to segment the zones with ceiling mounted antennas
that cover tiny areas. It's all a matter of user-to-cell ratio. The more
users you have, the more antennas you need. It is pretty much a known and
fixed ratio. Go to a big ball-game where tens of thousands of people are
present and the cell companies have rolled in temporary antenna arrays (and
a few million dollars in support equipment) to provide directionalized
coverage for each section in the stands. In almost all circumstances, the
roaming agreements do not apply to Picocells. There are now companies that
will be a single provider for these events and facilities and will sell
roaming access to the carriers on a contract basis.
5. Femtocells. I have one myself. These are smaller than a Picocell and is
designed to be placed in a home or office. To date, this has been used to
provide indoor coverage where signal strength is weak. The Femtocell plugs
into your own DSL connection and establishes a voice-over-packet (which is
how the cell-network works anyway) back to the MTSO. When you leave your
house, the phone should automatically switch over to the normal antennas.
More often than not, though, the current variety of Femtocells won't allow
an established call from outdoors to roam to the indoor connection. With a
Femtocell, any data traffic is automatically moved to the DSL connection
and kept off the cell-towers. Do a traceroute to see what your Femtocell
does.
6. WiFi. You'd think that the WiFi capability of your smartphone is a
convenience to you. Although it is, the primary reason is because the cell
companies want to get that data traffic off their network as much as
possible. It is a massive advantage for them to deload the network, which
means passing off the load onto ANYTHING else. Which leads to point 7.
7. Carrier provided WiFi. The dirty little secret is that one particular
cell company has been doing this for years. On the cell towers is also a
WiFi access point. The SSID is not broadcast and it is kept relatively
private. The handsets, though, are programmed to establish the link and use
it for data. It is hit-and-miss, but for some situations it works quite
well. Wardrivers know all about this. Many picocells are co-equipped with
WiFi. Much of this network was turned down when the latest varients of 3G
were released. This one carrier I'm referring to had to do this initially
because they acquired a huge 2G network of junk and couldn't afford to
upgrade everything to 3G.
I believe that in a year or so, we'll start seeing a massive deployment of
combination Femtocell/WiFi devices from the carriers. These will be sold as
an "advantage" to the user, but the primary advantage is to the carrier. I
most certainly would not discourage anybody from getting them, but
understand that the carriers are hugely concerned about deloading as much
as they can off their antennas.
I've been asked, recently, if 4G is much faster than 3G. Of course, this
depends on the technology and generation of 3G. But the reason why 4G is
faster isn't so much that there is more bandwidth in the access portion
(handset to tower), but because the backhaul has been heavily increased as
part of the 4G deployments (The carriers are going from 5-10mbps per
backhaul to 50-200 mbps per backhaul). Granted, it IS FASTER, but in the
access portion, this is totally a shared network. If you were the only
person on-line, you're speed would be the maximum that the radio allows.
But a second user doing the same thing you are will cut down the available
bandwidth to you. Basically, what it does is allow MORE people to have a
satisfactory experience. The fact is, today's 3G devices have maxed out the
3G network (which every AT&T user is painfully aware of). Going to 4G will
give everybody a little breathing room, but that's just temporary. More
customers and we'll be right back were we started from. Also note, that
voice ALWAYS takes priority.
My new cellphone (Nexus S 4G) has some capabilies my old Palm phone didn't.
One of them is the Android OS with the availability of some seriously
powerful apps for people like me who work in the industry. I've been
running speed tests on various towers. The EV-DO 3.75 site serving my
location is configured for about 3mbps speeds. Depending on time of day,
(user load), my throughput ranges from just shy of 3mbps to about 50kbps.
What 4G will give is a theoretical max of 128mbps to upward of 1Gbps in
extremely limited and isolated usages. Those speeds come at a fantastical
price in spectrum and distance limits. 2G (and varients) are time-slot
based and served out in 56kbps-ish chunks. But it doesn't matter if it's
2G, 3G or 4G if the backhaul is limited. 4G may be capable of some
incredible speeds, but the backhaul network is leased from other companies
and they charge dearly.
Anything else?
AG (and so on) Schnozz
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|