AT&T Increases Data Cap In Face Of $100 Million FCC Fine

[Math Geek]

so the engineers of the mobile company are lying to us about the company lying to us and the engineers are wrong about being able to handle what the engineers say they can handle?? so overall, does this double negative cancel out and AT&T is actually doing the right thing by (not) lying to us about arbitrary data caps they keep hidden from customers until they feel it is time to throttle.....

ok now my head hearts, imma go lie down for a while......

 

[InvalidError]

Wireless technology is moving towards directional wireless communications which can "focus in" on signals coming from a certain direction. MIMO is a good example. Using 2 or more antennas, it can use a different multipath route between source and destination as an extra transmission channel, even though it's at the same frequency spectrum.

Except that the computational complexity grows exponentially with the number of antennas and endpoints, so scaling is limited, assuming the matrix even has a solution that can simultaneously maintain adequate SNR to all endpoints without reducing the data rate.

And that's before accounting for environmental factors which can change much faster than they can be compensated for.

 

[itmoba]

This reminds me of parents who scream at telephone companies because their child racked up a $20,000+ bill by texting. Is it the telephone company's fault? Absolutely not.

I'd say phone companies deserve some of the blame for it:
1- they make it difficult if not impossible to set limits - service is an all-or-nothing offer, agree to unlimited billing or go shop elsewhere
2- they charge 100-10000X more than what data/text actually cost to provide in some of their pay-per-use plans and overage fees (for example, $1/5MB on the smallest data plans is $200/GB, easily 1000X more than costs, and $0.25/text is over $7500/MB, millions of times more than costs)
3- that sort of bill would be considered unconscionable by most people: nearly nobody in their right mind would agree to be held liable for that sort of phone bill if they actually knew there was a real possibility it could happen to them
4- most major bill shock cases get settled out of court shortly after they hit the news and social networks because carriers know they will most likely lose if the cases are allowed to get through court, which would set very embarrassing precedents against them

As far as I am concerned, users should not be liable for any extra costs when they are not provided the ability to put a ceiling on how much they can be billed.

Yes, the bill is practically the wet dream of usury and extortion -- however, service providers typically settle out of court because of the legal fees incurred. Think about it: If a bill comes to the lofty sum of $5000, then, why have the legal team effectively work for, say, 20 hours on a case at a rate of, say, $150 per hour? That totals to $3000. The court will side with the service provider because the person with whom they entered into agreement is assumed to be responsible for their mistakes. Furthermore, there're additional expenditures on PR, and the negative backlash alone fiscally outweighs the paltry sum of $5000.

 

[InvalidError]

however, service providers typically settle out of court because of the legal fees incurred. Think about it: If a bill comes to the lofty sum of $5000, then, why have the legal team effectively work for, say, 20 hours on a case at a rate of, say, $150 per hour? That totals to $3000. The court will side with the service provider because the person with whom they entered into agreement is assumed to be responsible for their mistakes.

They are also settling out of court because they know that even if the courts might not outright null the bill due to lack of adequate usage notification and control mechanisms, they will drastically reduce it to something more sensible: even if the carrier is willing to spend $3000 in legal fees to attempt defending their $5000 bill for SMS usage, it makes no sense for the court to uphold it when the carrier's very own unlimited SMS option is $10/month. If the court case went all the way through, the bill would likely get reduced to something like $100 - a $90 penalty for lacking the foresight to order unlimited text before texting like mad.

How many sane people would rack up $500 personal phone, text and data bills if any extra charges had to be authorized as they are incurred? Almost nobody. Back when 1-900 scams were a plague, people could call their phone company and opt out from being able to call 1-900 numbers. Wireless carriers do not offer that simplest yet extremely effective option because they are earning too much money from unintentionally incurred charges.

The only reason honest people get bill shock is because carriers are not providing the tools people need to properly protect themselves. Usage notifications are a nice start but they do you little good when the stats are running a few days late.

If I had the option to, I would not pre-authorize any charges beyond the wireless plan's nominal monthly fee.

 

[alidan]

Where are you going to find enough spectrum to provide everyone with wireline-like speed without having to make individual cells so small that you would need cellular masts on nearly every other building? The build and maintenance costs would be unsustainable. Nobody can do it even if they wanted to, it is practically and economically unfeasible.

Wireless technology is moving towards directional wireless communications which can "focus in" on signals coming from a certain direction. MIMO is a good example. Using 2 or more antennas, it can use a different multipath route between source and destination as an extra transmission channel, even though it's at the same frequency spectrum.

The ultimate will be phased array antennas, which allow a receiver (and transmitter if you know where the intended recipient is) to electronically "point" the antenna in a certain direction. Once you move to point-to-point wireless communications, most of the spectrum limitations vanish.

the engineers from phone/cable companies have admitted many times that data caps are nothing but a cash grab. they have explained numerous times that the network could easily handle multiple times more traffic than they face. about the only place the network is truly taxed is in very crowded cities like NYC. but they have even managed to get enough in place to handle the loads they ask for.

I can't speak for phone, but if you compare cable Internet prices with the price for an equivalent dedicated line, it's priced about right. e.g. An OC3 (155 Mbps fiber) costs about $20k-$45k/mo. Go with the lower price since we're talking wholesale. 155 Mbps * 1 month = 50.95 terabytes/mo. At a 300 GB/mo cap, that's 170 customers you can squeeze into that single OC3 line. $20,000/170 customers = $117.64 per month per customer.

Or put another way, you don't expect everyone to hit the 300 GB/mo cap. So you charge customers $80/mo and put 250 customers on that single OC3 with a 300 GB/mo cap. As long as average utilization stays below 204 GB/mo per customer, you're ok. (FWIW, I'm on a $60/mo 100 Mbps plan with a 700 GB/mo cap, so my cable company is actually charging me significantly less than this. If I were to actually use all 700 GB/mo, my consumption of an OC3's capacity would be priced at $275/mo)

So no, their lines can't handle multiple times more traffic. The price is about right. Where the cable companies are gouging you are in channel bundles and set top box rentals.

japan has gigabit up and down as the norm for 20$ i refuse to believe it costs as much as you say, we are getting screwed hard somewhere. i also know that municipal internet what it was made and not illegal, was able to provide internet for FAR cheaper than 100~$ with a 300gb cap and it ranks in the top 10 fastest in the world.

 

[InvalidError]

japan has gigabit up and down as the norm for 20$ i refuse to believe it costs as much as you say, we are getting screwed hard somewhere.

AFAIK, the cheapest 1Gbps access is HKBN (Hong Kong) and it is ~$35/month. However, although the MAC link speed might be 1Gbps, actual speeds to "on-net" sites are more like 200-300Mbps and once you leave sites that have direct peering or hosting with HKBN, speeds can drop well under 100Mbps.

If an ISP in North America tried to sell 1Gbps for $35/month where typical performance is around 30% of advertised even for sites directly connected to the ISP and 5-10X slower for everything else, people would be screaming bloody murder.

Having to overbuild the network to sustain higher worst-case speeds makes things substantially more expensive.

 

[alidan]

japan has gigabit up and down as the norm for 20$ i refuse to believe it costs as much as you say, we are getting screwed hard somewhere.

AFAIK, the cheapest 1Gbps access is HKBN (Hong Kong) and it is ~$35/month. However, although the MAC link speed might be 1Gbps, actual speeds to "on-net" sites are more like 200-300Mbps and once you leave sites that have direct peering or hosting with HKBN, speeds can drop well under 100Mbps.

If an ISP in North America tried to sell 1Gbps for $35/month where typical performance is around 30% of advertised even for sites directly connected to the ISP and 5-10X slower for everything else, people would be screaming bloody murder.

Having to overbuild the network to sustain higher worst-case speeds makes things substantially more expensive.

could be true, i just remember what charts said a few years back for service and its cost because that really stuck out to me. but i also remember a russian telephone company who made a contest of who could use the most bandwidth, duded used 419gb and the same companies cheapest plan (even the cheapest plan was unlimited internet) was 33$ a month, in america that would be what, a couple grand charge?

looking at the rest of the world i honestly believe we are getting screwed at every level and things do not cost as much as they say, they just refuse to spend money on upgrading infrastructure. i'm just thinking of things like how hollywood movies never make a profit due due to how they handle accounting, the distribution houses charge something like 100 billion a movie or along those lines, something absolutely insane so if they ever sucker someone into a % of the movies proffit, they can stiff them because the movie never made a profit, though the higher ups own both production and distribution.

 

[InvalidError]

looking at the rest of the world i honestly believe we are getting screwed at every level and things do not cost as much as they say, they just refuse to spend money on upgrading infrastructure.

Try searching for quotes on pulling fiber cables through places like downtown New-York - you should be able to find a few since NY had a public RFQ sessions for tons of fiber network building stuff a few years ago. That will give a pretty good idea of how ridiculously quickly network build costs can pile up. Just pulling the cables through neighborhood can cost over $300 per meter.

 

[uglyduckling81]

japan has gigabit up and down as the norm for 20$ i refuse to believe it costs as much as you say, we are getting screwed hard somewhere.

If an ISP in North America tried to sell 1Gbps for $35/month where typical performance is around 30% of advertised even for sites directly connected to the ISP and 5-10X slower for everything else, people would be screaming bloody murder.

Having to overbuild the network to sustain higher worst-case speeds makes things substantially more expensive.

Australians are all sold ADSL 2 over here and it's all advertised as 20-24Mbps. The average Australian speed actually obtained is 2Mbps. That's downloads. Uploads they advertise 1Mbps so you can probably guess what speeds we are actually getting generally.

Not a damn thing anyone can do about their BS advertising and actual speeds being delivered. I wouldn't be complaining too much about only getting 300Mbps for $35. My ADSL costs $60/month which is a cheap plan over here.

Australia started rolling out a National Broadband Network which was going to be fibre to the home for everyone but even that was capped at 100Mbps download.
Conservative employees of Rupert Murdock.... Oh I mean our right wing government squashed that plan though just after it started. God forbid Australians get Internet capable of streaming video, it might take a cut out of Murdock's pay TV empire.

 

[InvalidError]

Australia started rolling out a National Broadband Network which was going to be fibre to the home for everyone but even that was capped at 100Mbps download.

I like the NBN: it is a great cautionary tale for people who wish broadband infrastructure got nationalized in their country. It sounds great and perfectly logical in theory - nearly all infrastructure should be some form of regulated monopoly for cost efficiency reasons and much of it already is in most countries - but does not necessarily lead to more affordable and better service for everyone after political factors weigh in.

 

[Solandri]

Wireless technology is moving towards directional wireless communications which can "focus in" on signals coming from a certain direction. MIMO is a good example. Using 2 or more antennas, it can use a different multipath route between source and destination as an extra transmission channel, even though it's at the same frequency spectrum.

Except that the computational complexity grows exponentially with the number of antennas and endpoints, so scaling is limited, assuming the matrix even has a solution that can simultaneously maintain adequate SNR to all endpoints without reducing the data rate.

And that's before accounting for environmental factors which can change much faster than they can be compensated for.

The complexity grows linearly with the number of antennas. All you do is shift the phase of the arriving signal based on the linear offset from the adjacent antenna (assuming the antennas are arranged in a line or a plane). The amount of phase shift determines the angle at which you're "pointing" the antenna. This is 1950s technology. They used to do it with analog circuits. Computers just allow you to rapidly recalculate for a different phase offset, allowing you to quickly change the direction the antenna is "pointed" so you can "scan" to build up a 3D map of RF signals. Basically create a "photo" of RF signal sources.

https://en.wikipedia.org/wiki/Phased_array

japan has gigabit up and down as the norm for 20$ i refuse to believe it costs as much as you say,

Japan is much more densely populated than the U.S., meaning the amount of infrastructure needed per connection is substantially less. Hence the cost per connection is a lot less.

we are getting screwed hard somewhere.

If you really, truly believe it can be done a lot cheaper, then go for it. Start your own ISP and sell service for substantially less than the cable companies. If you're right, people will flock to your much-cheaper Internet service, and you'll put the cable company out of the Internet business. Put your money where your mouth is.

 

[InvalidError]

The complexity grows linearly with the number of antennas. All you do is shift the phase of the arriving signal based on the linear offset from the adjacent antenna (assuming the antennas are arranged in a line or a plane). The amount of phase shift determines the angle at which you're "pointing" the antenna.

That is only true when you need to phase the array to focus on a single point. In a cellular network, you need to find coefficients that will simultaneously optimize signal for every data stream and every receiver in the area.

 

[Co BIY]

I am skeptical of the big carriers and their motivations but in this case I am more skeptical of the FCC and their lack of transparency. Why are they not simply putting out a rule that lays out what carriers have to make clear in their TOS. (Like clearly posting the speed limit before police enforcement.)

The use of hyperbolic language (unlimited) in advertising is accepted with the understanding that reasonable people would not think it is actually 100% unlimited. If only a small percentage of users ever run into the caps and the service is essentially "unlimited" for most users then it looks pretty reasonable. The FCC could easily rule that "Unlimited" needs an * and attached fine print that says "monthly data capped at X service beyond that provided at Xmbs. An average of X% of users hit the cap each month over the last 12 months." They could also rule that "Unlimited" can't be used with any restrictions at all.

If the FCC made one clear rule that applied to all carriers it would be much better than fines after the fact. The situation right now stands that even if a carrier thought that "Unlimited" plans were bad practice or could subject itself to FCC action it is under intense competitive pressure to use the term because others are doing so. (Unlimited, Free, ect... are powerful marketing themes so one business can't easily unilaterally disarm.)

The fact that the FCC has levied the huge fine but that there is the strong possibility that it will be waived at the whim of the politically appointed FCC commissioner makes one wonder at the motives involved. Is there pressure on Verizon to make certain political donations to show that their "heart is in the right place"?

The FCC should be laying down clear rules that apply to all carriers to curb practices it sees as damaging to consumers rather than making a show of large fines that may never actually be levied depending on standards that are not clearly articulated.

AT&T's political contributions are shifting with the political winds. Is that a factor in the timing of these FCC fines?

http://www.opensecrets.org/orgs/summary.php?id=d000000076&cycle=A