Google fiber is about to roll out in Austin. Hubby and I are looking forward to getting free 60 thingies if we purchase a $300 google modem....kids are wondering what it would be like to play stupidonlinegame at 100megathings a second.

Suddenly my Timewarner rip-off 15mega whatever is up to 60 thingies. Kids asked me if I'd started paying more for the internet or I'd never have noticed.

If you were an ISP charging a flat rate then you're missing on the chance to stiff your customers for more money
(There are flat rate ISP's though...but I'm sure they keep their eye on the pricing models.)

The large ISP's often sue to prevent competition even when local governments are trying to introduce it.
For example Municipal Broadband.

http://consumerist.com/2014/08/28/ho...at-block-them/

Oh, I may miss out on the opportunity to gouge customers, but I most certainly wouldn't be missing out on the chance to beat the holy crap out of my competitors, take all their customers, and walk away with their profits.

That's called competition. It's the same thing that allows you to afford cars, flying on an airplane, "long distance" phone calls, computers, mobile phones, the internet, etc.. (all of which were originally very expensive and reserved solely for the elite, but you might be too young to remember that).

OK, that article seems to suggest it's an infrastructure issue. One company invests in the infrastructure in a city, and that's who owns broadband in that town, and therefore a monopoly without competition.

That answers my question. Thank you.

I wish

And...I do have my own company

I'm afraid it's not that simple, even if internet (and phone costs) in the US are a rip-off.

The usual relevant parameter for customer bandwidth, along with the stated connection speed, is "contention ratio", which is defined as the ratio of the potential maximum demand to the actual bandwidth (individual max speed x number of users / total system bandwidth). Cheap ISPs have contention ratios of, say, 50:1, and better ISPs have nearer to 20:1 or 30:1. Contention costs money, because the ISP has to have higher total system bandwidth with the same number of paying customers.

A new-entrant ISP in the US that offered unlimited usage pricing only would attract high-bandwidth users, and they would need low contention ratios to support such users.

It's an issue several ISPs in the UK faced when they were advertising higher individual bandwidths - they mainly attracted high-volume users, and they therefore often didn't have the income to keep upgrading the hardware at telephone exchanges and so keep contention ratios low enough for their speeds to stay as advertised. Many of them were bought out by larger telecoms companies as a result, often resulting in a further deterioration of service.

Yes, the UK and the US went down fundamentally different paths to create the broadband market. In the UK, BT must give other ISPs space on its twisted pairs into the home, so there's the direct competition you mentioned. In most (all?) US markets, the bone thrown to the cable providers was a monopoly in each market area (and regulation, of course). At best in any given market, you have your telco that may offer an ADSL service, and your cable offering DSL. And that's the extent of competition. Certainly, we have no other choices in our town.

I'm all for Net Neutrality but caps are not part of Net Neutrality. Internet backbones are provided by an array of private companies such as AT&T, Verizon, Deutsche Telekom, Level 3 Communications, Seabone, and many others. Usually backbone players sign contracts with each other to exchange services allowing all backbone players to use the backbone services provided by other backbone players. ISPs that aren't backbone players sign contracts with a backbone player and usually piggyback on main contract that the backbone player has with other backbone players. Since nothing is free, someone has to pay for the trillion dollar infrastructure of the internet.

Without knowing the details of an ISP/Backbone contract, I suspect that the ISP has to pay based on the amount of traffic to and from the ISP. That is part of the cost to an ISP.

If traffic is high to the ISP, they will be required to improve their infrastructure and connection capabilities to the internet to serve that traffic or provide poor service since their infrastructure and connection capabilities can't handle the traffic.

Since ISPs charge based on the maximum line speed, if two customers have the same speed line but one is a low usage customer and another is a high usage customer, the low usage customer is paying for high usage customer. Therefore caps with additional charges above the caps seems like a fair way to keep the costs down for the low usage customer.

Thank you both. this also helps explain why there is so much more bandwidth available in Europe at such low cost.

Here's how it appears to have played out:

In Europe, Telekoms used to be government owned - well into the 90's. Prior to that, the government-owned telekom monopolies rendered a network that was horribly limited, mostly to inner-cities and government institutions. Even into the late 60's, having your own phone in Europe was very expensive and considered rather "posh" and "elitist".

By comparison, US infrastructure at the time was far more advanced and ubiquitous, and almost everyone had a phone, if only available from the monopolistic AT&T. Also expensive, but unlike Europe, not beyond the reach of most homes, which is what made AT&T the giant it was.

By the 80's, it became critical that pathetic European infrastructure be revamped, and new technologies like fibre had just become viable, and so that's what it was all replaced with. Up to the end of the 90's, European Telekoms were laying fibre like madmen - to everywhere.

Then, in the mid-late 90's, European governments privatised the telekoms with the stipulation that the infrastructure had to be provided to all competitors at very low rates, to prevent "monopolies".

The US, in contrast, was stuck with their ubiquitous older infrastructure that had been laid in the 50's 60's. Because the "internet" hadn't yet emerged as a broad consumer product, a complete rebuild was never necessary or practical, and so, sufficient infrastructure just wasn't in place when the internet became a mass consumer product.

The big US firms were the only ones willing and able to fund installing new infrastructure - with the stipulation that they are legally able to enjoy a virtual "monopoly" on the infrastructure they install (and really, I'd expect same if it were me). Those agreements were granted by localities rather than as a nation-wide project. The result being a mixture of different infrastructures owned by various private entities, all of which enjoy a virtual monopoly over their patches, but not a nation-wide monopoly (e.g. facilitating competition in the grand market, but not in local markets), which keeps it all legal.

And that's what I've been able to reckon is the reason for the difference today.

Is that somewhat accurate?

Not quite correct.

Problem 1

In the 1960s, private companies started laying cable throughout the US for cable TV even to small rural areas. Cable technology was not very good at that time but good enough to carry analog signals and the cable could be directly connected to the TV without a set top box.

As subdivisions were built in the 1970 and 1980s, better cable was laid to those subdivisions but cable technology was still not that great.

When the internet came about in the mid 1990s and high digital compression was invented (MPEG-2), the cable companies decided replace all that cable with high speed fiber. They thought they might lose cable subscribers if they required them to use set top boxes and therefore decided to provide analog and digital on the same cable. The problem is that on fiber, analog channels are bandwidth hogs requiring about 20x the bandwidth of digital channels. This meant instead of laying inexpensive 200 MHz cable that only required one amplifier per mile, they laid much more expensive 640 MHz-890 MHz cable that required 3-4 amplifiers per mile.

AT&T was the leader in laying the cable but the debt load nearly drove them bankrupt and they sold the cable business to Comcast. After that, AT&T was a non player in the cable business and was bought by Southwestern Bell. Cable companies that survived the heavy debt load of the 1990s, started providing internet service, VOIP, and digital TV. Even though the analog channels took up about 3/4ths the bandwidth of a 640 MHz cable system, they could still provide several hundred channels of SD digital programming. However once HD programming appeared, they didn't have a lot of bandwidth on the cable so HD expansion moved slowly (not a big deal since there wasn't much HD channels available). To plan for future HD channels and higher speed internet services, the cable companies added more amplifiers in the street and when high gain amplifiers became available, they replaced the street amplifiers to get an additional 120 MHz of bandwidth on the cable.

However by 2009, the analog channels were restricting the cable companies from providing more HD channels and higher speed internet services but they feared that they would lose customers to satellite TV if they dropped analog channels and forced customers to the higher priced digital TV packages since satellite companies had low priced packages that competed with their analog packages. To get around the problem, the cable companies eliminated the analog channels in two stages (first stage was to eliminate extended basic which was about 30-50 analog channels) and provide those customers with free set top boxes to receive those channels plus the basic channels in digital. The second stage was to eliminate the basic analog channels (about the last 20 analog channels) and provide those customers a free set top box to receive those channels in digital.

Now the cable companies had plenty of bandwidth to do whatever they want but they still have a large debt load to pay off. Many cable companies now provide up to 150 mbps internet speed. There are currently inexpensive cable modems that can double that speed but the demand isn't there for cable providers to provide that speed. Their basic internet service is about 25 mbps. Most cable systems currently have about 120 HD channels, 400 SD channels, on-demand for hundreds of channels, VOIP, and internet service up to 150 mbps.

Problem 2

The US has significantly less population per square mile compared to western Europe and the cities are designed differently which are comprised mostly of detached suburban homes. Providing cable services to rural areas is very expensive where several hundred of miles of cable may have to be run (excluding the cable in the rural towns) to provide cable services to possibly 50,000 homes if all subscribe. In most US cities, I suspect that the amount of cable that needs to be laid is probably around 5x that of typical European cities to serve the same number of households.

Attempted Solutions

Both AT&T and Verizon had upgraded their phone systems to fiber in major cities. The bandwidth is small compared to cable systems but large enough to provide good TV and internet services. However the fiber only runs along the main streets and twisted pairs are run to the homes for phones service. Verizon decided that they were going to be the premier fiber provider and ran fiber all the way to the homes (cable companies use coax cables for the last 50-200 feet). AT&T after getting badly burned in the 1990s, decided on a much more cost effective solution by using the twisted pair for the phone as a solution for the last mile and used special equipment to condition the line to carry about 20 mbps. Everybody that has Verizon likes the service but it is expensive and Verizon can't make money off the service so they abandoned their plans for further expansion. AT&T is cheaper than cable TV and uses a switching network where only selected channels will come down the twisted pair. Therefore each home has a switcher on the far end of the twisted pair. Because they are using twisted pair for the last mile, bandwidth to homes is fairly limited (initially they could only carry one HD channel and one SD channel to the home simultaneously but that might have improved since then). As far as internet service, AT&T's top speeds are well below the lowest speed that cable provides. If 5G can get high speeds reliably at a low price, that may possibly be a solution for on demand services but the infrastructure required for mobile services in the US is far greater than western Europe per household so that doesn't appear to be a near term solution.

In terms of the reasons that the competitive landscape is regulated the way it, you have it right. Michael's explanation of the technical background is a terrific read. Is the day getting closer when high quality broadband is delivered by sattelite or microwave or otherwise than by a cable into the home? Is that where 4g (or whatever it is) phones are already? That would shake up the competitive landscape for the cable cos.

Again, thank you both. Good reading, and explains a lot.

Regarding wireless: Here in Spain, where infrastructure has never really enjoyed a major retrofit except in major metro areas (I'd suspect largely due to the Franco era), internet speeds are low, reliability is low, and costs have traditionally been very high (compared to most of western Europe).

But Wireless... something new in the past 3 or 4 years.. well, it's becoming a rather boom industry here, where dozens of little startup providers are kicking the daylights out of the old dinosaur telecom (Telefonica). It's cheaper, faster, more reliable, and there's actually fierce competition amongst them.

Naturally, the sleepy incumbent is screaming bloody murder, attempting to impose more regulation to "level" the playing field, but uncharacteristically for Spain, that hasn't happened yet (fingers crossed). But it's not as cheap as in Germany yet, and if you look up "protectionism" in the dictionary, you'll see a picture of Spain and France, so I wonder how long that may last, or how much competition will be allowed before regulation puts an end to it.

Anyway, if that's any sign of things to come, hopefully you stateside punters may get a deserved break if small wireless providers are allowed to kick the arses of the big dinosaurs.

It's hard to envision that home direct to and from satellite at a high speed would ever be possible. A satellite can possibly transmit several hundred channels to millions of customers at high speed (about 12 mbps for HD and 1.5 mbps for SD per channel) but each channel is just one transmission for all customers. However imagine a satellite receiving millions of high speed packets from the internet and transmitting those to millions of different customer's homes but each packet is for a different home. The airway frequency spectrum doesn't exist for that much data. Cell, WI-FI, or internet technology would likely be of no use to try to solve that problem since satellite technology works completely different than any of those technologies. Another problem would be that each home would need a high powered directional transmitter to transmit packets to the satellite.

I wondered how satellite could provide on demand services (I've always had cable so I never experienced satellite on demand service) until I discovered that on demand is not provided via the satellite but via a high speed internet connection to your home (typically fiber from a cable provider).

How Does DirecTV on Demand Work? | eHow

Cell technology is probably the most possible solution to the problem but that appears to be significantly in the future. The cellular frequency spectrum is limited as to how much data it can carry.

Currently I believe the fastest consumer cell speeds are when using 4G LTE networks. Although 4G has a theoretical peak speed of 1 gbps, in real life situations, it usually only produces about 5-12 mbps. .

I suspect the speed of 4G LTE can vary significantly from time to time depending on the traffic to and from the tower and this can cause serious problems with video streaming. As an example, the way Netflix works is that it tests the speed of the connection and if it determines that it is 8 mbps, it will likely assume that the connection can maintain an average speed of at least 5 mbps so it so it starts loading a 30 second video buffer in the device and then starts playing the movie with a quality that averages about 5 mbps. As the movie is playing, the Netflix server will try to keep the video buffer full in the device but if it empties, Netflix pauses and waits until buffer is full again before it resumes the movie. If the connection speed remains below the expected speed for the movie, pauses may possibly occur every few minutes.

Surprisingly, Wi-Fi can cause the same problem in an apartment or condo complex. Someone may have a cable speed that can maintain 25 mbps and Netflix will stream it's highest quality 14 mbps video but when people in the complex start using their computers, your Wi-FI speed can drop to as low as 1 mbps. Fortunately there is a way to get around the problem in apartment and condo complexes by purchasing a dual band router and use the new 5 GHz band which uses the faster "N" or "AC" protocols for Wi-Fi. Although the concepts used by WI-FI is drastically different than cellular networks, both are fighting for air time.

Although internet service is more expensive in the US than Europe, it is far cheaper, faster, and more reliable than cellular transmissions. With most cellular plans, they either limit the data transmissions to only a few GB or throttle transmissions to 3G or even 2G speeds after a certain amount of GBs have been transferred. AT&T and Verizon, the two largest carriers, do not offer un-throttled plans and suspect that is because they can't handle the traffic. If today, everybody cancelled their cable TV service, internet service prices would likely skyrocket unless customers upgraded to faster internet service.

Like cable service, the US has to provide a lot more infrastructure to provide cellular service throughout the US than western Europe and therefore is more expensive. Verizon has the best coverage in the US but from the following map, there are large areas without coverage. If you unclick 2G and 3G, the coverage gets significantly less since a lot of areas don't have 4G coverage.

Verizon Coverage Maps - 2G,3G and 4G LTE Coverage Map | OpenSignalMaps - OpenSignal

That's a very interesting post, thanks.

I do find it difficult to believe that the US population being more spread out can justify the higher costs, though. The US is more spread out, so I was willing to believe this when I first arrived. However, once I started to drive outside the cities, I found that rural coverage is just absent, and small towns are generally covered only by basic cell service. This is so much worse than the UK that I would expect the increased cost per customer from population density to be balanced by the reduced cost from absent or cheaper coverage.

First you have to understand how cell phone technology works. The cellular frequency spectrum is limited and if it was tried to carry all cell phone communications of a city on that spectrum by one tower, the best that could probably be accomplished would be to carry voice and text only and even then, establishing a cell connection may be difficult. Besides that, in order to communicate with the tower, a cell phone would have to have high transmitting power quickly draining the batteries and probably frying your brain. Therefore to solve the problem, they divide the frequency spectrum into bands and place towers throughout the city using low transmitting power at towers and on the cell phones and each band can be reused many times throughout the city but the bands of the same frequency can't overlap each other's service areas but different frequency bands do overlap the service areas of other bands. In this way as a person moves about a city, their cell connection is handed off to the next tower on a different frequency using a different cell in the new band without you knowing that anything changed. By having multiple bands within a city and the bands can be reused, the number of connections that can be served simultaneously can be significantly increased. Also since bands overlap the bands of other service areas, if all the connections (known as cells) are used on the nearest tower, an alternate tower will be used for the connection.

As you get into the suburbs, the towers can be placed further apart since there are less obstructions allowing signals to travel further. As you get into the country, towers are normally placed on hills and can be placed up to about 10 miles apart for good coverage.

However whether in a city, suburb, or the country, the closer the towers are together, the more reliable the coverage and this can be done because of the concept of breaking the cellular frequency spectrum into bands. Also in cities where areas are very densely populated, one tower may support multiple bands.

In the cities, there is normally fiber that the towers can connect to allowing access to the internet for voice and data but in the country, fiber many not near by the tower. If fiber is close enough to the tower, laying fiber to the tower may be the most cost effective way to provide connectivity but often that isn't an option in the US. Communications via satellite isn't a good solution since the frequency spectrum is limited, there is large latency which would make data communications very slow, and it wouldn't be very cost effective with all the equipment needed to transmit and receive data via satellite. Therefore they usually communicate via microwave towers. Microwave communications is a narrow beam line of sight communications between microwave towers and the towers can have multiple dishes allowing microwave towers to transmit and receive data to and from many towers. A microwave tower is not cheap but if there isn't fiber nearby, it is probably the most cost effective solution. In addition to the extra cost of microwave towers, sometimes getting power to the towers in the country isn't cost effective and the towers are run by diesel generators. If a small town does not have fiber and serves too few people to pay for microwave towers, I suspect they use landlines to get to the internet and then only voice and possibly text can only be used.

The UK and Oregon are approximately the same size in area but Oregon serves about 1/15th the number of people with approximately the same number of towers. However in the UK, there is probably a greater chance that towers can be connected to fiber than Oregon adding significant costs to the infrastructure in Oregon. Cellular providers may not cover Oregon 100% but if they have coverage in all the towns and along the major highways, the infrastructure costs will probably be greater than the infrastructure costs that serve the same number of people in the UK.

One extreme example of the use of microwave towers is that high frequency traders are usually located in NYC, the traders use programmed computers to trade, they trade securities which are located in NYC on the NYSE and NASDAQ exchanges but they also trade commodities and options that are located in Chicago on the CME and CBOE exchanges, and it takes about 15 milliseconds (about 1/65th of a second) to acquire data via fiber laid directly to Chicago. The speed is slower if the internet is used. The fiber that was laid has to go down streets, around corners, and follow roads so it is not a straight line to Chicago. A private company decided to spend billions of dollars to build microwave towers from NYC to Chicago with towers about 50 miles apart and sell the services to high frequency traders at about $35,000 per month per trading platform so that the traders can acquire CME or CBOE data in about 9 milliseconds (less than 1/100th of a second). For many high frequency traders, the cost was worth the 6 millisecond advantage over other traders using the fiber.

Here in Spain, we do have pretty good 3G coverage almost anywhere (as I think is true in most of western Europe), which suggests 4G may also have similar coverage eventually. But all 3G is throttled and capped, just like in the US.

It's WIMAX that's popular here as the de-facto alternative to an expensive "land line" which are very unreliable and never delivers the speeds it's supposed to. The typical, most basic land line consumer end-cost including basic internet, without any extras is at least €52 per month (roughly $70 USD), even though it's not uncommon to struggle to achieve 1Mbps, even with a 6Mbps contract. And it's pretty common that you will lose internet entirely for some period of time on a regular basis, particularly in bad weather, rain or other adverse conditions.

Whereas WIMAX is typically €25 - €40 ($30 - $50 USD) per month, uncapped, delivering precisely the speed you contract for (typically 6Mbps is the lowest you can get, and you actually get that download speed or better) and more importantly, it is far more reliable, works even in adverse weather conditions.

4G is only now coming on the market, and frankly, remains to be proven. I'd have to expect that as with all 3G, it's likely to be throttled or capped anyway, whereas WIMAX is not, which may explain the attraction.

But in Germany, for example, household internet is very cheap, and the infrastructure is rock-solid, and is indeed all by land-line infrastructure, and is completely uncapped. It would be extremely rare to ever lose your internet connection, unless it's something you caused yourself, such as unplugging your router.

Yet in Germany, 10Mbps or higher would typically carry an end-user cost between €5 and €20 (roughly $7 - $30 USD) per month.