On a personal note.

As with everything in this world, things change.

My wife and I have decided to leave our home of over a decade and move to a new area of the country. In all honesty, this is a little unnerving and somewhat frightening but we are also equally excited.

Unfortunately, this will cause a pause in my regular posts of a somewhat indeterminate length of time.

I look forward to a quick return and wish you all the best in our absence.

Is political policy fracturing the net?

One of the incredible things the Internet has clearly demonstrated is its ability to evolve faster than anyone can predict. Naturally, that hasn't prevented people like me from trying.

But what happens if we reach a stage where the international community no longer wants to cooperate? In the last year or two we are now beginning to see some indications that some countries are not satisfied with the Internet being controlled by ICANN and would prefer this authority be shifted to the UN or in the latest proposal to create an independent Internet separated from the current network.

The Public-Root organization is an association that wishes to augment the control ICANN exercises or more correctly create another set of DNS servers that would allow a parallel Internet to coexist with the current one. This article discusses one group in Turkey that is seriously considering creating another network that would operate outside of ICANN's oversight even though interoperability would be maintained.

A Turkish group seeking to improve communications technology is pushing the use of a splinter Internet network that might avoid control by the United States.

The Turkish Informatics Association's recommendation comes as the U.S. Commerce Department has announced it would indefinitely retain oversight of the Internet's main traffic-directing computers, known as root servers.

A federation of independent root operators, The Public-Root, is trying to become an alternative network with 13 root servers in 10 countries, including Turkey.

The Turkish Informatics Association and Amsterdam-based Unified Identity Technology want to use that alternative network to offer Internet addresses that end in a company or individual name--such as "www.yourname"--without ".com" or a country code like ".tr" for Turkey.

Incredibly, this was anticipated years ago as you can read here.

Despite its best efforts, the Internet Corporation for Assigned Names and Numbers (ICANN) has proven overall to be a failed experiment in Internet policy development, implementation, and management. ICANN's lack of meaningful representation, and its continuing pattern of drastic and seemingly arbitrary structural and policy changes (among other shortcomings), have created an unstable and suspicion-ridden environment that is detrimental to the interests of the vast majority of Internet users around the world. The resulting overly politicized situation not only threatens the stability of the Internet itself, but also invites drastic and undesirable interventions by a variety of vested interests.

In November of this year the World Summit on the Information Society (WSIS) will be conducting Phase 2 of their discussions on how Internet governance and financing mechanisms should be operated. The WSIS has published a Declaration of Principles that lays out their position. This document lays out exactly what the WSIS believes and charts a direction for their group to follow.

Will these actions fracture the Internet? What are the possible implications and benefits?

I don't think anyone can fully comprehend what the full effect of this action might be but I do believe that the current system controlled by ICANN is in serious need of restructuring. One of the issues facing many of us is how Domain Names are registered and the pricing that supports this structure. In 2007 Verisign will be up for renewal as the controlling body that oversees .com domain names. Recently, Verisign won the right to continue to oversee the .net domain structure against several competitive bids.

Among some of the questions that have arisen is will Verisign raise the price they charge and if so, by how much? While the price to register a domain name is inexpensive by US standards many places in the world would like to see a reduction in these fees.

In the first quarter of the year, about 6.7 million new domain names were registered worldwide, representing a 36 percent increase from the same period a year earlier and a 4 percent increase from the fourth quarter of 2004, VeriSign said.

In total, there were 76.9 million domain names registered at the end of the first quarter of 2005.

This is some serious money we are talking about here. More importantly, there are also issues that the current system is faulted by allowing people to tie up domain names purely for speculation by registering them purely for the purpose of resale, a term known as "Domain Squatting." Verisign themselves have also come under fire for "Traffic Squatting" as well as "Typo Squatting" along with Verisign's own version of domain squatting.

These actions have royally pissed people off over the last few years and many have made the decision that the oversight of the top level domain names needs to not only be removed from Verisign (the owners of Network Solutions) but from ICANN and the US government as well.

I don't know if this is necessarily a good or a bad thing but the discontent from the perceived mismanagement is rapidly starting to boil over.

As long as the Internet(s) all agree to exchange information in an open fashion we will not lose the potential benefit of a worldwide communications platform but what happens if this approach isn't taken? Will we lose one of the biggest opportunities we as a group have ever had to work through our problems? What could be the possible benefit for the US to keep control if we lose a huge portion of the world as part of that decision? What about the effect this would have on commerce? I would submit this would also have a chilling effect on science and research of all kinds as the worldwide scientific community (long known for being border agnostic) would lose its most effective tool for sharing and disseminating information.

This also leads me to wonder if we might also be looking at a number of other policy decisions causing the net to fragment in ways we might not even be considering.

Is there a potential for splinter networks to wish to be removed from the greater Internet as a whole? Certainly a coherent argument could be presented that the Supreme Court's latest ruling in the Grokster case might want to motivate people to form their own closed groups so that they can share whatever they choose without fear of retribution. In fact, I would think a case could be made that this would be legal as the sharing of materials between "friends" is perfectly allowable. I am not sure if this could be extended to the entire population of Philadelphia should the city decide to close off their proposed WiFi infrastructure as a privately shared resource between friends.

What does this say about pornography, even the illegal variety? If people set up closed networks where the data doesn't traverse the Internet, how could they be discovered?

What we are seeing here is the presentation of a series of very serious choices. Do we as a nation change the way we are doing things to meet the demands of a significant portion of the rest of the world or do we all become a little poorer? Can we admit that our oversight of Verisign has some serious issues and needs to be changed or will a fair portion of the world walk away from us?

Perhaps the most important question I can ask, what do we as a people envision the Internet to be and where do we want to take this technology? Are we so certain that the decisions we make today are what's best for our country and the world that we are willing to gamble we can continue to run the Internet?

If so, we are about to get a wakeup call that is going to be much harder to rectify than it was to create.

As hard as this may be for many to understand, we are not in control, I am not sure we ever were. Our continued belief that we will shape the future as far as technology and communications go is rapidly becoming a myth.

I believe it is time for us to take our place as partners in this endeavor and step away from the belief we are in the driver's seat. The reality is, we are now on autopilot and we will either go with the flow or be lost on the wayside.

Oversubscription - or - how I watched the gravy train dry up.

The dirty little secret in the ISP world is our ability to oversubscribe traffic on our networks. If we examine the typical WISP business model we find that quite a few concurrent users can be "pushed" down a single T1 line. If we then take into account that at no one period of the day will there ever be every single customer we have using the network we find that one T1 line (with a capacity of 1,54Mbps) can carry far more broadband users (broadband as defined by speeds of 200Kbps) than the straight math would lead you to suspect.

1554Kbps/200Kbps=7.77 concurrent users

Even if the numbers above were close to reality and we could count on a 10 to 1 over subscription rate we find that the maximum amount of customers per T1 would max out somewhere around seventy-seven! If we then assume that we are billing each customer at $39.95/month total gross revenue would come in at a little over $3K/month. Considering that in many parts of the country a T1 line is now well below $1K/month everything would look pretty rosy! Of course, there are other costs that would need to be accounted for, cost of equipment, payroll, insurance but no matter.

What happens when the oversubscription rate drops?

Two well documented trends we are seeing in the end users on-line usage is the length of time they are staying connected and the amount of bandwidth being consumed in a 24 hour period.

Among many of the contributing factors influencing bandwidth usage is streaming media as Internet radio moves into the mainstream but also as video advertisements soar to astronomical heights. Granted, the streaming music doesn't take up an enormous amount of bandwidth even though as the push for increased quality becomes more mainstream we will see a rise in this rate, the amount of people using this technology is quickly growing. It isn't unusual to see a large proportion of people in any given workplace using Internet radio and while each individual user is only consuming a very small amount of bandwidth relatively speaking, the entire organization is consuming a continuous stream of a 200Kbps on Internet radio traffic alone. Add VoIP calls (using a continuous stream of 100Kbps per call while rapidly being adopted by the business mainstream) along with the other day to day Internet traffic and you can easily see how a small to medium sized company will saturate a T1 line for an entire 10 to 12 hour workday.

But what happens when everyone goes home at night?

The study, conducted in partnership with Frank N. Magid Associates, surveyed 27,841 Internet aged 13 and over on 25 different publisher Web sites. It found 51 percent of respondents watch online video at least once a month; 27 percent watch Internet video at least once a week; and five percent watch it on a daily basis.

The full article this quotation is from can be found here:

This follows up on the reports that web pages themselves were becoming significantly larger in size, containing more graphics and advertisements per page which contributes to much longer load times.

I think we can clearly infer that the trend is towards using more bandwidth and using it in very different ways than we could count on in the past.

From a historical perspective we used to be able to account for a "burst and release" pattern where users might grab a web page or download a relatively speaking small file and then stop for a few minutes freeing us the portion of their connection for someone else to use. Those days are rapidly coming to a close.

What does this mean to an ISP or more importantly a WISP?

The immediate effect is that the infrastructure needs to be designed so as to make it easily scalable. This is critical to the continued growth while keeping the existing user base satisfied. The second variable is that the access point/base stations deployed MUST be able to handle a continuous stream of small packets - something that most equipment cannot handle.

Let's take a look at some of the specifications WiMAX is planning to provide. According to this article from Daily Wireless Alvarion has officially released information about their WiMAX equipment.

As you can see from the chart below, this is going to be some pretty impressive equipment.

WiMax System Performance

Range	< 4 miles	4-6 miles	> 6 miles
Base-station cost (’04 pricing)	$5k - $20k for WISP class $20k+ for carrier	same	same
CPE price	< $300	same	same
Adaptive modulation scheme	64 QAM	16 QAM	½ QPSK up to 16 QAM
Data throughput (20 MHz channel*)	75 Mbit/s	50 Mbit/s	17 Mbit/s to 50 Mbit/s depending on link quality
No. of business users (T1 level) 1	206	138	46 to 138
No. of residential users (512 kbit/s) 2	1,552	1,035	345 to 1,035
Source: Intel Assumes two 10MHz bands in the base station as benchmark for comparison purposes. Over-subscription rate is 5x for business and 12.5x for residential. Also takes into account overhead (efficiency), which for 802.16 is 85% independent of number of users.

Chart courtesy of Daily Wireless

What I have a slight problem with is the claims as to how many users can be services from a single basestation. According to the chart above, within the four mile range limit we should be able to see an awesome 75Mbps of throughput. I am not sure how this would translate into 206 T1 class business users based on the subscription rates I see as rapidly approaching. In fact, if you are willing to believe that we are near a one to one subscription rate the best we could expect to see would be somewhere in the neighborhood of 50 T1 class users per basestation at the theoretical best.

Somehow I don't think the equipment will be able to handle the flood of continuous packets these businesses will be expecting to be able to pass but I will say that Alvarion has surprised me on many occasions before and this might be one more time to add to my list. As far as the claim that one of these basestations will be capable of handling 1552 residential users (at 512Kbps service levels) that is something I find very hard to accept. I have no idea what oversubscription rates are being used to quote that but the term "optimistic" certainly comes to mind in much the same way 802.11b will deliver 11Mbps.

I can't wait to see what kind of effect IPTV and devices like the SlingBox are going to have on this model but I can tell you that as we move forward our customers are going to demand more bandwidth and in a continuous mode.

This is what Lightreading has charted and is predicting for the future.




Will you be ready?