<<< Date Index >>>     <<< Thread Index >>>

[IP] more on Hub & Spoke or P2P Mesh? That is the question.





Begin forwarded message:

From: Bob Frankston <Bob2-19-0501@xxxxxxxxxxxxxxxxxx>
Date: June 22, 2006 2:20:11 PM EDT
To: dave@xxxxxxxxxx, ip@xxxxxxxxxxxxxx
Cc: "'Jock Gill'" <jg45@xxxxxxx>, dpreed@xxxxxxxx, Gerry Faulhaber <gerry-faulhaber@xxxxxxxxx>
Subject: RE: [IP] Hub & Spoke or P2P Mesh? That is the question.

There is a very different option which is consistent with a marketplace dynamic. Jock, you seem to be confusing the anti- marketplace of the Regulatorium with a marketplace where the incentives are in alignment. The key, as I’m wont to emphasize, is not presuming the answer, it’s in preventing preventing alternatives. If we have pure infrastructure then we enable far more alternatives than if we mandate a particular high speed infrastructure.

The goal is to maximize opportunity not to maximize some arbitrary factor in the short term. Given then we can consider:

Infrastructure at various scales – private, community, region etc. It’s just wire and metal (and wireless). Use of the infrastructure for various networks built upon it. This is akin to having a lot of local Ethernets and using IP protocols at the app level to gain wider routing. Don’t worry about speed. The vital applications don’t need speed and we have today’s TV already. Speed is easy and you’ll get it but only if you don’t make it a requirement and sacrifice the future for a dollop of broadband. Capacity is a more interesting measure. You don’t want static IP addresses – that’s like putting a concrete base on your cordless phone. Look at Skype as an example – it creates its own stable addresses without tying them to the accidental topology of the network. You get P2P meshing as an accidental byproduct. Where I do agree is that local traffic tends to be local and peering is a business relic not a technology. Note that I put wireless in parenthesis because, though I see it as a major part of the future, I’m willing to focus on other aspects because connectivity is going to moot spectrum allocation far sooner than any explicit political agenda. After all, if you have abundant wireless connectivity from nearby access points why run your own transmitters or, worse, satellites, just to get coverage. Why use point to point radios when you can reach the world.

Here is where, I presume, Gerry and I agree, the marketplace is more powerful than Congress though Congress can do a lot of damage trying to help it.



-----Original Message-----


From:   Jock Gill <jg45@xxxxxxx>

Date:    June 22, 2006 11:51:34 AM EDT

To:       dave@xxxxxxxxxx

Cc:       Jock Gill <jg45@xxxxxxx>, Gerry Faulhaber <gerry-

faulhaber@xxxxxxxxx>, Frankston Bob <bobf@xxxxxxxxxxxxx>, “Reed David

            P.” <Dpreed@xxxxxxxx>

Subject:            Hub & Spoke or P2P Mesh? That is the question.



Dave,



The conversation on Muniwireless might want to be expanded a bit.

The first issue is: Do we want a connectivity environment that looks something like this:

1.       100 mbps architecture for stage 1

2.       1 Gbps architecture for stage 2 - within 3 years

3.       Full P2P mesh network at all levels

4.       Backhaul via I2, NLR or equivalent

5.       Local loop for local traffic

6.       exchange points, not peering points

7. symmetrical connectivity w/ static IP addresses for true “end-to- end” network

8.       seamless, 7x24, always on Big Broaddband connectivity





etc.

Or do we insist on remaining locked into the 20th century Hub & Spoke architecture [Master/Slave] in order to preserve the “customer” model supported by “billing”? IE, do we want to invent the future looking in the rear view mirror?

If we want a true P2P mesh network model for the 21st century, then we need to not only allow sharing of connectivity, but to actively encourage it. This suggests that we need a completely new billing model. More accurately, a no billing model!

If we want option 1, P2P mesh, then it would appear that public entities are the best, perhaps ONLY, entities to run it. Why? Because they do not have requirements to bill every user every month for every bit. We are not individually billed for police and fire protection, nor for street lights, nor public education and so forth. Thus government, which generally DOES work, has an proven and existing model for public services, such as a P2P mesh network, that can only work if sharing is leveraged for cooperative gain, not billed for private gain.

If I were a stock holder in one of today’s incumbent connectivity providers, would I want my company to invest its scarce human and financial capital in very low margin bit hauling or in high margin innovative products and services delivered over commodity bits? Do I want the cost of bits to be a barrier to my high margin products and services?

The conversation has been distorted by the recent notion that everything should be a market function and the false premise that anything a government does automatically and unfairly competes with the “sacred” private sector. Rubbish.

The private sector loves externalities that allow them to shift costs to others. Bit hauling should be seen as an opportunity for a magnificent externality that allows them to re-allocate resources to more profitable operations. In this light, it is, in fact, in the self interest of the incumbents to work with the municipalities to create the very best possible bit hauling externality. Consider: Did the proponents of the market object to our tax dollars building the Eisenhower Interstate Highway system that has allowed them to externalize so many transportation costs? Or that out tax dollars subside airports that benefit the business traveller the most? Or that our tax dollar created the internet in the first place?

So the win win here is a robust P2P mesh network operated by municipalities that will allow the private sector to externalize low margin operations in order to improve their bottom lines.

Morning thoughts from a motel room in KS.

Jock





-------------------------------------
You are subscribed as roessler@xxxxxxxxxxxxxxxxxx
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/