Bandwidth Aggregation and Conglomeration

I was reading todays (Monday Jan 16 Pg C6) New York Times and noticed
an article on "Sharing Broadband to Increase Speed". Two companies
with two different approaches.
http://www.WiBoost.com
http://www.MushroomNetworks.com
Install their contrivance at your house, get the neighbors to do the
same, and you all get to share the aggregate bandwidth of the entire
neighborhood via wireless. Not a bad idea, if it works.

As always, I'm sceptical. Neither company mentions that only the
incoming bandwidth is aggregagated, not the outgoing. WiBoost uses
5.8GHz which should not cause interference problems with existing
2.4GHz wireless LAN's or store and forward delays. I can't decode
what Mushroom Networks is doing. WiBoost expects users to pay $300
each for their technology, when the users can get about a years DSL
service for a 2nd DSL line for the same price, or several years
service at higher speeds, for much less. I dunno. Still, methinks
it's worth a look.
--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558 (E-Mail Removed)
# http://802.11junk.com (E-Mail Removed)
# http://www.LearnByDestroying.com AE6KS

Jeff Liebermann wrote:
> I was reading todays (Monday Jan 16 Pg C6) New York Times and noticed
> an article on "Sharing Broadband to Increase Speed". Two companies
> with two different approaches.
> http://www.WiBoost.com
> http://www.MushroomNetworks.com
> Install their contrivance at your house, get the neighbors to do the
> same, and you all get to share the aggregate bandwidth of the entire
> neighborhood via wireless. Not a bad idea, if it works.
>
> As always, I'm sceptical. Neither company mentions that only the
> incoming bandwidth is aggregagated, not the outgoing. WiBoost uses
> 5.8GHz which should not cause interference problems with existing
> 2.4GHz wireless LAN's or store and forward delays. I can't decode
> what Mushroom Networks is doing. WiBoost expects users to pay $300
> each for their technology, when the users can get about a years DSL
> service for a 2nd DSL line for the same price, or several years
> service at higher speeds, for much less. I dunno. Still, methinks
> it's worth a look.
Are US Telco's "Sharing,caring people"?

http://wiboost.com/technology/whitepapers/wp3root/
ISPs have several choices regarding how they support WiBoost (or don't
support WiBoost).
A. ISPs can entirely prevent WiBoost from working on their networks.

B. ISPs can restrict WiBoost so it allows connections only between nodes
of the ISP's choosing (such as the ISP's own customers or customers who
pay an additional charge).

Did not Novell have a similar product in their internet connection offerings
some 10 odd years ago? It was a two channel link and was a way of
effectively getting twice the thruput on a dialup system (pre DSL). The
principle is not new.

Peter

"Frank" <(E-Mail Removed)> wrote in message
news:dqiitt$cia$(E-Mail Removed)...
> Jeff Liebermann wrote:
> > I was reading todays (Monday Jan 16 Pg C6) New York Times and noticed
> > an article on "Sharing Broadband to Increase Speed". Two companies
> > with two different approaches.
> > http://www.WiBoost.com
> > http://www.MushroomNetworks.com
> > Install their contrivance at your house, get the neighbors to do the
> > same, and you all get to share the aggregate bandwidth of the entire
> > neighborhood via wireless. Not a bad idea, if it works.
> >
> > As always, I'm sceptical. Neither company mentions that only the
> > incoming bandwidth is aggregagated, not the outgoing. WiBoost uses
> > 5.8GHz which should not cause interference problems with existing
> > 2.4GHz wireless LAN's or store and forward delays. I can't decode
> > what Mushroom Networks is doing. WiBoost expects users to pay $300
> > each for their technology, when the users can get about a years DSL
> > service for a 2nd DSL line for the same price, or several years
> > service at higher speeds, for much less. I dunno. Still, methinks
> > it's worth a look.
> Are US Telco's "Sharing,caring people"?
>
> http://wiboost.com/technology/whitepapers/wp3root/
> ISPs have several choices regarding how they support WiBoost (or don't
> support WiBoost).
> A. ISPs can entirely prevent WiBoost from working on their networks.
>
> B. ISPs can restrict WiBoost so it allows connections only between nodes
> of the ISP's choosing (such as the ISP's own customers or customers who
> pay an additional charge).

I used a program by MidCore for a long time on two 56k modems. It works
with two dsl or cable or any connection combinations you can dream up too.
Browser asked from proxy cache. Everything downloaded into a proxy. Smart
enough to basically round robbin and attempt to keep all connections busy.
Set file split size minimum to starting at x bites and it reassembled in the
cache seeming instantly. But only worked downloading. You could pick and
choose which connection to upload easy enough if you didn't want to
specifically use the first or next avail. They priced themselves out of a
job... was too pricey. You can make one yourself with a bit of
imagination,, proxy and download accelerator.


"Pierre" <(E-Mail Removed)> wrote in message
news:dqincj$r7m$(E-Mail Removed)...
> Did not Novell have a similar product in their internet connection
> offerings
> some 10 odd years ago? It was a two channel link and was a way of
> effectively getting twice the thruput on a dialup system (pre DSL). The
> principle is not new.
>
> Peter
>
> "Frank" <(E-Mail Removed)> wrote in message
> news:dqiitt$cia$(E-Mail Removed)...
>> Jeff Liebermann wrote:
>> > I was reading todays (Monday Jan 16 Pg C6) New York Times and noticed
>> > an article on "Sharing Broadband to Increase Speed". Two companies
>> > with two different approaches.
>> > http://www.WiBoost.com
>> > http://www.MushroomNetworks.com
>> > Install their contrivance at your house, get the neighbors to do the
>> > same, and you all get to share the aggregate bandwidth of the entire
>> > neighborhood via wireless. Not a bad idea, if it works.
>> >
>> > As always, I'm sceptical. Neither company mentions that only the
>> > incoming bandwidth is aggregagated, not the outgoing. WiBoost uses
>> > 5.8GHz which should not cause interference problems with existing
>> > 2.4GHz wireless LAN's or store and forward delays. I can't decode
>> > what Mushroom Networks is doing. WiBoost expects users to pay $300
>> > each for their technology, when the users can get about a years DSL
>> > service for a 2nd DSL line for the same price, or several years
>> > service at higher speeds, for much less. I dunno. Still, methinks
>> > it's worth a look.
>> Are US Telco's "Sharing,caring people"?
>>
>> http://wiboost.com/technology/whitepapers/wp3root/
>> ISPs have several choices regarding how they support WiBoost (or don't
>> support WiBoost).
>> A. ISPs can entirely prevent WiBoost from working on their networks.
>>
>> B. ISPs can restrict WiBoost so it allows connections only between nodes
>> of the ISP's choosing (such as the ISP's own customers or customers who
>> pay an additional charge).
>
>

In article <(E-Mail Removed)>, (E-Mail Removed) wrote:
>I was reading todays (Monday Jan 16 Pg C6) New York Times and noticed
>an article on "Sharing Broadband to Increase Speed". Two companies
>with two different approaches.
> http://www.WiBoost.com
> http://www.MushroomNetworks.com
>Install their contrivance at your house, get the neighbors to do the
>same, and you all get to share the aggregate bandwidth of the entire
>neighborhood via wireless. Not a bad idea, if it works.
>
>As always, I'm sceptical. Neither company mentions that only the
>incoming bandwidth is aggregagated, not the outgoing. WiBoost uses
>5.8GHz which should not cause interference problems with existing
>2.4GHz wireless LAN's or store and forward delays. I can't decode
>what Mushroom Networks is doing. WiBoost expects users to pay $300
>each for their technology, when the users can get about a years DSL
>service for a 2nd DSL line for the same price, or several years
>service at higher speeds, for much less. I dunno. Still, methinks
>it's worth a look.

I suspect a lot of people are like me, I'm reasonably satisfied with my
downstream side (of course more would be better) it is upstream bandwidth
that is in short supply IMO at least on my connection. I'm on BS's best
tier for my are and the upstream promise is 385 (actually get a bit less,
DSL you know). So in my opinion without a boost in the upstream side, not
worth the bother. But that is me.

"Organization: Committee to Maintain an Independent Xenix"
Thought MS owned this?

fundamentalism, fundamentally wrong.

(E-Mail Removed) (Rico) hath wroth:

>"Organization: Committee to Maintain an Independent Xenix"
>Thought MS owned this?

There's a really long old story behind COMIX. In a previous life, I
sold and supported SCO Xenix and Unix. Although Microsloth started
Xenix somewhat before PCDOS/MSDOS, they sold most of the rights to SCO
when IBM appeared on the scene and wanted PCDOS.

Fast forward many years, and we have one of the numerous battles in
the Unix wars. The one that inspired COMIX was when AT&T invested in
Sun Microsystems, and precipitated a revolt among Unix license
holders, who claimed that Sun would get preferential treatment in Unix
releases. I'll spare you the details. I saw this as an opportunity
to establish a competing organization, where I would sell membership
subscriptions for exorbitant amounts. That's where the "independent"
came from. Absolutely nobody was interested and the idea died. I
missed the chance to register the domain somewhat later, but the idea
was dead by then. I also had a background in publishing pornographic
comic books while going to skool, which somewhat inspired the acronym.
--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

"Pierre" <(E-Mail Removed)> hath wroth:

>Did not Novell have a similar product in their internet connection offerings
>some 10 odd years ago? It was a two channel link and was a way of
>effectively getting twice the thruput on a dialup system (pre DSL). The
>principle is not new.

Dunno about Novell. Officially, it's called "channel bonding
multilink" and is supported by Windoze for dialup and ISDN. With
broadband, it's "multi-homing". However, these require support from
the ISP to bond the two data channels to a single IP address. This is
very commonly done with IDSN, where two 64Kbit/sec B1 channels are
combined into a single 128Kbit/sec channel. With multiple different
service providers, that isn't going to work. I read that multi-homing
DSL is commonly offered by UK ISP's, but not in the US.

Another way to do it is with a load balancing router:
http://www.edimax.com/html/english/p...-PRIrouter.htm
This gives the same benifits but suffers from the problem that a
single download is limited by the bandwidth of a single broadband
connection and does not magically aggregate the entire bandwidth for a
download. My guess(tm) is that the aformentioned aggregation schemes
might be a similar load balancing router with wireless links attached.

--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

> This gives the same benifits but suffers from the problem that a
> single download is limited by the bandwidth of a single broadband

Unless you're using a download manager that spawns multiple connections
and the host accepts them, then it works.

David.

David Taylor <(E-Mail Removed)> hath wroth:

>> This gives the same benifits but suffers from the problem that a
>> single download is limited by the bandwidth of a single broadband
>
>Unless you're using a download manager that spawns multiple connections
>and the host accepts them, then it works.

True. That also works for BitTorrent outgoing bandwidth. The system
will open as many streams as it thinks is necessary to saturate the
bandwidth in both directions. That's both outgoing and incoming. The
server could limit the number of streams per IP, but that won't work
if the load balancing router offers requests from multiple ISP's, each
with it's own IP address. Install one of these wi-fi bandwidth
conglomerators and you can easily hog all of the bandwidth in the
entire neighborhood both in and out. Oh, did I say "hog"? I mean't
"share". Hmmm, I didn't see any mention of QoS or bandwidth
management in either web pile.

--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On Wed, 18 Jan 2006, in the Usenet newsgroup alt.internet.wireless, in article
<dqincj$r7m$(E-Mail Removed)>, Pierre wrote:

>"Frank" wrote:

>>Jeff Liebermann wrote:

>>> Install their contrivance at your house, get the neighbors to do the
>>> same, and you all get to share the aggregate bandwidth of the entire
>>> neighborhood via wireless. Not a bad idea, if it works.

If you are hoping to be able to download a single file faster - no. The
host you are downloading to knows about one IP address, but not about the
second [,third, fourth], and therefore can't shovel it down all of the
pipes at the same time.

If you are downloading multiple files (or otherwise using the Internet
in separate streams/applications/functions), this _MAY_ increase your
effective bandwidth. You would appear to the world to be two (or more)
separate computers that have no relation to each other.

>>> As always, I'm sceptical. Neither company mentions that only the
>>> incoming bandwidth is aggregagated, not the outgoing.

I'm not exactly sure that would be the case. Technically, there is going
to be limitations in bandwidth somewhere along the chain from A to B,
but that's true of any connection.

>> Are US Telco's "Sharing,caring people"?

Of course. ;-)

>> ISPs have several choices regarding how they support WiBoost (or don't
>> support WiBoost).
>> A. ISPs can entirely prevent WiBoost from working on their networks.

I suspect this depends on how things are done.

>> B. ISPs can restrict WiBoost so it allows connections only between nodes
>> of the ISP's choosing

Likewise. Sharing of multiple streams of traffic (downloading a web page
that itself has dozens of items of eye-candy can be downloaded in parallel
without the ISP knowing.

> Did not Novell have a similar product in their internet connection offerings
>some 10 odd years ago?

Ten years ago... January 1996?

1717 The PPP Multilink Protocol (MP). K. Sklower, B. Lloyd, G.
McGregor, D. Carr. November 1994. (Format: TXT=46264 bytes)
(Obsoleted by RFC1990) (Status: PROPOSED STANDARD)

1990 The PPP Multilink Protocol (MP). K. Sklower, B. Lloyd, G.
McGregor, D. Carr, T. Coradetti. August 1996. (Format: TXT=53271
bytes) (Obsoletes RFC1717) (Status: DRAFT STANDARD)

Yup - looks like the protocol was there then.

>It was a two channel link and was a way of effectively getting twice the
>thruput on a dialup system (pre DSL). The principle is not new.

I suspect you are thinking about what they were calling Shotgun Modems,
which were two independent modems on a single card (though you could do
this with a lot of extra effort with individual cards/externals). I'm
not sure what the limitation is on the number of modems than can be
"joined" this way. By the same token, there was also a sharing
mode in ISDN where you could combine channels to gain extra bandwidth.

The minor problem with Multilink is that the peers have to support this
capability and you bet your bippy that they charge for this feature. But
this allow you to increase the download rate of a single stream - a file
transfer rate, because the stream is only divided between the ISP and you.
The other method - where you are sharing connections (which may or may
not even be to the same ISP) doesn't require that the ISP does anything,
but only works where you are downloading in parallel streams (that is,
not dragging a single file). Jeff, you may want to review the Linux
Adv-Routing-HOWTO document.

Old guy