Connecting Devices With Wireless Grids

Roland Piquepaille writes "A new concept is emerging in networking: wireless grids. These grids connect all kinds of wireless devices, such as sensors or cell phones, with each other and with more traditional wired grids. IEEE Internet Computing has devoted a very long and thorough article about these wireless grids which can deliver new resources, locations of use, and institutional ownership and control patterns for grid computing via ad hoc distributed resource sharing." (Read more below.)

"The article says that applications for wireless grids fall into three classes: the ones which aggregate information from the range of input/output interfaces found in nomadic devices, those which focus on the locations and contexts in which the devices exist, and those that leverage the mesh network capabilities of collections of nomadic devices. The authors add that these grids "emerged from a combination of the proliferation of new spectrum market business models, innovative technologies deployed in diverse wireless networks, and three related computing paradigms: grid computing, P2P computing, and Web services." If you're interested in the future of wireless networks, the original article is a must-read, but check this summary if your time is limited."

Comments

The Wireless model

[stecoop]

It would be nice if the Internet cloud bubble dissolves when there are enough wireless devices to remove the necessary Internet link via the high-speed backbone.

Re:The Wireless model

[Anonymous Coward]

Wouldn't it. Someone else said that ad hoc high speed/low latency/few hops networks don't get built by anti-establishment types with good intentions. Real work and plannings needs to be done.

Re:The Wireless model

[sean23007]

Hmm. You seem to think that in this situation your handheld would be serving everyone else around you, and all of them are just leeching, and you don't get any service from anywhere, it all originates with you. Well, I should probably point out to you that that's not really how sharing models work. When you're not using your resources, they are shared among your peers who might need them. When you are using resources, if you need more than your machine possesses, your peers share with you. And the wireless grid network spreads across all the devices. You are just a link in the chain. Not the start of the chain. Everyone is not leeching off you. And you aren't paying to make everyone else's experience better. Everyone is paying. And everyone gets more than they would if they were going it alone. That's the point.

Re:The Wireless model

[towerdave]

To boil down what sean23007 said, it's internet Communism. Due to human nature, it won't work, unless the system can't be changed to block sharing of resources. There will always be a few that will share, but many more that will leech. TowerDave

Re:The Wireless model

[putch]

while i do agree that for the forseeable future this is fantasy it is an exciting idea.

30-40 years ago the concept of a personal computer would have evoked a similar response. let alone some kind of "invisible global network" that people could access with computers carried in their pockets.

this could work if and when mobile cpus have enough power (both computing and electrical) and, i would imagine, that much of this "grid" would be stationary Access Points (or perhaps they'd be more like relay stations).

but yes, this probably isn't going to catch on anytime soon.

Re:The Wireless model

[Kristoph]

You could design the grid "participant" software such that it would only use the idle resources of your device.

So your network bandwidth and CPU resources would be 100% yours when you needed them but when your device was partially or fully idle the resources of your device would be made availlable to other users and devices.

Personally, security issues not withstanding, I would have no problem with someone making use of my idle resources as long as when I needed them they would be availlable to me instantly.

]{

um...

[Anonymous Coward]

"but check this summary if your time is limited." - my time is always limited, for I am a mortal man.

connecting...

[garcia]

Interesting research but I hope that their theories remain just that (at least as far as using CPUs from personal wireless devices).

Yes, plenty of people are using wireless devices, and yes they could be used together to encode a concert or whatever, but no, I wouldn't want to be sharing my devices CPU time without compensation (say that encoding's output for free).

I want devices to be smaller, faster, and use less power. This seems to promote a need for more CPU time and a bigger battery.

Is that a wireless grid device in your pocket or are you just happy to see me?

Re:connecting...

[Apathetic1]

If wireless grid computing takes off, battery and CPU time might be a good tradeoff for access to the massive computing power of the grid should one need it.

Some kind of a system where a client is given tit-for-tat access to grid CPU time might make a lot of sense.

not a great [scalable] idea

[Anonymous Coward]

wireless presents some issues such as limited bandwidth and high latency over long links or routes. distributed computing requires very low latency and very high bandwidths, and wireless just doesn't seem to me like a solution which will attain the same performance per dollar (unless it's a weird circumstance, like multiple sites separated by many many miles, but even then some kind of frame or T1 might do better than wireless or microwave).

Re:Health Implications

[HarveyBirdman]

Anyone worry that years from now they'll find out wireless causes cerebral cancer or something?

Not really, no.

What is the point?

[Sanity]

This sounds like just another attempt to coin a new term. Skimming through the article I don't see any new concepts, nor even new combinations of concepts. Grid computing instead of P2P, now wireless grids, what next - P2P XML?

Trust and Security

[Keitopsis]

Couple of thoughts:

1) Can we really do authentication for masses of "grid" members without eating up the bandwidth?

2) Is this the next market for spoofing-spam distributors?

Surely this is the way to go

[djsmiley]

Especially along the lines of free speach,

How ever, before any of this can REALLY take off, we first must insure that the underlying security is safe.

Dont build on quicksand if it has windows in it!.

Am I just being dense, or...?

[Deep Fried Geekboy]

...isn't this a very long-winded way of saying 'the internet will soon have a substantial wireless component'?

I can't see what's new here at all. Yes, there will have to be a few more technologies for managing ad-hoc networks. But that's about it.

As for us all sharing our resources in one warm fuzzy anarcho-syndicalist wireless IT hive, dream on. (Or, more precisely, give T-Mobile your first-born).