Paul Golding's Weblog on Wireless
These are just my occasional thoughts on wireless etc.
They do not represent the view of any company:













Subscribe to "Paul Golding's Weblog on Wireless" in Radio UserLand.

Click to see the XML version of this web page.

Click here to send an email to the editor of this weblog.

Join Paul's mail list
 

 

Monday, March 17, 2003
 

:: Grid processing on your personal area network ::

I forwarded info on the Nokia BT pen to Dave at Motorola and he came back with an interesting comment - why not use a PDA to do the complex hand-writing recognition and then push on the plain-text to the phone. This may or may not make sense, but the idea of collaboratively using all the processing capabilities in personally networked devices is an attractive one and various approaches could be conceived.

An extension from that is to use processing resources in the network in a grid manner. Now this seems like a statement of the obvious, afterall we already do that with the client/server architecture. But the problem remains one of delay and leads me to think again about an old idea, that of using the cellular base station as a proxy processor for the handset. For example, let's say I want to do natural voice processing on the device (the ultimate way to compose a text message) or near real-time translation from one language to another. These types of application are extremely sensitive to delays and could be killed by the round trip back to a server on the net. One only has to use GPRS to discover that the "always on" doesn't mean "instantly available".

Having powerful computing power available at the base station would possibly enable such applications to work in a networked fashion. Of course, there is no need to add hardware to the base station itself, a simple API would suffice to get data in and out of the base station and onto a co-located server. Such applications would be limited to pipelined operation only - taking something in and squirting something back out either to the phone or back onto the wire to the backhaul and onto the net itself (via the usual routes in the core network).

Having said all this, maybe the UMTS Quality of Service mechanisms will enable "excessive" delays to be overcome, but maybe not...what we are interested in here is a class of applications where user response time is "as if" it were locally processed on the device.

Needs more thought and discussion.


12:33:31 PM    

:: Marvel at sending 3D messages ::

Send one of these to a friend using ShoutMessenger (only in US though)


9:11:16 AM    


Click here to visit the Radio UserLand website. © Copyright 2003 Paul Golding.
Last update: 4/4/2003; 12:13:57 AM.
This theme is based on the SoundWaves (blue) Manila theme.
March 2003
Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          
Feb   Apr



Sign up to my email list

Click here to send an email to the editor of this weblog. - Send me email

Available for hire as consultant via Magic E Company who can also develop software for you.