(keitai-l) Re: Multimodal technology

From: David Davies <david_at_intadev.com>
Date: 11/09/01
Message-ID: <B3132D1A714FB24DBC192EB865EF56800AB01C@id-mail.INTADEV.intadev.com>
On Wednesday, 7 November 2001 11:46 PM Marc Printz wrote:

>> Now: that is interesting. How does the call setup work? Tedious? 
>> Or could you in principle with the press of one link get a voice 
>> connection to an IVR system up and at the same time an i-mode 
>> connection? What about i-appli, is it possible to have concurrent 
>> i-appli running? THIS would be excellent.


Without having exact details of FOMA's internal handset and base station
architecture (which I am guessing they keep secret) one cant say exactly
how it is implimented (anyone that knows specifics please sing out and
feel free to correct the figures that I guessed)  

But being based on WCDMA the ability to support multiple sessions is an
inherent feature of the architecture.  

WCDMA divides a Code Division frequency band into a number of time
slices in the same way that TDMA does.  
|--|--|--|--|--|--|--|--|

In every cycle a phone recieves a signal for only 1/8th (??) of the time
and other phones pick up the other 7 slots (actually I think a few are
used for curcuit info, but for the purpose of explanation we can ignore
that.)
|--|--|VV|--|--|--|--|--|

The system typically uses one for sending and a differend bandwidth for
receiving.

When operating in voice mode the phone only needs to use one timeslice
which would be about 32k (??) 
In packet mode in order to get higher bandwidth it just takes more
slices as it needs them.
|--|--|DD|DD|DD|DD|--|--|
Some architectures also allow the handset to use some of the sending
band for receiving which is why the download speed is usually faster
than the upload speed. For Packet data it doesn't keep a slice the whole
time but only uses it as it needs to send a packet.

This means the more people in an area using phones the less likely you
are to be able to get 384k (not really a problem just yet) And it may
also mean that if you're just unlucky and your particular frequency is
filled up then you cant get more frequency even though there may be more
available on different frequencies [ or code divisions ]

To handle voice and data at the same time it simply uses one timeslice
for voice and grabs others for the data packets as it nneds them.
|--|--|VV|DD|DD|--|--|--|


This model is very different to [OLD] GSM WAP networks where the data is
transferred over the voice line in the same manner that modems do.  In
such an architecture the data communication completely takes over the
voice line so the two cant exist at the same time. This is why GSM WAP
phones need to connect and disconnect to the data service while iMode
phones are allways connected.

The rest of the world is now getting GPRS which is a packet data service
similar to what DoPa is on PDC (the system that iMode runs over now) it
existed in Japan back in 98 (??)  
The main difference between PDC (TDMA) and FOMA (WCDMA) is that WCDMA
uses some cool algorithims to pack more data into the same frequency
band (takes a lot longer to explain)

Regards,
David Davies
http://www.intadev.com




[ Need archives? How to unsubscribe? http://www.appelsiini.net/keitai-l/ ]
Received on Fri Nov 9 06:48:51 2001