Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: HTSG Usage Model Special Committee



Title: Message
Hello Mike and Bjorn and all,
 
Thanks for these contributions.   At the meeting this week, we came out with an initial list
of deployment environments - not as detailed as Bjorn's.  We also discussed the
question as to how many simulation targets we wanted,  and agreed that "5"
was a suitable number.  So, when we create our scenarios,  we'll probably include
elements from more than one of the deployment environments listed below into
the one model. 
 
For example, we'll probably have a "home" model that includes a device in the
same room as the AP,  plus one in a different room.  This is both a realistic usage
model,  and exercises the technology with different channel conditions.
 
I like Mike's classification.   I think we'll start off with a large list of applications
and then realise that we can lump applications together into groups like this.
 
I think another thing we need to do is to agree on definitions for some terms.
I've variously used: "use case", "scenario", "simulation target".  Vinko has a
"channel model" and Bjorn has added: "deployment environment".  If someone
wants to come up with some definitions that relate these and throw out the
baggage, I'd be very grateful.
 
I spoke with Mary Cramer after the telecon. She will come up with a first draft of
a usage model document,  I'll have a first hack at it (and see if I can include Bjorn's
input),  and then put it out to those
who volunteered at the meeting (cc everybody) for their input.  Hopefully we
can complete this by the next meeting to give the meeting an input paper to
discuss.
 
The minutes from the previous meeting will be issued in due course (I'll notify
separately).
 
It appears that the IEEE 802 mailer daemon rejects headers with >1024 B.  My list of
Usage model participants transgresses this limit.  So, if you want to email to the group,
can I suggest that you "blind copy" the list of names from the minutes (which won't
go into the header) and put a smaller list in the other address fields.    FYI - I've done
this with this message which may explain why you get two copies.
 

Best Regards,

Adrian P Stephens
Intel Corporation

Tel: +44 771 276 3448 (Mobile)
Tel: +44 1223 763457 (Office)


 

-----Original Message-----
From: Mike Moreton [mailto:Mike.Moreton@synad.com]
Sent: 04 July 2003 08:12
To: Bjorn Andre Bjerke; Stephens, Adrian P; Youngsu Kim; Bobby Jose; Chiu Ngo; Chris Hansen; Colin Lanzl; Craig Hornbuckle; Dov Andelman; Eldad Perahia; Frank Howley; Garth Hillman; Irina Medvedev; Jason Ellis; Javier Delprado; jimlans@mobilian.com; Jim Tomcik; jrosdahl@ieee.org; kevin@proxim.com; Kotecha, Lalit; Law Choi Look (Choi); Majid Malek; Malik Audeh; Mary Cramer; PJohansson@acm.org; Paul Feinberg; Qinghua Li; Rahul Malik; Robert.Huang@am.sony.com; Roberto Aiello; Rolf Devegt; Sanjeev Sharma; Sean Coffey; Steve Halford; Timothy Wong; Tkashi Ishidoshiro; Tomer Bentzion; Tomoko Adachi; Vinko Erceg; Wayne King; Woo-Yong Choi; Xiaolin Lu; Yashuhiko Inoue
Cc: jrosdahl@ieee.org; Matthew B. Shoemake; Sean Coffey; stds-802-11@ieee.org; stds-802-19
Subject: RE: HTSG Usage Model Special Committee

Bjorn,

 

There are some very different requirements amongst the applications in your “Data” section, which got me wondering whether it might be helpful to split all the applications into six groups, depending on the combination of their bandwidth requirements, with how tight their QoS requirements are.

 

For example

 

Category 1 – High bandwidth, tight QoS

Interactive gaming

Video conferencing

 

Category 2 – High bandwidth, medium QoS

Video streaming,

File serving

 

Category 3 –  High bandwidth, loose QoS

File backup

 

Category 4 - Low bandwidth, tight QoS

Voice over IP

 

Category 5 – low bandwidth, medium QoS

Audio streaming

 

Category 6 – low bandwidth, loose QoS

Instant messaging

 

I’d venture to say that category 6 applications don’t need to be included in the models.

 

Mike Moreton

Synad Technologies Ltd.

 

-----Original Message-----
From: Bjorn Andre Bjerke [mailto:bbjerke@qualcomm.com]
Sent: 03 July 2003 21:02
To:
Stephens, Adrian P; Youngsu Kim; Bobby Jose; Chiu Ngo; Chris Hansen; Colin Lanzl; Craig Hornbuckle; Dov Andelman; Eldad Perahia; Frank Howley; Garth Hillman; Irina Medvedev; Jason Ellis; Javier Delprado; jimlans@mobilian.com; Jim Tomcik; jrosdahl@ieee.org; kevin@proxim.com; Kotecha, Lalit; Law Choi Look (Choi); Majid Malek; Malik Audeh; Mary Cramer; Mike Moreton; PJohansson@acm.org; Paul Feinberg; Qinghua Li; Rahul Malik; Robert.Huang@am.sony.com; Roberto Aiello; Rolf Devegt; Sanjeev Sharma; Sean Coffey; Steve Halford; Timothy Wong; Tkashi Ishidoshiro; Tomer Bentzion; Tomoko Adachi; Vinko Erceg; Wayne King; Woo-Yong Choi; Xiaolin Lu; Yashuhiko Inoue
Cc:
jrosdahl@ieee.org; Matthew B. Shoemake; Sean Coffey; stds-802-11@ieee.org; stds-802-19
Subject: HTSG Usage Model Special Committee

 

Hi all,

Here are some thoughts on how to start developing a set of usage models for 802.11n.

The definition of a usage model and its purpose as accepted by the participants in the June 17 teleconference are as follows:

1. Usage Model - a detailed model of expected realistic deployments and applications of 802.11n devices and networks.
2. Purpose - the purpose of the usage models is to provide a basis for the development of functional requirements and comparison criteria for proposals to the HTSG.

As the definition implies, a usage model is made up of two parts, namely
* a deployment environment, and
* a set of applications typically associated with the particular deployment environment.

As a first step in the process of defining a simple but adequately realistic set of usage models, I thought it would be useful to separately list all deployment environments and applications that have been mentioned during our discussions so far (plus some that perhaps haven't been mentioned).

The next steps would be:
* Map applications to deployment environments (channel and interference models for each environment to be developed by the channel modeling special committee)
* Develop detailed specifications for each application, e.g., traffic type, data rate, packet size, arrival model, load, delay requirements, etc.
* Derive requirements for the technology from the combination of deployment environments and application requirements

Of course, we will have to narrow the long list of possible environments and applications down to a manageable number of combinations, but rather than deciding on that number up front, let us first see how many distinctly different environments and applications we are faced with. In making the following two lists, I took the liberty of borrowing from earlier presentations given to the HTSG by Adrian Stephens and Javier del Prado.


Deployment environments

1. Residential:
* Intra-room communications
* Room-to-room communications
* Indoor-to-outdoor (e.g., for using a lap-top/TV on the patio, etc.)

2. Enterprise:
* Enclosed offices
* Sea of cubicles
* Meeting room
* Large factory floor
* Hospital
* Warehouse
* Classroom/lecture hall
* Campus (i.e., indoor-to-outdoor as well as outdoor access point for outdoor users)

3. Hot spot:
* Airport
* Library
* Convention center
* Hotel
* Shopping mall
* Train station/bus terminal
* Drive-in windows
* Sports stadium/concert hall
* City square (e.g., Verizon's plans for adding hot spots to Manhattan phone booths)
* Public park

4. "Wireless cable":
* Residential neighborhood (e.g., TV/phone/Internet connection via pole-top access points)

5. Mobility:
* Hot spots on trains, buses, air planes
* Curb-to-car communications
* Car-to-car communications



Applications

1. Video:
* SDTV
* HDTV
* Video conferencing
* Internet video streaming

2. Voice/audio:
* Wireless VoIP
* Audio/music (Hi-fi stereo and multichannel - 5.1, 7.1 etc)
* MP3/AAC audio

3. Data:
* Web browsing/content downloading
* E-mail
* E-commerce transactions
* Instant messaging/chat
* File backup
* File server
* Interactive gaming
* Telemetry

Please add, correct and/or modify the above list and circulate to the whole committee. Looking forward to the discussion!


Regards,
Bjorn



--
Bjorn A. Bjerke                  +1.781.276.0912 (direct)
Qualcomm, Inc.                 +1.781.276.0901 (fax)
9 Damonmill Sq., # 2A        bbjerke@qualcomm.com
Concord, MA 01742, USA   www.qualcomm.com