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

Re: [802.19] FW: [802.19] Definition of Coexistence

Title: Re: [802.19] FW: [802.19] Definition of Coexistence
Tolerable is a good replacement for minimal. However, the real issue is definition of “their tasks”. The presentation in Denver by Intel showing that 40MHz channel did not affect Bluetooth VoIP quality (note that this is not representative of voice operation on Bluetooth devices) claimed that no harm was done since the MOS score for voice quality was the same. However, this is only one task for Bluetooth devices and many are doing more than one task and expectations for latency (not measured in the Intel presentation) may not be possible when more than 50% of the entire 2.4 GHz unlicensed spectrum is being used by a single pair of devices or multiple devices connected to a single AP.

Bluetooth specifications (based on IEEE 802.15.1-2005) were designed to allow at least 16 devices in the same area to establish private piconets between a master device and multiple slave devices with the master device determining when slave devices can transmit. This was based on having 79 hopping channels in the 2.4 GHz spectrum. The 2005 version of IEEE 802.15.1 added Advanced Frequency Hopping (AFH) based on IEEE 802.15.2-2003 to avoid spectrum where other devices such as IEEE 802.11 were operating. This works fine for individual piconets, but will reduce the number of piconets that can operate in the same area since total available spectrum is reduced by 25-30% for each IEEE 802.11 AP operating in the same area. Using AFH, IEEE 802.15.1-2005 compliant devices have no impact on IEEE 802.11 devices operating in the same area.

With current IEEE 802.11 and Bluetooth devices operating in the same area, current coexistence mechanisms (only implemented by Bluetooth devices, nothing done by 802.11 devices) result in a 25-30% reduction in expected Bluetooth system behavior and no reduction in 802.11 expected behavior. With the introduction of 40 MHz channels by 802.11n, Bluetooth system behavior will be reduced by 50-60% and there still will be no reduction in 802.11 expected behavior. And this does not include the impact on other 2.4 GHz devices such as IEEE 802.15.4 (ZigBee) devices.

One of the claims made by the 802.11n supporters was that when multiple 802.11 APs are configured in the same area, up to three 20-22 MHz channels would be occupied with a larger impact that one 40 MHz channel. Usually this is done to increase availability of 802.11 service in buildings where distances to be served are greater than 100 feet. In one area, the intensity of 802.11 on all channels would vary allowing Bluetooth devices to operate with tolerable impact.

Due to the limited amount of available 2.4 GHz spectrum (80 MHz), I think it is unwise to allow one AP to use over 50% of that spectrum when other devices are also attempting to use that spectrum at the same time. It should be reasonable for an 802.11n device to first scan the 2.4 GHz spectrum and determine whether it is clean (no energy detected from other devices) and only use 40 MHz channels if it is clean. A better option would be to completely disallow 40 MHz operation in 2.4 GHz spectrum as the odds of finding another devices operating in the same area are very high with over 2 billion Bluetooth devices shipped and a billion more being shipped every year. The impact on 802.11n performance would be less that the >50% impact 40 MHz channels would have on other 2.4 GHz devices.

A better coexistence rule might be “if the impact on other devices operating in the same area is greater than the performance gain obtained by another device increasing spectrum allocation, the increased spectrum allocation should not be allowed”.

Regards, John


On 7/26/08 12:47 PM, "Shellhammer, Steve" <sshellha@xxxxxxxxxxxx> wrote:

Forwarding to the reflector.
 
Steve
 

From: Benjamin A. Rolfe [mailto:ben@xxxxxxxxxxxxxx]
Sent: Friday, July 25, 2008 5:45 PM
To: Shellhammer, Steve
Subject: Re: [802.19] Definition of Coexistence

Steve, all,

I think this is a good basic definition.  "Tolerable" is better than "minimal", as it implies various degrees of coexistence may exist.  I think the basic definition is sufficient for coexistence. When we do specific analysis we get into degrees of "good" coexistence.

The first task in analysis is determining what the coexistence impacts may be, then we can figure out what is tolerable and what is a problem :-).

-Ben



----- Original Message -----

From: Shellhammer, Steve <mailto:sshellha@xxxxxxxxxxxx>  

To: STDS-802-19@xxxxxxxxxxxxxxxxx

Sent: Friday, July 25, 2008 10:35 AM

Subject: Re: [802.19] Definition of Coexistence
Craig,
          I would say it is not so much as a “minimal” impact but a “tolerable” impact. The 802.15.2 definition is targeted at that level, since it says that the application can still perform its task.  Another way of saying it, is that the QoS performance meets an acceptable level of performance.
          Thanks for commenting.
Steve

From: Craig Warren [mailto:cwarren@xxxxxxxxxxxx]
Sent: Friday, July 25, 2008 9:07 AM
To: Shellhammer, Steve; STDS-802-19@xxxxxxxxxxxxxxxxx
Subject: RE: Definition of Coexistence
Hi Steve;
Should there be some statement in here like some minimal impact of the task being performed kind of wording?
kindest regards;
Craig


From: Shellhammer, Steve [mailto:sshellha@xxxxxxxxxxxx]
Sent: Friday, July 25, 2008 8:39 AM
To: STDS-802-19@xxxxxxxxxxxxxxxxx
Subject: [802.19] Definition of Coexistence
802.19 TAG,
          Recently with the coexistence discussions regarding 802.11 VHT60 I have been asked several times for a definition of coexistence.  It was pointed out at the closing EC meeting that one cannot evaluate coexistence without a proper definition.  I explained that the current definition that we have in 802 is the one that was standardized in 802.15.2 several years ago.  The 802.15.2 recommended practice was approved by both 802.11 and 802.15. Yesterday, Bruce Kraemer, the chair of 802.11, asked me if the 802.19 TAG supported that definition.  I have never asked the TAG that question.  So I thought I would send out an email on this question.
          The definition in the 802.15.2 recommended practice is,
3.1.2 coexistence:
The ability of one system to perform a task in a given shared environment where other systems have an ability to perform their tasks and may or may not be using the same set of rules.
 
           I would like to hear feedback on this definition.  The TAG could choose to support this definition or come up with its own definition.  There is a perception that one cannot evaluate the coexistence of two wireless networks without an agreed upon definition.
          If the TAG decides they want to support another definition we could later hold an email ballot on the definition if making it more formal is useful.
          Comments are encouraged.
Steve




--  
John R. Barr (John.Barr@xxxxxxxxxxxx)
Director, Standards Realization - <http://www.motorola.com>
Vice Chairman of the Board, Bluetooth SIG - <http://www.bluetooth.org>
(847) 576-8706 (office) +1-847-962-5407 (mobile) (847) 576-6758 (FAX)