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

RE: stds-80220-requirements: Spectral Efficiency (4.1.2)

Title: RE: stds-80220-requirements: Spectral Efficiency (4.1.2)

Philosophically, these two approaches are the same.  My proposal, made prior to yesterday's
conference call, was to establish that spectral efficiency be calculated for a fixed 5 MHz block
size under the condition that >= 99% of the emissions are contained within that block.
That is the same as saying that  < 1% of the emissions are allowed to be outside of the 5 MHz block.
Essentially, this is one form of spectral mask.  Alternatively, David's proposal is that, for a "5 MHz
licensed block",  the spectral efficiency (or aggregate throughput) would be caluculated for candidate
air interfaces operating within a 5MHz licensed block and hence meeting common out of band emission limits
(e.g. the US PCS limits in Part 24.238).  Both proposals would allow for a fair comparison of spectral efficiency 
(or aggregate throughput) for different air inteface proposals.  David's approach, that we discussed in detail
on yesterday's Requirements CG conference call is obviously more reflective of the conditions in a real world
deployment.   I support either approach, and now prefer the latter one as it appears more straight forward.
As we discussed, the Evaluation CG is responsible for working out the details.  I believe John Humbert will
be posting a new proposed edit on the definition of Block Assignment that is consistent with requiring the
use of a common spectral mask in the evaluation of air interface proposals and David Shively will be posting a
proposal for modifying the text on SE to reflect the discussion on the CG.  Let's respond to those when they
appear.  In principle, I think we are all on the same page.
Best regards,
-----Original Message-----
From: []On Behalf Of Rausch, Walter F [GMG]
Sent: Friday, December 05, 2003 10:50 AM
To: Marc Goldburg
Cc:; Shively, David
Subject: RE: stds-80220-requirements: Spectral Efficiency (4.1.2)

I agree with David on these points.
Walter Rausch
-----Original Message-----
From: Shively, David []
Sent: Thursday, December 04, 2003 3:37 PM
To: 'Marc Goldburg'; Rausch, Walter F [GMG]
Subject: RE: stds-80220-requirements: Spectral Efficiency (4.1.2)


I had proposed text back on Nov. 17 with the intention of including
guard bands; meaning that the RF emission bandwidth of the radio and
any required guard bands must fit within the licensed block.  Thus,
a licensed block is equivalent to the 802.20 "block size".

I think that using the block size (e.g. 5 MHz) to evaluate the various
proposals is completely quantitative and consistent as long as each
proposal is required to meet the same out of band emission limits. Here
I am assuming that a network operator has a licensed block of 5 MHz and
must adhere to the appropriate emission limits at the edges (such as the
US PCS limits in Part 24.238).

Best regards,
David Shively

-----Original Message-----
From: Marc Goldburg []
Sent: Thursday, December 04, 2003 2:22 PM
To: Shively, David;
Subject: RE: stds-80220-requirements: Spectral Efficiency (4.1.2)

David, Walter,

Joanne's proposal to calculate spectral efficiency relative to a 99%
emissions bandwidth is the only quantitative and broadly applicable
proposal I have seen to date for accounting for the real-world effects of
coexistence on spectral efficiency.  In the absence of any other proposals,
would your concerns be addressed by employing the 99.9% emissions bandwidth
instead of the 99% emissions bandwidth? 



------- start of forwarded message (RFC 934 encapsulation) -------
From: "Shively, David" <>
To: Requirements Reflector <>
Subject: RE: stds-80220-requirements: Spectral Efficiency (4.1.2)
Date: Thu, 4 Dec 2003 12:29:38 -0500


I agree your points.  For a realistic deployment in a licensed block of 5
(and assuming different network operators on either side of the block), the
99% emission bandwidth will have to be somewhat less than 5 MHz.

Using "block size" in this way and also in the spectral efficiency
will enable a consistent evaluation between proposals.

Also, the other block sizes (e.g. 1.25 MHz) should be treated in the same

Best regards,
David Shively

- -----Original Message-----
From: Rausch, Walter F [GMG] []
Sent: Thursday, December 04, 2003 10:50 AM
To: Joanne Wilson
Cc: Humbert, John J [NTWK SVCS]; Requirements Reflector
Subject: RE: stds-80220-requirements: Spectral Efficiency (4.1.2)


I have great concerns that while your proposal is likely suitable for
systems whose adjacent block operators are utilizing the same technology and
concur with the shared operation, it falls short of protecting such adjacent
block operations that may be utilizing other technologies. The reality of
field deployability dictates that emissions that fall outside of the
authorized channel block(s) must conform to applicable regulatory
constraints. As a consequence, guardbands must be incorporated into the
channel bandwidth AND the spectral efficiency calculations. Channel
bandwidth is then the sum of the "occupied bandwidth" (the 99% power
bandwidth in your proposal) plus the required guardbands.

Walter Rausch