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

Re: [802.21] [Mipshop] Re: Architectural Considerations for Handover InformationServices (was: Re: CARD Discussion Query Discussion)



On Mon, Oct 17, 2005 at 07:58:35AM +0200, Hong-Yon Lach wrote:
> Here are my thoughts...
> 
> "Extensibility, flexibility" are nice features, but they should be put into
> their proper perspective. XML is certainly convenient for content editing by
> human users. But what does extensibility and flexibility mean  for 802.21?
> What processes and procedures are involved when certain deployment would
> like to adopt and/or modify new contents in the 802.21 (e.g. IS) exchanges?
> Are such processes and procedures to benefit from XML? How?

Extensibility means that ability to support not only existing PHY/MAC
technologies but also any future PHY/MAC technology, without changing
the 802.21 specification itself every time a new PHY/MAC technology is
invented.  

Flexibility means that ability to support various types of queries to
be useful for different handover mechanisms, protocols, algorithms and
policies.  

> 
> TLV has been a well-established norm for payload encoding for signalling
> protocols. One could feel unsatisfied with TLV and propose enhancements.
> This is welcome. But I think it is fair to expect any new proposals to prove
> concretely the advantages of XML over the conventional TLV in the context of
> signalling protocols in 802.21.
> 
> I don't believe that there is an essential lack of understanding of XML and
> TLV. They are created and used for their targeted purposes. Is there a "one
> size fits all"? I don't think so. Personally, it is obvious to me that XML
> is not appropriate for signalling protocol with respect to payload size
> (over the air), extra processing cycles in encoding and decoding (irrelevant
> whether XML is supported in the handset), extra delay to the reactiveness of
> signalling protocol (which could be important if 802.21 is to facilitate
> handover process), etc. I could be wrong, and I would appreciate anyone
> explaining to me otherwise, and how XML could make things better than TLV in
> this specific context of 802.21.

I think your expection is fair.  But the major issue here is that we
don't have a solution for conventional TLV proposed for 802.21 to
compare with XML.  Some people claim that they like TLV, but without a
complete solution we cannot do concrete comparison.  I was asked to do
some comparison in July meeting and did it in September meeting
(21-05-0347-00-0000-XML_IS_Introduction.doc).  However, because no one
has came up with TLV-based solution, that was all what I was able to
provide.  If you need a concrete comparison, someone has to come up
with a complete TLV-based proposal.  In fact, I was further trying to
design a complete TLV-based solution in August for comparison purpose,
but I just gave up because I found a lot of proprietary work-around
for schema and query handling.

Yoshihiro Ohba

> 
> Cheers,
> Hong-Yon
> 
> > From: Yoshihiro Ohba <yohba@TARI.TOSHIBA.COM>
> > Reply-To: Yoshihiro Ohba <yohba@TARI.TOSHIBA.COM>
> > Date: Fri, 14 Oct 2005 19:08:21 -0400
> > To: <STDS-802-21@listserv.ieee.org>
> > Subject: Re: [802.21] [Mipshop] Re: Architectural Considerations for Handover
> > InformationServices (was: Re: CARD Discussion Query Discussion)
> > 
> > Several comments:
> > 
> > - Informaiton encoding is one design factor, but we should think more
> > about other factors such as extensibility, flexibility, and actual
> > volume of encoded contents (regardless of how it is encoded).
> > 
> > - In reality, 3GPP2 has XML-based method (e.g., XCAP) in its
> > dependency list.
> > 
> > Yoshihiro Ohba
> > 
> > 
> > 
> > On Fri, Oct 14, 2005 at 06:19:26PM -0400, Singh Ajoy-ASINGH1 wrote:
> >>      But it does not seem to be true given the current discussion.
> >> 
> >>  
> >> 
> >> Ajoy-> In IETF folks are trying to help 802.21 to get the work done
> >> better. 
> >> 
> >> I do not believe there is anything wrong if some folks do not agree with
> >> XML query. Honestly I do not believe XML query is  good for low power
> >> 
> >> mobile devices. Based upon our current experience folks are finding hard
> >> to cope up with SIP and SIP compression itself. It looks good on paper,
> >> 
> >> but somehow when you implement compression you may find that delay
> >> introduced by compression out performs the bandwidth saving gained from
> >> 
> >> compression. I do not believe that adding one more level of compression
> >> for XML will be better idea.
> >> 
> >> As a research topic it may look very attractive, but we have to look at
> >> today's reality. I would like to state that in one of my discussions
> >> 
> >> I suggested that even TLV encoding can be modified to carry XML script
> >> to address some use cases where XML may be a good choice.
> >> 
> >> I am afraid we are trying to push XML without considering reality at
> >> hand. 
> >> 
> >>  
> >> 
> >> Also, what is wrong about sharing view of 802.21 when MIPSHOP is going
> >> to define a protocol that would cater need of
> >> 
> >> 802.21 itself?  
> >> 
> >>  
> >> 
> >> Nothing wrong in sharing the view.  But do we need to  mention the straw
> >> poll result to establish our  technical reasoning?
> >> 
> >>  
> >> 
> >> Ajoy-> Perhaps you did not like the word straw poll.  I will make note
> >> of that.  I think enough reasoning was given to justify that
> >> 
> >> XML query complex query may not be good for low power mobile devices.
> >> 
> >>  
> >> 
> >>  
> >> 
> >>  
> 
> 
>