RE: stds-802-mobility: Hand-off question
My personal opinion is that users will find wireless devices
substantially more attractive if those devices are capable of roaming
seamlessly amongst heterogeneous networks. Connectivity would be
available over a larger coverage area and the device would be more
dependable as a means to access communication dependent services.
802 has not traditionally played in this space. Roaming within 802 has
tended to be confined to a particular MAC/PHY pair and within a
particular network. 802.11 being the most common example.
This does not mean it is unwise for 802 standards to anticipate
heterogeneous network roaming and either specify mechanisms or leave in
the necessary hooks. It is arguably unwise for new 802 standards to
ignore this issue, given what we know about the nature of the wireless
landscape today. Isolated roaming network technologies could find
themselves rapidly supplanted by networks that have effective
heterogeneous roaming capabilities since the benefit to the user is
obvious.
From the work in the IEFT such as SEAMOBY and TRIGTRAN, from 3GPP[2] and
from the numerous and generally inadequte/proprietary WiSP AAA systems
it is clear to me that any good heterogeneous roaming technology
requires implementations that impact not only layer3/IP/whatever, but
also the PHY and MAC and the network based authentication and billing
systems. Without such a broad solution, roaming can be neither seamless
nor ubiquitous in the general case.
It is probably clear to everyone else that I have come down on the side
of an 802 wide handoff effort, since I'm chairing the group. This should
not be interpreted as a an indication that I think it is the right place
to solve the problem. It is certainly the right place to solve a part of
the problem but there will definitely be a requirement for higher layer
parts of the puzzle and there might be a case for some work specific to
each of the 802 MAC,PHY pairs. This remains to be seen.
What I believe would be a very bad outcome is for network entities to
have to have separate implementations of roaming signalling to account
for handoff between each of the 802 wireless technologies. That would
not serve 802 well and would not be good for users.
A L1/2 to L3 conduit for roaming related signalling is something that
could apply to the general 802 framework but that might have specifics
for each of the wireless standards. 802.11k has shown the way here and
is a valuable contribution to filling in the handoff puzzle. It is
probably wise to learn from the 802.11k work and understand how it can
be extended or emulated elsewhere within 802 general case and how what
they have can be part of the larger solution. Similarly 802.11i is
bridging the gap between security and fast roaming. 802.16e is
addressing base to base negotiation of QoS stream handoff. This is all
good stuff. Providing access to this technology through a single general
802 roaming interface/mechanism/practice will be better for everyone who
has to implement roaming between different 802 systems or between 802
and non 802 systems.
So for the health of 802 wireless standards in general, I encourage
people the working groups to think in terms of what roaming mechanisms
should be local to a MAC/PHY pair and what mechanisms should be part of
a more general solution and work with the appropriate groups to effect
the optimal roaming solution. I also encourage people to to think how
best we can meet the needs of the higher layer roaming mechanisms. The
IEFT, 3GPP and others have made a lot more progress on roaming at their
end than 802 has done at its end. Now is a good time for 802 to adapt to
the changes in the broader networking world and roaming is one of the
top items on the list.
DJ
David Johnston
Intel Corporation
Chair, IEEE 802 Handoff ECSG
Email : dj.johnston@intel.com
Tel : 503 380 5578 (Mobile)
Tel : 503 264 3855 (Office)
> -----Original Message-----
> From: Lalit Kotecha [mailto:lalit123@yahoo.com]
> Sent: Tuesday, April 22, 2003 8:49 AM
> To: Joseph Cleveland; 'djames@arraycomm.com';
> arsha@optonline.net; stds-802-mobility@ieee.org
> Subject: RE: stds-802-mobility: Hand-off question
>
>
>
> There are so many wireless technologies, operation at different
> fequencies (900, 1800, 1900, 850, 2400, 3500 MHz) and standards
> available. I do not think these standards were designed to
> take care of
> inter-technology handoff. As market need changed, some of the
> standards
> made provision to support inter-technology handoff.
>
> We need to analyze even for existing standard that what is currently
> supported in 2G as well as 3G and what is roadmap for these standards.
> Also, this needs to be worked very closely with otehr standards bodies
> (3GPP, 3GPP2, etc.)
>
> Again if I understand correctly, 802.xx is mainly about developing PHY
> and MAC and rest is addressed differently. In the similar argument,
> 802.11e will suport QoS etc. But suppport of voice (codecs), VoIP (all
> signalling) and all new streaming media video application is out of
> scope over there.
>
> just my 2c,
>
> Thanks
> Lalit Kotecha
>
> --- Joseph Cleveland <JClevela@sta.samsung.com> wrote:
> > If handoff across air interfaces is not supported, users will need
> > multiple
> > handsets. Do we really want this situation?
> >
> > Joseph Cleveland
> >
> > -----Original Message-----
> > From: Dave James (UK) [mailto:djames@arraycomm.com]
> > Sent: Tuesday, April 22, 2003 12:55 AM
> > To: arsha@optonline.net; stds-802-mobility@ieee.org
> > Subject: RE: stds-802-mobility: Hand-off question
> >
> >
> > What could be the purpose ? MBWA is just that - truly
> broadband, not
> > 3G at
> > all.......
> >
> > -----Original Message-----
> > From: owner-stds-802-mobility@majordomo.ieee.org
> > [mailto:owner-stds-802-mobility@majordomo.ieee.org] On
> Behalf Of Prem
> > Pungaliya
> > Sent: Monday, April 21, 2003 6:17 PM
> > To: stds-802-mobility@ieee.org
> > Subject: stds-802-mobility: Hand-off question
> >
> >
> >
> > Will handoff be supported between 802.20 and public 3G?
> >
> >
>