On Friday, October 12, 2001, at 02:42 , Curt Sampson wrote:
> Well, speaking of roaming, Lucent claims to have solved all of this:
>
> http://www.theregister.co.uk/content/5/22178.html
Neither does Lucent make that claim nor does COPS solve anything. It is
not more than a publicity stunt repackaging an old technology that is
already out there, slightly modified and under a new name.
Logica have built this quite a while ago. It's called Ulysses Roaming
Interworking Gateway (RIG) ...
http://www.logica.com/pdf/mobile_net/ulysses_rig.pdf
... and Concert are operating it with success ever since ...
http://www.logica.com/company_info/news/press_releases/press_releases.asp?
display=detail&id=64&page=4&sec=0&exp=10&loc=0&lan=0
Logica have since extended it to WAP and UMTS.
All Motorola and Lucent have done is to take the same functionality and
integrate it into an HLR. This is not a new idea though. Logica went for
a central gateway simply because no mobile operator wanted to replace
all their HLRs and VLRs. They prefer to sign up with Concert's gateway
service instead.
Note, that it is the VLRs in the visited networks which communicate with
HLRs in home networks, therefore all VLRs have to be updated or
replaced. Many operators have hundreds of them and replacing or
upgrading them would cost them dearly for little benefit.
Likewise, the ZEBRA design has a signalling interceptor called an XLR.
This too was originally intended for deployment into each network.
Again, mobile operators rejected this and the design was altered to have
a central XLR operated by a roaming centre to provide a service to
mobile operators.
Even with traditional GSM roaming the trend has long been to outsource
as much as possible to clearing houses and to deploy as little functions
as possible into individual networks. Especially now where most
companies seek to reduce cost it is very unlikely that this trend will
reverse anytime soon.
Besides, even if Lucent can convince operators to spend money on this,
what good would it do ?
The problem with roaming is not that HLRs and VLRs of different
standards cannot communicate. Logica's Ulysses gateway has long solved
this problem and Concert's service is well established.
The problem with roaming is that there are too many parties who have to
make a buck on the service. The home network carries the risk yet has no
real-time visibility of spending, hence they require a cut. The visited
network provides the network infrastructure, hence they are entitled to
revenue. Clearing houses are required to settle huge amounts of
micro-transactions, they too are entitled to revenue. Then there are too
many calls that get send around in circles over international circuits,
which adds further cost.
COPS provides no solution to this mess at all. Quite to the contrary, it
adds further complexity and cost.
The only approach that can solve the problem is to *reduce* complexity
and cost by finding a less complicated way to provide roaming services
and cut out too many middlemen.
The "visited network charges" roaming model (i.e.ZEBRA) is the only
proposed solution so far that does this.
rgds
BK
[ Need archives? How to unsubscribe? http://www.appelsiini.net/keitai-l/ ]
Received on Fri Oct 12 10:33:46 2001