(keitai-l) Re: [link] wlan/plan

From: Benjamin Kowarsch <benjk_at_mac.com>
Date: 06/07/02
Message-Id: <74CD9798-79D3-11D6-BA3F-003065FB21DC@mac.com>
On Friday, June 7, 2002, at 12:13 , wimjam wrote:

>  but the issue might change when mobile operators are becoming WLAN
> operators, and they start integrating the WLAN gateways with their 
> mobile
> backbones (Nokia and Ericsson are offering this solution already
> commercially).

I doubt that. There is no reason to assume that roaming surcharges come 
down if both visited and home network infrastructure is in the same 
hands. The experience in the mobile phone market shows that this is not 
the case. You often pay higher roaming charges in a visited network 
which is owned by the same group as your home operator. They charge as 
much as they possibly can.

Besides, why operate a multimillion dollar database (HLR/AuC) with an 
expensive semi-proprietary protocol (MAP/SS7) if you have to go through 
an SS7/IPsec gateway eventually, while you could do the same on a 5000 
USD per 1U scalable BSD server and free MySQL or PostgreSQL ?

It's not like HLR queries are complex transactions.

It's not that MAP/SS7 would add any value in this particular application.

The only reason behind this is the avoidance of competition. Once 
de-facto standards are based on boxes that only make sense for mobile 
telcos to have, then WLAN roaming will be tied to billing by mobile 
telcos and nobody else. And prices will go through the roof again.

> As you said authentication would be done through SIM or through
> username/password. But this information could be encrypted through 
> IPSEC.
> Some of these models are already working with for instance Telenor and 
> Telia
> in the Nordics.

I am not saying it cannot be done. I am saying it is going to be 
expensive.

And the expense is not just the cost for whatever technical solution is 
employed.

Roaming charges are about risk, logistics and greed.

Let's assume the WLAN operator receives 50 yen for one minute of access 
to their network ...

You may safely add another 50 yen to cover the risk, another 100 yen for 
logistics and another 100 yen or more for greed. That's how you get 300 
yen per minute when roaming.


Trust me, if roaming onto a WLAN is modeled after mobile phone roaming 
and a mobile phone company will be in charge of billing (and thereby 
setting the roaming surcharges), then WLAN roaming will become as 
outrageously overpriced as mobile phone roaming is now.

I'd rather trust the credit card companies to run the backbone of a 
roaming infrastructure and the logistics. They would probably charge 
somewhere between 3 and 5%, initially with an annual membership fee of 
100 or 200 USD, and it would be truly global, universal and affordable.

You'd go to your local bank to apply for a MasterRoam or RoamVisa SIM 
card. This could be used in traditional style internet cafes by sliding 
the card into a reader of a terminal or you could use it in your own 
notebook to connect via WLAN in any signed up hot spot.

The local network provider, internet cafe or WLAN hotspot would 
advertise their rate (say 500 yen per hour) and there may be a minimum 
purchase of connection time, but you would be assured by the credit card 
company's MasterRoam or RoamVisa logo in the window which guarantees you 
a surcharge of not more than -say- 5% on top of whatever the shop 
charges. This would make a bazillion times more sense than anything that 
the mobile phone industry has come up with so far - their roaming track 
record stinks.

regards
benjamin
Received on Fri Jun 7 08:00:31 2002