[LINK] SMS - shortcomings in emergency services

[Glen Turner]

Hi Richard,

I read the paper and I'm not happy.  The paper seems to test
a worst-case design:
  - *not* sending to geographic areas, and thus needing to
    find the home location register of handsets.
  - requiring message storage on failure, despite late
    emergency messages being confusing.
  - ignoring the penetration of mobiles. This solves the
    authentication problem. When one mobile beeps out an
    emergency message then it is a scam, when everybodies'
    phone goes 'beep' within a minute then it's for real.

A real design would augment the GSM design. It would send
emergency messages geographically, running through all mobiles
in a Mobile Switching Centre and sending them a message,
discarding failures.

There would be no use of the Short Messaging Service Centre
or the Home Location Register, the facets of the GSM design
which prevent geographic message broadcasts from scaling.

Plenty of other GSM services augment the basic GSM design
(Internet access being one), so it's not an unreasonable
thing to ask. It is however, expensive.

I'm left wondering if the lack of imagination in the paper
serves a lobbying purpose. If SMS broadcast appears technically
unfeasible then the GSM telcos avoid an expensive augmentation
of their network for a service they gain no revenue from.

Regards, Glen

-- 
  Glen Turner   <http://www.gdt.id.au/~gdt/>