> (1)
There's also the issue how the network and application layers have been
designed. WLAN has been designed to be "dumb", like the wired LAN, very
easy to use but any sophistication (e.g. security issues) has been difficult
to implement. BT is an good example of overdesigned protocol, resulting
increased time to implement and prone to compatibility issues. BT problems
between different vendors are way beyond the problems of the early days of
WLAN.
> (2)
> for example, 200 people packed in a railroad car and enjoying
> music/video using Bluetooth, or 200 students typing on their
> Bluetooth keyboard in a classroom? how will the 79 frequencies
> and 8 MAC addresses provide the capacity? will the low power
> radio with frequency hopping be enough for robust and fast
> response radio links?
> (please tell me how if Bluetooth can solve the problem.)
Take any wireless technology without broadcasting options and 200 people in
a railroad car watching music/video will be an issue. I would guess that
200 keyboards and mice might even work with Bluetooth. But then again, is
there are a reason why the video should go over wireless link? Typically
the screen and receiving unit are attached. Is there are advantage of
having a classroom full of wireless keyboards and mice? Not really, pretty
much the opposite. Without cables the students will be more likely to steal
the keyboards and mice.
It's really a question of where wireless and wired communication makes most
sense. It would make sense to have e.g. "wireless usb", "wireless firewire"
and "wired bluetooth" to make the implementations easier. WLAN wouldn't
have been a success if it wasn't so well compatible with the wired LAN.
> one problem with WPAN is that the traffic is way higher than
> WWAN on a per node basis. this is easy to understand if you
> compare the traffic over your Internet connection, to/from
> your peripherals, and on the system bus.
Not exactly true, there's more traffic on my LAN/WLAN than through the
Internet connection but the peripherals are a different issue. Most of the
high-volume peripherals are already connected through the LAN and would
never benefit from e.g. Bluetooth and there are a number of low-volume or
low-bandwidth peripherals that would benefit from a single, very local
wireless connection. The peripherals also have different traffic patterns,
a scanner would probably benefit more from WLAN connection than BT. WLAN
for keyboard or mice would be an overkill.
Petri
Received on Thu Jul 25 14:36:29 2002