WPA - is it worth it?

This is with reference to home use only.

The question is basically as subject says - on a home network is WPA
really necessary. I've read all the arguments about its advantages
over WEP time and again. Undoubtedly it provides a more secure system
than WEP - but is WEP really that bad?

I ask this because both my router and clients are able to make use of
WPA-PSK so the facility is available to me and several times I've
considered switching over to it from the 128bit WEP I use currently.
However, I am conscious that the continually changing encryption
cypher chews up more processing power, creates bigger packets and
therefore presumably uses more bandwidth and slows down
communications. I imagine this will also reduce the operating time on
battery powered laptops.

Does anyone know how significant these performance issues actually
are? WEP may be easier to crack but as my network contains nothing
that remarkable is anyone actually even going to bother trying? It
strikes me as being like locking a broken down car. You can't steal
the car and if you can't see anything through the window which looks
interesting why would you bother breaking in at all? Sure if it was
unlocked you might have a snoop (if you're that way inclined) in case
there's anything there, but if locked, would you go to the effort of
breaking in just for a quick look around?

I'd be interested to know peoples' thoughts :-)

Simon Pleasants schrieb:
> Does anyone know how significant these performance issues actually
> are?

they wrote in a german computer magazin (c't 13 -> www.heise.de/ct) last
week, that you should choose the best encryption which all of your
computers can manage. The loss of speed between WEP and WPA was between
the measurement inaccuracy.

Herbert

On Tue, 06 Jul 2004 13:25:04 +0100, Herbert Pommer
<(E-Mail Removed)> wrote:

>Simon Pleasants schrieb:
>> Does anyone know how significant these performance issues actually
>> are?
>
>they wrote in a german computer magazin (c't 13 -> www.heise.de/ct) last
>week, that you should choose the best encryption which all of your
>computers can manage. The loss of speed between WEP and WPA was between
>the measurement inaccuracy.
>
>Herbert

Thanks Herbert. I understand the point about using the best you can.

I followed the link you provided but as the magazine's website did not
require me to ask for directions to the town hall (the useful things
you learn at school) or order a currywurst (the useful things you
learn at the top of an Austrian mountain) there was not a lot I could
understand.

I also don't quite understand "The loss of speed between WEP and WPA
was between the measurement inaccuracy".

To change my original analogy slightly, I kind of figured no WEP is
like not locking the door to your house, anyone can come and go as
they please. WEP is kind of like locking the door, no-one can get in
unless they really want to and they have to have the time, skills,
technology and inclination to break in. If you had lots of valuables
inside suddenly (and someone outside knows it) breaking past that
locked door now becomes worthwhile. WPA is kind of like armour
plating the locked door and putting in motion sensors, perimeter
alarms and so on. Getting in suddenly becomes a MASSIVE amount more
difficult but it all uses up power and needs maintenance etc. Surely
a locked door would be a sufficient deterent if there's nothing inside
anyone would want anyway?

Simon Pleasants schrieb:
> [...Currywurst...]
School wasn't that helpful ;-)

> I also don't quite understand "The loss of speed between WEP and WPA
> was between the measurement inaccuracy".

They couldn't measure the differences of speed, processor-power or
battery use in their laboritories. The differences have been so small,
that other factors (like temperature) could have change them.

Your analogy is quite right. but if you have the possibility to "armour
plate the locked door" without having much more trouble - why you won't
do that?
on the other hand I think WPA(-PSK) is more friendly to install and to
maintain, because you set up the access and the encryption in one step.

Herbert

On Tue, 06 Jul 2004 14:18:40 +0100, Herbert Pommer
<(E-Mail Removed)> wrote:

>Simon Pleasants schrieb:
>> [...Currywurst...]
>School wasn't that helpful ;-)

Ah but I learned that by hanging around too many "imbiss" in my youth
:-) "Ein bier bitte" I have found to be equally useful.

>> I also don't quite understand "The loss of speed between WEP and WPA
>> was between the measurement inaccuracy".
>
>They couldn't measure the differences of speed, processor-power or
>battery use in their laboritories. The differences have been so small,
>that other factors (like temperature) could have change them.

Which would seem to indicate that the performance overheads are
minimal and scarcely worth mentioning - yet in all the reviews /
reports about WPA they are mentioned nevertheless.

>Your analogy is quite right. but if you have the possibility to "armour
>plate the locked door" without having much more trouble - why you won't
>do that?

My reluctance was based primarily on the possible loss of performance
of the network as a result - hence my interest in the "performance
overheads" being mentioned. If they are really as little as that then
it would seem that WPA is a win-win situation.

On 06 Jul 2004, in uk.telecom.broadband, Simon Pleasants wrote:
re http://www.heise.de/ct
>there was not a lot I could understand.

<http://translate.google.com/translate?hl=en&u=http%3A%2F%2Fwww%2Eheise%2Ede%2F ct%2F>

might help, though I didn't spot a comparison/discussion of WEP vs WPA

(Google has a translate feature on its toolbar, from which the above URL was
extracted, there's also <http://babelfish.altavista.com/>, from memory)