Wireless Detection and Connection Delays

Hi

What are the factors that determine the signal detection and actually
get a connetion to a wireless hot-spot.
What is the "acceptable" time in which an application - say a wifi
signal sniffer ... should be able to detect the signal and establish a
connection when in vicinity of a wireless hotspot.
If anyone has done any profiling about this or have any idea about
this, please help.

Thanks
Sandeep

On 18 Nov 2005 06:34:15 -0800, (E-Mail Removed) wrote:

>What are the factors that determine the signal detection and actually
>get a connetion to a wireless hot-spot.

The minimum connection time is the interval between SSID broadcasts.
You can't connect unless you can first identify. Typical broadcasts
are 10 times per second so you should be able to identify an access
point in less than 100msec. However, the actually connection requires
considerable back and forth traffic. This includes associations
requests (6 exchanges), key exchange (6 ??? requests), DHCP requests
(4 exchanges), ARP requests (2 exchanges), and initial DNS lookup. If
there's some security involved (802.11x) there's even more packets
going back and forth. It's difficult to calculate unless one knows
the setup.

>What is the "acceptable" time in which an application - say a wifi
>signal sniffer ... should be able to detect the signal and establish a
>connection when in vicinity of a wireless hotspot.

Passive sniffer (Kismet) = 1 management packet.
Active sniffer (Netstumbler) = 2 packet exchanges.

This is why sniffers are more sensitive and faster at finding access
points than an actual connection and login. They only need one or two
packets to produce a useable data point, while a full connection and
login requires many more.

>If anyone has done any profiling about this or have any idea about
>this, please help.

I suspect that this is your homework assignment. If so, please use
the resources at the skool and do your own research. If not, you did
not state what you are trying to accomplish and what you have to work
with. It's difficult to answer such theoretical questions without
context.


--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Jeff Liebermann wrote:
> On 18 Nov 2005 06:34:15 -0800, (E-Mail Removed) wrote:
>
> >What are the factors that determine the signal detection and actually
> >get a connetion to a wireless hot-spot.
>
> The minimum connection time is the interval between SSID broadcasts.
> You can't connect unless you can first identify. Typical broadcasts
> are 10 times per second so you should be able to identify an access
> point in less than 100msec. However, the actually connection requires
> considerable back and forth traffic. This includes associations
> requests (6 exchanges), key exchange (6 ??? requests), DHCP requests
> (4 exchanges), ARP requests (2 exchanges), and initial DNS lookup. If
> there's some security involved (802.11x) there's even more packets
> going back and forth. It's difficult to calculate unless one knows
> the setup.
>
> >What is the "acceptable" time in which an application - say a wifi
> >signal sniffer ... should be able to detect the signal and establish a
> >connection when in vicinity of a wireless hotspot.
>
> Passive sniffer (Kismet) = 1 management packet.
> Active sniffer (Netstumbler) = 2 packet exchanges.
>
> This is why sniffers are more sensitive and faster at finding access
> points than an actual connection and login. They only need one or two
> packets to produce a useable data point, while a full connection and
> login requires many more.
>
> >If anyone has done any profiling about this or have any idea about
> >this, please help.
>
> I suspect that this is your homework assignment. If so, please use
> the resources at the skool and do your own research. If not, you did
> not state what you are trying to accomplish and what you have to work
> with. It's difficult to answer such theoretical questions without
> context.
>


Thanks a lot for your response. BTW, It is not my homework assignment.
I am work on a product which provides wifi connectivity taking control
over windows wireless zero config. My objective was to make a
comparision between windows wireless management and ours. Being new to
the field, i was looking for ideas. Yours were a good ideas giving me a
starting point. Let me search for more. Thanks again.
Sandeep

On 19 Nov 2005 08:23:23 -0800, (E-Mail Removed) wrote:

>
>Jeff Liebermann wrote:
>> On 18 Nov 2005 06:34:15 -0800, (E-Mail Removed) wrote:
>>
>> >What are the factors that determine the signal detection and actually
>> >get a connetion to a wireless hot-spot.
>>
>> The minimum connection time is the interval between SSID broadcasts.
>> You can't connect unless you can first identify. Typical broadcasts
>> are 10 times per second so you should be able to identify an access
>> point in less than 100msec. However, the actually connection requires
>> considerable back and forth traffic. This includes associations
>> requests (6 exchanges), key exchange (6 ??? requests), DHCP requests
>> (4 exchanges), ARP requests (2 exchanges), and initial DNS lookup. If
>> there's some security involved (802.11x) there's even more packets
>> going back and forth. It's difficult to calculate unless one knows
>> the setup.
>>
>> >What is the "acceptable" time in which an application - say a wifi
>> >signal sniffer ... should be able to detect the signal and establish a
>> >connection when in vicinity of a wireless hotspot.
>>
>> Passive sniffer (Kismet) = 1 management packet.
>> Active sniffer (Netstumbler) = 2 packet exchanges.
>>
>> This is why sniffers are more sensitive and faster at finding access
>> points than an actual connection and login. They only need one or two
>> packets to produce a useable data point, while a full connection and
>> login requires many more.
>>
>> >If anyone has done any profiling about this or have any idea about
>> >this, please help.
>>
>> I suspect that this is your homework assignment. If so, please use
>> the resources at the skool and do your own research. If not, you did
>> not state what you are trying to accomplish and what you have to work
>> with. It's difficult to answer such theoretical questions without
>> context.

>Thanks a lot for your response. BTW, It is not my homework assignment.

My appologies, but I wasn't sure and really didn't want to ruin some
students learning experience.

>I am work on a product which provides wifi connectivity taking control
>over windows wireless zero config.

I would welcome such a product. There are ocassional queries in this
newsgroup for replacement wireless client managers that offer more
control and more diagnostics than Windoze Wireless Zero Config.
Specifically, the deficiencies (in order of importance) are:
1. An inability to select a specific access point by MAC address when
the SSID's are all identical.
2. Diagnostics that will disclose WEP/WPA/authentication errors and
generate a log file showing the negotiation progress. At this time,
it is impossible to use WZC for troubleshooting connection failures.
3. Site survey tools for searching for access points that clearly
identify the access point to avoid spoofed and fake access points.
4. More detailed status and signal reports (signal strength, signal
to noise ratio, error rate, connection speed, and MAC address).
5. Lack of a traffic monitor (bits/sec in both directions).
6. Lack of ability to adjust how agressively the client will stay
connected to an access point before giving up and searching for a
better signal. This is needed for roaming until 802.11r is released.
Intel recently added this feature to their Proset client.
7. WPA support for Windoze 2000.

>My objective was to make a
>comparision between windows wireless management and ours.

I can run some simple timing tests for you next week. However, these
are easy to do and you can probably do them yourself. You'll find
that simple initial discovery and connection is almost instantaneous
with anyone's client. However, when you add encryption and RADIUS
authentication to the puzzle, the times will increase. Add an
automatic VPN connection (SecureNet), and the connect times increase
even more. If there is a software firewall installed on the client
computah, add some more time.

I'm not sure why connect time is an issue. Most users have become
familiar with what has become an ordeal process for using wireless.
From start, it might be as bad as:
1. View available networks.
2. Select SSID of network.
3. Supply WPA phrase.
4. View splash screen. Click OK on legal disclaimer.
5. Supply login and password.
6. Start VPN client.
7. Select terminating system.
8. Supply login and password.
9. Display and select available servers.
10. Start user application.

Sounds to me like an ideal candidate for some scripting or login
automation.

>Being new to
>the field, i was looking for ideas. Yours were a good ideas giving me a
>starting point. Let me search for more. Thanks again.
>Sandeep

Good luck.

--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

> I'm not sure why connect time is an issue. Most users have become
> familiar with what has become an ordeal process for using wireless.
> From start, it might be as bad as:
> 1. View available networks.
> 2. Select SSID of network.
> 3. Supply WPA phrase.
> 4. View splash screen. Click OK on legal disclaimer.
> 5. Supply login and password.
> 6. Start VPN client.
> 7. Select terminating system.
> 8. Supply login and password.
> 9. Display and select available servers.
> 10. Start user application.
>
> Sounds to me like an ideal candidate for some scripting or login
> automation.
>

Thanks for you comments again! Would you say I use WMI and VBscript for
writing the Automation Scripts as you suggested ? Also, if you have
some sample scripts that can help me or point me towards the right
direction, I would really appreciate this.

Thanks again
Sandeep

In article <(E-Mail Removed) .com>, (E-Mail Removed) wrote:
>> I'm not sure why connect time is an issue. Most users have become
>> familiar with what has become an ordeal process for using wireless.
>> From start, it might be as bad as:
>> 1. View available networks.
>> 2. Select SSID of network.
>> 3. Supply WPA phrase.
>> 4. View splash screen. Click OK on legal disclaimer.
>> 5. Supply login and password.
>> 6. Start VPN client.
>> 7. Select terminating system.
>> 8. Supply login and password.
>> 9. Display and select available servers.
>> 10. Start user application.
>>
>> Sounds to me like an ideal candidate for some scripting or login
>> automation.
>>
>
>Thanks for you comments again! Would you say I use WMI and VBscript for
>writing the Automation Scripts as you suggested ? Also, if you have
>some sample scripts that can help me or point me towards the right
>direction, I would really appreciate this.
>
>Thanks again
>Sandeep
>

If only MS would expose the objects above to work with. WMI ought to be
where these objects exist, but ...

fundamentalism, fundamentally wrong.