Discussion:
[OpenWireless Tech] Open secure wireless
Mitar
2015-06-17 11:33:16 UTC
Permalink
Hi!

Reading this old post:

https://www.eff.org/deeplinks/2011/04/open-wireless-movement

I wanted to point some research done on this some time ago:

http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf

And also some progress:

http://www.riosec.com/articles/open-secure-wireless-20

If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.

(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)


Mitar
--
http://mitar.tnode.com/
https://twitter.com/mitar_m
Diderik van Wingerden
2015-06-18 07:18:47 UTC
Permalink
Hi Mitar,

Thanks for sharing. I am no expert on the subject, but it sounds like a
great addition to open wireless (and wireless networking in general). So
would it be possible to implement this in LibreCMC (or OpenWRT) for
example? And would it then require something on the client's end? Like a
new driver or certificate, as you mention? I mean, the solution would of
course be adopted much faster if a client install/config of some sort
would not be necessary, or at least be super easy.

best regards,
Diderik
Send Tech mailing list submissions to
To subscribe or unsubscribe via the World Wide Web, visit
https://srv1.openwireless.org/mailman/listinfo/tech
or, via email, send a message with subject or body 'help' to
You can reach the person managing the list at
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Tech digest..."
1. Open secure wireless (Mitar)
----------------------------------------------------------------------
Message: 1
Date: Wed, 17 Jun 2015 04:33:16 -0700
Subject: [OpenWireless Tech] Open secure wireless
Content-Type: text/plain; charset=UTF-8
Hi!
https://www.eff.org/deeplinks/2011/04/open-wireless-movement
http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf
http://www.riosec.com/articles/open-secure-wireless-20
If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.
(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)
Mitar
--
Warm regards, hartelijke groet,

Diderik van Wingerden
+31621639148
http://www.think-innovation.com/

"Do what is right."
Russell Senior
2015-06-18 07:55:13 UTC
Permalink
Does this idea require a keeping a private key on the router? If so,
that's a problem, since routers are often quite vulnerable to physical
access. If an entire community network relied on a single certificate
for authentication across all of their infrastructure (based on their
extended SSID), then losing one AP could mean complete compromise.

On Thu, Jun 18, 2015 at 12:18 AM, Diderik van Wingerden
Post by Diderik van Wingerden
Hi Mitar,
Thanks for sharing. I am no expert on the subject, but it sounds like a
great addition to open wireless (and wireless networking in general). So
would it be possible to implement this in LibreCMC (or OpenWRT) for
example? And would it then require something on the client's end? Like a
new driver or certificate, as you mention? I mean, the solution would of
course be adopted much faster if a client install/config of some sort
would not be necessary, or at least be super easy.
best regards,
Diderik
Send Tech mailing list submissions to
To subscribe or unsubscribe via the World Wide Web, visit
https://srv1.openwireless.org/mailman/listinfo/tech
or, via email, send a message with subject or body 'help' to
You can reach the person managing the list at
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Tech digest..."
1. Open secure wireless (Mitar)
----------------------------------------------------------------------
Message: 1
Date: Wed, 17 Jun 2015 04:33:16 -0700
Subject: [OpenWireless Tech] Open secure wireless
Content-Type: text/plain; charset=UTF-8
Hi!
https://www.eff.org/deeplinks/2011/04/open-wireless-movement
http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf
http://www.riosec.com/articles/open-secure-wireless-20
If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.
(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)
Mitar
--
Warm regards, hartelijke groet,
Diderik van Wingerden
+31621639148
http://www.think-innovation.com/
"Do what is right."
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
Christopher Byrd
2015-06-18 21:22:02 UTC
Permalink
WPA-Enterprise requires a private key on the authentication server, but the
AS doesn't have to run on the access point. In a distributed scenario like
a community network, it's likely there would be a centralized AS. If you
use good practices and don't use the same RADIUS key for all routers, then
loosing one AP would likely compromise only sessions on that AP. The
exception here is if fast credential roaming (like 802.11r) is used, in
which case other sessions may be cached on the AP as well. Some distributed
wireless systems use a lightweight AP and centralized controller (split-MAC
architecture); in those instances not much of value (other than the
hardware) is lost when an AP is compromised.

Christopher
Post by Russell Senior
Does this idea require a keeping a private key on the router? If so,
that's a problem, since routers are often quite vulnerable to physical
access. If an entire community network relied on a single certificate
for authentication across all of their infrastructure (based on their
extended SSID), then losing one AP could mean complete compromise.
On Thu, Jun 18, 2015 at 12:18 AM, Diderik van Wingerden
Post by Diderik van Wingerden
Hi Mitar,
Thanks for sharing. I am no expert on the subject, but it sounds like a
great addition to open wireless (and wireless networking in general). So
would it be possible to implement this in LibreCMC (or OpenWRT) for
example? And would it then require something on the client's end? Like a
new driver or certificate, as you mention? I mean, the solution would of
course be adopted much faster if a client install/config of some sort
would not be necessary, or at least be super easy.
best regards,
Diderik
Send Tech mailing list submissions to
To subscribe or unsubscribe via the World Wide Web, visit
https://srv1.openwireless.org/mailman/listinfo/tech
or, via email, send a message with subject or body 'help' to
You can reach the person managing the list at
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Tech digest..."
1. Open secure wireless (Mitar)
----------------------------------------------------------------------
Message: 1
Date: Wed, 17 Jun 2015 04:33:16 -0700
Subject: [OpenWireless Tech] Open secure wireless
<
Content-Type: text/plain; charset=UTF-8
Hi!
https://www.eff.org/deeplinks/2011/04/open-wireless-movement
http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf
Post by Diderik van Wingerden
http://www.riosec.com/articles/open-secure-wireless-20
If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.
(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)
Mitar
--
Warm regards, hartelijke groet,
Diderik van Wingerden
+31621639148
http://www.think-innovation.com/
"Do what is right."
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
demos
2015-06-18 22:03:57 UTC
Permalink
Hi all,

Another example for a secure wireless mesh can be GNUnet.
It does not require a central instance that could be compromised.
Every AP/client has his own (public)key.
Filesharing is anonymous already, they are implementing Onion Routing
without the need of central authorities.
The standard GNUnet although has Perfect Future Secrecy.
Right know i don't know wether the tiny packaging will include that
feature too.
It's still in process:
https://github.com/openwrt/packages/tree/master/net/gnunet

Cheers
Demos
Post by Christopher Byrd
WPA-Enterprise requires a private key on the authentication server, but
the AS doesn't have to run on the access point. In a distributed
scenario like a community network, it's likely there would be a
centralized AS. If you use good practices and don't use the same RADIUS
key for all routers, then loosing one AP would likely compromise only
sessions on that AP. The exception here is if fast credential roaming
(like 802.11r) is used, in which case other sessions may be cached on
the AP as well. Some distributed wireless systems use a lightweight AP
and centralized controller (split-MAC architecture); in those instances
not much of value (other than the hardware) is lost when an AP is
compromised.
Christopher
On Thu, Jun 18, 2015 at 2:55 AM, Russell Senior
Does this idea require a keeping a private key on the router? If so,
that's a problem, since routers are often quite vulnerable to physical
access. If an entire community network relied on a single certificate
for authentication across all of their infrastructure (based on their
extended SSID), then losing one AP could mean complete compromise.
On Thu, Jun 18, 2015 at 12:18 AM, Diderik van Wingerden
Post by Diderik van Wingerden
Hi Mitar,
Thanks for sharing. I am no expert on the subject, but it sounds
like a
Post by Diderik van Wingerden
great addition to open wireless (and wireless networking in
general). So
Post by Diderik van Wingerden
would it be possible to implement this in LibreCMC (or OpenWRT) for
example? And would it then require something on the client's end?
Like a
Post by Diderik van Wingerden
new driver or certificate, as you mention? I mean, the solution
would of
Post by Diderik van Wingerden
course be adopted much faster if a client install/config of some sort
would not be necessary, or at least be super easy.
best regards,
Diderik
Send Tech mailing list submissions to
To subscribe or unsubscribe via the World Wide Web, visit
https://srv1.openwireless.org/mailman/listinfo/tech
or, via email, send a message with subject or body 'help' to
You can reach the person managing the list at
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Tech digest..."
1. Open secure wireless (Mitar)
----------------------------------------------------------------------
Post by Diderik van Wingerden
Message: 1
Date: Wed, 17 Jun 2015 04:33:16 -0700
Subject: [OpenWireless Tech] Open secure wireless
Content-Type: text/plain; charset=UTF-8
Hi!
https://www.eff.org/deeplinks/2011/04/open-wireless-movement
http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf
Post by Diderik van Wingerden
http://www.riosec.com/articles/open-secure-wireless-20
If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.
(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)
Mitar
--
Warm regards, hartelijke groet,
Diderik van Wingerden
+31621639148 <tel:%2B31621639148>
http://www.think-innovation.com/
"Do what is right."
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
--
Echt Dezentrales Netz - EDN:
The goal of EDN is to verify the applicability of existing technologies
and solutions,
and to integrate them in a comprehensive product.
High level security communication via an Open Wireless Meshnet including
several services.

https://wiki.c3d2.de/Echt_Dezentrales_Netz/en
Key here: https://pgp.mit.edu/pks/lookup?op=get&search=0x9B365E2DBF83D308
Mitar
2015-06-18 22:31:57 UTC
Permalink
Hi!

I think the main approach I would take is to not have any special
private key on the router, but just do session-based encryption. WPA2
is already doing that. (Of course, it is not perfect, if you listen to
initial frames you can decrypt traffic.) You should need a private key
only to prevent MITM attacks. But for example for mesh networks there
are so many other ways to do MITM that it is questionable how much
would be worth to do try to prevent it on the client connection.


Mitar

On Thu, Jun 18, 2015 at 12:55 AM, Russell Senior
Post by Russell Senior
Does this idea require a keeping a private key on the router? If so,
that's a problem, since routers are often quite vulnerable to physical
access. If an entire community network relied on a single certificate
for authentication across all of their infrastructure (based on their
extended SSID), then losing one AP could mean complete compromise.
On Thu, Jun 18, 2015 at 12:18 AM, Diderik van Wingerden
Post by Diderik van Wingerden
Hi Mitar,
Thanks for sharing. I am no expert on the subject, but it sounds like a
great addition to open wireless (and wireless networking in general). So
would it be possible to implement this in LibreCMC (or OpenWRT) for
example? And would it then require something on the client's end? Like a
new driver or certificate, as you mention? I mean, the solution would of
course be adopted much faster if a client install/config of some sort
would not be necessary, or at least be super easy.
best regards,
Diderik
Send Tech mailing list submissions to
To subscribe or unsubscribe via the World Wide Web, visit
https://srv1.openwireless.org/mailman/listinfo/tech
or, via email, send a message with subject or body 'help' to
You can reach the person managing the list at
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Tech digest..."
1. Open secure wireless (Mitar)
----------------------------------------------------------------------
Message: 1
Date: Wed, 17 Jun 2015 04:33:16 -0700
Subject: [OpenWireless Tech] Open secure wireless
Content-Type: text/plain; charset=UTF-8
Hi!
https://www.eff.org/deeplinks/2011/04/open-wireless-movement
http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf
http://www.riosec.com/articles/open-secure-wireless-20
If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.
(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)
Mitar
--
Warm regards, hartelijke groet,
Diderik van Wingerden
+31621639148
http://www.think-innovation.com/
"Do what is right."
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
--
http://mitar.tnode.com/
https://twitter.com/mitar_m
demos
2015-06-18 23:45:35 UTC
Permalink
Hi:)

Am 19.06.2015 um 00:31 schrieb Mitar:> Hi!
Post by Mitar
I think the main approach I would take is to not have any special
private key on the router, but just do session-based encryption. WPA2
is already doing that. (Of course, it is not perfect, if you listen to
initial frames you can decrypt traffic.)-> that would be prevented having the public key of your communication partner.
well and the meta-data-protection feature? :)
meta data are the context for content, they are the index of a book.
they are sensitive data.

I forgot to mention that it does authentication too and has a friend to
friend
mode- for a friend to friend darknet.

You should need a private key
Post by Mitar
only to prevent MITM attacks. But for example for mesh networks there
are so many other ways to do MITM that it is questionable how much
would be worth to do try to prevent it on the client connection.
Page 40 examines possible attack scenarios on GNUnet.
http://dotnetlabs.org/Content/pdf/GNUnet.pdf
Are these attacks considered here?(Index page 4, the very helpful
metadata :))


good night.
Demos
Post by Mitar
On Thu, Jun 18, 2015 at 12:55 AM, Russell Senior
Post by Russell Senior
Does this idea require a keeping a private key on the router? If so,
that's a problem, since routers are often quite vulnerable to physical
access. If an entire community network relied on a single certificate
for authentication across all of their infrastructure (based on their
extended SSID), then losing one AP could mean complete compromise.
On Thu, Jun 18, 2015 at 12:18 AM, Diderik van Wingerden
Post by Diderik van Wingerden
Hi Mitar,
Thanks for sharing. I am no expert on the subject, but it sounds like a
great addition to open wireless (and wireless networking in general). So
would it be possible to implement this in LibreCMC (or OpenWRT) for
example? And would it then require something on the client's end? Like a
new driver or certificate, as you mention? I mean, the solution would of
course be adopted much faster if a client install/config of some sort
would not be necessary, or at least be super easy.
best regards,
Diderik
Send Tech mailing list submissions to
To subscribe or unsubscribe via the World Wide Web, visit
https://srv1.openwireless.org/mailman/listinfo/tech
or, via email, send a message with subject or body 'help' to
You can reach the person managing the list at
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Tech digest..."
1. Open secure wireless (Mitar)
----------------------------------------------------------------------
Message: 1
Date: Wed, 17 Jun 2015 04:33:16 -0700
Subject: [OpenWireless Tech] Open secure wireless
Content-Type: text/plain; charset=UTF-8
Hi!
https://www.eff.org/deeplinks/2011/04/open-wireless-movement
http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf
http://www.riosec.com/articles/open-secure-wireless-20
If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.
(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)
Mitar
--
Warm regards, hartelijke groet,
Diderik van Wingerden
+31621639148
http://www.think-innovation.com/
"Do what is right."
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
--
Echt Dezentrales Netz - EDN:
The goal of EDN is to verify the applicability of existing technologies
and solutions,
and to integrate them in a comprehensive product.
High level security communication via an Open Wireless Meshnet including
several services.

https://wiki.c3d2.de/Echt_Dezentrales_Netz/en
Key here: https://pgp.mit.edu/pks/lookup?op=get&search=0x9B365E2DBF83D308
demos
2015-06-19 08:41:08 UTC
Permalink
Post by demos
Hi:)
Am 19.06.2015 um 00:31 schrieb Mitar:> Hi!
Post by Mitar
I think the main approach I would take is to not have any special
private key on the router, but just do session-based encryption. WPA2
is already doing that. (Of course, it is not perfect, if you listen to
initial frames you can decrypt traffic.)-> that would be prevented having the public key of your communication partner.
well and the meta-data-protection feature? :)
meta data are the context for content, they are the index of a book.
they are sensitive data.
I forgot to mention that it does authentication too and has a friend to
friend
mode- for a friend to friend darknet.
You should need a private key
Post by Mitar
only to prevent MITM attacks. But for example for mesh networks there
are so many other ways to do MITM that it is questionable how much
would be worth to do try to prevent it on the client connection.
Page 40 examines possible attack scenarios on GNUnet.
http://dotnetlabs.org/Content/pdf/GNUnet.pdf
Are these attacks considered here?(Index page 4, the very helpful
metadata :))
To be more clear with these i mean the MITM attacks in mesh networks you
mentioned.
Do you think they apply to GNUnet too?
Post by demos
good night.
Demos
Post by Mitar
On Thu, Jun 18, 2015 at 12:55 AM, Russell Senior
Post by Russell Senior
Does this idea require a keeping a private key on the router? If so,
that's a problem, since routers are often quite vulnerable to physical
access. If an entire community network relied on a single certificate
for authentication across all of their infrastructure (based on their
extended SSID), then losing one AP could mean complete compromise.
On Thu, Jun 18, 2015 at 12:18 AM, Diderik van Wingerden
Post by Diderik van Wingerden
Hi Mitar,
Thanks for sharing. I am no expert on the subject, but it sounds like a
great addition to open wireless (and wireless networking in general). So
would it be possible to implement this in LibreCMC (or OpenWRT) for
example? And would it then require something on the client's end? Like a
new driver or certificate, as you mention? I mean, the solution would of
course be adopted much faster if a client install/config of some sort
would not be necessary, or at least be super easy.
best regards,
Diderik
Send Tech mailing list submissions to
To subscribe or unsubscribe via the World Wide Web, visit
https://srv1.openwireless.org/mailman/listinfo/tech
or, via email, send a message with subject or body 'help' to
You can reach the person managing the list at
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Tech digest..."
1. Open secure wireless (Mitar)
----------------------------------------------------------------------
Message: 1
Date: Wed, 17 Jun 2015 04:33:16 -0700
Subject: [OpenWireless Tech] Open secure wireless
Content-Type: text/plain; charset=UTF-8
Hi!
https://www.eff.org/deeplinks/2011/04/open-wireless-movement
http://www.riosec.com/articles/Open-Secure-Wireless
http://www.riosec.com/articles/Open-Secure-Wireless/Open-Secure-Wireless.pdf
http://www.riosec.com/articles/open-secure-wireless-20
If you are not doing that already, I think EFF should get on board of
supporting those changes to the standard.
(BTW, originally, as presented in 1.0 paper, WiFi standard does allow
open and secure connections, just no operating system really
implements it because they all first prompt for the password, before
trying to connect to the encrypted WiFi network to figure out the
password is really required.)
Mitar
--
Warm regards, hartelijke groet,
Diderik van Wingerden
+31621639148
http://www.think-innovation.com/
"Do what is right."
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
_______________________________________________
Tech mailing list
https://srv1.openwireless.org/mailman/listinfo/tech
--
Echt Dezentrales Netz - EDN:
The goal of EDN is to verify the applicability of existing technologies
and solutions,
and to integrate them in a comprehensive product.
High level security communication via an Open Wireless Meshnet including
several services.

https://wiki.c3d2.de/Echt_Dezentrales_Netz/en
Key here: https://pgp.mit.edu/pks/lookup?op=get&search=0x9B365E2DBF83D308
Loading...