Security Basics mailing list archives
RE: 802.1x and PEAP
From: shankarnarayan.d () netsol co in
Date: Sat, 6 Mar 2004 15:01:17 +0530
I agree with David when he says that the Cisco cards with TKIP and MIC don't work with other cards - the basic reason may be enumerated as follows 1. Cisco uses the pre-standard version of TKIP/ MIC combine 2. Cisco's TKIP works as follows a. A base key is hashed with the IV to provide a per packet key. The same is used to derive a WEP key to encrypt the data b. When all the IV's for a particular base key are complete, the base key is changed dynamically (based on a time defined by the sys admin) c. In this way, the regularly changing base key and the ever-changing IV are hashed to provide new keys regularly d. MIC is provided using Neils Ferugeson's algorithm This TKIP is not the same as the now revised draft standard for 802.11i and hence it is not compatible with other Wireless LAN vendor cards (who may use the same or alternate algorithms) The TKIP key mixing function that the standards are pursuing include mixing the 802MAC address and a temporal key to index and S-box. The MAC ensures different key for different work stations. This provides a 16-byte output that resembles 13-byte WEP key and 3-byte corresponding to the IV. The process is pretty complex and runs through multiple equations / mixing functions etc This is one among the reasons why TKIP on Cisco may not be compatible with those of other vendors when using Cisco AP's. Shankar -----Original Message----- From: Rosenhan, David [mailto:David.Rosenhan () swiftbrands com] Sent: Friday, March 05, 2004 7:41 PM To: Rosado, Rafael (Rafael); Camillo Bucciarelli Cc: security-basics () securityfocus com; shankarnarayan.d () netsol co in Subject: RE: 802.1x and PEAP True, Cisco was one of the first vendors to even make it available. WiFi or the IEEE hadn't come out with any true standard yet so Cisco basically built their own, this is why you can't use another vendors card and have TKIP with MIC and WEP enabled on a Cisco AP. But I am right when I say that no other card (that I have tried or that I know of) then a Cisco card will work with a Cisco AP running WEP+TKIP+MIC. However with the new IBM T40 built in wireless card you can use the new Cipher suite Cisco implemented in the IOS code on the 1200's, 350's and 1100's (after a free download of the software from the IBM website which includes IBM's Application software and driver updates) This integrates WPA with TKIP using the IEEE standard, it also works with Cisco LEAP and with regular EAP. There is an option above the Cipher suite option that is Cisco proprietary MIC and TKIP used with WEP in the IOS code. I have spent countless hours testing all of this so if you need more info then let me know. David Rosenhan, CCNP Information Technology -----Original Message----- From: Rosado, Rafael (Rafael) [mailto:rarosado () lucent com] Sent: Friday, March 05, 2004 6:32 AM To: Rosenhan, David; Camillo Bucciarelli Cc: security-basics () securityfocus com; shankarnarayan.d () netsol co in Subject: RE: 802.1x and PEAP David, I disagree with your comment about TKIP and MIC being proprietary. TKIP and MIC are part of the Wifi Alliance's interim solution to WEP deficiencies which are a subset of the Wifi Protected Access solution of IEEE 802.11i (still in draft, expected to be ratified sometime 2nd-3rd QTR 2004). Cisco has a proprietary version of TKIP, but it is based on the framework estabished by the WiFi Alliance. Rafael Rosado, CISSP, CISA IT Security Manager Lucent Technologies IT Infrastructure - Network Design 2400 SW 145th Avenue Miramar, Florida 33027 Office: 954-885-2176 Facsimile: 954-885-3861 Email: rarosado () lucent com This electronic mail message contains information belonging to Lucent Technologies, which may be confidential and/or legal privileged. The information is intended only for the use of the individual or entity named above. If you are not the intended recipient, you are hereby notified that any disclosure, printing, copying, distribution, or the taking of any action in reliance on the contents of this electronically mailed information is strictly prohibited. If you receive this message in error, please immediately notify us by electronic mail and delete this message. -----Original Message----- From: Rosenhan, David [mailto:David.Rosenhan () swiftbrands com] Sent: Thursday, March 04, 2004 3:18 PM To: Camillo Bucciarelli Cc: security-basics () securityfocus com; shankarnarayan.d () netsol co in Subject: RE: 802.1x and PEAP Camillo, Broadcast key rotation can only be done with an authentication server. TKIP and MIC are Cisco proprietary, if you have an AP running VXWorks and not IOS they you won't get a different vendors card other then a 340 or 350 card to work with TKIP and MIC, period, even if you upgrade to IOS a different vendors card will not work with TKIP and MIC, but there are other options with IOS. If you upgrade to IOS on your AP (1200's and 350 AP's are up-gradable to IOS) then you have some new options, you can now use new IEEE standards like WPA, the problem is the manufacturers card has to support it. WPA is really new, even with Cisco 340 and 350 cards you have to use a separate piece of software (Like the Funk Odyssey client) to use WPA pre-shared keys. IEEE also included TKIP with WPA and you don't need a server to use it with the new IOS software on the 1200 and 350 AP's. Plus there are options for EAP with WPA and broadcast key rotation with authentication to a RADIUS server (Cisco has doc's that talk about how the ACS server works with all of this on their website). Thanks! David Rosenhan, CCNP Information Technology -----Original Message----- From: Camillo Bucciarelli [mailto:camillobucciarelli () yahoo it] Sent: Thursday, March 04, 2004 8:43 AM To: shankarnarayan.d () netsol co in Cc: security-basics () securityfocus com Subject: RE: 802.1x and PEAP Can I use these features(Enhanced MIC verification for WEP, Temporal Key Integrity Protocol, Broadcast WEP Key rotation) with a non-cisco wireless adatpter? Such as a 3com wireless PCMCIA? Actually I've tried a cisco aironet 340 wireless card. Regards, Camillo Bucciarelli --- shankarnarayan.d () netsol co in ha scritto: > This can be done best on the wireless networks
having AP's from Cisco. The others are still in the process of accomplishing the same on their Access Points (most have done it, some are yet to accomplish the same). The broadcast key is negotiated for the first time and then the
same is changed at periodic intervals (configurable by an administrator). The old broadcast key is used to encrypt the new key and the same is broadcast out to all the clients on the access point at the expiry of the administrator defined time limit. On a Cisco you would use the following commands on the Aironet 1100/ 1200 (with IOS) in order BM_1036542configure terminal BM_1036548 interface dot11radio { 0 | 1 } broadcast-key change seconds BM_1036574 end BM_1036580 copy running-config startup-config Rgds, Shankar -----Original Message----- From: Camillo Bucciarelli [mailto:camillobucciarelli () yahoo it] Sent: Wednesday, March 03, 2004 3:03 PM To: shankarnarayan.d () netsol co in Subject: RE: 802.1x and PEAP Thanks, this is what I need to know. I have another question: I need to use 802.1x in order to enable the "broadcast key rotation"? Camillo shankarnarayan.d () netsol co in wrote: The Lines below have been pulled straight from the PEAP working draft.
This clearly defines that the initial negotiation of the PEAP is as in
the TLS - thus providing the necessary security. Hope this answers your question OR have I got it wrong - If you believe this is not the information that you were looking for request you to please rephrase your question Shankar Protected EAP (PEAP) Version 2 is comprised of a two-part conversation: [1] In Part 1, a TLS session is negotiated, with server authenticating
to the client and optionally the client to the server. The negotiated key is then used to encrypt the rest of the conversation. [2] In Part 2, within the TLS session, zero or more EAP methods are carried out. Part 2 completes with a success/failure indication protected by the TLS session or a protected error (TLS alert). The PEAP conversation typically begins with an optional identity exchange. The initial identity exchange is used primarily to route the
EAP conversation to the EAP server. Since the initial identity exchange is in the clear, the peer MAY decide to place a routing realm
instead of its real name in the EAP-Response/Identity. In short, the first exchange is based on TLS where certificates are used much in the same way as that used in the EAP-TLS. The remaining information of identity etc is then pumped through the TLS tunnel. Hence, EAP-TLS may be one of the methods (actually the most common method) used to establish the tunnel (using certificates) Shankar -----Original Message----- From: Camillo Bucciarelli [mailto:camillobucciarelli () yahoo it] Sent: Tuesday, March 02, 2004 3:46 PM To: security-basics () securityfocus com Subject: 802.1x and PEAP Good morning, I'm looking for detailed information about the Protected EAP. I can't understand what the supplicant and Access Server use to establish the TLS tunnel. Here's an example: Authenticating Peer Authenticator ------------------- ------------- <- EAP-Request/ Identity EAP-Response/ Identity (MyID) -> <- EAP-Request/ EAP-Type=PEAP, V=0 (PEAP Start, S bit set) EAP-Response/ EAP-Type=PEAP, V=0 (TLS client_hello)-> <- EAP-Request/ EAP-Type=PEAP, V=0 (TLS server_hello, TLS certificate, [TLS server_key_exchange,] [TLS certificate_request,] TLS server_hello_done) EAP-Response/ EAP-Type=PEAP, V=0 ([TLS certificate,] TLS client_key_exchange, [TLS certificate_verify,] TLS change_cipher_spec, TLS finished) -> <- EAP-Request/ EAP-Type=PEAP, V=0 (TLS change_cipher_spec, TLS finished) EAP-Response/ EAP-Type=PEAP -> TLS channel established (messages sent within the TLS channel) They exchange a server_key_exchange and a client_key_exchange used to derive the session key. It seems to me that the key exchange between the client and the server
is done in clear text, but this means that I can actually sniff this exchange. Now, this seems not logical to me. Anyone here has any idea about "where" I am wrong ? Do the two elements hash in some way the keys ? Or, another possibility, do we actually have the client key encrypted with the public key that belongs to the server - that is of course available - and we have the server key *only* that is transmitted in clear text ? In the TLS protocol of course the two key are encrypted with the ublic key of the "other end". But in PEAP ? Thanks in advance, Camillo ===== Camillo Bucciarelli
______________________________________________________________________
Yahoo! Mail: 6MB di spazio gratuito, 30MB per i tuoi allegati, l'antivirus, il filtro Anti-spam
http://it.yahoo.com/mail_it/foot/?http://it.mail.yahoo.com/
------------------------------------------------------------------------ ---
Free 30-day trial: firewall with virus/spam protection, URL filtering,
VPN, wireless security Protect your network against hackers, viruses, spam and other risks with Astaro Security Linux, the comprehensive security solution that combines six
=== message truncated === ===== Camillo Bucciarelli ______________________________________________________________________ Yahoo! Mail: 6MB di spazio gratuito, 30MB per i tuoi allegati, l'antivirus, il filtro Anti-spam http://it.yahoo.com/mail_it/foot/?http://it.mail.yahoo.com/ ------------------------------------------------------------------------ --- Ethical Hacking at the InfoSec Institute. Mention this ad and get $545 off any course! All of our class sizes are guaranteed to be 10 students or less to facilitate one-on-one interaction with one of our expert instructors. Attend a course taught by an expert instructor with years of in-the-field pen testing experience in our state of the art hacking lab. Master the skills of an Ethical Hacker to better assess the security of your organization. Visit us at: http://www.infosecinstitute.com/courses/ethical_hacking_training.html ------------------------------------------------------------------------ ---- ------------------------------------------------------------------------ --- Ethical Hacking at the InfoSec Institute. Mention this ad and get $545 off any course! All of our class sizes are guaranteed to be 10 students or less to facilitate one-on-one interaction with one of our expert instructors. Attend a course taught by an expert instructor with years of in-the-field pen testing experience in our state of the art hacking lab. Master the skills of an Ethical Hacker to better assess the security of your organization. Visit us at: http://www.infosecinstitute.com/courses/ethical_hacking_training.html ------------------------------------------------------------------------ ---- --------------------------------------------------------------------------- Ethical Hacking at the InfoSec Institute. Mention this ad and get $545 off any course! All of our class sizes are guaranteed to be 10 students or less to facilitate one-on-one interaction with one of our expert instructors. Attend a course taught by an expert instructor with years of in-the-field pen testing experience in our state of the art hacking lab. Master the skills of an Ethical Hacker to better assess the security of your organization. Visit us at: http://www.infosecinstitute.com/courses/ethical_hacking_training.html ----------------------------------------------------------------------------
Current thread:
- 802.1x and PEAP Camillo Bucciarelli (Mar 02)
- <Possible follow-ups>
- RE: 802.1x and PEAP shankarnarayan . d (Mar 03)
- RE: 802.1x and PEAP Camillo Bucciarelli (Mar 03)
- RE: 802.1x and PEAP Camillo Bucciarelli (Mar 04)
- RE: 802.1x and PEAP Rosenhan, David (Mar 04)
- RE: 802.1x and PEAP Rosado, Rafael (Rafael) (Mar 08)
- RE: 802.1x and PEAP shankarnarayan . d (Mar 08)
- RE: 802.1x and PEAP Rosenhan, David (Mar 08)
- RE: 802.1x and PEAP Camillo Bucciarelli (Mar 08)
- RE: 802.1x and PEAP Rosenhan, David (Mar 08)
- RE: 802.1x and PEAP shankarnarayan . d (Mar 08)
- RE: 802.1x and PEAP Jason Humes (Mar 08)
- Re: 802.1x and PEAP balinsky (Mar 29)