[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [RAI] RAI review of draft-ietf-speechsc-mrcpv2-19



Dan,

I prefer the text that recommends SRTP (It is a SHOULD and not a MUST). The text we currently have is based on the security reviews we got for RTP payload specifications, and as you can see it addresses the issue of why not to mandate SRTP.

Roni

 

From: Dan York [mailto:dyork at voxeo.com]
Sent: Wednesday, July 15, 2009 12:11 AM
To: Roni Even
Cc: 'Roni Even'; 'Daniel Burnett'; speechsc at ietf.org; 'Saravanan Shanmugham'; rai at ietf.org
Subject: Re: [RAI] RAI review of draft-ietf-speechsc-mrcpv2-19

 

Roni,

 

So as the RAI reviewer, are you okay with the text I suggested:

------

12.3. Media session protection 
Sensitive data is also carried on media sessions terminating on MRCPv2 servers (the other end of a media channel may or may not be on the MRCPv2 client). This data includes the user's spoken utterances    and the output of text-to-speech operations. MRCPv2 servers MUST support a security mechanism for protection of audio media sessions. MRCPv2 clients that originate or consume audio similarly MUST support a security mechanism for protection of the audio. 

------

 

Or would you prefer this text that includes the recommendation of SRTP?  (Which I noticed you did in the RTP payloads spec - and it makes sense to me to provide some basic guidance.):

------

12.3. Media session protection 
Sensitive data is also carried on media sessions terminating on MRCPv2 servers (the other end of a media channel may or may not be on the MRCPv2 client). This data includes the user's spoken utterances    and the output of text-to-speech operations. MRCPv2 servers MUST support a security mechanism for protection of audio media sessions. MRCPv2 clients that originate or consume audio similarly MUST support a security mechanism for protection of the audio. If appropriate, usage of the Secure Real-time Transport Protocol (SRTP) [RFC3711] is recommended.

------

 

Regards,

Dan

 

Regards,

Dan

 

On Jul 14, 2009, at 4:55 PM, Roni Even wrote:



Dan,

This is the general idea. The major reason is that there are various ways to protect the data and if you are not mandating one for interoperability then it can be more general

 

For example we have the following text when discussing security in the RTP payloads specifications.

 

RTP packets using the payload format defined in this specification

   are subject to the security considerations discussed in the RTP

   specification [RFC3550] and any appropriate RTP profile.  The main

   security considerations for the RTP packet carrying the RTP payload

   format defined within this memo are confidentiality, integrity, and

   source authenticity.  Confidentiality is achieved by encryption of

   the RTP payload.  Integrity of the RTP packets is achieved through a

   suitable cryptographic integrity protection mechanism.  Such a

   cryptographic system may also allow the authentication of the source

   of the payload.  A suitable security mechanism for this RTP payload

   format should provide confidentiality, integrity protection, and at

   least source authentication capable of determining if an RTP packet

   is from a member of the RTP session.

 

   Note that the appropriate mechanism to provide security to RTP and

   payloads following this memo may vary.  It is dependent on the

   application, the transport, and the signaling protocol employed.

   Therefore, a single mechanism is not sufficient, although if

   suitable, usage of the Secure Real-time Transport Protocol (SRTP)

   [RFC3711] recommended.  Other mechanisms that may be used are IPsec

   [RFC4301] Transport Layer Security (TLS) [RFC5246] (RTP over TCP);

   other alternatives may exist.

 

Roni Even

 

From: rai-bounces at ietf.org [mailto:rai-bounces at ietf.org] On Behalf Of Dan York
Sent: Tuesday, July 14, 2009 11:16 PM
To: Roni Even
Cc: 'Daniel Burnett'; speechsc at ietf.org; 'Saravanan Shanmugham'; rai at ietf.org
Subject: Re: [RAI] RAI review of draft-ietf-speechsc-mrcpv2-19

 

Roni,

 

------
12.3. Media session protection  
Sensitive data is also carried on media sessions terminating on MRCPv2 servers (the other end of a media channel may or may not be on the MRCPv2 client). This data includes the user's spoken utterances and the output of text-to-speech operations. MRCPv2 servers MUST support SRTP for protection of audio media sessions. MRCPv2 clients that originate or consume audio similarly MUST support SRTP. Alternative media channel protection MAY be used if desired (e.g. IPSEC).

------

 

Based on your comments and the srtp-not-mandatory draft (which was just revised to http://tools.ietf.org/html/draft-ietf-avt-srtp-not-mandatory-03 ), my understanding would be that you are advocating something more like this:

 

------

12.3. Media session protection 
Sensitive data is also carried on media sessions terminating on MRCPv2 servers (the other end of a media channel may or may not be on the MRCPv2 client). This data includes the user's spoken utterances    and the output of text-to-speech operations. MRCPv2 servers MUST support a security mechanism for protection of audio media sessions. MRCPv2 clients that originate or consume audio similarly MUST support a security mechanism for protection of the audio. 

------

 

Is that an accurate summary of your feedback?  Would that text be acceptable?

 

Regards,

Dan

 

On Jul 9, 2009, at 4:56 PM, Roni Even wrote:




Eric,
My comment is that in this case in AVT we say that you do not need to
mandate SRTP but mandate a security mechanism that can be  not only SRTP but
in a different layer like ipsec. This is why I gave a reference to the
srtp-not-mandatory draft

Roni



-----Original Message-----

Sent: Thursday, July 09, 2009 11:28 PM

To: Roni Even

Cc: Saravanan Shanmugham; Daniel Burnett; speechsc at ietf.org;

Subject: Re: RAI review of draft-ietf-speechsc-mrcpv2-19

 

The reality is that NO ONE has implemented any security to date. The

GENART reviewer raised the same issue, and so far the work group has

the same response: MRCPv2 (the speechsc work group) is not planning on

figuring out which of the seven key exchange mechanisms to use in

SIP.  We are counting on the community publishing something, and

people using it.  After all, we are the "using SIP for media resource

control" work group, not the "media resource control work group using

something like SIP for control."

 

Does this work for you?

 

On Jul 7, 2009, at 3:40 PM, Roni Even wrote:

 

[snip]

 

 

18.   In section 12.3 the suggestion is to use SRTP as the mandatory

interoperability mode. If the reason for mandating SRTP is for a

common mode you should also decide on a key exchange mechanism. I

not-mandatory-02

for discussion on media security.



_______________________________________________
RAI mailing list
RAI at ietf.org
https://www.ietf.org/mailman/listinfo/rai

 

-- 

Dan York, Director of Conversations

Voxeo Corporation   http://www.voxeo.com  dyork at voxeo.com

Phone: +1-407-455-5859    Skype: danyork  

 

 




 

 

 

 

 

 

 

-- 

Dan York, Director of Conversations

Voxeo Corporation   http://www.voxeo.com  dyork at voxeo.com

Phone: +1-407-455-5859    Skype: danyork