Re: [rtcweb] 答复: Fwd: I-D Action: draft-westerlund-rtcweb-codec-control-00.txt

[Magnus Westerlund <[e-mail]>]

On 2012-05-21 03:44, 邓灵莉/denglingli wrote:
> Hi, Magnus
> 
> It seems to me that there may be another security threat in multi-party
> applications of COP, where an entity needs to combine multiple sets of
> requested parameters, than the one discussed in the draft. 
> That the initial downgrading of the combined potential ceiling for collected
> parameters for media quality (codec capabilities plus COP parameters as
> stated in Section 5) through SDP transaction by a malicious participant.
> Unlike the one stated in Section 8, the latter behavior only happens once
> and could neither been distinguished afterwards as "actively harmful" nor to
> be ignored in order to serve actually poorly-equipped users.
> Would that be an issue?
> 

Yes, this is clearly a security threat to the complete solution. Not
that it is specific to codec control. It is a threat to all things
expressed in the SDP, like which codecs being used, security mechanism
is negotiated etc.

In the WebRTC security architecture my undestanding is that it so far
are a deliberate choice of allowing the JavaScript and the web browser
to be allowed to modify the SDP if desired by them. Thus a security
model based on hop by hop security for the JSEP/SDP messages has been
selected. For example the usage of HTTPS / Websocket over TLS can
provide the security to prevent third parties not directly addressed
from seeing and affecting the JSEP/SDP messages.

Cheers

Magnus Westerlund

----------------------------------------------------------------------
Multimedia Technologies, Ericsson Research EAB/TVM
----------------------------------------------------------------------
Ericsson AB                | Phone  [phone]
Färögatan 6                | Mobile [phone]
SE-164 80 Stockholm, Sweden| mailto: [e-mail]
----------------------------------------------------------------------