Re: [Spud] No. Operators don't need SPUD for mobile network management

["Fossati, Thomas (Nokia - GB)" <[e-mail]>]

On 27/07/2016 04:05, "Spud on behalf of Tom Herbert"
<[e-mail] on behalf of [e-mail]> wrote:
>On Mon, Jul 25, 2016 at 3:49 AM, Smith, Kevin, (R&D) Vodafone Group
><[e-mail]> wrote:
>> Another operator here:  I like the notion of PLUS being used to hint to
>>the TCP sender regarding network radio cell conditions, i.e. mobile
>>throughput guidance[1].  So network information being sent out to help
>>tune cwnd.
>>
>I am not a big fan of having middleboxes purposely parsing an
>modifying TCP options, but I do like security considerations of mobile
>guidance and hope that PLUS might adopt some of the same concepts.
>Specifically:
>
>"The protocol specified in this document assumes that a trustful
>relationship between the Throughput Guidance Provider and the TCP
>server has been formed"
>
>and
>
>"Throughput guidance is considered confidential information"
>
>and
>
>"The identity of the Mobile Throughput Guidance provider that injects
>the throughput guidance header must be explicitly known to the
>endpoint receiving the information."

These security & privacy considerations are MTG-specific.  If (as I do
wish) MTG is ported to PLUS, MTG will have to define how security &
privacy of its own vocabulary is achieved.

Section 6.3 of draft-trammel-spud-reqs already has text that takes this
scenario into consideration:

   [...] SPUD should support different levels of
   trust than the default ("untrusted, but presumed honest due to
   limitations on the signaling vocabulary") and fully-authenticated;