Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: [802.21] AW: [802.21] Ad hoc telecon for Dec 13th

a few notes for clarification. What we mean by NAT/NAPT/FW traversal is the fact that, depending on the transport used (e.g. TCP or UDP) there may be issues that IETF needs to consider. FW traversal is a recognized problem for several signaling protocols, even if as you correctly say it is a matter of configuration by the SPs. However, for any solution to be generic, it must be able to traverse FW in the worst scenario possible (i.e. FW with strict configurations), since we cannot foresee up front what the various SPs' policies are for their FWs. The problem in particular is when , after a "registration" between the MIHF in the terminal and the MME in the network took place, there is some inactivity for a period of time between the two, and then the MME needs to send a command to the terminal. Such command is a packet on the downlink that may not be able to pass through the FW. This is the case with UDP being used as transport, since the FW may have deleted the UDP state cre!
 ated when the terminal initiated the communication. Even with TCP, it is difficult to expect that a single TCP session is used for the whole duration of the ES/CS "registration session", since you would have keep-alive issues for long sessions with little activity, therefore you would need a new TCP session to be able to send the command from the MME to the terminal. This would be an incoming TCP session for the terminal, and most FW would not allow such session to pass through (e.g. this is the common case today in cellular networks). 
The main point here is that, as a requirement for the IETF work of developing the transport, we need to ask IETF to ensure that NAT and FW traversal are enabled. If IETF decides there are no issues, then it's even better, otherwise they need to ensure NAT/FE traversal is possible.


>-----Original Message-----
>From: ext Kalyan Koora [mailto:kalyan.koora@BENQ.COM] 
>Sent: Tuesday, December 13, 2005 2:28 AM
>Subject: [802.21] AW: [802.21] Ad hoc telecon for Dec 13th
>Hi Srini,
>I find the slide set far more better, especially the discovery 
>I have a small clarification need to the new point added by 
>you. I will just point it out here and can be discussed later 
>in telconf.
>Regarding firewall traversal, this is a set of rules specified 
>by a service provider itself (i.e. blocking couple of sites, 
>IPs or MAC addresses).
>It is not clear to me how we are going to address this in our 
>-----Ursprüngliche Nachricht-----
>Von: Srinivas Sreemanthula [mailto:Srinivas.Sreemanthula@NOKIA.COM]
>Gesendet: Dienstag, 13. Dezember 2005 05:24
>Betreff: Re: [802.21] Ad hoc telecon for Dec 13th
>Hello all,
>Here is the slideset that is built on top of last meeting and 
>some email discussions. We can use these topics for open 
>discussions and draw some conclusions.
>>-----Original Message-----
>>From: ext Srinivas Sreemanthula
>>Sent: Monday, December 12, 2005 3:09 PM
>>Subject: Re: [802.21] Ad hoc telecon for Dec 13th
>>Here is telecon bridge info for  Dec 13th 9am-11am EST about
>>L3 ES/CS requirements.  An agenda will follow soon.
>> US Phone Number: 972-894-6500
>> EU Phone Number: +358 7180 71870
>> Conference ID: 37494, PIN: 561988
>> Type of reservation: Single reservation: 13.12.2005
>> (GMT-06:00) Central Time (US & Canada)  Number of
>>participants: 20  Instructions language: English