Thanks Gary, needed a sanity check of some kind. Not implementing Lawful Intercept per se - we won't be implementing the various mediation repositories for data ready to be pushed to a LEA domain etc., instead we would be focused on a narrower area of monitoring certain signalling events and then if trigger conditions occur e.g. bandwidth, user identities, media endpoints/type, attempt to reroute the media stream and perhaps attempt possible filtering of the contents of the media stream.
No, you're right, our design would lead to a man-in-the-middle point in the media flow, and thus be detectable (but then i haven't seen any LI that isn't detectable in some way) but it wouldn't detrimental in our case, as the user policies will contain a monitoring provision. Cathal On 3/16/07, Gary Cote <[EMAIL PROTECTED]> wrote: > > What you're describing is conceptually feasible. The AS would be a B2BUA > that > terminates one leg of the call and originates the second leg. If the > AS hosts the > RTP proxy, it would insert its address into the SDP. Or the RTP proxy > could be > off-board. > > You mention Lawful Intercept. Are you in fact trying to implement LI, > or just do > something similar? > > The reason I ask is that I've always been under the impression that one of > the > requirements for lawful intercept is that the feature should be > "invisible" to the > client, otherwise you potentially tip off the bad guys to the fact > that they're being > monitored. Not sure where I picked that up, so it could be right or > wrong. Anyway, > the scheme you're describing might not meet that requirement, as the > signaling > and media paths would be different for a monitored call vs. a > non-monitored call. > > -- > Gary Cote > www.awardsolutions.com > _______________________________________________ Sip-implementors mailing list [email protected] https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
