There are some very serious problems with the use of encryption in multiple domains of trust on the Public Internet. Getting beyond the boundaries of each domain (sphere of control) is a dilemma that the folks in the IETF are now grappling with. Some have suggested that secure voice on the Internet (as we know the public Internet, anyway) is years off. But most current deployments of voice services that involve IP are not on the Internet, yet. They are on private IP backbones for the moment.
Securing voice is a more manageable prospect in the private IP backbone domains than it is on the 'net. The private IP b-b's will also be used in connection with initial Cable, DSL and other variations of ITSP offerings.
They simply use IP instead of PCM, for the most part. But the device controls are still centralized within the network, as opposed to being supported by the IP protocol itself, as would be characterized by end point control.
The current flavors of voice using IP are not true Internet Telephony architectures, they are instead Voice over IP, a.k.a., IP Telephony. There is a difference.
VoIP/IP Telphony emulates, to a great extent, the PSTN, using the IP streams to optimize bandwidth utilization on the long hauls. And in the newer short haul voice realms, such as Cable and DSL, they will be used for the same purposes: To optimize the number of voice channels that can be achieved while still allowing ample capacity for data in an integrated manner.
Internet Telephony, on the other hand, uses Internet protocols through and through. It is not intended to be ready until certain transitions are achieved through several migration stages of VoIP/IP Telephony, first. FWIW. |
