Several wireless technical forums, e.g., 3GPP, 3GPP2, and MWIF, have agreed upon SIP as the basis of the session management of the mobile Internet. It seems that SIP will certainly be an integral part of the mobile Internet's protocol architecture. Thus, it would be desirable to use SIP to provide means of terminal, service as well as personal mobility for all applications. The underlying rationales for seeking such a solution are

1. the expected growth of Internet multimedia services, particularly, the expected migration of voice telephony (i.e., VoIP) onto the Internet,
2. strong likelihood of using SIP for supporting service and personal mobility, and
3. the belief that SIP can also support terminal mobility with minimal extensions.

The advantages of using SIP for supporting mobility are that it

• allows users to depend on their appliances rather than the network for supporting mobility on an end-to-end basis without reliance on and knowledge about abilities of network elements for packet interception and forwarding, i.e., mobile users can roam across SIP environments without concern about whether they support network layer mobility or not,

• provides a means of route optimization and improved performance for real-time services via SIP signaling messages for address binding, registration, etc., and
• allows dealing with mobility at a semantic level above IP terminals (e.g., moving of a media stream from one terminal to another).

In the short term, supporting mobility with SIP at the application layer complements other approaches that rely on network layer mobility protocols (e.g., mobile IP, GPRS) though it may replace them in the longer term.